iris/ai-app
1
0 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Compare commits

base: iris/ai-app:643daf56377c5030e4448fa2f061e0e07dc116f0
iris/ai-app:main iris/ai-app:rustify iris/ai-app:worktree-agent-ace98b0bdaf33ffff iris/ai-app:worktree-agent-a6e37a2335f436d08 iris/ai-app:worktree-agent-a9002910a315fe719 iris/ai-app:worktree-agent-a16b22e34539b810e iris/ai-app:worktree-agent-a1ff0294b6c29127e iris/ai-app:worktree-agent-a27094a7db775552a iris/ai-app:worktree-agent-afe80868604fef704
..
compare: iris/ai-app:main
iris/ai-app:main iris/ai-app:rustify iris/ai-app:worktree-agent-ace98b0bdaf33ffff iris/ai-app:worktree-agent-a6e37a2335f436d08 iris/ai-app:worktree-agent-a9002910a315fe719 iris/ai-app:worktree-agent-a16b22e34539b810e iris/ai-app:worktree-agent-a1ff0294b6c29127e iris/ai-app:worktree-agent-a27094a7db775552a iris/ai-app:worktree-agent-afe80868604fef704
114 Commits
643daf5637 .. main
Author SHA1 Message Date
iris-aiandClaude Opus 5 53fc59a946 Draw a model's thinking as markdown 
A model reasons in the same headings, lists and fenced code it answers in,
so plain text put rows of hashes and asterisks around the working the reader
opened the card to read. The card now draws `MarkdownText`, live while the
block is still open so a delta costs a parse of its last block rather than
of the whole reasoning, and the words answer the card's own tap through
`LocalMarkdownTap` the way a memory note's do.

A settled block is warmed with the rest of a page; an open one deliberately
is not, since warming a prefix per delta is a parse of the block per delta
held for ever. Echo's `/think` fixture is markdown now, because a fixture of
flat prose exercises none of this.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-21 22:43:45 -04:00
iris-ai 3dbf04f5ec One control per kind of setting, and no paragraphs under any of them 
Three explanatory paragraphs were left on the session settings screen: what
auto-resume does, what Move costs, what changing the thinking level costs.
The last two are consequences of an action, so they are `RestartDialog` now --
the same question the machines tab already asks before it reloads a model, and
now in one file rather than private to that screen. The first is gone; the
switch beside it says what it is.

Thinking was two controls: a row of chips for a llama session, where it is a
provider-declared choice, and a picker button for a Claude session, where it
is the session's own effort. One kind of information drawn two ways, decided
by which code path the value came down. A choice is `PickerRow` in a list of
settings and `ChipGroup` on a form being filled in, and which of the two is
the screen's to say rather than the provider's.

`warnAboutRestart` went with them. It marked a control "(on restart)" and
wrote a sentence under the form, and it had nothing to mark: every session
param is `restart: false` and every model param is `restart: true`. The field
now decides whether saving a model's settings stops to ask, which is the
question it was always about.

The wait's bar keeps the status row's own margin instead of running to the
edges of the glass.

Checked on the emulator against the sandbox: an echo, a claude-cli and a
llama session, each with no paragraph left under a setting and thinking drawn
the same way in all three; "Think high?", "Move to /tmp?" and the model
settings dialog all stop to ask. ktfmt, compile, lint and the unit tests are
clean.
 2026-09-21 12:25:49 -04:00
iris-ai 5626a7d595 A setting is a title and a box, and a wait's bar spans the screen 
A field's hint had a grey line of its own between the label and the box, so a
form of a dozen settings was mostly explanation. It is the empty box's
placeholder now, where it costs no height and says the same thing to anybody
who has not typed yet.

The wait's bar was drawn in whatever width the status row had left beside its
words, which is where the context figure goes -- so a session loading a model
or reading a prompt stopped saying how full it was, at the moment somebody is
watching it fill. The bar is the whole width above the row instead, the
compaction's indeterminate one with it, and the percentage joins the status
word ("reading prompt 30%") rather than replacing the context. The row keeps
the bar's height when there is no bar: it sits above the transcript and the
box, and one that came and went would move both under the reader's thumb
every time a turn started.

Checked on the emulator against the sandbox's progress rig: /loading and
/reading both draw a full-width bar with "context unknown" still beside them,
and the settings form has no grey lines left between a label and its box.
ktfmt, compile, lint and the unit tests are clean.
 2026-09-21 12:13:08 -04:00
iris-ai 1aac22bfc9 Say a transcript's size, name a model by something, and ask before a reload 
Five things asked for on the phone, and one trap behind the first of them.

A llama prompt carries `<__media__>` where a picture was, and llama.cpp
pairs each marker with a decoded image when it tokenizes -- so a marker in
words nobody attached a picture to fails the turn, and then fails every
later one, since the conversation is folded out of a transcript that holds
it for good. A model saying the marker back is enough to do it. Every
message with words in it now goes through `without_marker`.

A model's `general.name` is filled in by whatever converted the file, and
`convert_hf_to_gguf.py` fills it from the directory it converted: Prism ML's
Bonsai publishes `general.name = "Hf"`, which is unique and so passed the
label cascade and told a reader nothing. A name is now used only where it
shares a word with the repo or the file it came from; one that does not
drops to the file name.

The model settings dialog and the server card said what a save would cost in
a paragraph under the control, read after the decision if at all. Both ask
instead, in the shape the rest of that screen already uses for Stop and
Delete -- and the dialog's question is asked over the edits, so Cancel comes
back to them.

A field's label was `labelMedium` in the variant colour while every setting
beside it was body text, which on one form read as two ranks of setting.

`GET /sessions/{id}`'s `transcriptFile` now carries the file's size, and the
settings screen draws it beside what this phone has cached.

Checked on the emulator against the sandbox: the labels line up, the row
says "14 kB · 14 kB cached", and both confirmations appear over a loaded
Qwen3-0.6B. `cargo test` 275 passed, clippy and fmt clean, lint clean.
 2026-09-21 11:59:26 -04:00
iris-ai 4b5ed6e398 Review fixes: keep the keyboard off the settings form, and key the waits apart 
The settings dialog was a dialog, so the platform moved it off the
keyboard; a screen is not, and the lower half of the form was under it.

A tool call registered its abandonable wait under the call's own id, which
the model chooses -- one named `generating` would have taken the turn's
entry and left the other wait with nobody to answer it.
 2026-09-21 03:39:06 -04:00
iris-ai 7278a58387 Settings as a screen with two tabs, and fields that cost one line 
The settings dialog had outgrown a dialog: it scrolled inside itself,
covered the session it is about, and had nowhere to put a second tab. It
is a screen now, drawn over the session like the file explorer so the
session under it stays composed, with the back gesture to leave it. The
second tab is ProviderScreen itself -- the same composable the machines
tab opens -- so a provider's settings have two ways in and one
implementation.

Every text field in the app goes through LabelledField: the label is a
line above the box rather than a thing floating inside it, the hint says
what leaving it blank means, and the padding is one line's worth.
Material's outlined field spends the height of three lines to hold one,
which on a form of a dozen settings is a screen and a half of scrolling.
The value's own text is unchanged -- the framing was what cost.

A session also gets a system prompt, which for llama.cpp is one entry in
the params table and no app change: it rides in front of the conversation
on every request rather than being recorded as the first thing in it, so
changing it takes effect on the next message. ParamKind::Prose is new
because a paragraph in a one-line box shows six words of itself.
 2026-09-21 03:35:54 -04:00
iris-ai 386c1c4def Pause a llama turn at once, and let Stop take the model with it 
A turn waits on three things that look nowhere at all: a permission
question, a tool call with a minute to run, and the completion itself,
which says nothing while the prompt is read -- tens of seconds on a long
conversation. Setting a flag left the turn exactly where it was until
whichever it was came back.

The wait is now what ends, not the work. Each of those runs on a thread of
its own and the interrupt answers the wait; the abandoned thread finishes
into a channel nobody is reading. 43ms to end a turn in every state,
measured against a real model -- including mid prompt-processing, which
used to be a twenty-second wait. That makes cancellation a token per turn
rather than a flag on the session: the abandoned thread wakes up some time
later, and a flag the next turn had reset would let it write into a
conversation it is no longer part of. The open thinking block moves to
Shared for the same reason -- the thread that knows one is open is no
longer the thread that ends the turn.

Stop, meanwhile, did nothing at all to a llama session: it signals the
session's recorded process and process::stop refuses a Shared one, which
is the whole point of that record -- so the session sat at idle. A Shared
record routes to the driver now, because what stopping means for a session
that borrows the machine's process is the driver's to say. It ends the
turn, says exited itself, and gives up its claim on the model; each live
session claims the model it is on, and the model is unloaded when the last
claim goes. A model another session is using stays where it is.
 2026-09-21 03:16:55 -04:00
iris-ai 849c3b599f Don't draw a wait's bar at nothing 
A fraction of zero is the absence of a sample, not a measurement of the
work: a bar at nothing for two minutes says the load has not started,
which of one 8 GB into a 12 GB file is false.

Measured raw off the router's event stream today: llama-server reports a
model's load as 0 and then 1 with nothing in between, on both the 0.6B and
the 27B, and reports prompt processing once a batch. So the bar now draws
for a long prompt -- the wait worth watching -- and a load that says
nothing draws none, which is what not knowing should look like.
 2026-09-21 03:01:02 -04:00
iris-ai 66b3403a71 Read a session's transcript as the file it is 
The conversation on screen is a drawing of the record, and when the two
disagree -- or when something in the record is what has gone wrong -- there
was no way to see the record itself from the phone.

View raw in the session settings dialog opens the file explorer on the
transcript, which is its third caller and needed no new screen: the file
name and path in the header, a back button, and the lines as code. The
session says where the file is, because only the backend knows, and it
names this backend's machine rather than the session's -- for a remote
session those are two different filesystems. Back from the file lands in
the session's own directory, where the log and the process record are.

The paragraph under Reload goes, and both buttons move to a line of their
own: two buttons and a measurement do not fit a phone's width.
 2026-09-21 02:51:19 -04:00
iris-ai 049780fda6 Say how far a llama session's wait has got 
Both of its waits are measured somewhere and neither reached the phone: a
model coming off disk, which the router publishes on its event stream and
nowhere else, and a prompt being read, which the generation stream will
report when asked. A session now answers GET /sessions/{id}/progress with
{of, fraction, stage?}, one thread per router keeping the load's fraction
per model, and the session screen asks twice a second while it is drawing
a wait that has one.

Asked for rather than emitted: a load reports five times a second, and an
event is a transcript line for ever. The sample says which status it
measures, so one that outlived its wait cannot be drawn under another
word. The phone puts the bar in the status row's free width and the
percentage where the context figure sits -- a row of its own would move
the transcript every time a turn started -- and names the stage where a
model loads more than one file, because the fraction starts again for
each. /loading and /reading in an echo session are the rig.
 2026-09-21 02:44:03 -04:00
iris-ai 78f2fe3b79 Make Pause end a llama turn that is running a tool 
A llama turn's interrupt set a flag that the streaming loop and the tool
loop check, and nothing else. A tool call looks nowhere at all while it
runs: llama-server runs a shell command to its own timeout, up to a
minute, and the turn sat there for all of it with the phone showing a
Pause that had done nothing.

The wait is now a channel with two writers -- the call's own thread and
the interrupt -- so whichever speaks first decides what the model is
told. The call is left running on the machine and its answer dropped,
since nothing in that protocol takes one back. The same release covers a
permission question, so interrupt, detach and stop share one method.
 2026-09-21 02:19:36 -04:00
iris-ai 6ed896f1e1 Record the install trap behind a llama.cpp build that never answers 
`llama-server` is a 16 KB launcher against `libllama-server-impl.so`, so a
build installed without a working runpath dies at exec and reaches the phone
as a model that never became ready. `GNUInstallDirs` picks `lib64` on some
distributions while the recorded runpath says `lib`, which is how the two
flags come apart.
 2026-09-21 02:05:24 -04:00
iris-ai 7e7910083c Let a machine have more than one llama.cpp 
A model whose kernels are not upstream needs the fork that has them, and
the ordinary models still want the ordinary build. Anything under
`~/.local/share/ai-app/llama/<name>/` -- `llama-server`, or the
`bin/llama-server` a `cmake --install --prefix` leaves -- is now
discovered beside the one on PATH and becomes a provider called
`llama-cpp-<name>`, with its own router, preset and model settings.

That keeps the module's security property rather than bending it: the
phone still names no command, because what runs is still decided by what
somebody put on the machine. Each probe answer is tagged with what was
asked for, since two of these are now the same program under different
paths.

Flash attention joins the model settings (`flash-attn` in the preset).
llama.cpp's `auto` stays the default; the control is for a model whose
publisher asks for `on` outright, which Prism ML's ternary Bonsai does.

Verified against the fork built into that directory: discovery answers
`llama-cpp-prism`, the child server is started with `--flash-attn on`,
and Ternary-Bonsai-2-27B PTQ1_0 loads and answers through a session.
 2026-09-21 01:37:36 -04:00
iris-ai df48a334f7 Draw a background task's command as code, and frame the card it opens 
A backgrounded command in a session's panel was a terminal glyph beside its
words. The glyph said "command" and so did the monospace face, which is the
same fact twice -- so the mark goes and the command is drawn the way every
other verbatim thing in this app is: highlighted, monospace, on the raw
surface. The glyph stays for the kinds whose description is prose, and for a
command a provider never named.

Tapping one landed the tool card against the bottom edge of the screen, since
a reversed list anchors an item by its bottom -- so a tall card arrived at its
last line. A card that fits the viewport is centred now, and one that does not
has its top put at the top, which is where reading it starts. Only for a
journey the reader asked for: a restore still puts them back exactly where
they stopped.

The panel's own close chevron is gone -- the drag, the scrim and Back all
close it -- and the count is drawn even when it is zero, so "nothing is
running" and "nobody has asked yet" stop looking identical. There is then
nothing to expand, so that heading carries no chevron either.

Checked on the emulator against the sandbox: the card centred in a
mid-transcript tap, sat at the newest end where the list clamps, and the
zero heading drew with no control.
 2026-09-20 19:40:07 -04:00
iris-ai cedb18e8c1 Draw a background task as what it ran, and go there on a tap 
A card in the session's panel said "background command" under every
description -- and for Codex, which names a terminal by a process id and
gives no description at all, that phrase was the whole of every card.

Both halves of the answer are in the transcript rather than in what the
provider says: a driver now reports which tool call its task belongs to
(Claude's `task_started` carries the `tool_use_id`, Codex's terminal list
the `itemId`), and `LiveSession::background_tasks` resolves those ids
against the transcript into a sequence number and, where the provider said
nothing, the command the call was made with. So the card draws the command,
and the kind shrinks to a mark beside it whose name is what a screen reader
is given.

Tapping one goes to that call in the transcript, opened, which is where a
backgrounded command's output already lands -- rather than drawing a second
copy of it beside the panel. The journey is the one a reopened session
already makes to put a reader back where they stopped, now one function
(`travelTo`). It has to release the held backlog first: events arriving
while the reader is away from the newest end are held rather than applied,
so a task started since they scrolled back was in no row at all and the tap
looked like it had done nothing.

Verified against the sandbox on the emulator: the panel draws
`sleep 120 && echo done` for an echo session's `/background`, and tapping
it lands on that Bash card with its output showing.
 2026-09-20 18:46:22 -04:00
iris-ai 3b309766d7 Keep a typed path as typed, and expand ~ where it is used 
A llama session's tools all answered "failed to spawn process
[exit code: -1]": `llama-server` takes the working directory as an
`x-tool-cwd` header and `chdir`s to it with no shell in the way, so a
`~/…` cwd named a directory of that name. `files::resolve_blocking`
asks the machine that will serve the session what the path is, and the
driver does that once at launch.

The other half is the storing. `~` and `/home/someone` are a path and a
snapshot of where it pointed, and it is the snapshot that breaks when an
account is renamed -- so `machines::tidy` no longer expands one and
`set_cwd` no longer contracts one (`shorten_home` is gone with it). The
identity file is expanded at the point `ssh` is invoked instead.

Spawn now asks the same question of a typed working directory that
`set_cwd` already did: absolute or home-relative, and actually there on
the machine that will run it. It accepted anything, so a typo became a
session whose process could not start, reported later and pointing at
nothing.

Also: a provider written before `mcp_servers` existed adopts the
defaults a probe would give it now, so a machine discovered before
2026-09-19 stops silently having no web search.

Exercised end to end against a real llama session spawned with
`cwd: "~/repos/ai-app/server"`: `exec_shell_command` with `pwd` answered
`/home/bob/repos/ai-app/server`, exit 0.
 2026-09-20 17:06:13 -04:00
iris-ai bd9596d782 Let a llama session be shown a picture where the model reads one 
A multimodal model is loaded with the `mmproj` found beside its weights --
which is how a repository publishes the pair -- and an attached image rides
in the request as an `image_url` data URI, so it reaches a model on another
machine without the file going there. Nothing is done for a model without a
projector: no captioning, no OCR, no second model.

Whether a session takes pictures is measured rather than assumed:
`/props`'s `modalities.vision` from the server that loaded the model, in
three states, because a model still coming off disk has genuinely not said.
Unknown is offered rather than refused -- a control withheld because nobody
could ask goes missing from sessions that would have taken it. The answer
reaches the phone twice per model as `Event::Images`, so the photo button is
withdrawn the moment a model with vision is left rather than at whatever
later point the session row is fetched again.

A message carrying an image a model cannot read is stopped rather than
stripped: `llama-server` refuses the whole request over one image part, and
a message sent without its picture would be answered as though the picture
had never been mentioned. The phone will not attach one, and the driver
refuses it again at the three moments the answer can first exist -- at the
door, when a message queued behind a loading model is read, and at the tool
boundary a steer enters by. An earlier turn's image folds into a line of
words for a model without vision, so switching a conversation onto one does
not end it.

A projector is filtered out of the models a provider *offers*, since a
session started on one is a server that cannot load it; it stays in the
machine's own model list, where a file on a disk is managed.

Verified against ggml-org/SmolVLM-256M-Instruct-GGUF, local and over ssh:
"In this picture there is a red circle." Switching that session to
Qwen3-0.6B reports `refused`, refuses the next picture with the reason, and
still answers an ordinary message.
 2026-09-20 16:19:53 -04:00
iris-ai b7fd18b195 Let the reader put the session list in its own order 
Nothing sorts the sessions tab any more. The order is the server's
`sessions` list, which is the reader's arrangement: holding a row puts the
screen in selection mode -- the same gesture and the same bottom bar as the
import tab -- and each card grows a burger handle at its right edge that
drags the row to a new place, with a tick of haptic feedback for each one it
passes.

The two attempts this replaces, sorting by activity and then by when each
agent was turned on, were both looking for an order a session could not move
itself out of; no rule computed from what a session is doing can be one.
`POST /sessions/order` rewrites the config's order, so it is the same on
every device and survives a backend restart, and `SessionConfig::started`
goes with the sort that needed it.

Rearranging is independent of the selection: the handle moves the row it is
on, picked out or not. The click moved off the card and onto its contents so
that a press landing on the handle cannot also select the row it is about to
move. Selection's one action is Delete, which now takes the whole set.

Two traps in `Reorder.kt`, both measured on the emulator and written down in
`this-machine-android`: a crossing is decided from how far the finger has
travelled, because a lazy list animates an item into its new place and its
`offset` reports the old one for several frames; and the viewport is pinned
with `requestScrollToItem` around each move, because a lazy list keeps its
place by the key of the top item and would otherwise follow the row being
dragged.

Verified on the emulator against the sandbox: the order survives an app
restart and a backend read-back, a two-row drag moves exactly two rows, a
drag to the bottom edge scrolls the list and lands the row last, pressing the
handle without moving changes nothing, and deleting two selected sessions
leaves the rest in place.
 2026-09-20 01:06:15 -04:00
iris-aiandClaude Opus 5 942edd6b31 List a session's background tasks above its subagents 
The count beside the status said how much work was going and never what,
so "3 bg tasks" was a number with no way to find out what it was about.

Drivers now report the tasks themselves rather than a size:
`Driver::background_tasks` returns `Vec<BackgroundTask>` -- id, the
provider's own description, and a kind -- served by
`GET /sessions/{id}/background`. It is runtime state, never persisted,
and `null` is "nobody has said", which is what a session with no process
answers and what the panel says in words rather than drawing as an empty
list. `description` is optional because Codex names a background terminal
by a process id, and a number drawn as a name is worse than admitting
there is none.

Claude's `background_tasks_changed` entries turn out to be objects
carrying `task_id`, `task_type` and `description`, so each is read rather
than counted -- and an `ambient` one is now dropped from the list and the
count alike, on the CLI's own instruction: a live-update watcher is not
activity, and counting one left a session reading `waiting` with nothing
to wait for.

The phone draws them in the right-hand panel above the subagents,
collapsed to "2 bg tasks running" and pushing the subagents down when
opened. Both lists are items of one lazy column, so neither can run off
the panel, and the section is refetched whenever the live count moves --
a card for work that has finished is exactly the stale measurement the
count exists not to be.

Verified against the real Claude CLI (2.1.261): a backgrounded `sleep 120`
came back as `{"id":"br16327wr","description":"Sleep for 120 seconds",
"kind":"command"}`, and on the emulator against the echo rig the section
appeared, expanded, and dropped a card as its task finished.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-19 23:29:40 -04:00
iris-ai c8bfc958ad Slide the main screen over a session with a right swipe 
Switching conversation was a step back to the list and a step down into
another, which disposed the session being left and refetched its whole
transcript over the tunnel on the way back. A right swipe now pulls
MainScreen itself over the open session -- the screen Back would have
shown, moved over the session instead of replacing it -- and swiping it
back off returns to a live stream, an unsent draft and the scroll
position it had. Tapping the session already open is that same swipe
back; tapping another is a screen of its own; deleting the one
underneath closes the screen, since there is nothing left to return to.

One gesture drives both this and the subagent panel (SidePanels.kt, now
the home of the drag and animation SubagentPanel had): two draggables
over the same content cannot share a horizontal drag, so the position is
a single signed reveal, negative left and positive right, which also
makes it impossible to have both open. The panels exist only inside a
session, so nothing on the main screen swipes anywhere.

Full width and no tonal step for this one, because a screen standing in
for another must be the same colour as it; the subagent panel keeps its
88% and its sliver. The list keeps its rows while it asks again -- the
panel refetches on every open, and blanking it each time handed the
reader an empty screen about something never in doubt -- with a bar over
the top while an answer is outstanding.

Where the panel has got to is read from draw lambdas only: it changes
every frame of a drag, and a body that reads it recomposes the session
beneath once per frame. Composition sees booleans that change twice per
gesture, the same correction the keyboard inset needed.

Verified on the emulator against the sandbox with ui-trace: the panel
opens and closes on the two swipes, tapping another session replaces the
screen, deleting the open one leaves for the list, the subagent panel is
unchanged, and neither swipe does anything on the main screen. ktfmt,
compile, lint and the unit tests are clean.
 2026-09-19 22:15:14 -04:00
iris-aiandClaude Opus 5 ef788b0405 Queue a llama message sent while its model loads 
A message sent into a loading session was recorded as *read* the moment it
arrived: the phone drew it as sent, nothing read it for the next minute, and
the turn then folded the conversation out of a transcript that by then held
that same message and appended it again -- so the model was sent it twice.
It queues now, exactly as a message sent into a running turn does: drawn as
waiting, takeable back, and opening the first turn when the model arrives.
The conversation is read before the message is announced, which is what makes
"everything before this message" true rather than a race against the pump.
`await_ready` is left for the one case that still needs it, a turn whose model
was changed under it, and the `idle` that used to close a load is now decided
beside that first turn rather than racing it.

Two silent endings found while reproducing it, both of which look on the phone
like a message that was sent and never answered: an `{"error": ...}` chunk
arriving mid-stream on an otherwise successful response (the GPU out of memory
mid-decode), and a stream that stops without its `[DONE]` (the model unloaded
under the session). Neither is an ordinary end; the turn fails for both, and
keeps whatever arrived before it.

Ran against a real llama session on this VM's Qwen3-0.6B: a message sent
during the load now queues and is answered when the model lands, and
unloading the model mid-reply now says so instead of going quietly idle.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-19 21:20:58 -04:00
iris-aiandClaude Opus 5 c3c6ab0ecf Steer a llama turn at its next tool boundary 
A message typed into a running llama session waited for the turn to end and
then opened one of its own, so a turn spending minutes on a chain of tool
calls read nothing sent during it -- which is the one moment steering is for.
It now goes into the request the loop is about to build, prefixed with the
same note every other driver's steer carries.

The boundary being ours rather than the CLI's has two consequences worth
keeping: a waiting message can be taken back right up to the moment it is
read, and an interrupted turn deliberately takes nothing, since a request
that is not going out must not record a message as read.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-19 19:30:35 -04:00
iris-aiandClaude Opus 5 81c30dcda1 Download a model onto the machine that will serve it 
The Models tab was about this backend's own disk, which is the wrong disk
for every session that runs anywhere else: llama.cpp reads the file where
it runs. So the models of a machine live under that machine's llama.cpp
provider now, beside the settings deciding how each is loaded, and the
download that produces one happens there.

A download is a detached `curl` on that machine, started by a script this
server writes and never spoken to again. Its state is a file beside the
partial, so nothing about it is held here: it survives the app closing,
this backend restarting and a second device watching, and the progress is
`wc -c` of the partial against the size HuggingFace published rather than
anything remembered. A run whose process is gone is reported failed, since
`kill -0` is asked at each listing, and there is no "finished" state -- a
download that finished is a model, in the list beside the ones still
going. Resuming is guarded by the published sha256, which is also checked
before the file takes its real name.

Two other things the same screens wanted:

A provider is drawn as a card rather than as a line of text, bordered
against the machine card it sits in -- the tint it had was one step along
the surface ladder and rendered as one flat block -- with room to tap and
no chevron.

Nothing in a raw block wraps any more; the block scrolls sideways
instead, one offset for all its lines, so a diff or a column-aligned test
run still reads as one.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-19 18:55:51 -04:00
iris-ai 8c323fc7a9 Serve a machine's models from one shared llama-server 
A llama.cpp session had its own `llama-server`: two sessions on one model
held two copies of it in memory, a model change bought a load only that
session benefited from, and the process was a session's to end. A machine's
models are now served by one `llama-server` in **router mode** -- no `-m`,
a preset file naming models and their flags, a child server per model asked
for, and each request routed by its `model` field. So one server per model
with that model's own settings is what a machine runs, while this backend
has one process, one port and one record per machine to keep track of.

The record is the mechanism every other driver already uses, so a restart
adopts it; a session records the same pid in its own directory as
`Detail::Shared`, and `process::signal` refuses to signal one of those --
which is what keeps stopping, deleting or cleaning up after one session
from unloading a model every other session is using. Nothing stops a router
on its own. That is deliberate (a loaded model is minutes of disk) and it is
why the machines tab now has a card per provider that opens its own screen:
how each model is loaded, how many stay in memory, Unload, and Stop.

How a model is *loaded* therefore belongs to the model on its machine rather
than to a session -- context size, GPU layers, threads, slots, speculative
decoding -- written into the preset as llama-server's own argument names.
Saving them re-reads that file, which unloads the model; that is the change
taking effect, and the dialog says so before you save. What stays a
session's is everything that rides on a request, including which tools it
offers: the router hosts one set for the machine and the choice is a filter
applied here, so it costs no reload (2,181 tokens of prompt with all seven,
698 with none).

Verified end to end against the scratch backend and the emulator: two
sessions sharing one loaded model with one child process, a second session
joining it with a 26ms prefill, a backend restart adopting the router and
answering with the prompt cache intact, the same over ssh to this VM, a
model's settings reaching the running server, Unload, and Stop leaving every
session `exited` with no error line.
 2026-09-19 17:37:31 -04:00
iris-aiandClaude Opus 5 74cda485e5 Give a llama session a thinking level, asked of the model 
A `thinking` param on the llama driver: "auto", "off", or a level, applied as a
chat-template argument on the next request -- `reasoning_effort`, or
`enable_thinking: false` for off -- so unlike the server flags it costs no
reload. It lands in the session settings dialog beside the other model
settings, which is what declaring it in `DriverKind::params` buys.

Which levels exist is the model's answer rather than a constant, because the
vocabularies disagree: the 27B here takes low, medium and xhigh and **raises**
on high and max, so a fixed list is a turn that fails on send. The driver asks
the loaded server (`thinking_options`) -- `chat_template_caps.
supports_reasoning_effort` for whether levels mean anything at all, which is
the gate that stops the control silently doing nothing on a template that
ignores the argument, then `/apply-template` per level, one cheap render each
at load time. Off is a separate argument and a separate question: honoured when
turning it off renders a different prompt, and both renders have to have
worked, since a template that refuses it also renders differently.

A level the loaded model cannot take is dropped from the request and said in
the transcript, naming what it does take. What is *not* said is anything about
a model nobody has asked yet: the answer is `Option<Vec<String>>`, where None
is "no server has been up" and an empty list is the model that genuinely takes
none.

Verified against the 27B on the GPU: "low" thought for 697ms and 79 characters,
"off" produced no thinking block at all, and "high" answered `this model does
not take "high" -- it takes off, low, medium, xhigh.` The picker wraps to two
rows in the settings dialog and shows the session's current value.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-19 16:14:35 -04:00
iris-aiandClaude Opus 5 369b8f7e52 Report what a reply spent reading its prompt, and pin the clock right 
`UsageDelta` gains `prefillMs`, llama-server's own `timings.prompt_ms`, so the
footer under a finished reply is "read 9.5s · 50.3 tok/s · 3:00 PM". Prefill is
the half of a turn that was invisible and is often the larger: measured on the
0.6B here, 1m 4s for the first turn after a model loads against 22ms for the
next, whose prompt the server still had cached.

The clock moves to the end of the line. Everything in front of it is a
provider's own measurement, so a session on another provider has fewer of them
or none, and a reader who has learned where the time is should not have to find
it again because the model changed. The costs grow leftwards into the space
instead, and a test asserts every shape of the line ends with the same thing.

Verified on the emulator against a real llama session: three replies reading
"read 1m 4s · 193 tok/s · 3:54 PM", "read 25ms · 308 tok/s · 3:54 PM" and
"read 22ms · 194 tok/s · 3:54 PM", with the clock in one column.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-19 15:57:56 -04:00
iris-aiandClaude Opus 5 b660905098 Say which half of the wait a llama turn is in 
A turn has two waits in front of the first token and they were one word.
`SessionStatus::Loading` was already the model coming off disk; this adds
`SessionStatus::Reading` for llama-server processing the prompt -- emitted when
the request goes out, cleared by the first thing the model says of any kind, so
it covers every generate in a tool loop rather than only the first.

Prefill is the expensive half on this machine: measured 9.5s for 6,068 tokens
and 22s for 14,068 on the 27B with the GPU to itself. Reported as `running`
that was indistinguishable from a model thinking, which is the thing the reader
is waiting for. The phone draws both with the working spinner and its own
words -- "loading model" and "reading prompt" -- and the session screen's
status row now spins for all three busy states instead of only `running`,
which is also how `loading` stops being a bare word with nothing moving.

Measured while checking the tok/s figure, and recorded in the rigs skill: the
27B holds 55.5 to 50.3 tok/s between 1.5k and 14k of context, so decode decays
gently, while the 0.6B on the CPU falls 30.1 to 11.5 over 6k. A shared GPU is a
different failure -- the model does not load at all.

Verified on the emulator against a real llama session: "loading model" while
the server started, then "reading prompt" with the spinner through prompt
processing, then the thinking card.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-19 15:44:04 -04:00
iris-aiandClaude Opus 5 bb5ac1a242 Draw a model's thinking, and what a reply cost to produce 
A llama.cpp session's `reasoning_content` becomes `Event::Thinking` deltas
closed by an `Event::ThinkingDone` carrying the span the driver measured, and
the phone draws it as a card of its own: "Thinking" with the spinner a running
command has, then "Thought for 12.4s". Deliberately not a tool call, so a run
of calls cannot collapse the reasoning into "Called 6 tools"; the reasoning is
also kept out of the next prompt, which `conversation` already ignored.

`UsageDelta` gains `tokensPerSecond`, the provider's own figure or nothing --
llama.cpp reports `timings.predicted_per_second` and the coding CLIs report no
such thing -- and a finished reply carries a small line under it saying when it
was sent and, where there is one, how fast it came out: "3:00 PM · 149 tok/s".

The compact usage bar drops the provider's name for the window and puts its
length after the time left instead: "42% · 3h 20m left / 5h".

Three things that had to come with it: the transcript coalesces runs of
thinking deltas as it does reply deltas, so one block is one row of a page
rather than a page of its own; `joinPages` welds a block cut by a page boundary
(`healSplitThinking`), since the half with no ending spun for ever; and
`UsageDelta` now reaches the fold, which is what carries the rate to the reply.

Verified on the emulator against a real Qwen3-0.6B session and the echo rig's
new `/think [seconds]`: the spinner while it runs, "Thought for 1.4s" and
"2:54 PM · 149 tok/s" after, the reasoning on tapping the card, and the usage
bar reading "42% · 3h 19m left / 5h".

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-19 15:07:25 -04:00
iris-ai 45f249ae91 Even out the composer's row, and put its two pickers in settings 
The composer's row gave the model and the permission mode whatever width
their words asked for, after three word-shaped action buttons had taken
theirs. A llama session's model names run long, so the permission mode
was squeezed to a chip too small to tap.

The three actions are circles now -- one diameter, the platform's minimum
touch target -- so they take what their glyph needs and nothing more, and
the two pickers share what is left evenly rather than by the length of
what they say. Every gap on the row is the same.

Both pickers are also rows in the session settings dialog, for every
provider that offers them: the dialog has a line each, so the whole model
name is readable there. Choosing goes through the same two functions as
the composer's pickers, so the model switch warning cannot be skipped by
picking from one of the two places -- and it closes the dialog rather than
stacking a question behind it.

Looked at on the emulator against the sandbox: a llama session with the
27B loaded, and a claude-cli one, both composer and dialog, with the
keyboard up and down.
 2026-09-19 14:28:38 -04:00
iris-aiandClaude Opus 5 81ab564a09 Declare provider settings, and give the context figure a denominator 
Two things a session could not say, and one it was saying wrongly.

**Every provider setting is reachable.** `-np 1`, the MTP draft depth, the
tool set, the sampling parameters -- most were hardcoded to what measured
best on this machine, which is right as a default and wrong as a constant:
the next machine has a different GPU and a different core count, and
nobody running this app can edit the source. `DriverKind::params` now
declares what a provider takes -- key, label, shape, what blank means, and
whether a change waits for a restart -- and the phone renders whatever
arrives, on the spawn form and in the session settings dialog. Adding a
setting to a driver is one entry in that table and no app change.
`POST /sessions/{id}/params` takes the whole map, so an absent key is the
instruction to unset; the sampling half applies at once and the session is
told in words which of the rest are waiting for a restart.

`tools` is one of them, because it is the biggest lever on a tight
context: the seven built-in definitions are ~1,300 tokens of every prompt
(2,191 against 887 with none). `"none"` omits the flag rather than passing
it on, since `--tools none` is `unknown tool "none"` and a server that
exits.

**The context figure has a denominator.** `Event::ContextWindow` carries
it, read from `llama-server`'s `/props` once the model is up -- the
measurement rather than the request, since a session that named no context
size gets the model's own. Neither coding CLI states its window, so those
keep the bare figure: "2,042" and "2,042 / 8,192" are deliberately
different-looking, and a missing ceiling is never drawn as a proportion of
an assumed one.

**And the numerator was wrong**, by the length of the last reply: it was
the prompt alone, so a five-word answer reported 2,042 against a slot
holding 2,355. It is the turn's total now, which matches `llama-server`'s
own `n_tokens` to within a token.

Two defects the review found, both of which would have shipped: changing
settings on a *stopped* session reported "no process running, so it can't
take new settings", when a stopped session is exactly when you would set
them for the next start; and `GET /tools` answers **403** rather than an
empty list on a server started without `--tools`, so reading it as a
failure made the no-tools session one that never started.

Verified against real models: settings spawned and changed live, the
restart note, a session with two tools and one with none, and the counter
checked against the server's own slot occupancy each time.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-19 13:58:08 -04:00
iris-aiandClaude Opus 5 ac476ab0c9 Give llama.cpp sessions tools, web search and a model picker 
A llama session was a chat box: no tools, a fixed model, no permission
mode, and a model name drawn as the path the file sits at. It now runs the
agent loop itself, which is what the pieces below all hang off.

Tools are `llama-server`'s own (`--tools all`), which that server both
publishes and runs -- `GET /tools` for the definitions, `POST /tools` to
call one. Web search is Exa's MCP server, reached from this backend rather
than from the machine serving the model: that is what llama.cpp's own web
UI does, and it puts the search on the machine with a route out instead of
the one with the GPU. `llama-server`'s `--mcp-servers-json` can only spawn
local commands, so using it would have meant a Node bridge on every
machine that serves a model.

Driving the loop is what makes the permission gate ours. Two modes,
`manual` and `bypassPermissions`, which is what the mechanism has: the web
UI asks before every call and remembers the tools you say "always" to. The
allowances fold back out of the transcript's own answers, so they survive
a restart and a model change without being stored anywhere else.

Also here, because tools made each of them matter:

- **Loading is a state.** A 12 GB model takes twenty seconds to reach
  memory and refuses everything until it has; the session used to report
  `running` for that whole time, and a message sent meanwhile came back as
  an error. It is `loading` now, and the message waits.
- **The model can be changed.** A `llama-server` holds one model, so this
  stops it and starts another. The conversation survives because it was
  never in the server.
- **Models are named, not pathed.** `general.name` read out of the file
  itself -- over ssh too, in the round trip the spawn was already making.
  Where two models share a name the file name breaks the tie.
- **`-np 1`, and the MTP draft head where the file has one.** Measured on
  the 27B here: 41.5 tok/s plain, 61.4 with `--spec-type draft-mtp` at one
  slot, and 28 with it at four -- speculating against a split KV cache is
  worse than not speculating. The flag is conditional because asking for a
  head that is not there makes `llama-server` exit.
- **A refusal says what to do.** Tool results are thousands of tokens, so
  an overrun context is now ordinary; it was "http status: 400" and is now
  the server's own "exceeds the available context size, try increasing it".

`GET /machines/{id}/models` is gone: the provider models route answers the
same question, and two answers to one question is how a picker comes to
offer a model the spawn screen does not.

Verified end to end against real models: a tool call asked and allowed, an
Exa search, a shell command, a 27B loaded while a message waited on it, a
model switch mid-session, a second message queued behind a running turn,
and the whole of it again on a session running over ssh.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-19 08:11:49 -04:00
iris-aiandClaude Opus 5 392cc5413d Recover Claude sessions with a stale resume token 
A resume token the CLI will not accept made the session unrecoverable
rather than merely failed: every later start passed the same `--resume`,
died the same way, and nothing ever forgot it, so the chat could not be
opened again from the phone.

The reader now recognises both refusals the CLI gives, forgets the token
and emits `Cleared`, which is what the Codex path already does for a
thread whose rollout has gone. The transcript is this server's and
survives; only the model's context restarts.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-19 06:56:41 -04:00
iris-ai 03376af446 Fix panel drag release flicker 2026-09-17 14:13:05 -04:00
iris-ai 0a2f0eed5f Move subagents into session side panel 2026-09-17 13:41:24 -04:00
iris-ai cd0229bed6 Show elapsed-time cursor on usage bars 2026-09-17 13:06:00 -04:00
iris-ai 84f978f16d Open a call from its foot upward, as a row already does 
A closed call inside a group opened downward wherever it was pressed: the
anchor asked for the group's top in every case, which is right for a tap on the
call's heading and wrong for one at its foot, where what the reader wants held
is the edge under their finger. It is the rule every other row has had since
the anchoring went in, applied one level down.

An open in the call's lower half now asks for no scroll at all, which is the
same answer a row gets and for the same reason: the list holds the group's
bottom edge, a Column keeps the calls below the one growing at their distance
from it, so the growth comes off the call's top. The closing behaviour is
untouched -- the arithmetic is the same shift, written as the one term it
cancels down to, since where the call sits in the group and where the group
sits in the viewport drop out of it.

Checked with ktfmtFormat, compileDebugKotlin, lintDebug and testDebugUnitTest,
and on the emulator against a sandbox group of six calls, with enough
conversation behind it for the list to actually scroll -- a transcript shorter
than the viewport pins to the bottom and gives every anchor the same answer,
which is how this was missed. A call closed at 1321..1447: opened from 1430 it
leaves the call below it at 1485 to the pixel and takes the growth off its top;
opened from 1360 it leaves the calls above it where they are and moves the one
below down by the full 327; closed again from 1400 it lands at 1338..1464,
centred on the tap within a pixel.
 2026-09-16 03:40:34 -04:00
iris-ai ee5bef3686 Centre a closed call on the tap, not the group around it 
A call inside an open group was the one case still anchored by an edge: the
group held its top, which is right when a call is opened -- the heading under
the finger is the edge being pressed -- and wrong when one is shut by however
far down the open card the reader pressed. On a card of output that is most of
the screen, and what it looks like is the card collapsing into its own top, a
long way from the hand. It is the same rule as every other close now: what is
left of the call lands centred on the finger that shut it.

A call is not a row, so the scroll is still asked for against the group and
merely aimed at the call. What makes that possible is the group reporting how
far down its own top edge the call is drawn and how tall that card is, which is
the part only it knows; the calls above the one toggled do not move, so shifting
the group by the difference puts the call where the finger wants it.

Checked with ktfmtFormat, compileDebugKotlin, lintDebug and testDebugUnitTest,
and on the emulator against a real imported conversation: a call opened from its
heading inside a group of twelve leaves that heading where it is, and closing it
again from the middle of its output at 1800 lands the closed call at 1738..1860
-- centred on the tap to the pixel. The group's own close still centres on its
heading (1287..1413 for a tap at 1350).
 2026-09-16 03:22:55 -04:00
iris-ai 914985b8b2 Ask for the scroll in the gesture, not from the layout 
A correction made from the layout is a frame late whatever phase it is made in:
from placement it is never picked up by another measure and does nothing at all,
and from measure it lands on the next frame with the uncorrected one drawn
first. That is the flick when a card is opened, and it is why a close could
finish somewhere other than where it was aimed -- the two are the same fault.

Asked for at the tap instead, the request is consumed by the same measure pass
that first lays the row out at its new size, so the resize is drawn once, in its
right place. What makes that possible is that none of the three positions needs
to know the new height. A top edge holds by placing the item *above* the row
where it already is -- that item's bottom edge is the row's top edge whatever
becomes of the row, and it does not have to be composed for the list to place
it. A close places the row itself against the height it had at the moment it was
opened, which is the height it is going back to.

So the measure-phase hold, its modifier and the list's `afterMeasure` hook are
all gone, and this is 75 lines shorter than the version that could not do it.

Checked with ktfmtFormat, compileDebugKotlin, lintDebug and testDebugUnitTest,
and on the emulator against a real imported conversation as well as the sandbox:
a group closed by its heading lands centred on the tap (1467..1593 for a tap at
1530), a card closed at 1500 and at 1800 lands centred on each, opening by a
heading holds the heading to the pixel, and opening or closing a real call
inside a group of two leaves that group's heading exactly where it was.
 2026-09-16 03:08:17 -04:00
iris-ai 581e07624f State where a resized row goes instead of walking it there 
Correcting by the error each pass could see, and asking again on the pass that
answered, was a frame per pass with the ones in between drawn: opening a card
visibly stepped. It also still missed, because two of the passes were spent
finding out what the list would do rather than telling it.

`requestScrollToItem` against the row itself says it outright, in one pass: the
row is placed wherever it has ended up and whether or not it is still on screen,
and a negative offset -- which is the list being asked for the rows below a card
that has just given the screen its whole height back -- is exactly what a close
needs and works. Everything else follows from where that puts it.

A call opened or shut *inside* a group is a third case, and it was being treated
as the second: the group is not the thing opening, it is the container, and
centring it on the tap threw a group of six the length of the screen. What keeps
the call under the finger is holding the group's top edge, so that everything
above the change -- that call's own heading included -- stays where it is.

Checked with ktfmtFormat, compileDebugKotlin, lintDebug and testDebugUnitTest,
and on the emulator against the sandbox with sampling as fast as the device will
report it, so an intermediate frame would show: a 2,785px card closed at 600,
1200 and 1800 lands centred on 600, 1200 and 1800 to the pixel, each in one
step; opening by the heading holds the heading still; and opening or closing a
call inside a group of five leaves the group's heading exactly where it was.
 2026-09-16 02:53:51 -04:00
iris-ai 1b38579b97 Land a closed card centred on the tap that closed it 
Two things were wrong with the hold, and each hid the other.

It held a *share* of the row's height: the point the finger was on stayed, in
proportion, which is the same miss in miniature as holding an edge. Tap away
from the middle of a long card and the heading landed most of a card's height
from the finger, and off it. A shut card is a heading, and the only place it
belongs is centred under the hand that shut it, wherever down the card the tap
was.

And the correction was worked out from the change in height, which needs the
list to behave the way the arithmetic assumed. It does not: which item it holds
still across a resize depends on what it has composed -- a row taller than the
screen is anchored on itself -- and a scroll it cannot honour in full is
honoured in part with nothing said. Measured rather than predicted now: each
measure pass asks for the error it can see, and the pass that answers is where
the rest becomes askable. Three passes is the worst seen, including the one
where a card that reached past the bottom of the screen has shut, left the
viewport entirely, and has to be asked back to the bottom edge before there is
anything to measure at all.

The pass has to be the *measure* one. A scroll asked for during placement is
never picked up by another measure and does nothing whatever -- which is what
the first version of this did, and why a close moved nothing -- so the list
took an `afterMeasure` hook and the correction lives there.

Checked with ktfmtFormat, compileDebugKotlin, lintDebug and testDebugUnitTest,
and on the emulator against the sandbox, against a 2,785px card in a
conversation with room on both sides: closed at 1450 it lands 1387..1513, at
1700 it lands 1637..1763, at 1950 it lands 1887..2013 -- centred on the tap to
the pixel each time. Opening by the heading still holds the heading still, and
a group closed from its footer bar lands on the bar. Where the conversation
runs out -- a card at the very start with nothing above it to scroll -- it
lands as close as the list can put it, which is what it could always do.
 2026-09-16 02:38:48 -04:00
iris-ai b86a5dc37a Hold an open call out of its run without taking one out of a group 
Being open did two things to grouping, and only one of them was wanted. It held
a call standing on its own out of the run it belongs to, so a command finishing
behind the card being read no longer shuts it and folds it away mid-sentence.
It also took a call *out* of the group it was already inside, and that is what
made collapsing jump: grouping is what gives a row its identity, so one tap
rebuilt the rows around the finger -- opening a call inside a group split the
group into two pieces with mismatched keys, and closing one replaced three rows
with one, which no anchor survives. Measured at 450px of jump, with the card
that was closed going with it.

So the held-out set is now the screen's, not the transcript's: a call that has
never been drawn inside a group and is open stands out of its run, and a call
that has been in one stays in it whatever the reader does to it. Being inside a
group once is a fact about what the reader has been shown, which is why the
screen is what remembers it.

Checked with ktfmtFormat, compileDebugKotlin, lintDebug and testDebugUnitTest,
and on the emulator against the sandbox: opening a call inside an open group of
six leaves it one group of six and closing it returns every row to the pixel it
came from; a call opened while standing alone survives a reply landing behind
it, and folds back into "Called 3 tools" when it is closed without moving the
rows below it.
 2026-09-16 01:57:50 -04:00
iris-ai 463acb28fa Close a card on the point that was touched 
Collapsing held one of the row's edges -- whichever the tap was nearer -- which
is right for opening and wrong for closing: the row that shuts leaves a heading
where a screenful of card was, and both its old edges can be a screen's length
from the finger that shut it. It now keeps the touched point itself, which for a
closed card is the same thing as landing under the hand that closed it. Opening
is unchanged and deliberately so: those rows are small, every point in them is
within a heading's height of both edges, and the edge pressed is what the reader
wants held rather than a fraction of an unbounded expansion.

One number carries both readings -- the share of the row's height above the
touch, spent as it is on a close and rounded to the nearer edge on an open.

The scroll offset the correction asks for goes negative on a close, and has to:
that is the list being asked for the rows below what it has composed, which is
where the newer content comes from when a card gives a screenful back. It was
clamped at zero, which was invisible while every correction was a row growing
and is what left closes uncorrected.

Checked with ktfmtFormat, compileDebugKotlin, lintDebug and testDebugUnitTest,
and on the emulator against the sandbox: a 1441px card closed at a quarter of
its height put the collapsed card's top at 743px against 743 predicted, and
opening a card by its heading still holds the heading still.
 2026-09-16 01:50:33 -04:00
iris-ai 827a30768c Count Codex background terminals 2026-09-16 01:07:40 -04:00
iris-aiandClaude Opus 5 cbae7ee8c0 Order the session list by when each agent was turned on 
A running session no longer moves: the ones with a process come first,
oldest start first, so starting one appends it to the bottom of that
group and nothing it goes on to do -- beginning a turn, finishing one,
asking a question -- can shift it. Sorting by activity with the
awaiting-answer ones floated to the top is what this replaces; the
status word and its colour already say which session wants something
without the row having to move to say it. Stopped sessions are a group
below, most recently active first.

The order is the server's: `SessionConfig::started` is written each time
a process is started for a session and reported as `started`, so it is
the same on every device and survives a backend restart -- which adopts
processes rather than starting them, and so could not work the times out
for itself. Applied on the phone, because presentation order is a
display decision.

`LiveSession::info` takes the session's config entry rather than a
parameter per field read from it, which is what `AutoResumeView` existed
to bundle; that goes.

Verified on the emulator against the sandbox: three echo sessions kept
their order while the newest-active one was messaged; a stopped and
restarted session moved below one started after it; a stopped session
dropped below every running one; and after a backend restart the
recorded times came back unchanged.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-15 23:54:36 -04:00
iris-ai a9cfea89e5 Show Codex background task counts 2026-09-15 23:26:37 -04:00
iris-ai 947ea8ecf2 Keep tool group keys unique across transcript 2026-09-15 23:20:47 -04:00
iris-ai f00a178cf0 Ignore replayed tool starts 2026-09-15 23:14:11 -04:00
iris-ai 9bcf0f1a48 Keep open tool cards out of groups 2026-09-15 22:43:45 -04:00
iris-ai 33b130b6bb Let failed messages be discarded 2026-09-15 19:03:44 -04:00
iris-ai 3d1b1e304d Render whole-file patches from change metadata 2026-09-15 16:18:18 -04:00
iris-ai 06bf1c8f81 Preserve messages during Codex thread recovery 2026-09-15 15:30:22 -04:00
iris-ai 3f94eeb6d6 Create Codex threads on first message after clear 2026-09-15 14:43:23 -04:00
iris-ai 1c60e78b55 Keep Claude commands out of subagents 2026-09-15 14:29:18 -04:00
iris-ai 8262ceb786 Show live background task counts 2026-09-15 13:44:32 -04:00
iris-ai 9fd21af4e8 Reconcile Claude background task state 2026-09-15 12:49:22 -04:00
iris-ai f0661919bb Offer Claude sign-in from failed sessions 2026-09-15 12:24:25 -04:00
iris-ai 0be15adbee Post the drawer's row behind the app's own banner 
A notification arriving while the app was open was shown as a banner and
nowhere else, so a moment that happened while the phone was face-up on a desk
left nothing behind at all -- the banner is seconds long and reaches only
somebody already looking.

The two are not two versions of one thing: a banner interrupts and a row
records. Both go up now, and the banner having done the interrupting is what
makes the row a silent one (`setSilent`), so one moment is worth a noise once.
What keeps the drawer from filling up is the other end rather than suppression,
and already was: opening a session clears whatever is posted about it, whichever
way the reader got there.

Checked with ktfmtFormat, compileDebugKotlin, testDebugUnitTest and lintDebug,
and on the emulator against the sandbox, reading the posted record out of
dumpsys: app on the session list gives a banner and flags=AUTO_CANCEL|SILENT;
app backgrounded gives flags=AUTO_CANCEL; opening the session leaves nothing
posted about it in either case.
 2026-09-15 02:06:12 -04:00
iris-ai 1e52b2910c Keep the last tool call outside its group once it finishes 
A call left its group only while it was running, so the moment a command ended
it vanished behind "Called 3 tools" -- and a session that has run its last
command and is composing its answer, or has finished the turn entirely, spends
most of its time in exactly that state. What folds a call back into its run is
therefore not finishing but being overtaken: anything arriving behind it, a
reply included, makes it history.

Standing outside the run is the call's place in the list as it is now rather
than something recorded on the call, so it is asked of the list while grouping
it, where the rest of that decision already lives.

Checked with ktfmtFormat, compileDebugKotlin, testDebugUnitTest and lintDebug,
and on the emulator against the sandbox: "/tools 3 1" settles as "Called 2
tools" with the third Bash card beneath it, and folds to "Called 3 tools" the
moment the next reply lands.
 2026-09-15 01:52:51 -04:00
iris-ai 036eb375aa Keep the running tool call outside its group 
A run of adjacent calls is drawn as one collapsed card, which hid the one
thing worth seeing without opening anything: the command the session is
running right now. It is a row of its own while it runs and folds back into
the run when it ends.

Grouping stays a display decision, so the pieces a running call cuts a run
into are keyed there. The first piece keeps the run's name -- that name is
what survives a page of history landing in front of it -- and later pieces
take their own first call's id behind it, since the call a run was named
after can itself be the one running.

The echo rig's /tools gap now runs between a call's start and its end rather
than between one call and the next, which is where a real session's time goes
and what makes the running state observable at all.

Checked with ktfmtFormat, compileDebugKotlin, testDebugUnitTest (new
ToolRowsTest) and lintDebug, cargo fmt/clippy/test, and on the emulator
against the sandbox: "Called 2 tools" with the live Bash card beneath it.
 2026-09-15 01:13:03 -04:00
iris-ai b9b777acaf Release messages after Codex recovery 2026-09-14 15:16:05 -04:00
iris-ai 46831520e3 Recover Codex sessions with missing rollouts 2026-09-14 15:01:47 -04:00
iris-aiandClaude Opus 5 3af2502982 Tell the model when a message was a steer 
A message typed during a turn reaches the model at the next model call if
the turn has one left, and otherwise as the opening line of the next turn --
Claude's read out of the fifo after the turn ended, Codex's requeued when
turn/steer is refused. Read there it is indistinguishable from a reply, so
the model treats the answer it just gave as seen.

Both drivers now compose the text the CLI receives through
driver::message_body, which prefixes a note saying the message was written
without having seen the rest of that turn. The transcript still holds the
words that were typed; only the CLI's copy carries the note.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-13 22:30:32 -04:00
iris 59ebd75b46 Route general lessons to the code-lessons skill 
The routing note under "Things that have bitten" pointed at
~/.claude/TOOLCHAIN.md, which no longer exists -- its contents were folded
into the this-machine-* skills. Name the destinations that do exist, and add
the third case: a lesson that would bite any project anywhere now has a home
in the code-lessons skill rather than defaulting back to here.
 2026-09-13 15:54:00 -04:00
iris 579689cbb8 Keep a thinking level the settings dialog set 
The dialog held the level for as long as it was open and read it back
from the frozen row the session screen was opened with, so reopening it
showed the old level until a return to the list refetched the row. The
level is the session screen's own datum now, like the title.
 2026-09-13 12:34:21 -04:00
iris fe25108c51 Count every Codex model request, not just the last 
App-server sends `thread/tokenUsage/updated` once per *model request*, and a
Codex turn makes as many as it made tool calls. The translator held the last
one until `turn/completed`, so a turn's cost was reported as its final
request alone -- measured against the real rollout, 28,878 tokens for a turn
that spent 51,399 -- and the gap grows with how much work the turn did. The
context figure also stood still for the whole turn, which is exactly when it
is moving most.

Reported as each arrives instead: `tokens` now adds up to what the turn
spent, and the context figure climbs during the turn (28,921 -> 33,190 ->
35,978 on a two-file read here, matching Codex's own `last_token_usage`
exactly at every step).
 2026-09-13 03:58:47 -04:00
iris 898e6b92d0 Clarify subagent coordination cards 2026-09-13 02:48:21 -04:00
iris cad0cbcfbe Keep subagent delivery out of assistant text 2026-09-13 01:05:35 -04:00
iris 83b113ef0f Support Codex subagent transcripts 2026-09-13 00:41:59 -04:00
iris 7d9df5d572 Rename setups and add provider reauthentication 2026-09-12 22:56:43 -04:00
iris e9a0f1b9da Do not enlarge images on open 2026-09-12 21:52:47 -04:00
iris 6d765ff6e4 Make image viewer truly full screen 2026-09-12 21:33:29 -04:00
iris 559e6c9226 Suppress errors for requested session stops 2026-09-12 20:49:38 -04:00
iris 76895bc644 Remove image viewer touch ripple 2026-09-12 20:28:28 -04:00
iris 0b4da64062 Fix image zoom focal point 2026-09-12 20:16:25 -04:00
iris 62cb6c91d5 Navigate explorer back toward project 2026-09-12 19:38:18 -04:00
iris 2ff0b13950 Return from files to explorer 2026-09-11 12:40:27 -04:00
iris 57e1cec09c Render Codex web searches as common tools 2026-09-11 02:31:21 -04:00
iris 6226a1cb43 Keep compact transcript history loading 2026-09-11 01:18:45 -04:00
iris 22f263ccce Restore directory navigation on Android back 2026-09-11 00:28:55 -04:00
iris 59965d314f Keep compact Codex history loading 
Restart the history observer after every successful page so a collapsed tool page cannot consume the only layout invalidation that would request the next one.\n\nVerified with a cold 365-event tool-heavy sandbox transcript: without scrolling or expanding a group, cache coverage advanced continuously to sequence 1. Android format, compile, lint, and JVM tests pass. Transcript bench: 26 rows/26 units loaded, transcript draw 0.72 ms per frame (debug emulator).
 2026-09-10 18:51:29 -04:00
iris e3cca97cdd Open transcript file links in explorer 2026-09-10 18:08:05 -04:00
iris 3c6e6778fd Keep sent messages visible until received 2026-09-10 02:17:11 -04:00
iris 3c19b5a9bb Fix Codex transcript convergence 2026-09-10 01:24:16 -04:00
iris f9c8f640ce Fix quoted Bash tool titles 2026-09-10 00:46:09 -04:00
iris 4c15150338 Return from files to explorer 2026-09-09 22:48:52 -04:00
iris cbdd8493ed Unwrap double-quoted Codex Bash commands 2026-09-09 22:17:14 -04:00
iris 26fe9895e7 Unwrap rendered Codex Bash commands 2026-09-09 22:11:52 -04:00
iris b00e89795e Parse Codex app-server patch payloads 2026-09-09 21:30:15 -04:00
iris 10ce1a216b Defer Codex patches until their diff arrives 2026-09-09 20:58:48 -04:00
iris b507656abd Normalize shell and patch tool cards 2026-09-09 20:30:52 -04:00
iris 4dc3e3d784 Fix Codex transcript streaming and images 2026-09-09 15:14:24 -04:00
iris 14dd520719 Fix explorer back and session usage selection 2026-09-09 13:01:27 -04:00
iris 8c88a7e991 Use native Codex steering and transcript deletion 2026-09-09 12:19:11 -04:00
iris 00538cc19b Show separate Codex usage pools 2026-09-08 00:41:06 -04:00
iris 7ee88dfd9c Make model and permission choices provider-specific 2026-09-08 00:08:09 -04:00
iris 6a0202b1b5 Add Codex JSON sessions and usage limits 2026-09-07 23:29:15 -04:00
irisandClaude Opus 5 0862b47f76 Record the loose end the taskNote outage exposed, and the last of its checks 
A session whose transcript will not parse is skipped with only a log line, so
from the phone it is indistinguishable from an idle unresponsive one. That is
why the outage needed a report from Bryan rather than showing itself. The
cause is fixed; the class is not, and it is the "design the unknown state
first" rule rather than a bug in one code path.

Also rustfmt on the parse path, which the fix landed unformatted.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-06 22:43:01 -04:00
irisandClaude Opus 5 fd71d876e1 Never let one unreadable line take a transcript down 
Removing `Event::TaskNote` hours after adding it made every transcript that
had recorded one unreadable. `Transcript::open` parses every line, so `launch`
failed for those sessions and `SessionManager::new` logged
"couldn't relaunch session <id>" and skipped them -- and a skipped session has
no pump and no driver. On the phone that is no status, no history and nothing
sendable, for every live session that had run a background task. One
unfamiliar word took down every conversation it appeared in.

A transcript is append-only and permanent, so the set of kinds one can hold
only ever grows: what this build writes is not what it may have to read. A
line can come from a newer server, or from an older one that wrote a kind
since dropped, and neither may be able to end the file.

`Indexed::parse_at` degrades a line it cannot make sense of to
`Event::Unreadable { kind }` instead of failing the whole read. It keeps the
line's seq -- the cursors, the page bisection and the next-seq counter are all
addressed by it, and dropping the line would hand out a seq the file already
contains -- and carries the word the line called itself, so the phone can say
what is missing rather than that something is. A line with no readable seq is
still an error: that one cannot be placed at all.

`Event::TaskNote` comes back retired rather than deleted: deserializable,
never constructed, dated, with the reason on it. The phone folds it to no row,
which is the point -- an unreadable line correctly draws a placeholder, and
one per background task is the wall the row was removed for in the first
place.

Found while diagnosing a report that live sessions had lost their status and
could not be sent to. 173 server tests pass, including the new one, which
fails on the old code within a second.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-06 22:23:45 -04:00
irisandClaude Opus 5 9cc52beb09 Report a backgrounded command into the card that launched it 
A backgrounded command has no subagent, so there is no second transcript for
its report to live in and its own tool card is the only record of it anywhere
-- and until the task notification arrives that card is showing the launch
result, which says the command is running. It was left saying that for ever.

The report now updates the call's own row (`Event::ToolUpdate` against its
tool_use id), so the card ends up holding what became of the command instead
of a claim nothing was ever going to correct. That includes the endings that
carry no summary: those are exactly the ones that went wrong, and a stale
"running in background" reads worst on them, so they say the status word
rather than nothing. A task with a subagent behind it is untouched and its
report stays where it was, in that subagent's own transcript.

Echo grew `/background [seconds]` for the shape end to end: the Bash call, the
launch result, a turn that ends `waiting`, and the completion arriving later
to correct the card and start a second turn.

Verified on the emulator: the card reads `Background command "sleep 5 && echo
done" completed (exit code 0)` where it had said "Command running in background
with ID: ...". 172 server tests, ktfmt, clippy, rustfmt, Android lint and the
JVM unit tests clean.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-06 21:47:54 -04:00
irisandClaude Opus 5 1bbb642973 Take subagent reports out of the main transcript, and separate turns with a rule 
A row per finished background task is a screenful of dividers about work the
reader was not asking after, and one of them turned out to be a whole shell
command drawn as centred prose, because its words came from somewhere with no
reason to keep them short. `Event::TaskNote` is gone entirely, along with the
row that drew it. A subagent's closing report is recorded as that subagent's
own transcript's closing text and is read in the subcard, which is where it
was already going; what the parent gets a row for is a message a subagent
genuinely sends it, which arrives by the peer path and has had one all along.

What remains is the actual defect and the smallest thing that fixes it. The
fold still refuses to grow a settled reply, so a turn boundary is always a
message boundary, and where two replies then abut it puts a `TurnBreak`
between them: a hairline, no words, no colour. Made by the fold rather than
sent by the server, because it is not something that happened -- it is the
boundary between two things that did. `joinPages` puts one in at a page seam,
which the fold never gets to see.

The task notification is still what closes a task in `Status::Waiting`'s
bookkeeping, and the registry lookup that recognises one this translator never
saw start is what makes that work for a session adopted across a restart.

Verified on the emulator: three replies, three rules, and nothing about the
helpers anywhere in the parent. 170 server tests, 85 JVM tests, ktfmt, clippy,
rustfmt and Android lint clean.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-06 21:39:19 -04:00
irisandClaude Opus 5 ef1aad8776 Keep a subagent's words in its own transcript, and count the ones already running 
Two corrections to the previous commit.

A subagent's closing report belongs in the subagent's transcript, which is
where it already is; drawing it as a card in the parent's put the same
paragraph in two places for a reader who did not ask for it. The row is a
divider now -- a boundary, which is what the transcript actually needed there
-- closed, saying only what reported and how it went. Opening it shows the
report anyway, since leaving the conversation to read one line has its own
cost, and a backgrounded command has no transcript of its own so this is the
only place its report exists at all: that one names itself from its summary
and has nothing left to open. `TranscriptDivider` grew a `trailing` slot for
the chevron rather than the row growing its own copy of the rules.

And the status was wrong for a session that was already running before the
update, which is every session when the backend is replaced under it.
Adoption picks a session's stdout back up from a recorded offset, so the
`task_started` lines for subagents launched earlier are behind it and the
translator never saw them -- it started with an empty set and reported `idle`
with a subagent plainly still working. `Subagents::any_open` reads the
directory instead, which is a measurement rather than bookkeeping and is right
for a session this process did not start. Both sources are kept and neither
subsumes the other: the translator's own set is the only thing that knows
about a backgrounded *command*, which has no subagent to be found. The same
pair decides whether an ending has already been reported, so a task that began
before the restart still gets its divider.

Echo's helpers now record their report as their own subagent's closing text,
the way the real driver does, so the fixture has the shape being tested.

Verified on the emulator: three dividers closed, one opened to its report, and
each reply drawn as its own message. 170 server tests, ktfmt, clippy, rustfmt,
Android lint and the JVM unit tests all clean.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-06 19:54:36 -04:00
irisandClaude Opus 5 5711c2568a Never run two turns into one, and say when a session waits on its own work 
A turn started by something with no row of its own -- a subagent reporting
back, a peer message the CLI only owns up to at the end -- met the previous
reply with nothing between it, and the fold grew that reply rather than
starting a new one. Two answers were drawn as one paragraph, running together
mid-sentence with not even a space between them. The fold now refuses to grow
a settled reply, and `joinPages` carries the same rule across a page boundary.

The other half is the row. `Event::TaskNote` records a background task
reporting back -- a subagent that finished, or a backgrounded command -- with
its title, how it ended and what it said; `TaskNoteRow` draws it as a card,
since somebody said this, and its own row rather than an update to the Task
call's, which is above everything the session has said since. Reported once
however many of the CLI's two lifecycle shapes arrive.

`SessionStatus::Waiting` is a session whose own turn is over while work it
started is not. `Idle` means "waiting for a person" and this means the
opposite, so reporting it as idle sent a "finished" notification at the one
moment that was untrue. Drawn as "waiting" in `waitingColor`; the queue and
the held-command boundary release on either end-of-turn status, so a message
sent while a subagent runs is not held until it finishes.

And a usage limit the account hits inside a subagent now reaches the session
as well as the subagent's transcript. `resume.rs` can only schedule against a
session, and a background Task outliving its parent's turn is the ordinary
case, so auto-resume was doing nothing at all for it.

The status word and its colour were two `when`s on two screens, and the second
missed `waiting` silently; they are `sessionStatusWord`/`sessionStatusColour`
now. Echo's `/subagent n` reproduces the whole shape, staggered a second
apart. Verified on the emulator against the sandbox: 169 server tests, ktfmt,
clippy, rustfmt, Android lint and the JVM unit tests all clean.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-06 18:57:30 -04:00
irisandClaude Opus 5 74c07d687a End subagents on the CLI's own task lifecycle, and detect a limit two ways 
Subagents were showing "running" long after they had finished. Measured
against 2.1.237 by running a session that launched one Task agent and
reading its stdout: a subagent's lines carry no `stream_event` at all --
they are whole `user`/`assistant` lines with a null `stop_reason` -- and no
`result` line is sent for one. So `ends_a_turn`, which watches for a raw
`message_delta` saying `end_turn`, could never fire for a subagent, and
nothing finished one until its session's process exited.

What the CLI does send is a task lifecycle, as top-level `system` lines:
`task_started` (with the tool_use id), `task_progress`, `task_updated`
(status, naming the task only) and `task_notification` (tool id, status, and
the agent's own summary). `translate_task` keeps the task -> tool mapping,
records the summary as the subagent's closing text -- the run showed its
child lines stop at its last tool_result, so without this a finished
subagent reads as stopping mid-tool -- and ends it. A `completed` update is
deliberately not the end, since its notification carries the summary; any
other terminal status is, because the failure to avoid is a subagent nothing
ever finishes. `ends_a_turn` stays as a second detector and must never be
the only one again. Verified by replaying the captured stream through the
server as a fake CLI: running, prompt, Bash call, output, report, exited.

`finish_all` now reads the directory rather than the live map, which is what
clears the ones already stuck: a subagent left running by an earlier run of
the server is exactly the one this process never touched, so it read
"running" again every time its session was started.

Auto-resume gets the same treatment on its own single point of failure. The
only thing that scheduled a resume was the CLI's error sentence at the end
of a failed turn; the CLI also sends `rate_limit_event` lines saying where
the account stands, and this server ignored them entirely. Both are read
now. Anything that is not an `allowed...` status counts as refused and is
logged if unfamiliar -- being wrong that way costs one question to the usage
meter, which is still what decides whether anything is sent, and being wrong
the other way is the feature silently not existing.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-06 13:29:23 -04:00
irisandClaude Opus 5 13d2d11c2d Order a session's subagents by activity, and delete finished ones by holding 
The subcards were oldest first, which buried whatever is working now. They
are ordered on the phone -- still running first, then most recently active --
over the server's stable oldest-first answer, since presentation order is a
display decision and a subagent that is thinking reports nothing meanwhile.

Holding a subcard selects it and several at a time, the import list's gesture
and its confirmation, so selecting is learned once. The selection bar sits
inside the session's card rather than at the bottom of the screen: it belongs
to one card, and one Delete is one request against one parent, so picking a
row in another card moves the selection rather than adding to it. Delete is
disabled, with the reason in words, while anything selected is still running
-- its transcript is still being written to and its process is the session's
to stop, so the server refuses that batch outright.

`POST /sessions/{id}/subagents/delete` takes the batch and checks every id
before removing any, so a set naming a running one is left exactly as it was
rather than half-deleted. It is `Subagents::start`'s path out. What counts as
running is shared with the list route through `has_a_process`, so the two
cannot disagree. On success the phone takes those rows out of that one card
and off the session's count, purges its cached copies, and drops the
expansion when nothing is left -- nothing else is refetched.

Driven on the emulator against the sandbox with ui-trace's new hold-by-name:
selecting two, the dialog, the rows going, a running one holding Delete
disabled, and the expander leaving with the last subagent.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-06 13:10:21 -04:00
irisandClaude Fable 5.1 cf10b17c5b End a subagent on its own end_turn, not the parent's tool_result, and give the expander a touch-sized row 
The Agent tool runs subagents in the background, so the parent's result
arrives at launch while the subagent works on for minutes; finishing on it
read a running agent as finished with a transcript cut off at launch. A
subagent now ends on its own message_delta end_turn, and a later line for a
finished one reopens it, since a background agent can be messaged again.

The card's expander row was only the chevron's height, so a tap for it
landed on the first subcard; it is the platform's 48dp minimum now.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-09-05 14:44:00 -04:00
irisandClaude Fable 5.1 9fa09b0af1 Show a session's subagents as subcards, each with a read-only transcript 
A subagent is a second transcript owned by a session, in the same event
model, with no process and no controls. The claude translator routes lines
carrying parent_tool_use_id to a per-subagent translator and transcript
under <session>/subagents/<tool_use_id>; three routes expose the list, a
transcript page and the SSE stream. Echo grows /subagent [n] as the rig.

On the phone a card with subagents ends in a chevron expander, collapsed by
default, opening to outlined subcards styled like dev-updater's components;
a subcard opens SessionScreen in read-only form, addressed through
TranscriptAddress so paging, cache and stream are shared.

Design in SUBAGENTS.md; choices awaiting review in DECISIONS.md.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
 2026-09-05 13:41:15 -04:00
iris eff5c8b0c0 Let the machine's own CLI refresh an expired token, and retry once 
A 401 from the usage endpoint means the stored access token has expired.
Refreshing it here is not an option: Anthropic's OAuth rotates the refresh
token, so a second refresher invalidates the CLI's copy and forces a
re-login on a machine that usually has a live session on it. So run the CLI
there instead and re-read what it wrote.

`doctor` rather than `auth status`: probed against 2.1.258 with an invalid
token, `auth status` answers loggedIn:true from the file alone and never
reaches the network. The same probe showed a failed refresh blanks both
tokens, which is why this stays on the 401 path.

Also gives ProviderConfig one program() so the CLI's default path is not
written down twice.
 2026-09-05 12:07:34 -04:00
iris 7b63330aaa Say when a usage 401 is an expired login, not an unreachable endpoint 
A 401 is the endpoint answering and refusing the stored OAuth token, which
Claude Code refreshes as it runs -- so a machine whose CLI has been idle
hands us a stale one. Reporting it as "usage endpoint unreachable" pointed
at the network instead of at the one thing that fixes it.
 2026-09-05 11:58:17 -04:00
irisandClaude Opus 5 6bdec6e785 Let a session resume itself when its usage limit lifts 
Off by default and per session: it spends quota the moment quota exists,
with nobody watching, which is not a thing a default may decide. Switched
on from the session settings dialog, with the message it sends editable
("continue" unless something else is typed).

Running out of quota becomes a state rather than an error. The Claude
driver recognises its dialect's sentence -- `Claude AI usage limit
reached|1788546972` -- and reports `LimitReached` with the reset time it
gave; nothing above a driver matches on a string. The transcript draws it
as a divider, like a clear or a compaction.

The schedule is a plan to *ask*, never a plan to send. Both reset times
available are untrustworthy in the direction that matters -- the dialect's
is written when the turn fails, the endpoint's moves when the window does
-- so the wait ends in a question to the usage meter, and only `ok` with
no window at 100% sends anything. A window still spent reschedules to its
own reset time, which is what makes a limit that lifts late wait longer
and one that lifts early resume sooner. A meter that cannot be asked is a
longer wait too, never a send. A day after the limit was hit the wait
gives up and says so in the transcript, so a machine that can never be
asked is not retried for ever.

The schedule is persisted on the session: a five-hour window outlasts a
backend restart, and a wait forgotten across one never comes back.

Driven end to end with echo, never a real account: `/limit [minutes]`
reports the same event a real driver does and `/usage` sets what the meter
answers, deliberately separate so the two can disagree. The wait moved
from the dialect's two minutes to the meter's seven when the meter changed
its mind, and the message went out on the first check after the meter came
back under the limit.

Also makes the settings dialog scrollable, which these two controls made
necessary: at a 1.5x system font it clipped the last of them with nothing
on screen to say so.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-05 05:21:43 -04:00
irisandClaude Opus 5 4821a02bd3 Default thinking level for new sessions, and move the rigs out of AGENTS.md 
`Config::default_effort` is what a session starts at when nothing chose one,
applied in `spawn_session` rather than filled in by the spawn screen so it
holds for an import and a bare API call too. It is set by the spawn screen's
own picker, whose label says so: one control, where new sessions are made,
rather than a settings page for a single value. Not on a provider, because
providers are discovered and the next rediscovery would erase it; not on the
phone, because a second device would then spawn at a level nobody there
chose. `GET`/`POST /defaults` carry it as a struct, so the permission mode --
still hardcoded to `auto` on the spawn screen -- can move there later without
a second route.

Only drivers that read a level are given one: an echo session was storing a
`--effort` it never passes to anything, which is a config file answering a
question about itself wrongly.

Separately, `AGENTS.md` is 35 KB sent with every request in this repo, and 12
KB of it was rigs and reference measurements that only matter once you are
running one. Those are the `ai-app-rigs` skill now -- the same text, still the
only copy, read when the work touches it. 35,198 -> 20,813 chars.

Verified on the emulator against the sandbox: the spawn screen pre-fills from
the server, picking `low` spawned a session at `low` and left `/defaults` set
to it, and an echo session spawned afterwards took no level at all.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-04 21:42:05 -04:00
irisandClaude Opus 5 1ff662c7c3 Let a session choose how hard it thinks 
Output is about an eighth of what a session costs and thinking is nearly
all of it -- prose is ~1.5% of output tokens, measured over 27,015 requests
of this account's own transcripts -- so the level is the largest saving
available short of shortening the conversation itself.

Shaped like the working directory rather than like the model: the CLI's
only two setting control requests are `set_model` and `set_permission_mode`
(checked against the 2.1.258 binary), so `--effort` is read when the process
launches and cannot be asked of a running one. `set_session_effort` records
the level and stops the process; the next message or Start launches one that
has it. That is also why the picker is in the session settings dialog beside
Move, and not on the bar beside the model and the mode, which take effect
mid-turn.

`None` is a level in its own right -- the CLI's own default -- so the picker
can return to it, and a blank is normalized to it at the boundary rather
than stored as a level the CLI would reject.

Offered only where it means something: `DriverKind::takes_effort` reports
the capability and the phone leaves the row out entirely, rather than the
session-type branch this app does not have anywhere else. A llama session
would otherwise get a control whose only effect is stopping its process.

Verified on the emulator against the sandbox's fake CLI: the picker sets it,
the server reports it, and an echo session's dialog is unchanged.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-04 21:23:58 -04:00
irisandClaude Opus 5 e4f0935f98 Keep the second auth test under a subscriber, so the tripwire is not flaky 
`gates_every_route_and_never_logs_the_token` failed about one full-suite
run in ten, on the assertion that a rejection *was* logged. Its sibling
ends with an unauthenticated request of its own, made with no subscriber
on that thread -- and tracing caches a callsite's interest process-wide
the first time it is reached, so whichever test got there first decided
whether the warning would ever be recorded.

That is the rule already written at the top of "Things that have bitten",
applied to one member of a set: the combined gating+logging test exists
because of it, and the enrollment test added later did not get it.
Twenty runs clean since.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-04 17:58:11 -04:00
iris 3c0214ece8 Merge branch 'main' of git.arirex.me:iris/ai-app 
# Conflicts:
#	AGENTS.md
#	PLAN.md
#	app/androidApp/src/main/kotlin/com/example/aiapp/SessionUsageBar.kt
#	app/androidApp/src/main/kotlin/com/example/aiapp/SpawnScreen.kt
#	server/src/config.rs
#	server/src/main.rs
#	server/src/routes.rs
#	server/src/session/echo.rs
#	server/src/session/llama.rs
#	server/src/session/transport.rs
#	server/src/ssh.rs
#	server/src/usage.rs
 2026-09-04 17:56:50 -04:00
irisandClaude Opus 5 127b25e60a Meter a session by its provider, and let llama.cpp run over ssh 
The rate-limit bar answered a question about an account, and picked the
answer by machine. One machine runs echo, the Claude CLI and a local
model side by side, so every echo session on it drew the CLI's five-hour
window: a quota that session cannot spend and could never run down. A
session now names its meter (`usageProvider`, from
`DriverKind::usage_provider`, which `usage::providers_for` reads too so
the two lists cannot disagree), and the phone matches on machine *and*
provider. Nothing meters echo or llama, and nothing at all is drawn --
including while the first fetch is out, since "checking" under a session
that turns out to meter nothing is a row the screen then withdraws.

Echo gets a meter it can be *told* about instead: `/usage 42`,
`/usage 95 20`, `/usage 42 never`, `/usage notloggedin`,
`/usage unreachable`, `/usage failed`, `/usage off`. Those states cost
real quota to arrange, which is why none of them had been looked at.

And llama.cpp runs wherever a setup says, which was the last of phase 5.
`Transport::reserve_port` is the second half of what a transport is --
"run this" plus "reach this port" -- returning the port the server binds
there and the port that reaches it here, and `Launch::reaching` puts the
`-L` tunnel on the connection that already carries the command. Three
things that came out of building it:

- A forwarded launch gets a pty and every other one keeps `-T`. Killing
  the ssh client ends a CLI by closing the stdin it reads; llama-server
  never reads its stdin, so the same kill left it running on the far
  machine with the model loaded -- one orphan per stopped session.
- The model is looked for on the machine that will serve it, at that
  machine's own models directory, so `GET /setups/{id}/models` is what
  the spawn screen offers rather than the backend's own downloads.
- The readiness poll watches the process, not only the port: a model
  that will not load exits in a second and would otherwise have been
  reported as "gave up after 300s". The failure carries the log's tail.

Exercised end to end against this VM over ssh to itself: spawn, load,
answer, outlive a backend restart, be adopted, answer again, and stop --
with both the ssh client and the far llama-server gone afterwards. The
local path, the Claude bar and the spawn screen checked on the emulator.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
 2026-09-04 17:45:32 -04:00
223 changed files with 31737 additions and 23443 deletions
Expand all files Collapse all files

No files matched your search

.agents/skills/ai-app-rigs
Symlink
+1
View File
@@ -0,0 +1 @@
../../.claude/skills/ai-app-rigs
.claude/skills/ai-app-rigs/SKILL.md
+380
View File
@@ -0,0 +1,380 @@
---
name: ai-app-rigs
description: ai-app's test rigs, harness scripts and reference measurements - ui-sandbox.sh, debug-transcript.sh, transcript-bench.sh, stream-bench.sh, trace-draw.sh, the /usage fixture vocabulary, the fake CLI, the rule that no UI-driving script may tap a coordinate, how to test llama.cpp and ssh on this machine, how importing behaves, and the scroll/stream/explorer numbers not worth re-measuring. Read before running or writing a benchmark, driving the app's UI from a script, exercising the session lifecycle, testing a llama or remote session, or touching the import screen.
---
# ai-app: rigs, harnesses and measurements
Moved out of `AGENTS.md` on 2026-09-04 so it is read when it is relevant
rather than sent with every request in this repo -- it was 12 KB of the 35 KB
that file cost on every one. Unchanged in the move, and still the only copy.
## The rigs
Each exists because something was invisible without it.
- **`app/ui-sandbox.sh`** — a second `ai-server` with its own `$HOME`, config
and data directory, holding eight invented Claude Code transcripts and a
`claude` that is two lines of shell. **That isolation is the point**: the
import screen lists whatever is in `~/.claude/projects`, which in this VM is
real agent transcripts, so exercising *delete* against the ordinary server
deletes somebody's conversation and exercising *import* starts a real
`--resume` on the owner's account.
Its port and root derive from the checkout's name, so two checkouts'
sandboxes cannot reach each other, and its token is generated once into
`~/.config/ai-app/sandbox-token` and carried across restarts along with any
the enrolment flow appended — so the emulator app is enrolled **once** (the
start banner prints the command) and stays enrolled. It shares the real TLS
certificates, because the installed APK pins that CA.
Driving verbs, so none of this is re-derived per session:
`./ui-sandbox.sh spawn [title]` (an echo session, prints its id),
`./ui-sandbox.sh send SID text|@file`, and
`./ui-sandbox.sh api /path [curl args]`.
`./ui-sandbox.sh keep` restarts the server without wiping the sessions and
enrolment already there — for when the fixture under test was expensive to
build; plain `start` wipes them, which is right for the list-screen
fixtures and wrong for that.
It passes `--delay` by default, and `AI_SANDBOX_BIG_MB` puts one large
transcript among the small ones while `AI_SANDBOX_SPAWN_DELAY` makes the
fake CLI slow to start. Both exist because operations that finish in
milliseconds have states on the way that nothing can observe, and an
unobservable state is one where broken and working look identical.
It also builds a fixture tree at the sandbox home's `~/files` for the
explorer, holding the states otherwise only reachable by finding a real
machine in one: an empty directory, a name with a tab and one with an
apostrophe, a binary file, one over `FILE_LIMIT`, one `chmod 000`, a
symlink to a directory and a broken one, a source file per language, and
the three sizes the limits were measured against (`edit-32k.rs`,
`edit-128k.rs`, `big-source.rs`). Point a session at it with
`./ui-sandbox.sh api /sessions/<id>/cwd -X POST -H 'content-type: application/json' -d '{"cwd":"~/files"}'`.
The explorer's 409 is produced by editing the file on the machine
(`printf … > file`) between pressing the pencil and pressing save.
- **`app/debug-transcript.sh`** — a real conversation on the emulator. The
echo driver is the right rig for most things and the wrong one for anything
whose cost scales with what was actually written: a real reply is longer,
is real markdown, and carries tool calls whose input and output are
kilobytes. Two faults were invisible until a real transcript was loaded — a
page of history landing mid-fling threw the reader back to the newest end,
and parsing one real reply took 51ms against 4.6ms for a synthetic one.
`-b` takes the biggest conversation on the machine rather than the newest,
which is what a scrolling test wants; `--stop` takes it down.
It copies the transcript into `/tmp` and gives the server a `HOME` of its
own, so the import can only see the copy — importing spawns `claude
--resume`, and against the real file that is a second CLI writing to a
conversation somebody may still be in. **A transcript never goes in this
repository**: they hold whatever was said, read and written in that
session, and `~/repos` is shared with the host besides.
- **`/usage` in an echo session puts up an invented meter**, which is how the
rate-limit screens' states are reached without spending quota: `/usage 42`,
`/usage 95 20` (minutes left), `/usage 42 never` (the between-blocks window
with no reset time), `/usage 42 unreadable`, `/usage notloggedin`,
`/usage unreachable`, `/usage failed`, `/usage off`. The vocabulary is
`usage::Fixture`'s, since those are its states. With none set an echo
session meters nothing, which is the ordinary case and draws no bar.
- **A fake CLI exercises the process lifecycle without a token.** Point a
`claude_cli` provider's `command` at a script that ordinarily runs
`cat > /dev/null` and it behaves the way the lifecycle code cares about:
it holds the fifo open, records a real pid, writes nothing, and dies on a
signal. So adopt, stop, restart and start are all drivable without a real
`--resume` and without spending a turn on somebody's account. Reach for
this when what is under test is *whether a process is running*, and for
`debug-transcript.sh` when it is *what the transcript draws*. The sandbox's
version also handles `auth login`: it prints an inert Anthropic-shaped URL,
rejects any code except `sandbox-code`, and exits successfully for that one.
- **`/think [seconds]` in an echo session puts up a thinking card**, long
enough to watch it spin before it closes with the span it actually took.
The rest of the turn is the ordinary echo reply, so it is also the rig for
a block and a reply meeting.
- **`app/transcript-bench.sh`** is the standard scroll measurement: it opens
the first session (or `-k` keeps the current screen), scrolls a fixed
gesture loop, and prints the app's render report — the same one the in-app
copy button produces, whose `on screen:` line names what the viewport was
holding. Compare two runs with the same gestures; the emulator's absolute
frame times transfer nothing, the report's accounting does. Run it either
side of any change under `Markdown*.kt`, `Transcript*.kt` or
`SessionScreen.kt`'s list, and put the report in the commit. The numbers
that move first are the worst `record: one block`, the reparse mean while
streaming, and the draw phase's accounting line.
- **`app/stream-bench.sh [-k] FILE`** is that measurement for a reply still
arriving. It taps "Jump to latest" so the list is pinned to the newest end,
resets the report, sends FILE, waits for the transcript to stop growing,
and prints. Both of those are corrections to a first version that measured
nothing: a transcript parked further back never redraws while a reply
streams into it, and a session is idle at *both* ends of a turn, so polling
for idle answers before the turn has started.
- **`app/trace-draw.sh`** names what a scrolling frame spends inside the
framework, from `atrace` text output with no trace processor needed. It is
how the cost of a layout node per link was attributed to the framework
rather than guessed at.
### Driving the UI
**No script that drives this app's UI presses a coordinate.** Every control
is found by the name it already carries for assistive technology —
`ui-trace record --do "tap 'Session settings'"` — which resolves the label
against the screen at the moment of the gesture and fails the whole run when
it is not there. `app/bench-lib.sh` is what the bench scripts share for it. A
coordinate is a position measured once by hand, and anything that moves the
control makes the tap land on whatever now sits there — the bench then
reports a number that was never measured, which reads exactly like a result.
Both bench scripts pressed the render report at `tap 723 205` until that
button moved into the session settings dialog on 2026-09-03. The check that
none has crept back:
grep -n "tap [0-9]" app/*.sh
Swipes are still coordinates, deliberately: a gesture across a scrolling area
is a distance rather than a control.
**Two traps in the emulator bench loop**, each of which cost a run.
`adb shell pm clear` removes the enrolment and the notification permission
along with the saved anchors, so the next run measures a permission dialog —
re-enrol with the command `ui-sandbox.sh` prints, and
`pm grant … POST_NOTIFICATIONS`. And a saved scroll anchor is per session id,
so the only way two builds start a scroll from the same place is a *fresh
session for each*.
**The emulator is `~/repos/emulator-tools`' business, not this repo's.**
`emu up` creates and boots the AVD named after this checkout — whatever `emu
name` prints, never a name typed out here, since this file is the same in
every clone. `run-android.sh` is that plus a build and an install. The `adb`
on `PATH` after sourcing `android-env.sh` is that repo's wrapper, which fills
in `-s` from the same rule. Gradle does not go through it, so a Gradle init
script from `emulator-tools` runs `emu check` before `installDebug`,
`uninstallDebug` and `connectedAndroidTest` and fails rather than fanning out
to every attached device; when it refuses, say which device you mean at the
moment you use it — `ANDROID_SERIAL=$(emu serial) ./gradlew …`.
### Testing llama.cpp and ssh here
**Both are set up here** and need nothing typed. The prebuilt llama.cpp lives
outside the repo at `~/.local/opt/llama.cpp-vk` — a **Vulkan** build as of
2026-09-19, replacing the CPU one that was there before — and is symlinked as
both `~/.local/bin/llama-server` and `/usr/local/bin/llama-server`. The second
is what makes **discovery find it over ssh**: `~/.local/bin` is not on the
PATH a non-interactive ssh session gets. It resolves its own libraries through
`$ORIGIN`, so no `LD_LIBRARY_PATH` is needed.
Two models are downloaded under `~/.local/share/ai-app/models`:
- `unsloth/Qwen3-0.6B-GGUF/Qwen3-0.6B-Q8_0.gguf`, 639 MB, loads in ~4s. It
calls tools correctly and is the right rig for the driver's shape. Do not
judge *answers* by it — asked for the second line of a file it read from
line 2 and then named the third.
- `ISTA-DASLab/Qwen3.8-27B-GSQ-RCO-GGUF/Qwen3.8-27B-GSQ-RCO-IQ3_S-mtp.gguf`,
12 GB, ~20s to load, and the only one here with a multi-token-prediction
head. It is the rig for anything about `loading` being a state of its own,
since 20s is long enough to send into.
- `ggml-org/SmolVLM-256M-Instruct-GGUF/SmolVLM-256M-Instruct-Q8_0.gguf`,
175 MB, plus the `mmproj-…` beside it, downloaded 2026-09-20 as the rig for
**vision**: it is the only model here that reads pictures, it loads in
seconds on the CPU, and it described a red circle correctly. The pair is
also what exercises the projector being found beside the weights, and the
projector being kept out of the models a provider offers. Qwen3-0.6B beside
it is the other half of that rig -- the model that answers `refused`.
**A second llama.cpp is installed here, and it is the rig for a custom
build.** `~/.local/share/ai-app/llama/prism/` is Prism ML's fork
(`prism` branch, `~/repos/llama.cpp-prism`, Vulkan, `cmake --install
--prefix`), so discovery finds it as a provider called `llama-cpp-prism`
beside the ordinary `llama-cpp`. It is what exercises that mechanism at all,
and it serves `prism-ml/Ternary-Bonsai-2-27B-gguf` -- ternary packings stock
llama.cpp rejects as unknown types. Rebuild it with
`cmake -B build -DCMAKE_BUILD_TYPE=Release -DGGML_VULKAN=ON
-DCMAKE_INSTALL_LIBDIR=lib -DCMAKE_INSTALL_RPATH='$ORIGIN/../lib'`, about six
minutes at `-j8`.
**Both of those install flags are load-bearing, and the failure is a session
that never becomes ready.** `llama-server` is a 16 KB launcher against
`libllama-server-impl.so`, so a build whose libraries it cannot find dies at
`exec` with `error while loading shared libraries` -- which reaches the phone
as the model never answering. The runpath has to be set, *and* the libraries
have to be where it points: `GNUInstallDirs` chooses `lib64` on some
distributions (Gentoo's amd64 profiles among them) while the runpath above
says `lib`. `readelf -d bin/llama-server | grep RUNPATH` and
`ldd bin/llama-server | grep 'not found'` are the two-second check after any
install here.
**Which Bonsai packing runs on the GPU is the backend's question, not the
model's.** Measured 2026-09-21 with `llama-bench -p 512 -n 64 -r 2 -fa 1
-ngl 99` on the free card:
| packing | backend | pp512 | tg64 |
| --- | --- | ---: | ---: |
| `PTQ1_0`, 5.53 GiB | Vulkan | 519 t/s | 7.5 t/s |
| `PQ2_0`, 7.21 GiB | **CPU**, 8 cores | unfinished after 9 min | -- |
The fork's Vulkan port covers `PTQ1_0` only -- shaders, a `mul_mat_vec` and
the FWHT included -- so `PQ2_0` has no kernel there and every matmul falls
back to the CPU. That reads exactly like a stuck load: the process sits at
700% CPU for minutes with the card idle. On CUDA and HIP it is the other way
round, since `mmq.cu` guards `PTQ1_0` out of the HIP build (it wants Turing
MMA) and leaves `PQ2_0` in. **ROCm cannot be tested in this VM**: there is no
`/dev/kfd`, because the GPU here is virtio-gpu rather than a passed-through
card.
7.5 tok/s is the honest speed of that Vulkan kernel, against 42 for the
IQ3_S 27B beside it -- smaller weights, slower decode. Nothing is
misconfigured; the fork's fast kernels are CUDA and Metal.
**Do not test with a 2-bit quant**: the IQ2_XXS of the 0.6B produces fluent
nonsense, which reads exactly like a broken driver — `llama-cli` produces the
same from the file directly, which is how to tell the two apart in a hurry.
**The GPU is shared and llama-server dies loudly when it runs out.** A second
server loading a model while the 27B holds VRAM fails with `radv/amdgpu:
Failed to allocate a buffer` / `MESA: error: buffer allocation failed` and
exits mid-request. `-ngl 0` runs it on the 8 cores instead, which is the way
to test the driver while something else holds the card -- through the app, that
is the model's "Layers on the GPU" set to 0 in the machines tab's provider
view, and `--models-max` above 1 is how two models come to be loaded at once
in the first place.
**Testing tools and MCP without the app**: `llama-server --tools all` publishes
its built-in tools at `GET /tools` and runs one at `POST /tools` with
`{"tool": …, "params": …}` and an `x-tool-cwd` header — so a whole agent loop
is drivable with `curl` and no model at all. The Exa MCP server at
`https://mcp.exa.ai/mcp` answers **without an API key** and needs a
`User-Agent` header (Cloudflare answers 403 without one, which reads as a
refusal rather than a missing header).
There is no second machine, so **ssh this VM to itself**. That is set up
too: the key is `~/.config/ai-app/ssh-self` (its public half is in
`~/.ssh/authorized_keys`, labelled removable), and the real config carries a
machine called **"this vm over ssh"** — `bob@127.0.0.1` with that
`identityFile` plus
`options: ["StrictHostKeyChecking=no", "UserKnownHostsFile=/tmp/ai-app-known-hosts"]`
so it touches nothing real — offering `claude-cli` and `llama-cpp`. It is the
whole rig for "does a remote llama session work", since the far machine is
this one and the model file is the same file. For a throwaway machine of your
own, point a provider's `command` at something harmless like `/bin/echo`
rather than at `claude`: the transport is what is under test, the process
exiting immediately is the signal, and it costs no tokens. The remote login
shell here is **fish**; the
remote script and `ssh.rs`'s POSIX quoting happen to mean the same thing in
both, but that is luck rather than design, and a shell that is neither is the
thing to suspect first if a remote spawn ever mangles an argument.
## Importing
The import list reports each session's **size as well as its line count**,
because the two disagree in the way that matters: these transcripts embed
screenshots as base64, so one line can be a megabyte. On this machine a 69 MB
session has 3,427 lines and a 44 MB one has 6,792 — nothing about a line
count tells you what continuing a session will cost. Shown, not warned about;
importing a large session is a choice somebody is entitled to make.
**Never import a Claude Code session that is open in a terminal.** The app
refuses it — see PLAN.md for the incident that made that a refusal rather
than a warning.
**One Claude Code session id can name two files, and the listing offers it
once.** Resuming from a different working directory makes the CLI write a
second transcript with the same id under that directory's project folder — an
ordinary state of a machine, not corruption. Everything downstream addresses
a session by id, and the phone keyed its list on it, so two rows sharing one
**closed the app** on a Compose duplicate-key throw. `parse_listing` keeps
the copy with the most lines, because the other is usually a few-hundred-byte
stub and is often the *newer* of the two, so recency is the wrong key.
Deleting removes every copy rather than the first, or the row came back after
a delete that reported success. The phone's half is `uniqueItems`, which
every list keyed on a server-chosen id goes through: a repeat there must
never be able to close the app, whatever produced it.
**Deleting a session offers to take the machine's own transcript with it** —
`DELETE /sessions/{id}?deleteForeign=true`, behind a switch in the
confirmation, and only where the driver keeps a record of its own
(`keepsOwnTranscript`, currently Claude Code or Codex). Off by default,
because leaving that copy is what makes an ordinary delete recoverable — and
the dialog's paragraph is rewritten when it is on rather than appended to,
since the sentence promising the conversation "should still be there to
import again" is exactly the one the switch makes false. The server deletes
the machine's copy *first*, so a machine it cannot reach leaves the session
where it was instead of half-deleted.
## Measurements worth not re-taking
- **`-np 1` is what makes the MTP draft head pay.** Taken 2026-09-19 on the
27B above, decode speed for a 300-token reply, from `llama-server`'s own
timings rather than the clock:
| flags | tok/s |
| --- | --- |
| plain, any `-np` | 41.5 |
| `--spec-type draft-mtp -np 1` | 61.4 |
| `--spec-type draft-mtp -np 2` (n-max 2) | 65.9 |
| `--spec-type draft-mtp`, default `-np` (4 slots) | 28 |
Draft acceptance is 0.53–0.73 in every case, so the head is working in all
of them: what changes is that speculating against a KV cache split four ways
is slower than not speculating. A model's preset gets `parallel = 1` unless
its settings say otherwise (the machines tab's provider view, since
2026-09-19), so this is recorded for whoever next sees MTP look broken or
next raises the slot count to answer two sessions at once. `--spec-draft-n-max 2` was
worth another 7% in a single sample and is deliberately *not* passed — one
sample on a virtualised GPU is not a number to hardcode.
- **Prompt processing is the expensive part of a llama turn here, and decode
speed falls only slowly with context.** Taken 2026-09-19 on a free GPU, the
27B with `--spec-type draft-mtp -np 1`, generating 160 tokens each time:
| context | decode | prefill of that prompt |
| --- | --- | --- |
| 88 | 43.4 tok/s (cold) | 21s |
| 1,569 | 55.5 tok/s | (model still warming) |
| 6,068 | 53.2 tok/s | 9.5s |
| 14,068 | 50.3 tok/s | 22s |
So a turn on a long conversation spends tens of seconds before the first
token, and that is what `SessionStatus::Reading` exists to say. The same
sweep on the 0.6B **on the CPU** falls much harder -- 30.1 tok/s at 44
tokens of context to 11.5 at 6,024 -- which is the shape somebody means by
"it gets slower as the conversation goes on". The figure the app draws is
`timings.predicted_per_second`, decode only, so prefill is never mixed into
it.
- **A busy GPU is a model that will not load at all**, not a slow one:
`radv/amdgpu: Failed to allocate a buffer` and `failed to load model` while
something else holds VRAM. A 0.6B that had been decoding at 149 tok/s ran at
16.7 in that window before its server died, so a tok/s figure taken while
the card is shared says nothing about the model.
- **Asking for the head when the file has none is fatal**, not ignored:
`context type MTP requested but model doesn't contain MTP layers` and the
server exits. Without the flag the same file logs `unused tensor
blk.N.nextn.* — ignoring` and runs normally, which is the state to look for
when MTP is silently not happening.
- **What the transcript screen costs to scroll.** Taken 2026-08-30 on the GPU
emulator against a real imported transcript with the server at
`--delay 120`. Settled and flinging fast, both into fresh history and back
through rows already drawn: **5.2–5.9% janky frames, 99th percentile
29–32ms, 0–2 slow UI-thread frames.** The stock Settings app on the same
device is 3.3% and 38ms, so this is at the platform floor. The number that
is *not* at the floor is the first few seconds after opening a session,
where every row on the way is being composed for the first time; that is
inherent to a lazy list and it is why a measurement taken before the screen
settles reads three times worse. **Settle first, then reset `gfxinfo`.**
- **The reset path is not reachable by reopening a session.** Measured
2026-09-04 against a session streaming at 20 events a second: reopening one
with an anchor 1,800 events back connects **87–119 events behind**, well
under `CATCH_UP_LIMIT`'s 200, because the restore is two requests — the
opening page, then one span covering the whole distance. To exercise the
reset at all you have to lower `CATCH_UP_LIMIT` in a throwaway build; at 5
the app takes the reset on a live connection, clears, refills and carries
on without reconnecting.
- **The session screen's stream survives backgrounding here** — 20 seconds at
the launcher while 415 events were produced brought no reconnect at all,
which is not what the comment above that loop expects, and is most likely
this emulator being headless rather than the phone's behaviour.
- **Reopening a cached session costs one request for one event** (the probe),
and scrolling the whole conversation back costs nothing more; a cold open
of the same 500-event session is two pages, 100 events. Measured
2026-09-04 on the emulator against the sandbox.
- **Reading is cheap and editing is not.** The viewer handles a 1 MiB,
28,000-line file because it draws one row per line; the editor is one
`BasicTextField`, which costs two seconds a frame at 128 kB and stops the
app at 1 MiB, so `EDIT_LIMIT` caps it at 32 kB with the reason said on
screen. If you make the editor faster, that number is what to move.
EXPLORER.md's "What the measurements said" has the rest.
.gitignore
-3
View File
@@ -21,6 +21,3 @@ certs/
config.ron
config.json
sessions/
# iris, the in-house UI library, is vendored at iris/ and built by cargo.
iris/target/
AGENTS.md
+437 -227
View File
@@ -1,6 +1,6 @@
# ai-app
A phone interface to AI coding sessions (Claude Code and llama.cpp),
A phone interface to AI coding sessions (Codex, Claude Code and llama.cpp),
replacing the Claude app for daily use. Rust/Axum backend on the desktop,
Kotlin/Compose Android app, WireGuard + pinned self-signed TLS + bearer token
between them.
@@ -9,7 +9,15 @@ between them.
its rationale, and what was rejected. Read it before changing anything
structural, and update it in place when a decision changes rather than
letting this file and the plan become two versions of the truth. This file is
the working notes layer: layout, commands, rigs, and things that have bitten.
the working notes layer: layout, commands, and things that have bitten.
**The rigs are the `ai-app-rigs` skill** — the sandbox and bench scripts, the
rule that no UI-driving script may tap a coordinate, how to test llama.cpp and
ssh here, how importing behaves, and the measurements not worth re-taking.
They moved there on 2026-09-04 because they are 12 KB that only matter once
you are actually running one, and this file is sent with every request. Read
it before writing or running a benchmark, driving the UI from a script, or
touching the import screen.
The central design point, worth not undoing by accident: **a session is a
child process, translated into one common event model.** A new session type
@@ -26,10 +34,188 @@ Module-by-module intent is in PLAN.md's "Backend layout".
- `server/` — the Rust backend (`ai-server`). `routes.rs`'s module doc
comment is the HTTP table and the surface's source of truth.
**A machine's models are served by one shared `llama-server`** (2026-09-19,
`session/llama/router.rs`): started with no `-m`, which makes it a
**router** — it reads a preset file naming models and their flags, starts a
child server per model asked for, and routes by the `model` field in each
request. So a session has no process of its own, two sessions on one model
share one copy of it in memory, and a backend restart adopts one process
rather than one per session. Four things fall out of it and are easy to get
wrong again — a session records the router's pid in its own directory as
`process::Detail::Shared`, and `process::stop` refuses to signal a `Shared`
record, which is what keeps one session ending from unloading everybody's
model; **nothing stops a router on its own**, and the only thing that does
is the machine's provider view (`POST /machines/{id}/providers/{p}/stop`);
how a model is *loaded* is per model on its machine
(`ProviderConfig::model_settings`, `LLAMA_MODEL_PARAMS`) rather than per
session, and saving those settings rewrites the preset, which **unloads**
that model; and the preset is read back before every edit, because a router
adopted from an earlier run is serving sections this process has never seen
and rewriting without them unloads those.
**A machine can have more than one llama.cpp** (2026-09-21,
`machines.rs`): anything at `~/.local/share/ai-app/llama/<name>/llama-server`
or `.../<name>/bin/llama-server` is discovered beside the one on PATH and
becomes a provider called `llama-cpp-<name>`, with its own router, preset
and model settings. That is how a model whose kernels are not upstream is
served -- Prism ML's ternary Bonsai is the one here, built from the
`prism` branch into `~/.local/share/ai-app/llama/prism` -- without the
phone ever naming a command, which is the property this module exists for.
Both providers offer the machine's whole models directory, since which
build reads which packing is not answerable from the file.
**A llama.cpp session runs on its configured machine** (built
2026-09-04, the last of phase 5): `Transport::reserve_port` returns the
port the server binds *there* and the port that reaches it *here*, and
`Launch::reaching` puts the `-L` tunnel on the connection already carrying
the command. Three things fell out of it and are easy to get wrong again —
a forwarded launch gets a pty (`-tt`) and every other one keeps `-T`,
because `llama-server` never reads the stdin whose closing ends a CLI and
the same kill left it loaded on the far machine; the model is looked for on
the machine that will serve it, so the spawn screen offers
`GET /machines/{id}/providers/{p}/models` rather than any list of this
backend's own; and the
readiness poll watches the process as well as the port, since a model that
will not load exits in a second and was being reported as "gave up after
300s". See PLAN.md's "Transport" and "llama-server management".
**A llama session has tools and runs the loop itself** (2026-09-19):
`--tools all` gives the router `llama-server`'s built-in set, which it also
*runs* (`GET /tools` for the definitions, `POST /tools` to call one), while
web search comes from an MCP server this backend connects to directly
(`session/llama/mcp.rs`, Exa preset in a discovered provider's
`mcpServers`). Driving the loop is what makes the permission gate ours:
`manual` asks before every call and remembers a tool you answer
"Always allow …" to, `bypassPermissions` never asks, and the allowances are
folded back out of the transcript. Which tools a *session* offers is a
filter applied to those definitions here, not a flag over there: one shared
server has one set, and the filter costs no reload (2,181 tokens of prompt
with all seven, 698 with none). Three more things fall out of it and are
easy to get wrong again — a model change **asks for another model** and
stops nothing, since the one being left may be another session's;
`parallel = 1` unless that model's settings say otherwise, and it is what
decides whether the MTP draft head is a 50% speed-up or a 33% loss; and
`spec-type = draft-mtp` is conditional on the file actually having a head,
because asking for one that is not there makes `llama-server` **exit**.
**A llama session takes a picture only where the model natively reads one**
(2026-09-20): a multimodal model is loaded with the `mmproj` found beside
its weights (overridable per model, `off` included), an attachment rides in
the request as an `image_url` data URI, and nothing at all is done for a
model without a projector. Four things fall out of it and are easy to get
wrong again -- whether a session takes pictures is `/props`'s
`modalities.vision` from the loaded server and never a guess from this side,
with three states because a loading model has not answered yet
(`Images::Unknown` is *offered*, since a control withheld because nobody
could ask is missing from sessions that would have taken it); a message
carrying an image a model cannot read is **stopped rather than stripped**,
refused at the door, at the queue and at the steering boundary, because
`llama-server` refuses the whole request over one part and a message sent
without its picture is a different message; an earlier turn's image folds
into a line of words for a model without vision, so switching models does
not end the conversation; and a projector is filtered out of the models a
provider *offers*, while staying in the machine's own model list.
**A llama session's thinking is drawn** (2026-09-19): `reasoning_content`
becomes `Event::Thinking` deltas closed by an `Event::ThinkingDone` carrying
the span the *driver* measured, and the phone draws a card that spins while
the block is open and says "Thought for 12.4s" once it is not. The reasoning
is deliberately not part of the next prompt (`conversation` ignores it), and
`timings.predicted_per_second` and `timings.prompt_ms` off the same stream
become `UsageDelta`'s `tokensPerSecond` and `prefillMs`, which is the
"read 9.5s · 50.3 tok/s · 3:00 PM" under a finished reply — nothing else here
measures either, so every other driver sends `None`, and the clock is last so
that it does not move when a provider reports fewer of them.
**A wait that can be measured says how far along it is** (2026-09-21):
`GET /sessions/{id}/progress` answers `{of, fraction, stage?}` and `null`
for a session that is not in one -- runtime state, asked for twice a second
by the session screen while it is drawing a wait, and deliberately never an
event, since a load reports five times a second and every event is a
transcript line for ever. The two sources are the router's `/models/sse`
stream, which is the **only** place a model's load progress appears (`GET
/models` says "loading" and no more), and `prompt_progress` chunks that
`"return_progress": true` adds to the generation stream. The sample says
which status it measures so it cannot be drawn under another one, and
`/loading [seconds] [stages]` / `/reading [seconds]` in an echo session are
the rig for the phone's half. **Zero is never reported**, being the absence
of a sample rather than a measurement -- which matters because
`llama-server` reports a model's load as 0 and then 1 with nothing between
(measured 2026-09-21 on both the 0.6B and the 27B), and reports prompt
processing once a batch; so the bar usually draws for a long prompt and not
for a load.
**A turn's wait has two halves and says which** (2026-09-19):
`SessionStatus::Loading` is the model coming off disk and
`SessionStatus::Reading` is `llama-server` processing the prompt -- emitted
when the request goes out and cleared by the first thing the model says, of
any kind. Prefill is the expensive half here (~10s at 6k tokens, ~22s at
14k), and as `running` it looked exactly like thinking. The phone draws
both with the working spinner and its own words, "loading model" and
"reading prompt".
**Thinking effort is a param, and which levels exist is the model's answer**
(2026-09-19): the `thinking` param rides on the request as a chat-template
argument (`reasoning_effort`, or `enable_thinking: false` for `off`), so it
needs no restart -- and the driver asks the loaded server which levels its
template actually takes rather than trusting the offered list, because the
27B raises on `high` and answers to `xhigh`. A level it cannot take is
dropped and said in the transcript, naming the ones it can.
**Every one of those is a default rather than a constant** (2026-09-19):
`DriverKind::params` declares what a provider takes — key, label, shape,
and whether a change waits for a restart — and the phone renders whatever
arrives, on the spawn form and in the session settings dialog. Adding a
setting to a driver is one entry in that table and no app change. `tools`
is in there too, because the seven built-in definitions are ~1,500 tokens
of every prompt, which on a small window is the difference between a usable
session and one that overruns. `DriverKind::model_params` is the same table
for a provider's **models**, drawn in the machines tab's provider view —
the settings that decide how a model is loaded, which belong to the machine
because one loaded copy answers every session using it.
**A model is downloaded onto the machine that will serve it** (2026-09-19,
replacing the fetch this backend used to do onto its own disk, and the
Models tab that went with it). `models.rs` writes a script and a detached
`curl` runs it *there*; the state of a run is a file beside the partial
(`x.gguf.download`), so nothing about it is held in this process — it
survives the phone closing, this backend restarting and a second device
watching, and `kill -0` at each listing is what stops a machine that was
rebooted from leaving a download claiming to be running. The progress is
`wc -c` of the partial against the size HuggingFace published, the sha256
it publishes is what makes a resume safe, and a finished download is not a
state: it is a model, in the list beside the one still going.
Codex is one persistent `codex app-server --stdio` process per session; its
driver uses native turn steering and interruption, persists the protocol
state and thread id, and reads subscription limits through the same CLI
protocol.
- `app/` — the Compose app, package `com.example.aiapp`, label "AI Sessions".
`AppRoot.kt` is the navigation `when`; `MainScreen.kt` the root's four tabs
(sessions, import, models, setups); `Api.kt`/`EventStream.kt` the REST + SSE
clients; `Events.kt` the event model mirror; `ServerConfig.kt` settings and
**Every text field is `LabelledField`** (`Field.kt`): the label is a line
above the box rather than a thing floating inside it, and the padding is one
line's worth. Material's outlined field spends the height of three lines to
hold one, which on a form of a dozen settings is a screen and a half of
scrolling. What is *not* shrunk is the value -- the framing is what was
expensive. A field's `hint` is what leaving it blank means, drawn inside the
empty box: **a setting is a title and a control and nothing else**, so no
explanatory line sits between them and no paragraph sits under them.
**What a setting costs is asked rather than written down** -- `RestartDialog`
is that question wherever it comes up (a model's settings, the shared
server's, moving a session's directory, changing its thinking level), because
each of them ends something that is running, and a sentence beside the control
is read after the decision if at all.
**One kind of information gets one control.** A choice is `PickerRow` in a
list of settings and `ChipGroup` on a form being filled in, and which of the
two is the screen's to say (`ProviderParamFields`' `choices`) rather than the
provider's -- a llama session's thinking level and a Claude session's have to
look the same.
**Session settings is a screen with two tabs** (`SessionSettingsScreen.kt`),
drawn over the session like the file explorer so the session stays composed.
The second tab is `ProviderScreen` itself -- the same composable the machines
tab opens -- so a provider's settings have two ways in and one
implementation; it takes `onBack = null` there, since the screen around it
has one.
`AppRoot.kt` is the navigation `when`; `SidePanels.kt` the one drag that
slides the whole main screen over a session from the left (`MainPanel.kt`)
and what it has running beside the turn -- its background tasks over its
subagents (`BackgroundTasks.kt`, `SubagentPanel.kt`) -- from the right, both keeping
the session composed underneath; `MainScreen.kt` the root's three tabs
(sessions, import, machines); `Reorder.kt` the drag that moves a row of a
lazy list, used by the session list's handles -- **the order of that list is
the reader's own and nothing sorts it** (`POST /sessions/order`);
`MachineModels.kt` the models on one machine
and the downloads putting them there, drawn inside `ProviderScreen.kt` for a
provider that serves files off that machine's disk; `Api.kt`/`EventStream.kt`
the REST + SSE clients; `Events.kt` the event model mirror; `ServerConfig.kt` settings and
the Keystore-sealed token.
- `wg-app-link/` — a **git submodule** shared with dev-updater: the pinned CA
and leaf (`certs`), QR enrollment and the bearer token (`enroll`), wg0
@@ -39,17 +225,23 @@ Module-by-module intent is in PLAN.md's "Backend layout".
build without it, since it is a path dependency, which is what keeps the two
projects version-locked to the commit this repo pins. What deliberately did
**not** move is the API surface and the config *schema*: routes, drivers,
sessions and setups are what makes this project itself.
sessions and machines are what makes this project itself.
- `SUBAGENTS.md` — a session's subagents as transcripts of their own
(`server/src/session/subagent.rs`, the subcards in `SessionListScreen.kt`
and the read-only form of `SessionScreen.kt`); `DECISIONS.md` holds the
choices made there that are still awaiting review.
**The transcript file itself is readable from the session settings dialog**
(2026-09-21): *View raw* opens the file explorer on it, from
`transcriptFile` on `GET /sessions/{id}` -- which names the machine *this
backend* runs on, not the session's. It is the explorer's third caller and
needed no new screen; a transcript past `FILE_LIMIT` (1 MiB) is refused the
way any other large file is.
- `EXPLORER.md` — the file explorer's design (`server/src/files.rs` and
`FilesScreen.kt` / `FileViewer.kt` / `FileEditor.kt`).
- `TRANSCRIPT_CACHE.md` — the phone's copy of what it has been sent. Read it
before touching `TranscriptCache.kt`, `TranscriptSource.kt`, or the opening
and stream effects in `SessionScreen.kt`.
- `TODO.md` — the working list.
- `RUST.md` — the plan for moving the app to Rust (on the `rustify`
branch of the `ai-app-2` clone): what has to be reproduced, the
framework decision, and the ordered experiments with their pass
conditions. Read it before touching anything under that branch.
- `.dev-updater.ron` — what Dev Updater builds here: the server (run as
`service: Managed(…)`, supervised by Dev Updater's own implementation
rather than a script kept here) and the APK, in parallel. It points at
@@ -142,153 +334,6 @@ two icon buttons the same width without either being given one — and why
genuine handshake against 10.66.0.1 with pinned TLS, no router or phone
involved. That is how to verify the wg0-only posture.
## The rigs
Each exists because something was invisible without it.
- **`app/ui-sandbox.sh`** — a second `ai-server` with its own `$HOME`, config
and data directory, holding eight invented Claude Code transcripts and a
`claude` that is two lines of shell. **That isolation is the point**: the
import screen lists whatever is in `~/.claude/projects`, which in this VM is
real agent transcripts, so exercising *delete* against the ordinary server
deletes somebody's conversation and exercising *import* starts a real
`--resume` on the owner's account.
Its port and root derive from the checkout's name, so two checkouts'
sandboxes cannot reach each other, and its token is generated once into
`~/.config/ai-app/sandbox-token` and carried across restarts along with any
the enrolment flow appended — so the emulator app is enrolled **once** (the
start banner prints the command) and stays enrolled. It shares the real TLS
certificates, because the installed APK pins that CA.
Driving verbs, so none of this is re-derived per session:
`./ui-sandbox.sh spawn [title]` (an echo session, prints its id),
`./ui-sandbox.sh send SID text|@file`, and
`./ui-sandbox.sh api /path [curl args]`.
`./ui-sandbox.sh keep` restarts the server without wiping the sessions and
enrolment already there — for when the fixture under test was expensive to
build; plain `start` wipes them, which is right for the list-screen
fixtures and wrong for that.
It passes `--delay` by default, and `AI_SANDBOX_BIG_MB` puts one large
transcript among the small ones while `AI_SANDBOX_SPAWN_DELAY` makes the
fake CLI slow to start. Both exist because operations that finish in
milliseconds have states on the way that nothing can observe, and an
unobservable state is one where broken and working look identical.
It also builds a fixture tree at the sandbox home's `~/files` for the
explorer, holding the states otherwise only reachable by finding a real
machine in one: an empty directory, a name with a tab and one with an
apostrophe, a binary file, one over `FILE_LIMIT`, one `chmod 000`, a
symlink to a directory and a broken one, a source file per language, and
the three sizes the limits were measured against (`edit-32k.rs`,
`edit-128k.rs`, `big-source.rs`). Point a session at it with
`./ui-sandbox.sh api /sessions/<id>/cwd -X POST -H 'content-type: application/json' -d '{"cwd":"~/files"}'`.
The explorer's 409 is produced by editing the file on the machine
(`printf … > file`) between pressing the pencil and pressing save.
- **`app/debug-transcript.sh`** — a real conversation on the emulator. The
echo driver is the right rig for most things and the wrong one for anything
whose cost scales with what was actually written: a real reply is longer,
is real markdown, and carries tool calls whose input and output are
kilobytes. Two faults were invisible until a real transcript was loaded — a
page of history landing mid-fling threw the reader back to the newest end,
and parsing one real reply took 51ms against 4.6ms for a synthetic one.
`-b` takes the biggest conversation on the machine rather than the newest,
which is what a scrolling test wants; `--stop` takes it down.
It copies the transcript into `/tmp` and gives the server a `HOME` of its
own, so the import can only see the copy — importing spawns `claude
--resume`, and against the real file that is a second CLI writing to a
conversation somebody may still be in. **A transcript never goes in this
repository**: they hold whatever was said, read and written in that
session, and `~/repos` is shared with the host besides.
- **A fake CLI exercises the process lifecycle without a token.** Point a
`claude_cli` provider's `command` at a two-line script — `#!/bin/sh` and
`cat > /dev/null` — and it behaves the way the lifecycle code cares about:
it holds the fifo open, records a real pid, writes nothing, and dies on a
signal. So adopt, stop, restart and start are all drivable without a real
`--resume` and without spending a turn on somebody's account. Reach for
this when what is under test is *whether a process is running*, and for
`debug-transcript.sh` when it is *what the transcript draws*.
- **`app/transcript-bench.sh`** is the standard scroll measurement: it opens
the first session (or `-k` keeps the current screen), scrolls a fixed
gesture loop, and prints the app's render report — the same one the in-app
copy button produces, whose `on screen:` line names what the viewport was
holding. Compare two runs with the same gestures; the emulator's absolute
frame times transfer nothing, the report's accounting does. Run it either
side of any change under `Markdown*.kt`, `Transcript*.kt` or
`SessionScreen.kt`'s list, and put the report in the commit. The numbers
that move first are the worst `record: one block`, the reparse mean while
streaming, and the draw phase's accounting line.
- **`app/stream-bench.sh [-k] FILE`** is that measurement for a reply still
arriving. It taps "Jump to latest" so the list is pinned to the newest end,
resets the report, sends FILE, waits for the transcript to stop growing,
and prints. Both of those are corrections to a first version that measured
nothing: a transcript parked further back never redraws while a reply
streams into it, and a session is idle at *both* ends of a turn, so polling
for idle answers before the turn has started.
- **`app/trace-draw.sh`** names what a scrolling frame spends inside the
framework, from `atrace` text output with no trace processor needed. It is
how the cost of a layout node per link was attributed to the framework
rather than guessed at.
### Driving the UI
**No script that drives this app's UI presses a coordinate.** Every control
is found by the name it already carries for assistive technology —
`ui-trace record --do "tap 'Session settings'"` — which resolves the label
against the screen at the moment of the gesture and fails the whole run when
it is not there. `app/bench-lib.sh` is what the bench scripts share for it. A
coordinate is a position measured once by hand, and anything that moves the
control makes the tap land on whatever now sits there — the bench then
reports a number that was never measured, which reads exactly like a result.
Both bench scripts pressed the render report at `tap 723 205` until that
button moved into the session settings dialog on 2026-09-03. The check that
none has crept back:
grep -n "tap [0-9]" app/*.sh
Swipes are still coordinates, deliberately: a gesture across a scrolling area
is a distance rather than a control.
**Two traps in the emulator bench loop**, each of which cost a run.
`adb shell pm clear` removes the enrolment and the notification permission
along with the saved anchors, so the next run measures a permission dialog —
re-enrol with the command `ui-sandbox.sh` prints, and
`pm grant … POST_NOTIFICATIONS`. And a saved scroll anchor is per session id,
so the only way two builds start a scroll from the same place is a *fresh
session for each*.
**The emulator is `~/repos/emulator-tools`' business, not this repo's.**
`emu up` creates and boots the AVD named after this checkout — whatever `emu
name` prints, never a name typed out here, since this file is the same in
every clone. `run-android.sh` is that plus a build and an install. The `adb`
on `PATH` after sourcing `android-env.sh` is that repo's wrapper, which fills
in `-s` from the same rule. Gradle does not go through it, so a Gradle init
script from `emulator-tools` runs `emu check` before `installDebug`,
`uninstallDebug` and `connectedAndroidTest` and fails rather than fanning out
to every attached device; when it refuses, say which device you mean at the
moment you use it — `ANDROID_SERIAL=$(emu serial) ./gradlew …`.
### Testing llama.cpp and ssh here
The prebuilt CPU llama.cpp lives outside the repo at
`~/.local/opt/llama.cpp` (the 15 MB `ubuntu-x64` release asset). It needs its
own directory on `LD_LIBRARY_PATH`, so start the server as
`LD_LIBRARY_PATH=~/.local/opt/llama.cpp ai-server …` and point a provider's
`command` at `~/.local/opt/llama.cpp/llama-server`. A 0.6B Q8_0 answers at
usable speed on this VM's 8 cores. **Do not test with a 2-bit quant**: the
IQ2_XXS of that model produces fluent nonsense, which reads exactly like a
broken driver — `llama-cli` produces the same from the file directly, which
is how to tell the two apart in a hurry.
There is no second machine, so **ssh this VM to itself**: generate a
throwaway key, append the public half to `~/.ssh/authorized_keys`, and
configure a host of `bob@127.0.0.1` with `identityFile` pointing at it plus
`options: ["StrictHostKeyChecking=no", "UserKnownHostsFile=…"]` so it touches
nothing real. Point a provider's `command` at something harmless like
`/bin/echo` rather than at `claude`: the transport is what is under test, the
process exiting immediately is the signal, and it costs no tokens. **Take the
key back out afterwards.** The remote login shell here is **fish**; the
remote script and `ssh.rs`'s POSIX quoting happen to mean the same thing in
both, but that is luck rather than design, and a shell that is neither is the
thing to suspect first if a remote spawn ever mangles an argument.
## Where things run (host vs this VM)
The machine itself — the two boxes, the shared `~/repos` mount, and why the
@@ -297,7 +342,7 @@ means here:
- **`ai-server` belongs on the host in production.** That is where the LAN
address the phone can reach is, and where WireGuard terminates.
`wg-setup-host.sh` sets that up (keys, `wg0.conf`, the phone's QR); run it
`wg-machine-host.sh` sets that up (keys, `wg0.conf`, the phone's QR); run it
there with `sudo WG_ENDPOINT=<ddns name>`.
- **The tunnel and the real phone can never terminate in the VM**, because
nothing outside can open a connection into it. Phone bring-up is host work.
@@ -330,47 +375,99 @@ day to day:
keep what a development server spawns. The flag decides only what **new**
sessions are marked as; what happens on the way out is decided by the
**mark**.
- **A llama.cpp router is not cleaned up by any of that**, throwaway sessions
included: it belongs to the machine rather than to a session, and a
development server that has loaded a model leaves it loaded — gigabytes of
VRAM — after `pkill ai-server`. Stop it from the machines tab's provider
view, or `pkill -f "[l]lama-server"` when testing.
- Each session directory holds `process.json`, `stdin.fifo`, `stdout.log` and
`stderr.log`. `stdout.log` is the driver's input, read from the byte offset
in `process.json`; removing either by hand while the session is live loses
output or replays it.
## Importing
## Auto-resume
The import list reports each session's **size as well as its line count**,
because the two disagree in the way that matters: these transcripts embed
screenshots as base64, so one line can be a megabyte. On this machine a 69 MB
session has 3,427 lines and a 44 MB one has 6,792 — nothing about a line
count tells you what continuing a session will cost. Shown, not warned about;
importing a large session is a choice somebody is entitled to make.
**A session switched to it sends itself a message once the account's usage
limit lifts** — off by default, per session, in the session settings dialog.
PLAN.md's "Auto-resume" is the design; day to day:
**Never import a Claude Code session that is open in a terminal.** The app
refuses it — see PLAN.md for the incident that made that a refusal rather
than a warning.
- **The schedule is a plan to ask.** `resume.rs` wakes at the scheduled time,
asks `GET /usage`'s meter for that machine and provider, and only sends when
it answers `ok` with nothing at 100%. Anything else — still spent, logged
out, unreachable — is a longer wait, and a still-spent window reschedules to
the reset time the *meter* now gives.
- **Test it with echo, never with a real account.** `/limit [minutes]` reports
the same `limitReached` event a real driver does, and `/usage 100 5` sets
what the meter answers. They are deliberately separate: the two disagreeing
is the case the design exists for. `/usage 20` is the limit lifting.
- The wait is on the session in `config.ron` (`resume`), so it survives a
backend restart. A day after the limit was hit it gives up and says so in
the transcript.
**One Claude Code session id can name two files, and the listing offers it
once.** Resuming from a different working directory makes the CLI write a
second transcript with the same id under that directory's project folder — an
ordinary state of a machine, not corruption. Everything downstream addresses
a session by id, and the phone keyed its list on it, so two rows sharing one
**closed the app** on a Compose duplicate-key throw. `parse_listing` keeps
the copy with the most lines, because the other is usually a few-hundred-byte
stub and is often the *newer* of the two, so recency is the wrong key.
Deleting removes every copy rather than the first, or the row came back after
a delete that reported success. The phone's half is `uniqueItems`, which
every list keyed on a server-chosen id goes through: a repeat there must
never be able to close the app, whatever produced it.
## A session waiting on its own work
**Deleting a session offers to take the machine's own transcript with it** —
`DELETE /sessions/{id}?deleteForeign=true`, behind a switch in the
confirmation, and only where the driver keeps a record of its own
(`keepsOwnTranscript`, which today means Claude Code). Off by default,
because leaving that copy is what makes an ordinary delete recoverable — and
the dialog's paragraph is rewritten when it is on rather than appended to,
since the sentence promising the conversation "should still be there to
import again" is exactly the one the switch makes false. The server deletes
the machine's copy *first*, so a machine it cannot reach leaves the session
where it was instead of half-deleted.
Since 2026-09-06 a session whose turn ended with a **backgrounded subagent or
command still running** reports `waiting` rather than `idle` — its own status,
drawn as the word "waiting" in `waitingColor` on both screens. `idle` means
"waiting for a person" and this means the opposite, so it also suppresses the
"finished" notification, which used to arrive at the one moment it was untrue.
Two things fall out of it and are easy to get wrong again: the queue and the
held-command boundary release on **either** end-of-turn status, so a message
sent while a subagent runs is not held until the subagent finishes; and
`sessionWorking("waiting")` is deliberately **false** — nothing is being
written, and the fold uses that same predicate to decide a reply is settled.
- **Nothing subagent-specific goes in the main agent's transcript** unless a
subagent sends it a real message that wakes it — which is the peer path, and
already has a row. A row per finished background task was tried and was a
screenful of dividers about work nobody was asking after, one of them a whole
shell command. A subagent's report is its own transcript's closing text and
is read in the subcard.
- **A backgrounded command has no subagent, so its report lands in the tool
card that launched it** — a `ToolUpdate` against the call's own id, replacing
the launch result that says it is still running. `/background [seconds]` in
an echo session is that shape end to end.
- **Two replies that meet are separated by a `TurnBreak`** — a hairline, no
words. The reply that follows a turn boundary is a **new** message: the fold
refuses to grow a settled reply, and without that the two ran together
mid-sentence. `./ui-sandbox.sh` plus `/subagent 3` or `/background 5` in an
echo session is the whole rig; the helpers stagger a second apart so each
reply is its own.
- **Claude's background-task level is authority; two edge sources are the
fallback.** Since Claude Code 2.1.261,
`background_tasks_changed { tasks: [...] }` replaces the live set and repairs
a missed ending edge. Its array size is also the measured `backgroundTasks`
count exposed on the session row and event stream; the phone draws a nonzero
count beside the status rather than deriving one from `waiting` or from the
subagent directory. **What those tasks are is `GET /sessions/{id}/background`**,
listed in the session's right panel above the subagents: runtime state, so it
is never persisted and `null` -- not an empty list -- is what a session with
no process answers. **A task is drawn as the command it ran, and tapping it
goes to the call that started it** -- both from the transcript rather than
from the provider: a driver reports which tool call its task belongs to and
`Session::background_tasks` resolves that id into a seq and, for a provider
that says nothing (Codex names a terminal by a process id), the command on
the call. The phone travels there with `travelTo`, the same journey a
reopened session makes to put a reader back where they stopped. An `ambient` task is dropped from both the list and the
count, on the CLI's own instruction: a live-update watcher is not activity,
and counting one leaves a session `waiting` for ever. An adopted CLI is sent a repeated `initialize` to ask
for the current set. Reconcile only between turns or at a result boundary:
a foreground agent is legitimately absent from a background-only snapshot.
Older CLIs still need both edge sources: `open_tasks` knows about a
backgrounded command, while `Subagents::any_open` finds a subagent whose
`task_started` is behind an adopted stdout offset.
- **A usage limit a subagent hits reaches the session**, not just the
subagent's own transcript; auto-resume can only schedule against a session.
That is the case where the main agent is idle and a background Task is
still burning quota.
- **Codex's count is two id sets added together.** Open child thread ids come
from the subagent registry; live background command process ids come from
app-server's experimental `thread/backgroundTerminals/list`. The command set
is runtime state, refreshed at lifecycle edges and once a second while
nonempty. Never decrement it from an unmatched completion.
- **The status word and its colour are `sessionStatusWord` /
`sessionStatusColour`**, shared by the list and the session screen. They
were two `when`s, and the second one silently missed `waiting`.
## Shared appearance
@@ -383,10 +480,136 @@ where it was instead of half-deleted.
swallowed the drag along with the tap, so a list could not be scrolled
while anything in it was busy.
- **A rate-limit bar belongs to a session's provider, not to its machine.**
One machine offers echo, the Claude CLI and a local model at once and only
the CLI spends anything, so a session says which meter reports on it
(`usageProvider`, from `DriverKind::usage_provider`, which
`usage::providers_for` reads too so the two lists cannot disagree) and the
phone matches a snapshot on machine *and* provider. Nothing meters a llama
or echo session, and the phone draws **nothing** for one — not a zero, and
not "unknown". Nothing while the first fetch is out either: "checking"
under a session that turns out to meter nothing is a row the screen then
has to withdraw.
## Things that have bitten
Project-specific only — a lesson that would bite any project on this machine
belongs in `~/.claude/TOOLCHAIN.md` or `~/.claude/MACHINE.md` instead.
- **A server started with no `--tools` answers 403 at `GET /tools`, not an
empty list.** The route is off rather than empty, so reading that as a
failure made "no tools" — the one setting whose entire purpose is to have
none — a session that never started. The router is always given
`--tools all` now and the choice is a filter here, so this is a trap for
whoever next changes how the server is started.
- **`POST /models/load` answers 400 for a model that is already loaded**, and
that is the *ordinary* case once one server is shared: a second session
naming a model somebody else loaded. The router driver asks what is loaded
first and treats "it is there" as the answer whatever the request said.
- **Starting a process from a blocking thread needs the runtime.** Loading a
model is minutes of disk, so it runs on a `std::thread` — and tokio's
`Command::spawn` registers the child with the reactor, so calling it with no
runtime context panics. The panic kills only that thread: the session said
`loading` for ever and nothing appeared in the log. `Routers` holds a
`tokio::runtime::Handle` and enters it around the spawn.
- **A llama session reports `loading`, and a message sent into it queues.**
Before 2026-09-19 the session showed `running` from the moment the process
started, so a minute of reading a model off disk was indistinguishable from
a minute of thinking -- and anything sent in that window came back as an
error, because `llama-server` refuses everything until the model is in
memory. `SessionStatus::Loading` is the state. A driver that reports
`Loading` owes the holding as well as the word, and **the queue is where it
holds**: held inside the turn instead (until 2026-09-20) the message was
recorded as read on arrival, so the phone drew it as sent while nothing was
reading it, and the turn then folded it out of the transcript *and*
appended it, sending it to the model twice. `Shared::await_ready` is now
only for a turn whose model was changed under it.
- **A llama turn that says nothing said something that was thrown away.** Two
silent endings were found on 2026-09-20 and both looked, on the phone, like
a message that was sent and never answered: an `{"error": ...}` chunk
arriving mid-stream on an otherwise successful response (a GPU that ran out
of memory mid-decode), and a stream that simply stops without its `[DONE]`
(the model unloaded under the session). Neither is an ordinary end, and
`generate` now fails the turn for both -- a reply that stops early is not a
reply, and the transcript keeps whatever arrived before it.
- **`<__media__>` in a llama prompt is a picture, wherever it came from.**
llama.cpp takes an image out of the request and leaves that marker in the
rendered text, then pairs each marker with a decoded image at tokenize time
-- so one in words nobody attached a picture to fails the turn with `number
of media markers in text (1) exceeds number of bitmaps (0)`, which reaches
the phone as "Failed to tokenize prompt". A model saying it back is enough,
and then *every* later message fails too, since the conversation is folded
out of a transcript that now holds it. `Message::new` and
`Message::from_user` take it out of anything that is text (`without_marker`),
which is every message with words in it.
- **A cancel flag is only as prompt as the next place somebody looks.** A
llama turn waits on three things that look nowhere at all: a permission
question, a tool call `llama-server` is running (a shell command there runs
to its own timeout, up to a minute), and the completion itself, which says
nothing for as long as the prompt takes to read -- tens of seconds on a long
conversation. Setting `cancel` left the turn exactly where it was until
whichever it was came back, so Pause did nothing on screen for all of it.
**The wait is what ends, not the work**: `awaiting` runs each of those on a
thread of its own and `Shared::abandon_turn` answers the wait, so the turn
ends in milliseconds (measured 43ms in every state) and the abandoned thread
finishes into a channel nobody is reading. Nothing here can stop a shell
command or a model mid-reply, and pretending otherwise is what the old code
did.
**Which is why cancellation is a token per turn** (`Cancel`), not a flag on
the session: the abandoned thread wakes up some time later, and a flag the
next turn had reset would let it write into a conversation it is no longer
part of. Its own token stays set for ever, so it says nothing -- and the
open thinking block is closed by `Shared::abandon_turn` rather than by that
thread, since the one that knows is not the one that ends the turn.
- **Stop ends a llama session and takes the model with it, if nobody else
wants it** (2026-09-21). Until then Stop did *nothing at all* to one:
`stop_session` signals the session's recorded process, and `process::stop`
refuses a `Shared` record -- so the session sat at `idle` with no sign
anything had happened. A `Shared` record now routes to `Driver::stop`,
because what stopping means for a session that borrows somebody else's
process is the driver's to say. The llama driver ends the turn, says
`exited` itself (nothing else will -- there is no process of its own to
die), and asks `Router::release`: each live session claims the model it is
on, and the model comes out of memory only when the last claim goes. A model
another session is using stays.
- **A path is stored as it was typed, and `~` is expanded where it is used.**
`~/repos/x` and `/home/someone/repos/x` are a path and a snapshot of where it
pointed, and the snapshot is what breaks when an account is renamed or the
value is read on another machine -- so nothing at the boundary rewrites one
in either direction (`machines::tidy` used to expand and `shorten_home` used
to contract; both are gone). Expansion belongs to the machine the path is on:
`ssh::quote_path` and `files::PATH_PRELUDE` for a remote one,
`ssh::expand_home` for one here. The exception that proves it is
**`llama-server`'s tools**, which take the working directory as an
`x-tool-cwd` header and `chdir` to it with no shell in the way: a `~` arrives
there as a directory of that name and *every* tool using one answers "failed
to spawn process\n[exit code: -1]", which on the phone looks like a session
whose tools are all broken. `files::resolve_blocking` is what the llama
driver resolves it with at launch, on the machine that will serve the
session.
- **A transcript outlives the enum.** Removing `Event::TaskNote` hours after
adding it made every transcript that had recorded one unreadable, so
`launch` failed for those sessions and `SessionManager::new` skipped them —
no status, nothing sendable, no new messages, for every live session that
had run a background task. **The set of kinds a transcript can hold only
ever grows**: a line may come from a newer server or from an older one that
wrote a kind since dropped, and one unfamiliar word must never be able to
end the file. `Indexed::parse_at` degrades a line it cannot read to
`Event::Unreadable { kind }`, keeping its seq — which is what everything
downstream is addressed by — and the phone draws it as a placeholder saying
which kind. Never delete a variant instead of retiring it; `Event::TaskNote`
is what retiring looks like, and the phone folds it to no row.
Project-specific only. A lesson that would bite any project on this machine
belongs in `~/.claude/MACHINE.md` or the `this-machine-*` skill for its
subject; one that would bite any project anywhere belongs in the
`code-lessons` skill, under the admission test at its end.
- **tracing caches callsite interest process-wide.** A test that hits a
`tracing::warn!` with no subscriber installed can poison the interest cache
@@ -480,6 +703,19 @@ belongs in `~/.claude/TOOLCHAIN.md` or `~/.claude/MACHINE.md` instead.
the reader hit the end of what was loaded on every swipe and stood there
for a round trip. It is `HISTORY_SCREENS` viewports now, counted from what
is actually on screen.
- **A page landing while the history observer was fetching it must trigger its
own successor.** The observer once collected only `LazyListState.layoutInfo`;
while its collector was suspended in `loadOlderPage`, a compact page could
be composed and laid out without leaving another change to observe afterward.
Keying the effect on `oldestSeq` still missed the opening prefetch: that key
changed while `loadingHistory` was true, so the restarted effect declined to
overlap it and never noticed the flag returning to false. Codex exposes both
failures because a page full of calls collapses into one tool group: loading
stopped until expanding that group forced a layout. The observer now collects
the cursor, loading, restoring and failure state with the layout, so returning
to not-loading always rechecks the settled height. A failed page turns the
history boundary into a Try again control rather than retrying in a loop or
requiring another scroll.
- **Only `fetchTranscript` was off the main thread; the fold was not.**
`foldEvent` returns a new list per event, so a page is that many copies of
a growing list — fine at 80 events and about 300,000 element copies at 800,
@@ -487,38 +723,12 @@ belongs in `~/.claude/TOOLCHAIN.md` or `~/.claude/MACHINE.md` instead.
shape: the `markdownIn` scan that decides *what* to parse ran before the
hop to `Dispatchers.Default`. The shape to watch for is a `withContext`
that wraps the *fetch* and leaves the work done with the result outside it.
## Measurements worth not re-taking
- **What the transcript screen costs to scroll.** Taken 2026-08-30 on the GPU
emulator against a real imported transcript with the server at
`--delay 120`. Settled and flinging fast, both into fresh history and back
through rows already drawn: **5.2–5.9% janky frames, 99th percentile
29–32ms, 0–2 slow UI-thread frames.** The stock Settings app on the same
device is 3.3% and 38ms, so this is at the platform floor. The number that
is *not* at the floor is the first few seconds after opening a session,
where every row on the way is being composed for the first time; that is
inherent to a lazy list and it is why a measurement taken before the screen
settles reads three times worse. **Settle first, then reset `gfxinfo`.**
- **The reset path is not reachable by reopening a session.** Measured
2026-09-04 against a session streaming at 20 events a second: reopening one
with an anchor 1,800 events back connects **87–119 events behind**, well
under `CATCH_UP_LIMIT`'s 200, because the restore is two requests — the
opening page, then one span covering the whole distance. To exercise the
reset at all you have to lower `CATCH_UP_LIMIT` in a throwaway build; at 5
the app takes the reset on a live connection, clears, refills and carries
on without reconnecting.
- **The session screen's stream survives backgrounding here** — 20 seconds at
the launcher while 415 events were produced brought no reconnect at all,
which is not what the comment above that loop expects, and is most likely
this emulator being headless rather than the phone's behaviour.
- **Reopening a cached session costs one request for one event** (the probe),
and scrolling the whole conversation back costs nothing more; a cold open
of the same 500-event session is two pages, 100 events. Measured
2026-09-04 on the emulator against the sandbox.
- **Reading is cheap and editing is not.** The viewer handles a 1 MiB,
28,000-line file because it draws one row per line; the editor is one
`BasicTextField`, which costs two seconds a frame at 128 kB and stops the
app at 1 MiB, so `EDIT_LIMIT` caps it at 32 kB with the reason said on
screen. If you make the editor faster, that number is what to move.
EXPLORER.md's "What the measurements said" has the rest.
- **A transcript snapshot cannot survive a suspension and then be assigned.**
`loadOlderPage` joined its page to `items`, suspended while `warm` parsed
markdown, and then assigned the joined snapshot. An SSE event arriving in
that gap appeared and vanished; reopening brought it back because the
transcript and cache had it all along. Warm against a candidate if needed,
then join against the current `items` and assign without another suspension.
Also keep the page's original `oldestSeq`: a stream reset while the fetch or
warm is suspended makes the page stale, and it must be discarded rather
than joined into the reset window.
DECISIONS.md
+44
View File
@@ -0,0 +1,44 @@
# Decisions awaiting review
Choices made while working autonomously, for Bryan to keep or change. Each
says what was picked and why; the detail is in the design doc it names.
Delete an entry once it has been looked at.
## Subagent views (2026-09-05, `SUBAGENTS.md`)
Made on my own judgement, limited blast radius:
1. **A subagent is a transcript, not a session.** It has no process,
controls or settings; it is addressed as `/sessions/{id}/subagents/{sub}`
and stored under the session's directory, so deleting the session takes
it. Alternative rejected: registering it as a session of its own, which
would give it a card in the main list and a driver that can do nothing.
2. **Read-only view is the session screen minus its controls**, rather than
a second, simpler transcript screen. Keeps paging, caching, selection
and rendering in one place. Cost: a `readOnly` mode threaded through
`SessionScreen`.
3. **The list only carries a count.** Each session row says how many
subagents it has; their titles and statuses are fetched when the card is
expanded. Keeps `GET /sessions` from reading every subagent transcript.
Consequence: an expanded card's statuses refresh with the list, not live.
4. **Expanded/collapsed is remembered per session on the phone**, not on
the server. Collapsed by default, per the transcript convention that new
things arrive collapsed.
5. **Subagents of imported sessions are not shown.** The import path still
skips `isSidechain` records; the CLI's own `subagents/agent-*.jsonl` files
are not read. Only subagents run while this backend was watching exist.
6. **Echo grows `/subagent [n]`** as the test rig, so nothing here needs a
paid turn to exercise.
Deferred, because they reach further than this feature:
- **Live status on the list.** Whether the session list should follow a
stream at all (it refreshes on demand today) decides whether subagent
status can ever be live there. Not changed.
- **Nested subagents.** A subagent's own Task calls are shown as tool calls
in its transcript and are not given transcripts of their own. Supporting
that is the same mechanism one level down, but the UI would need nested
expanders.
- **The subagent status row says "context unknown".** Nothing measures a
subagent's context; the row could leave it out rather than admit it.
EXPLORER.md
+48 -19
View File
@@ -14,7 +14,7 @@ AGENTS.md. `server/src/files.rs` is the backend and `FilesScreen.kt` /
## What it is, in one paragraph
A machine's filesystem, seen from the phone through the backend. The explorer
belongs to a **setup** (a machine), not to a session: a session only says
belongs to a **machine** (a machine), not to a session: a session only says
where to start. Every operation — list, read, write, create — is one shell
script run through `Transport`, exactly the way the import listing and the
usage fetch already work, so the local and the ssh case are one
@@ -25,18 +25,30 @@ message. The phone draws what came back.
### 1. Keyed on the machine, opened from the session
Routes live under `/setups/{id}/…`, beside `importable`, because a filesystem
Routes live under `/machines/{id}/…`, beside `importable`, because a filesystem
is a property of a machine. The session screen's folder button opens the
explorer with the session's setup and its `cwd`; a session with no `cwd`
explorer with the session's machine and its `cwd`; a session with no `cwd`
opens at the machine's home, which the **machine** resolves (`cd` with no
argument and `pwd -P`), never a path the phone guessed. Nothing in the
explorer knows what a session is, so a later entry point from the setups tab
explorer knows what a session is, so a later entry point from the machines tab
is one more caller and no new code.
Rejected: routes under `/sessions/{id}/`. The session would be a detour to
find the setup, and "browse this machine" from anywhere else would need a
find the machine, and "browse this machine" from anywhere else would need a
session to exist first.
The third caller arrived 2026-09-21 and cost no code here, which is the
property this decision was made for: **View raw** in the session settings
dialog opens the explorer on the session's own transcript file
(`fileTarget`), so the record can be read as it is on disk rather than only
as the conversation drawn from it. The session says where the file is
(`transcriptFile` on `GET /sessions/{id}`) because only the backend knows --
and it names **this backend's** machine rather than the session's, which for
a remote session are two different filesystems. Back from the file lands in
the session's own directory, where the log and the process record are.
A transcript past `FILE_LIMIT` is refused the same way any other large file
is, which is the known limit of this as a debugging tool.
### 2. One shell script per operation, over `Transport`, on both transports
Each operation is a small POSIX script handed to `sh -c script sh "$path" …`
@@ -68,7 +80,7 @@ Elsewhere the phone picks an **id** and the server resolves which file it
names, so an enrolled token cannot become "read me an arbitrary file". The
explorer's whole purpose is the path, so it takes one. Recorded in PLAN.md's
Security section in these terms: the token already gates spawning a
bypass-permissions agent in any directory on any machine a setup names, and
bypass-permissions agent in any directory on any configured machine, and
that agent can already read and write every file its user can. The explorer
is a shorter path to authority the token already holds, not new authority.
The import rule stands where it is, because there a path was unnecessary and
@@ -82,13 +94,14 @@ writing are fixed scripts; the phone chooses only the path and the bytes.
Same rule as `POST /sessions/{id}/cwd`, with the same wording, because where
a relative path would be depends on something the reader cannot see. Every
listing answers with `pwd -P` of the directory it listed, so the phone
navigates on a resolved absolute path — the parent is a string operation on
that, and a `~` the session was spawned with is shown as what it turned out
to be. The phone never resolves `..` itself.
navigates on a resolved absolute path. The phone also resolves `~` through the
same route, then shortens that directory and every path beneath it back to
tilde notation for display; it never guesses where a local or ssh user's home
is. The phone never resolves `..` itself.
### 5. A read is capped and typed, and every state it can be in has a word
`GET /setups/{id}/file` answers with one of `text` (content, size, mtime,
`GET /machines/{id}/file` answers with one of `text` (content, size, mtime,
sha256), `binary` (not UTF-8; size reported, nothing shown), `tooBig` (over
`FILE_LIMIT`, 1 MiB; size reported so the reader knows what they are looking
at), or the machine's own error.
@@ -101,7 +114,7 @@ what it is.
### 6. A write is conditional on what the reader saw
`PUT /setups/{id}/file` carries the sha256 the read reported. The script
`PUT /machines/{id}/file` carries the sha256 the read reported. The script
compares it against the file as it is now and exits distinctly if it differs;
the server answers **409**. Agents edit files while people read them; this is
the common case, not the exotic one, and silently overwriting an agent's edit
@@ -122,9 +135,9 @@ precondition is fresh without a second read.
### 7. Create refuses to overwrite
`POST /setups/{id}/file` runs under `set -C` (noclobber) and `: > "$1"`, so a
`POST /machines/{id}/file` runs under `set -C` (noclobber) and `: > "$1"`, so a
name that exists fails with the shell's own message rather than truncating
somebody's file; `POST /setups/{id}/dir` is `mkdir --` with the same
somebody's file; `POST /machines/{id}/dir` is `mkdir --` with the same
property. The modal names one thing in the current directory and has a switch
for "directory"; a created file opens straight into edit mode, because an
empty file is not something to look at.
@@ -208,17 +221,21 @@ absence the signal. Back with unsaved changes asks, and says the edits will
be lost. The explorer draws over the session, which deliberately has no
`imePadding`, so the explorer's own box adds it.
### 10. The explorer draws over the session, and back closes it first
### 10. The explorer draws over the session, and back follows what is open
`Screen.Session` in `AppRoot` gains a `files: FilesTarget?`. When set, the
`FilesScreen` is composed **on top of** the session in the same `Box`, and
the session stays composed under it: its event stream keeps flowing, its
scroll position and draft stay where they were, and returning from a file
costs nothing. Back — the button and the platform gesture — clears `files`
when set and goes to the list otherwise. Inside the explorer the same back
steps one level: editor → viewer (with the unsaved question) → listing →
parent directory, and only from the starting directory does it close. "Back
returns; it does not exit."
costs nothing. From an open file, both the header's back button and Android back
return to its containing directory. From a directory, the header's back button
clears `files` and returns to the session. Android back instead walks toward the
session's project directory: upward to the common ancestor, then down one path
segment per press, and at the project it returns to the session. This makes
Back from `/etc` visibly travel through `/`, `/home`, and onward to a project
under `~/repos`, rather than leading away from it. The `..` row remains explicit
parent navigation. An editor with unsaved changes asks before either route
discards them. "Back returns; it does not exit."
Rejected: a `Screen.Files` beside `Screen.Session`. Every route back from a
leaf screen goes to Main today, and a session disposed and re-created on each
@@ -267,6 +284,18 @@ The speedometer went; the report is a "Copy render timings" row in
already are. **Moving it is where the no-coordinate-taps rule got enforced**
(Bryan, 2026-09-03) — see AGENTS.md's "Driving the UI".
### 14. File links in a session open in the explorer
A markdown destination that is an absolute path or a local `file:` URI opens that document in the
session's explorer, on the session's machine. A trailing editor line and optional column are removed;
the viewer opens the file but does not yet scroll to a line. Web links, relative links and `file:`
URIs naming another host keep their ordinary external behaviour. The distinction is deliberately
narrow: a relative link might be a web reference, and the phone must not silently reinterpret it as
a path on another machine.
The markdown link handler is provided around the session rather than taught about machines. That
keeps the renderer reusable and makes the explorer's existing machine target the one navigation path.
## HTTP surface
In `routes.rs`'s module doc with the rest. Bodies use `deny_unknown_fields`
IRIS.md
-79
View File
@@ -1,79 +0,0 @@
# iris: notable public API changes
For Iris to read on her own time. Each entry is a change to iris's public
surface that a widget author or app author would notice: a trait method
added, removed or re-shaped; a type that callers construct differently; a
capability that moved. Small and trivial changes do not go here.
An entry gives the date, what changed, why, and a short before/after where
it helps judge the change without the session that made it. Newest first.
## 2026-09-04: `Widget::draw` reports the size it used; `desired_width`/`desired_height` are gone
A widget used to implement three methods (`draw`, `desired_width`,
`desired_height`); it now implements one, `fn draw(&mut self, painter: &mut
Painter) -> Size`, which draws into `painter.region()` and returns how much
of it was used. Why: the two extra methods routinely re-simulated what
`draw` was about to do anyway (`Span::desired_ortho` copied its own draw
loop to get cross-axis sizing right) — one visit per widget per frame
instead of up to three. A container that needs a child's size before
placing it (alignment, centering) draws the child once at a provisional
region, reads the returned `Size`, and calls the new `Painter::reposition`
to move it into its final spot — an O(1) offset write, not a second draw. A
widget whose drawn output never depends on the size it's given (a
fixed-size `Rect`, a decoded `Image`) overrides the new `fn
is_size_independent(&self) -> bool { false }` to `true`, which skips
redrawing it when only its offered region changes shape.
```rust
// before
fn draw(&mut self, painter: &mut Painter) { /* ... */ }
fn desired_width(&mut self, ctx: &mut SizeCtx) -> Len { /* ... */ }
fn desired_height(&mut self, ctx: &mut SizeCtx) -> Len { /* ... */ }
// after
fn draw(&mut self, painter: &mut Painter) -> Size { /* ... */ }
```
`SizeCtx` and `Cache` are gone with it — see `LAYOUT.md` for the full
design, the move-offset mechanism this shipped alongside, and the file
list.
## 2026-09-04: texture pipeline rebuilt off the binding array
`Textures`/`TextureHandle`, `GlyphPrimitive`, and `UiRenderNode::new` all
changed shape. Why: the old pipeline bound every texture ever drawn in one
`binding_array<texture_2d<f32>>` and asked every device, unconditionally,
for `VK_EXT_descriptor_indexing` — a real share of Android GPUs lack it,
and it failed outright on the Android emulator's software Vulkan. See
TEXTURES.md's "Recommended shape" and "Implemented, 2026-09-04".
- **`UiRenderNode::new` drops its `limits: UiLimits` parameter, and
`UiLimits` is gone.** Before: `UiRenderNode::new(&device, &queue,
&config, UiLimits::default())`. After: `UiRenderNode::new(&device,
&queue, &config)`. Nothing replaces it — there are no more
binding-array limits to size.
- **`src/default/render.rs`'s device request asks for no features and no
binding-array limits.** Before: `required_features:
Features::TEXTURE_BINDING_ARRAY | Features::PARTIALLY_BOUND_BINDING_ARRAY
| Features::SAMPLED_TEXTURE_AND_STORAGE_BUFFER_ARRAY_NON_UNIFORM_INDEXING`
plus two `max_binding_array_*` limits. After: `Features::empty()` (the
`DeviceDescriptor` default) and only `max_buffer_size` set, which was
never about the binding array.
- **`TextureHandle` has no `primitive()` method any more**; a caller
outside `iris` shouldn't have been calling it (it fed the old renderer's
internals), but if something did: use `image_index()` for a standalone
image's bind-group index. There is no equivalent for a page — a page has
no bind group of its own now, see below.
- **`GlyphPrimitive` has no public constructor from a struct literal.**
Before: `GlyphPrimitive { uv_min, uv_max, view_idx, sampler_idx, color,
flags }`. After: `GlyphPrimitive::new(uv_min, uv_max, layer, color,
flags)` — one `layer` (the shared atlas array's layer) instead of a
`view_idx`/`sampler_idx` pair, since a page is now a layer of one array
texture rather than its own bound texture.
- **A widget author drawing images is unaffected**: `Painter::texture`/
`texture_at`/`texture_within` and `Textures::add` keep their signatures.
What changed underneath is that each standalone image now gets its own
`wgpu::BindGroup` and draw call instead of a slot in the shared array —
invisible from the widget API, visible only in `UiRenderNode`'s internals
and in `iris`'s device requirements.
IRIS_TODO.md
-66
View File
@@ -1,66 +0,0 @@
# iris: known problems and things still to build
Iris's own list for the library, recorded 2026-09-04 in her words where it
matters, so the agents working through RUST.md pick these up in a sensible
order rather than rediscovering them. Each item says where it sits in the
order and what "done" looks like. Tick and date them in place.
## Fix
- [x] **Input does not fall through by input type (2026-09-04).**
`SensorUi::run_sensors` (`src/default/sense.rs`) used to set "consumed,
stop checking lower layers" from mere hover — a widget registered for
nothing but `click()` blocked a `Scroll` meant for whatever was behind
it, since "the cursor is over this widget" and "this widget handled the
event" were the same check. Fixed by judging consumption per input
kind: with no button transition and no scroll happening this frame
("momentary" activity), the topmost hovered widget still wins, same as
before; when something momentary *is* happening, only a widget whose
registered senses actually include a matching non-hover one (checked
via a new `TypeEventManager::registered`, which lists what a widget
registered without running anything) consumes it, so a widget with only
`Hovering`/click handlers can no longer block a scroll from reaching a
list underneath. `iris/src/sense_tests.rs` builds a button-over-a-list
`Stack` with a plain `HasEvents` impl (no GPU or window) and checks both
directions: a scroll over the button reaches the list, and a real click
still reaches the button — confirmed to fail on the pre-fix code and
pass after.
## Build
- [ ] **Benchmarks**, not unit tests, run on demand (a `benches/` or a
script under `iris/`, never in `cargo test`). The scenario that matters
most is a **message list** — chat apps and this app's transcript alike —
stressed with many messages and many images. One case in particular:
**resizing an input box** (typing enough text to grow it) that pushes a
long list of messages above it must stay very fast and recalculate
almost nothing — a move of everything above, not a re-layout. That is
exactly the O(1) move chain in LAYOUT.md; the benchmark is what proves
it. Done when the numbers are in this file with the command, and the
input-box case reports draws re-run, not just frame time.
- [ ] **Masks defined relative to each other.** Wanted: mask A multiplies
by something *and also* applies mask B — a mask can reference a parent
mask, the way the move chain references a parent offset. Today masks
are independent regions. Design it beside the move chain (same shape:
a parent index and a bounded walk in the shader); do it when a real
widget needs it, not before.
- [ ] **Positions as a single float per scroll.** Iris raised, and half
rejected, letting a scroll update one float rather than positions:
input handling cares about most elements in a list, so absolute
positions must be computed on the CPU anyway. LAYOUT.md's design
already lands here (GPU walks the chain, CPU resolves on demand for
hit tests). Keep the CPU resolution lazy and per query; do not
materialise every row's absolute position per frame.
- [ ] **Animations, last.** Cosmetic, so after everything above. Must be
**modular — a piece of the library rather than a core part forced into
everything, the same way input is**. Whatever the mechanism, a widget
that does not animate must pay nothing and import nothing for it.
## Reconsider
- [ ] **`WidgetView`.** Iris is unsure of it: what she wants is an easy way
to compose a widget from others (a button is the main case). With
sizing folded into `draw`, composing may be easy enough that `View` is
redundant. Decide after the layout change lands, by writing a button
both ways and keeping the one that is shorter to explain; delete the
other rather than keeping two ways.
LAYOUT.md
-897
View File
@@ -1,897 +0,0 @@
# iris: one `draw` that reports a size
Preference stated by Iris, 2026-09-04, on the `rustify` branch. Recorded before
any design or code so that it survives a cleared session. **Status: implemented
2026-09-04, against every pass condition in §8** (measured, not assumed — see
that section). Every widget listed in §7 was migrated in one change; none
kept `desired_width`/`desired_height`. Five points needed correction or
refinement beyond what this file originally specified — see "Deviations
found during implementation" below, added right before "For IRIS.md" — read
that section before touching `Aligned`, `Sized`, `MaxSize`, `Scroll`, or the
move-slot lifecycle in `render_state.rs`, since each of those five is a real
bug this file's first draft would have reproduced if implemented literally.
## What Iris asked for
> I don't like that widgets need both a draw and size functions. I'd much
> rather them have a single draw that reports a size, and if it needs to be
> moved then that can be done after the fact efficiently, or resized just
> done after as well. This should be done efficiently like everything else
> tries to do right now.
She added, a few minutes later: "single draw is not a requirement. It
just seems more efficient from what I've heard. Feel free to override any
decision I've made if you can find a genuinely better & still clean
alternative." So the single-draw model is the default to design against,
and the design below may reject it, but only with a written comparison
showing the alternative does less work per frame and is no harder to use.
Standing constraints from RUST.md still apply: no DSL, plain Rust, do as
little processing as possible per frame, but the model must cover every
layout need a real app has (the transcript's virtualised list, wrapped
text whose height depends on width, rows and columns that size to their
children, overlays, masks).
## What exists today
`Widget` (`iris/core/src/widget/mod.rs`) has three methods: `draw(&mut
self, &mut Painter)`, `desired_width(&mut self, &mut SizeCtx) -> Len` and
`desired_height`. A parent asks `SizeCtx::width/height` for a child, which
is memoised per widget id and axis in `Cache.size` keyed on the outer
size, then places the child with `Painter::widget_within(region)`. So a
child is visited twice (sized, then drawn), every widget implements sizing
twice (one per axis), and a widget whose size depends on what it draws
(wrapped text, a laid-out paragraph) does the layout in the size pass and
again in the draw pass unless it caches by hand.
Primitives are already positioned by `UiRegion` values whose scalars have
a `rel` and an `abs` part, resolved against the window in the vertex
shader (`core/src/render/shader.wgsl`), and `Primitives::region_mut`
exists to rewrite one instance's region in place. That is the mechanism a
"move after the fact" can build on.
## What the design must answer
1. **Parent-before-child ordering.** A row has to know each child's width
to place the next one, but under "one draw" the child's size only
exists after it has drawn. The answer is meant to be: the child draws
at a provisional origin, reports its size, and the parent *moves* it.
The move must be O(1) per moved subtree, not O(primitives in the
subtree). One way: every instance carries an index into a small
per-widget offset buffer, so moving a widget writes one entry and the
vertex shader adds it. Other ways may be better; the design should say
what was considered.
2. **Move vs resize are different costs and must be kept apart.** A move
never re-runs `draw`. A resize re-runs `draw` for exactly the widgets
whose size input changed, and a widget whose output does not depend on
its size (an icon, a fixed rect) must be able to say so and be skipped.
3. **Size-dependent content.** Wrapped text is the hard case: its height
is a function of its width. A single `draw` receives the available
size (what `SizeCtx.outer` is today) and reports what it used, so the
two-pass "measure then draw" collapses into one for the common case.
The design must say what happens when a parent wants the child's
height *before* deciding the width it will offer (rare; say whether it
is supported, or is done by drawing twice as an explicit, opt-in cost).
4. **Caching.** Today's `Cache.size` memoises by (id, axis, outer). The
replacement should memoise the whole draw result by (id, available
size) so that an unchanged subtree costs nothing on the next frame,
which is what makes a virtualised list cheap.
5. **Everything currently written against `desired_width`/`desired_height`
moves over in one change**, per the code rules: two names for one
concept is not an intermediate state to leave behind. The widgets are
in `iris/src/widget/` (`ptr`, `mask`, `image`, `rect`, `trait_fns`, and
whatever else is there when the change is made).
## Order relative to the texture work
TEXTURES.md's redesign touches the render core (shader, `GpuTextures`,
`Primitives`, `Painter`'s texture calls). This change touches the widget
trait, `SizeCtx`, `Cache`, `Painter`'s widget calls, and any offset
mechanism the vertex shader needs. They overlap in `Painter` and the
shader, so they are done **in sequence, textures first**, and the layout
design here is written (not implemented) while the texture work is in
progress, then implemented on top of it.
## Design
### 1. The new `Widget` trait
```rust
pub trait Widget: Any {
/// Draw within `painter.region()` (the space the parent offered) and
/// report how much of it was actually used, per axis.
fn draw(&mut self, painter: &mut Painter) -> Size;
/// True if `draw`'s output (both the primitives it writes and the
/// `Size` it returns) is the same for any `painter.region()` of the
/// same *content* -- an icon, a fixed-size rect, an already-decoded
/// image at its natural size. Default `false` (redraw on any change to
/// the offered region) because assuming independence wrongly produces
/// a stale draw; a widget must opt in.
fn is_size_independent(&self) -> bool {
false
}
}
```
No `available` parameter: `Painter` already carries the region the parent
handed down (`Painter::region()`, `core/src/ui/painter.rs:137`) and already
exposes the pixel-resolved form (`px_size()`, `:156`) and the output surface
size (`output_size()`, `:152`). Passing it again would be the same value
under a second name. `desired_width`/`desired_height` (`core/src/widget/mod.rs:20-21`)
and `WidgetAxisFns::desired_len` (`:24-35`) are deleted outright — not
deprecated, not kept as a fallback — because a widget that implements both
`draw` and `desired_*` for the same thing is exactly the "two names for one
concept" the code rules call out, and it is what today's `Span::desired_ortho`
(`iris/src/widget/position/span.rs:98-152`) already complains about in its
own comment: "this literally copies draw so that the lengths are correctly
set in the context, which makes this slow and not cool." Folding sizing into
`draw` deletes that duplicate simulation, not just moves it.
**No single-draw alternative was found that does less work per frame.** The
two-method trait was checked against three properties a real screen needs —
a row placing children in sequence, a widget centering on its own content,
and wrapped text — and in every one, `draw` already has to visit the child
to get a size that is *this specific one's* answer, which today's
`desired_width`/`desired_height` re-derive by re-running (a shrunk copy of)
the same layout the draw pass will do again. So the two-method trait is not
"measure once, draw once" in the general case; it is "measure once per axis,
then draw once," i.e. up to three visits per widget per frame, against one
under the design here. The single-draw model is therefore adopted as
proposed, not merely accepted as a preference.
### 2. Move: O(1) per moved subtree, via a per-widget offset chain
**What exists today, and why it is not O(1).** `UiRenderState::mov`
(`core/src/ui/render_state.rs:156-168`) fires when a widget's region keeps
its *size* but changes *position* (`draw_inner`, `:85-100`:
`active.region.size() == region.size()` after excluding the exact-match
case). It rewrites every primitive's `region` field via
`Primitives::region_mut` (`core/src/render/primitive.rs:176-179`) for the
widget's own primitives, then recurses into every child — O(primitives in
the subtree). Both call sites that trigger it today, `Scroll::draw`
(`iris/src/widget/position/scroll.rs:29-31`) and `Offset::draw`
(`iris/src/widget/position/offset.rs:9-11`), are "translate this subtree by
an abs pixel amount, `rel` framing unchanged" — a transcript scroll
re-touches every glyph in every visible row, every frame of the drag, and
I3's target is 800 rows on screen.
**Recommendation: a per-widget offset slot forming a parent-linked chain,
resolved in the vertex shader.**
- `UiData` (`core/src/ui/mod.rs:14-20`) gains
`pub move_offsets: TrackedArena<MoveOffset, u32>`, the same arena shape
already used for `masks: TrackedArena<Mask, u32>` on the line above it.
- `render/data.rs` gains `pub struct MoveOffset { pub delta: [f32; 2], pub
parent: u32 }` (`Pod`/`Zeroable`, `parent = u32::MAX` = "no ancestor,
add nothing more"). A pure abs-pixel translation, not a general
`UiRegion` remap — sufficient for every existing call site (above).
- `PrimitiveInstance` (`render/data.rs:11-18`) gains `pub move_idx: u32`,
a vertex attribute at `@location(7)` beside `mask_idx` at `6` — the same
kind of per-instance handle.
- `ActiveData` (`core/src/ui/active.rs`) gains `pub move_slot: MoveIdx`,
assigned **when the widget is first drawn** (`draw_inner`, beside
`active.insert`), with `parent` = the drawing widget's parent's slot.
`Painter` threads a `move_slot` field down exactly as it already threads
`mask` and `layer` (`painter.rs:9-20`), so a freshly-drawn descendant is
correct from its first frame — nothing is ever retrofitted onto an
already-active primitive. An unmoved widget's slot just stays `[0, 0]`.
- `Painter::primitive_at` (`painter.rs:23-38`) writes `move_idx:
self.move_slot`, matching how it already writes `mask_idx: self.mask`.
- `mov(id, delta)` becomes: look up `id`'s slot, write
`move_offsets[slot].delta += delta`. One write — no primitive touched, no
recursion, since descendants already reference this slot transitively.
- `shader.wgsl`'s vertex stage, after computing `top_left`/`bot_right` in
pixels (after `:106`, before the clip-space divide at `:113`), walks
`move_idx → move_offsets[i].parent` for a bounded number of steps (a
small constant, e.g. 16, with a CPU-side debug assertion that no chain
exceeds it), summing `delta` into both corners. Cost is O(chain depth),
paid every frame regardless of whether anything moved — negligible next
to the per-fragment texture sampling TEXTURES.md already measures this
GPU as not bound by.
**Why the chain, not the flatter thing first proposed.** Iris's own
phrasing — "every instance carries an index into a small per-widget offset
buffer" — describes a flat table: one slot per subtree *declared* movable,
no parent link. It breaks the moment two such subtrees nest — a row inside
a scrolling list, itself later given its own animated offset (a
swipe-to-delete mid-scroll) — because the row's primitives would have to
pick one slot and lose the other's contribution. The chain costs one extra
field and a bounded shader loop in exchange for no such gap, and since
every `ActiveData` gets a slot unconditionally rather than lazily, it costs
no more at the common depth of one than the flat version would.
**Against `region_mut` as the steady-state mechanism**: rejected for being
O(primitives in the subtree) — the cost this section removes — but kept
for a resize that changes a region's `rel` component (a genuine reflow,
§3) and for a size-independent widget's resize (§3), where the content's
shape doesn't change and one field write already suffices.
### 2b. Two more readers of "where is this widget," and masks
Moving the offset into the vertex shader means `ActiveData.region` is no
longer the on-screen truth once a widget has been moved — it is where the
widget was *drawn*, before any `move_offsets` delta. Two things read it as
if it still were, and both must move to a resolved query or they silently
answer with the pre-move position: a click landing on a scrolled row would
be routed to whatever used to be there, with nothing on screen to say so —
exactly the "wrong answer that looks like a right one" case the code rules
single out.
**Hit-testing.** `SensorUi::run_sensors` (`src/default/sense.rs:154-200`)
does the actual pointer routing, and line 170 is the read in question:
`let shape = self.active.get(id).unwrap().region;` (`self: &UiRenderState`),
immediately turned into pixels and tested against the cursor at `:171-172`.
Under this design that region must be resolved through the same chain the
GPU walks before it means anything. Add to `UiRenderState`:
```rust
/// `active[id].region`, corrected by every `move_offsets` delta between
/// `id` and the root — the CPU-side twin of the vertex shader's chain
/// walk, over the same arena, so the two cannot disagree about where a
/// widget is. O(chain depth), not O(primitives): a plain Rust loop over
/// `move_offsets`, bounded by the same constant the shader loop uses
/// (name it once, e.g. `render::MOVE_CHAIN_LIMIT`, and reference it from
/// the WGSL loop bound in a comment, since WGSL cannot `include!` a Rust
/// const across the language boundary).
pub fn resolved_region(&self, id: WidgetId) -> UiRegion;
```
`window_region` (`core/src/ui/render_state.rs:264-267`), the public
coordinate query already used outside hit-testing
(`src/default/attr.rs:15,17,70`, e.g. positioning one widget relative to
another's on-screen box), is reimplemented to call `resolved_region(id)`
before `.to_px(...)` instead of reading `.region` directly — one change
covers both call sites listed there. `sense.rs:170` changes to
`let shape = self.resolved_region(*id);`. Both are required the moment §2
lands, not an optional follow-up: an unmoved widget's chain is empty and
`resolved_region` costs one arena read to find that out, so there is no
version of this design where skipping the fix is a legitimate
optimization — it is a correctness gap, not a performance one.
**Masks.** `Painter::set_mask` (`core/src/ui/painter.rs:49-52`) bakes the
painter's *current* region into a `Mask` pushed onto
`masks: TrackedArena<Mask, u32>` (`core/src/ui/mod.rs:19`), and the
fragment shader clips every primitive against `masks[in.mask_idx]`'s raw
`rel`/`abs` fields, unaffected by any move (`shader.wgsl:147-157`). If the
widget that called `set_mask` — `Masked::draw`,
`iris/src/widget/mask.rs:7-11`, `painter.set_mask(painter.region()); ...` —
is itself later moved, its clip rectangle stays where it was drawn while
its content moves out from under it: a visibly wrong clip, immediately on
screen, not a latency question.
Fix: `Mask` (`core/src/render/data.rs:46-49`) gains `pub move_idx: u32`,
written from `Painter::set_mask` as `self.move_slot` — the identical slot
the mask-owning widget's own primitives already get (§2), not a second
mechanism. Resolution happens in the **fragment** shader, not the CPU, and
not the vertex shader either: `shader.wgsl`'s mask check (`:147-157`)
currently computes the mask's `top_left`/`bot_right` inline from
`masks[in.mask_idx]`; that computation is extended to walk the same
move-offset chain §2 added, via one shared function —
```wgsl
fn resolve_move(idx: u32) -> vec2<f32> { /* the bounded parent walk, used by both stages */ }
```
— called from `vs_main` for a primitive's own corners and from `fs_main`
for its mask's corners, so the walk is written once and the two stages
cannot drift apart (the sibling-rule from the code rules: one loop, not a
hand-copied second one in the other shader stage).
**Why the fragment shader, not a CPU-side mask rewrite at move time.** A
primitive's mask is frequently owned by a *different* widget than the
primitive itself — often several levels up a subtree, with its own,
independent move slot — so a primitive's resolved offset and its mask's
resolved offset are two different chain sums, both needed, and only the
fragment shader has both `in.move_idx` (this fragment's own chain) and
`in.mask_idx` (indirecting to a second, possibly unrelated chain) already
in hand per-fragment. Resolving mask regions on the CPU at move time would
mean, for every `mov()` call, walking forward to every mask instance the
moved widget's slot could affect and rewriting its raw region — exactly
the O(subtree) cost §2 exists to remove, just moved from primitives to
masks. The fragment shader already re-reads `masks[in.mask_idx]` every
frame (`:148`); one more arena read to resolve its chain costs nothing
extra in kind.
**The scroll-container case, checked rather than assumed.** A masked,
scrollable region is built as a `Masked` wrapping a `Scroll`
(`iris/src/widget/position/scroll.rs`, `iris/src/widget/mask.rs`) — the
viewport border is drawn (and `set_mask` called) by `Masked`, which is
never itself the target of `mov()`; only `Scroll`'s inner content is,
every frame the user drags. Because each widget's move slot is its own
(§2: assigned per `ActiveData`, not shared), `Masked`'s mask references
its own, stationary slot, while the scrolled content underneath references
a separate, deeper slot whose `parent` chain passes through — but does not
write to — the viewport's slot. Moving the content therefore never touches
the mask's resolved position, and the mask staying still while its content
slides past it is what this design already produces with no special case,
not an extra rule that had to be added for it.
### 3. Resize scope
A resize is "the region a widget's parent offers it changes such that the
widget's draw might produce different output" — as opposed to a move, which
by construction cannot (§2 is scoped to pure translation). Two independent
narrowings apply, and both are real, measured properties of the code as it
stands rather than new machinery:
**(a) A window resize does not, by itself, require touching most widgets.**
`shader.wgsl:105-106` recomputes every primitive's pixel position from
`window.dim` and the primitive's stored `rel`/`abs` pair *every frame,
already, on the GPU*. A widget laid out purely in `rel`/`abs` terms (no
call to `px_size()`, `output_size()`, or anything else that reads a
concrete pixel count) is therefore already correct after a resize with zero
CPU work — the shader did it. `UiRenderState::needs_redraw_all`
(`render_state.rs:229-231`) currently ignores this and redraws the entire
tree on every `resized`, which was the safe default while sizing and
drawing were two passes; it should be narrowed to only the widgets that
*do* read a concrete pixel value. Track this the same way `needs_redraw`
already tracks per-widget dirtiness (`Widgets::needs_redraw`,
`core/src/widget/widgets.rs:9`): a widget's `draw` call marks itself
pixel-dependent by calling through `Painter` methods that read
`output_size`/`px_size` (both already funnel through `Painter`, so the
marking is one line at each), and `resize()` (`render_state.rs:32-35`)
walks only that set instead of unconditionally setting `resized = true`
for a full `redraw_all`. This turns "every resize redraws everything" into
"every resize redraws what depends on pixels" — a real behavior change
beyond what was asked, so verify it against the I0b `pre_present_notify`
resize regression (that fix depended on `redraw_all`'s completeness)
before narrowing this.
**(b) A widget's `available` (its parent's offered region) can change
without the widget's *content* changing — this is what
`is_size_independent` (§1) answers.** When a container's own layout shifts
(a sibling grew or shrank, changing this widget's offered box), a widget
that returns `true` from `is_size_independent` is not redrawn: its
primitives are unaffected by size, only by placement, so the parent
either (i) issues a move (§2) if only position changed, or (ii) rewrites
the primitive's `region` fields directly via `region_mut` if the box
changed shape too (still O(primitives owned directly by this widget, not
its subtree, since a size-independent widget by definition has no
size-dependent descendants worth distinguishing — in practice this is
always a leaf: `Rect`, `Image`, a fixed glyph). A widget that returns
`false` (the default) is redrawn in full whenever `available` changes,
which is correct always, just not free.
**Ancestor propagation** (a resized child changing its own reported size,
requiring its parent to re-lay-out) is unchanged in spirit from today's
`redraw` (`render_state.rs:270-305`), which already walks up exactly the
ancestors whose cached size differs from the new one and stops as soon as
a size is unchanged (`:274-286`). That loop moves from consulting
`Cache.size` to consulting `ActiveData.size` (§5) but keeps its shape.
### 4. Wrapped text, and "needs child height before choosing width"
**Wrapped text is not a special case any more; it already reads as one
draw.** `TextView::render` (`iris/src/widget/text/mod.rs:57-76`) already
does exactly what single-draw asks for: it reads `ctx.px_size().x` as the
wrap width, shapes once, and memoizes the shaped layout keyed on that width
plus a changed-flag on the buffer and attrs (`:63-69`) — a second call with
the same width is a hash-map-style cache hit, not a re-shape. Under the new
trait this collapses `Text::draw`/`desired_width`/`desired_height`
(`text/mod.rs:133-147`, three functions) into one `Text::draw` that calls
`self.view.draw(painter)` once, which internally still calls `render`
once, hits its own cache, and returns the size it already computed. No
new caching is needed here; the two now-redundant call sites
(`desired_width`/`desired_height` each separately calling `render`) simply
disappear, which is a second `render` avoided per frame per text widget
that is being measured by a parent.
**"Parent wants the child's height before deciding the width it will
offer"** — the genuinely circular case named in the brief, e.g. a column
that sizes its own width to its widest child, where that child is wrapped
text whose height (which the column's *own* height depends on) depends on
the width the column has not yet decided. This is not solvable in one pass
for the same reason it is not solvable in CSS shrink-to-fit with wrapped
content: the two axes' answers are mutually dependent. `Span::desired_ortho`
(`span.rs:98-136`) already hits exactly this today and already resolves it
by an explicit second, throwaway pass (its own comment: "this literally
copies draw ... which makes this slow and not cool"). The design keeps that
resolution, made explicit rather than accidental: `Painter` gets
```rust
/// Draw `child` at a provisional region to learn its size under one
/// axis's worth of assumption, discard everything it wrote, then draw it
/// again at the region that assumption produced. For the rare parent that
/// cannot pick an offered size without already knowing the answer.
/// Twice the cost of one `draw`; every other case in this file avoids it.
pub fn draw_twice(&mut self, child: &StrongWidget, first: UiRegion, second: impl FnOnce(Size) -> UiRegion) -> Size;
```
implemented as: draw at `first`, record `Size`, remove the widget and its
subtree the same way a resize-triggered redraw already does (`draw_inner`'s
"if not \[same region\], maintain resize and track old children," `:97-100`,
which already frees the old primitives before redrawing) — reusing that
path rather than adding a second one — draw again at `second(size)`, return
the final `Size`. It is opt-in and named for its cost, so a widget only
pays it if it is the one that needs it; `Span`'s cross-axis case is the one
call site converted to it, replacing the hand-rolled duplicate loop.
### 5. Caching and invalidation
`Cache.size` (`core/src/ui/cache.rs`) is **deleted, not replaced with an
equivalent** — the thing it memoized (a `desired_width`/`desired_height`
answer, independent of drawing) no longer exists as a separate query, so
there is nothing left to cache at that layer. What already provides "an
unchanged subtree costs nothing" is the check `draw_inner` performs before
touching a widget at all (`render_state.rs:85-90`): if the widget is active,
its region is unchanged, and it is not marked dirty, `draw_inner` returns
immediately — no `Painter` constructed, no primitive touched, no shader
work beyond what the GPU already redraws from the unchanged instance
buffer. That check is kept exactly as it is; it is the caching mechanism,
and it already operates at (id, region) granularity, which subsumes "(id,
available size)" once size *is* what a region change means.
What is added: `ActiveData` gains `pub size: Size` — the value `draw`
returned, stored the moment it is (`draw_inner`, alongside building the
`ActiveData` struct at `:134-143`). This is what a parent placing this
widget for a second frame without redrawing it (because nothing changed)
reads instead of recomputing — it replaces `Cache.size`'s role of "answer a
size question without a full draw" with "read the size of the last actual
draw," which is always available because `draw_inner`'s skip path is only
reachable once the widget has been drawn at least once. `Cache::remove`/
`Cache::clear` (`cache.rs:9-17`) are deleted with the type; `ActiveData`
already has an equivalent lifecycle (removed in `remove`/`remove_rec`,
`render_state.rs:171-198`, freed with the widget).
### 6. Before / after
**A leaf, `iris/src/widget/rect.rs`** — the size-independent case:
```rust
// before
impl Widget for Rect {
fn draw(&mut self, painter: &mut Painter) {
painter.primitive(RectPrimitive { color: self.color, radius: self.radius,
thickness: self.thickness, inner_radius: self.inner_radius });
}
fn desired_width(&mut self, _: &mut SizeCtx) -> Len { Len::rest(1) }
fn desired_height(&mut self, _: &mut SizeCtx) -> Len { Len::rest(1) }
}
```
```rust
// after
impl Widget for Rect {
fn draw(&mut self, painter: &mut Painter) -> Size {
painter.primitive(RectPrimitive { color: self.color, radius: self.radius,
thickness: self.thickness, inner_radius: self.inner_radius });
Size::REST // fills whatever it was given -- used == available
}
fn is_size_independent(&self) -> bool { true } // content never depends on region size
}
```
**A container that needs the child's size before placing it,
`iris/src/widget/position/align.rs`**:
```rust
// before
impl Widget for Aligned {
fn draw(&mut self, painter: &mut Painter) {
let region = match self.align.tuple() {
(Some(x), Some(y)) => painter.size(&self.inner).to_uivec2().align(RegionAlign { x, y }),
(Some(x), None) => { let x = painter.size_ctx().width(&self.inner).apply_rest().align(x);
UiRegion::new(x, UiSpan::FULL) }
(None, Some(y)) => { let y = painter.size_ctx().height(&self.inner).apply_rest().align(y);
UiRegion::new(UiSpan::FULL, y) }
(None, None) => UiRegion::FULL,
};
painter.widget_within(&self.inner, region);
}
fn desired_width(&mut self, ctx: &mut SizeCtx) -> Len { ctx.width(&self.inner) }
fn desired_height(&mut self, ctx: &mut SizeCtx) -> Len { ctx.height(&self.inner) }
}
```
```rust
// after
impl Widget for Aligned {
fn draw(&mut self, painter: &mut Painter) -> Size {
let full = painter.region();
// Draw once at the full region to learn the child's real size --
// this placement is provisional and corrected below without a
// second draw.
let used = painter.widget_within(&self.inner, full);
let region = match self.align.tuple() {
(Some(x), Some(y)) => used.to_uivec2().align(RegionAlign { x, y }).within(&full),
(Some(x), None) => used.x.apply_rest().align(x).within(&full),
(None, Some(y)) => used.y.apply_rest().align(y).within(&full),
(None, None) => full,
};
painter.reposition(&self.inner, region); // O(1): one offset write, no second draw
used
}
}
```
`Painter::widget_within`/`widget`/`widget_at` (`painter.rs:55-76`) change
return type from `()` to `Size`, carrying the child's `draw` result back —
the only signature change needed to let a parent see what its child used.
`Painter::reposition` is new, computing the delta between where a child
was actually drawn and where it belongs and calling the O(1) `mov` from
§2. `SizeCtx` and `Painter::size_ctx`/`size`/`len_axis` (`painter.rs:141-150,
180-182`) are deleted — nothing calls `desired_len` any more, so there is
nothing left for `SizeCtx` to answer; `draw_text`/`label`/`px_size`/
`output_size` already exist redundantly on both `SizeCtx` and `Painter`
today (compare `size.rs:71-90` against `painter.rs:152-174`) and this
deletes the `SizeCtx` copies, keeping the `Painter` ones.
### 7. Migration — every file and widget that changes
One change, in dependency order (rename-and-move-together, per the code
rules — no intermediate state with both trait shapes):
- `core/src/widget/mod.rs` — the `Widget` trait (§1), delete
`WidgetAxisFns`, update `impl Widget for ()`.
- `core/src/ui/size.rs` — delete `SizeCtx` (the type and all its methods).
- `core/src/ui/cache.rs` — delete `Cache` (§5).
- `core/src/ui/painter.rs` — `widget`/`widget_within`/`widget_at` return
`Size`; add `reposition`, `draw_twice`; delete `size_ctx`, `size`,
`len_axis`; `primitive_at` writes `move_idx`.
- `core/src/ui/render_state.rs` — `draw_inner` captures and stores
`ActiveData.size`; `mov` becomes the O(1) offset write (§2); resize
narrowing (§3a); `redraw`'s per-axis loop reads `ActiveData.size`
instead of `Cache.size`.
- `core/src/ui/active.rs` — `ActiveData` gains `size: Size`,
`move_slot: MoveIdx`.
- `core/src/ui/mod.rs` — `UiData` gains `move_offsets`.
- `core/src/render/data.rs` — `PrimitiveInstance` gains `move_idx`;
new `MoveOffset` struct.
- `core/src/render/primitive.rs` — thread `move_idx` through `PrimitiveInst`
and `Primitives::write`, matching `mask_idx`.
- `core/src/render/mod.rs` — bind the new `move_offsets` storage buffer
(group 2, beside `masks`) and its update path.
- `core/src/render/shader.wgsl` — `InstanceInput` gains `move_idx`;
`MoveOffset`/`UiScalar`-shaped storage binding; a shared `resolve_move`
function (§2b) called from both `vs_main` (a primitive's own corners)
and `fs_main` (its mask's corners, once `Mask` carries `move_idx`).
- `core/src/ui/render_state.rs` — additionally, `resolved_region` (§2b)
and `window_region` (`:264-267`) reimplemented on top of it.
- `src/default/sense.rs` — `run_sensors`'s hit-test read (`:170`) switches
from `self.active.get(id).unwrap().region` to `self.resolved_region(*id)`
(§2b) — the pointer-routing fix this design requires, not an optional
follow-up.
- `core/src/render/data.rs` — additionally, `Mask` (`:46-49`) gains
`move_idx: u32` (§2b).
- `core/src/ui/painter.rs` — additionally, `set_mask` (`:49-52`) writes
`move_idx: self.move_slot` into the `Mask` it pushes (§2b).
- Every widget with a two-method `impl Widget`, collapsed to one `draw`
(§1, §6), `is_size_independent` added where true: `core/src/widget/mod.rs`
(`impl Widget for ()`), `iris/src/widget/rect.rs` (`Rect`, → true),
`iris/src/widget/image.rs` (`Image`, → true — a decoded image's primitive
never depends on the region it is offered, same as `Rect`),
`iris/src/widget/mask.rs` (`Masked`), `iris/src/widget/ptr.rs`
(`WidgetPtr`), `iris/src/widget/text/mod.rs` (`Text`, §4),
`iris/src/widget/text/edit.rs` (`TextEdit`),
`iris/src/widget/position/scroll.rs` (`Scroll`, keeps its `mov`-shaped
offset, now O(1) automatically via §2), `iris/src/widget/position/align.rs`
(`Aligned`, §6), `iris/src/widget/position/max_size.rs` (`MaxSize`),
`iris/src/widget/position/layer.rs` (`LayerOffset`),
`iris/src/widget/position/pad.rs` (`Pad`),
`iris/src/widget/position/stack.rs` (`Stack`),
`iris/src/widget/position/offset.rs` (`Offset`),
`iris/src/widget/position/span.rs` (`Span`, §4's `draw_twice` for the
cross-axis case, deleting `desired_ortho`'s duplicate loop),
`iris/src/widget/position/sized.rs` (`Sized`).
This list was produced by `grep -rn "impl Widget for\|fn desired_width\|fn desired_height"`
across `core/` and `src/`; re-run it before starting, since it is the
authoritative check that nothing was missed, not this paragraph.
- `iris/examples/{minimal.rs,task.rs,view.rs,tabs/main.rs}` — no direct
`impl Widget` found in any example (verified by the same grep); they use
the builder DSL in `core/src/widget/trait_fns.rs` and should need no
source change, which is itself part of the pass condition below.
### 8. Pass conditions
1. **Every example under `iris/examples` renders identically.** Run
`iris/run-headless.sh EXAMPLE --shot PNG` for each of `minimal`, `task`,
`view`, `tabs` before and after, and diff the PNGs pixel-for-pixel — not
"looks right," since a subtle wrap or alignment regression is exactly
what a diff catches and a glance does not.
**Result (2026-09-04): pass, all four, 0 differing bytes.** No PNG
library is installed in this VM (no PIL, no ImageMagick, no pip), so the
diff is a from-scratch PNG decoder (`zlib` + the five filter types) at
`/tmp/layout-shots/pngdiff.py`, comparing decoded pixel bytes rather than
file bytes (`cmp` alone is not conclusive across two separately-encoded
PNGs, though it happened to agree here for `minimal`). Before-shots were
taken with `git stash` at the pre-change commit; `tabs` needed two real
fixes (deviations 1 and 2 below) before it stopped differing — the other
three matched on the first try.
2. **Unchanged-frame cost, measured, not assumed.** Add a counter beside
the existing `debug_layers`/`active_widgets` instrumentation
(`render_state.rs:241-262`) for (a) `Widget::draw` invocations and (b)
`Primitives::write`/`region_mut` calls, both per `update()` call. Drive
one example (`tabs`, since it already has multiple widgets and an
interactive element) through one frame with nothing changed and report
both counts — the pass condition is **0 draws and 0 primitive rewrites**
for a frame in which nothing was marked dirty, resized, or moved.
**Result (2026-09-04): pass, 0 and 0.** Implemented as
`UiRenderState::take_counters() -> (u64, u64, u64)` (draws, `region_mut`
rewrites, `move_offsets` writes — a third counter, for condition 3
below), reset on read. Measured in
`iris/src/layout_tests.rs::an_unchanged_frame_draws_and_rewrites_nothing`
against a `Scroll` over 500 fixed-height rects (not the `tabs` example —
see the note on condition 3 for why this runs as a plain unit test
instead).
3. **Single-moved-child cost, measured.** Same counters, one frame in
which exactly one widget is moved (not resized) with N primitives in its
subtree — the pass condition is **1 write to `move_offsets`, 0 calls to
`Widget::draw`, 0 calls to `region_mut`**, independent of N. Construct
the case with a `tabs`-style example holding a deliberately large text
block (hundreds of glyphs) inside a `Scroll`, so N is large enough that
an O(N) regression would show up as a non-trivial write count rather
than being lost in noise.
**Result (2026-09-04): pass — 0 draws, 0 rewrites, 1 move_offsets
write, N = 500.** Built with rects rather than glyphs
(`iris/src/layout_tests.rs::scrolling_moves_in_o1_without_a_redraw`):
`iris-core`/`iris` touch no GPU or window to lay out and move a tree, so
this runs as a plain `cargo test`, not through `run-headless.sh` — a
`Widgets`/`UiData` pair and a bare `UiRsc` impl are enough, and it is
faster and more precise than reading counters out of a real example's
stderr. Getting a clean single move took two follow-up fixes beyond the
design as written (deviation 3, the `parent_move_slot` threading; and
the `Scroll` design decision below about offering last frame's content
length) — without either, the count was in the thousands (every rect in
the subtree redrawing) rather than 1.
4. **Hit-testing follows the move, not just the render.** In the same
scrolled-`tabs` construction as condition 3, scroll the content, then
send a synthetic cursor position over a widget that moved and assert
`run_sensors` (`src/default/sense.rs:154-200`) routes to that widget's
id, not to whatever is now at its pre-scroll coordinates or to nothing.
This is a correctness check, not a timing one — §2b's fix is required
before §2 can ship at all, and this is what would fail silently
(nothing on screen indicates a missed or misrouted hit) if it were
skipped.
**Result (2026-09-04): pass**, but checked one level below
`run_sensors`: `iris/src/layout_tests.rs::hit_testing_follows_a_scrolled_widget`
scrolls a widget and asserts `UiRenderState::resolved_region` (the
query `run_sensors`'s hit-test and `window_region` both now go through,
per §2b) reports the moved, not the pre-scroll, position — within
0.01px of the exact expected delta. `run_sensors` itself needs a
`HasEvents`/window/cursor-state harness this pass did not build; the
coverage that matters (does the position query the router uses reflect
the move) is exercised directly instead.
5. **A mask moves with its subtree.** Render a `Masked`-wrapped `Scroll`
both before and after scrolling it (`iris/run-headless.sh` against a
small purpose-built example, or an addition to `tabs`), and diff the
two frames: the clipped edge of the content must have moved with the
scroll while the viewport's own border (drawn by `Masked`, not moved)
stays put — the specific case worked through in §2b. A mask rectangle
that stayed at its pre-scroll position while its content slid past it
is the regression this checks for, and it is visible in a single
screenshot, not just in a counter.
**Result (2026-09-04): pass, checked numerically rather than by
screenshot.** No example in this repository builds a `Masked`-wrapped
`Scroll` (`tabs`'s "text edit scroll" tab uses `TextEdit`'s own internal
scrolling, not this widget), so there was nothing to screenshot without
first authoring a new example. Checked instead in
`iris/src/layout_tests.rs::a_mask_stays_put_while_its_scrolled_content_moves`,
on the exact data the fragment shader's `resolve_move` reads: the
masked widget's own `move_offsets` slot delta is `[0, 0]` both before
and after scrolling its content, because `Masked` is never itself the
target of a move — only its child is, on a separate, deeper slot in the
chain (§2b's "scroll-container case, checked rather than assumed"). A
pixel-level screenshot check of this remains open; see RUST.md's next
step.
6. **`cargo test --workspace`, `cargo clippy --all-targets`, `cargo fmt`**
stay clean at the defaults (iris has no tests today per I0b, so this is
presently only clippy/fmt; add the first real widget-layer tests here if
the move-offset chain or `draw_twice` are non-trivial enough to want
one, per "match the codebase's testing posture" — judge that once the
code exists rather than pre-committing to a number of tests here).
**Result (2026-09-04): pass.** `cargo fmt --all -- --check`,
`cargo build --workspace --all-targets`, and `cargo clippy --all-targets`
are all clean (one pre-existing, unrelated warning about `naga`/`wgpu`/
`winit` future-incompatibility, from dependencies, not this change).
`cargo test --workspace`: the 14 pre-existing `TextEdit` tests plus 4 new
ones in `iris/src/layout_tests.rs` (conditions 2–5 above), 18 passed, 0
failed — the move-offset chain turned out non-trivial enough (three real
bugs found only by writing it) to clearly clear the "match the testing
posture" bar this section left open.
### 9. Rejected, and why
- **A flat (non-chained) per-subtree offset table**, Iris's literal
phrasing — rejected in §2 for breaking under nested independent moves
(a swiped row inside a scrolling list). Costs nothing extra to avoid: the
chain is the same mechanism with one more field.
- **Keeping `region_mut` recursion as the only move mechanism** — rejected
as the steady-state path (O(primitives in subtree), exactly what a
transcript scroll must not pay every frame) but kept for resize-shaped
changes (§3) where the content's own region field, not an ancestor
chain, is what has to change.
- **A second, size-only trait method kept alongside `draw`** (e.g.
`fn size_hint(&self) -> Option<Size>` as a fast path some widgets could
implement to skip a draw when a cheap answer exists) — considered and
rejected: it reintroduces exactly the "two names for one concept" split
this change removes, for a saving `is_size_independent` (§1, §3b)
already covers for the cases where it would actually help (fixed-size
leaves). A widget whose size is cheap to compute but whose *drawing* is
not (unlikely in this codebase's widget set, but conceivable) is better
served by that widget caching its own draw output internally — exactly
the pattern `TextView::render` already uses (§4) — than by a second
trait method every implementor has to reason about.
- **Passing `available` as an explicit parameter to `draw`** (mirroring
Masonry's `layout(&mut self, ctx, bc: &BoxConstraints) -> Size`, the
yardstick per AGENTS.md) — rejected as redundant with `Painter::region()`,
which already carries the same information into every widget that needs
it; adding a parameter would just be a second route to a value already
reachable, and would invite the two drifting apart.
- **Eagerly propagating a moved widget's delta into every descendant's own
offset value** (rather than chaining and resolving in the shader) —
rejected as O(descendant widgets), which is smaller than O(primitives)
but still not O(1), and the shader-side chain costs nothing extra to get
the better bound.
## Deviations found during implementation (2026-09-04)
Five corrections this file's first draft did not anticipate, each found by
`iris/run-headless.sh tabs --shot` disagreeing with a pixel-identical
pre-change screenshot (pass condition 1) and traced with `eprintln!` in
`draw_inner`/`reposition` — not by reasoning about the design in the
abstract. Recorded here rather than silently fixed in place, per the code
rules' escape-hatch requirement.
1. **`Aligned`'s provisional draw must call `painter.widget`, not
`widget_within(&self.inner, painter.region())`.** §6's original text drew
the sample as the latter. `widget_within` composes its `region` argument
as *local*, `UiRegion::FULL`-relative coordinates against
`painter.region()` (exactly what `UiRegion::FULL.within(&self.region) ==
self.region` relies on); handing it `painter.region()` itself —
already-resolved, window-relative coordinates — composes that frame a
second time. For the root widget this is silently the identity (its
region already is `[0,1]`), which is why it can look correct in a
trivial case and only breaks once something is nested — i.e. always, in
practice. Symptom: a centered child rendered at a wildly wrong offset
nested more than one level deep. Fixed by using `painter.widget`, which
hands the child `self.region` unmodified, with no second composition.
2. **A widget that reports a size smaller than its offered region must
actually paint at that size, anchored top-left of what it was given —
not fill the full offered region while merely *reporting* a smaller
number.** `Sized` and `MaxSize` both had exactly this bug: their
`desired_width`/`desired_height` predecessors capped the *reported*
value but their `draw` bodies called `painter.widget(&self.inner)`
unconstrained, which was harmless under the old two-pass model (a parent
always queried the size *before* drawing, so by the time `draw` ran the
offered region already matched) but wrong under `Aligned`'s new
provisional-draw-then-reposition pattern, which offers the *whole*
region on the first, learning pass. Symptom: a `.sized((100, 100))` rect
rendered stretched to fill its whole row instead of a 100×100 square.
Fixed by having both widgets carve the declared sub-region (`UiSpan`
sized to the axis's `Len`, anchored at `AxisAlign::Neg`) out of whatever
they were offered before drawing the child in it. `Image` needed the
same treatment from the start (`texture_within` at its own natural size,
not `texture()` at the full offered region) and was written that way in
the first pass, once this was understood; `Rect`'s "fill whatever I'm
given" is the one case where painting the *whole* offered region really
is the declared behavior, so it needed no change.
3. **The move-offset chain's `parent` link cannot be found by looking up
the parent's `ActiveData` in `draw_inner`, because the parent's
`ActiveData` does not exist yet while its own `Widget::draw` is still
running.** `ActiveData` is inserted only after `draw` returns
(`render_state.rs`, end of `draw_inner`), so a child drawn partway
through its parent's `draw` body — the ordinary case, since every
composite widget draws its children from inside its own `draw` — would
always read "no parent" from `self.active`, silently orphaning it at the
root of the chain. Fixed by threading the parent's `move_slot` down
through `Painter` (it already carries `mask`/`layer` the same way) and
passing it explicitly into `draw_inner` as `parent_move_slot`, rather
than deriving it from `self.active.get(parent_id)`. `move_parent_of`
(the `self.active`-based lookup) is kept, but only for `redraw()`, whose
target's parent genuinely is already active at that call site — the
doc comment on it says which is which. Symptom: `reposition` computed
the right delta and wrote it to the right slot, but the shader never
saw it, because the primitive doing the actual painting chained to
`u32::MAX` one level too early.
4. **`Painter::reposition` cannot reuse `active.region` as "where the
widget currently is," because for a widget offered more room than it
used, `active.region` is the *offered* box, not the *painted* one.**
This only matters for `reposition` (used by `Aligned`); `mov` (used by
`draw_inner`'s own same-size-different-position dispatch, for `Scroll`
and `Offset`) has no such gap, because there the offered region *is*
the visual footprint — content is sized to fill exactly what it is
given. `reposition` instead reconstructs "from" as `active.size`
(already tracked, per §5) anchored at `AxisAlign::Neg` within
`active.region` — i.e. it assumes the child painted itself top-left of
whatever it was offered, per point 2's convention — and **overwrites**
the slot's delta rather than accumulating it the way `mov` does, since
"from" is recomputed fresh from stable inputs every call and repeating
the same `reposition` (an unrelated redraw elsewhere re-running this
widget's parent) must not drift further each time. The one shape this
does not cover: `Aligned` wrapping `Aligned`, where the inner one's own
`reposition` may have moved its content away from top-left already. No
widget or example in this codebase builds that today; if one needs to,
`reposition` would need the child to report *where* it painted, not
just how big, which is a larger change than this pass's scope.
5. **A widget's `move_offsets` slot is allocated once, on its first-ever
draw, and reused in place — never reallocated — for every later redraw
of the same id, with its delta reset to `[0, 0]` on each reuse.** Not
spelled out in §2's original text, which only said slots are assigned
"when the widget is first drawn." Reallocating a fresh slot on every
redraw would leave any *retained* (not-redrawn) descendant's `parent`
link pointing at a now-orphaned old slot — a permanent leak, and worse,
a descendant that silently stops tracking its ancestor's future moves.
Resetting the delta on reuse (rather than carrying it forward) is
required because a full redraw bakes the widget's correct absolute
position into the fresh `region` argument directly; a stale delta left
over from before the redraw would double-offset it.
Two further points worth recording because they were *design decisions*
made while implementing, not bugs — `LAYOUT.md`'s own text left them
unspecified rather than getting them wrong:
- **`Scroll` offers its content a region sized by the *previous* frame's
measured content length, not a fresh one.** A fresh measurement would
require drawing the content once to learn its size and — since that
provisional size essentially never matches the previously active one —
redrawing it a second time at the real size, on every single scroll
tick, which is exactly the cost §2 exists to remove. Using the stale
length means an ordinary scroll (position changes, content does not)
offers the same *size* as last frame, only shifted, which is what makes
`draw_inner` dispatch it as the O(1) move. The cost: a real content-size
change lags one frame before the container's scroll range reflects it,
self-correcting the frame after (the content length itself, read from
what was actually drawn, is never stale — only the offered *region* used
for placement is). No example in this repository builds a `Scroll` yet,
so this could not be checked against a pixel diff; it is covered instead
by `iris/src/layout_tests.rs`'s three `Scroll`-based unit tests, which
build a tree and drive `UiRenderState` directly with no GPU or window
needed.
- **`redraw()`'s parent-relayout check draws the widget first, then
compares the fresh `ActiveData.size` the draw produced against the size
from before removal** — the mirror image of the old code's "query size,
compare, decide whether to draw," which no longer has a size query to
do the comparison with before drawing (§5 deleted `Cache`/`SizeCtx`
along with `desired_width`/`desired_height`). This can occasionally draw
a widget once more than the old code would have (if the parent it
bubbles up to ends up redrawing the same widget again as part of its own
relayout) — `draw_inner`'s own skip/move dispatch absorbs most of that
redundancy for free, and this path is not one of §8's measured
conditions, so the remaining slack was accepted rather than chased
further.
## For IRIS.md
When this lands, copy this entry into `IRIS.md` (newest first):
> **2026-09-04 — `Widget::draw` reports the size it used; `desired_width`/
> `desired_height` are gone.** A widget used to implement three methods
> (`draw`, `desired_width`, `desired_height`); it now implements one,
> `fn draw(&mut self, painter: &mut Painter) -> Size`, which draws into
> `painter.region()` and returns how much of it was used. Why: the two
> extra methods routinely re-simulated what `draw` was about to do anyway
> (`Span::desired_ortho` copied its own draw loop to get cross-axis sizing
> right) — one visit per widget per frame instead of up to three. A
> container that needs a child's size before placing it (alignment,
> centering) draws the child once at a provisional region, reads the
> returned `Size`, and calls the new `Painter::reposition` to move it into
> its final spot — an O(1) offset write, not a second draw. A widget whose
> drawn output never depends on the size it's given (a fixed-size `Rect`,
> a decoded `Image`) overrides the new `fn is_size_independent(&self) ->
> bool { false }` to `true`, which skips redrawing it when only its
> offered region changes shape.
>
> ```rust
> // before
> fn draw(&mut self, painter: &mut Painter) { /* ... */ }
> fn desired_width(&mut self, ctx: &mut SizeCtx) -> Len { /* ... */ }
> fn desired_height(&mut self, ctx: &mut SizeCtx) -> Len { /* ... */ }
>
> // after
> fn draw(&mut self, painter: &mut Painter) -> Size { /* ... */ }
> ```
>
> `SizeCtx` and `Cache` are gone with it — see `LAYOUT.md` for the full
> design, the move-offset mechanism this shipped alongside, and the file
> list.
PLAN.md
+1180 -85
View File
File diff suppressed because it is too large. Load diff
RUST.md
-1137
View File
File diff suppressed because it is too large. Load diff
SUBAGENTS.md
+289
View File
@@ -0,0 +1,289 @@
# Subagents
A session's subagents -- helpers started by Claude Code's Task tool or Codex's
collaboration tools -- each get a transcript of their own, listed in a panel
over the open session and readable in the same transcript view the session has.
Designed 2026-09-05; extended to Codex's multiplexed app-server threads on
2026-09-13. The decisions Bryan has not yet reviewed are in `DECISIONS.md`.
## What a subagent is here
**A subagent is a second transcript owned by a session, in the same event
model, with no process and no controls.** It is not a session: it cannot be
messaged, stopped or started, and it has no machine, model or usage of its
own. Everything it shares with a session -- the transcript file format, the
paging routes, the SSE stream, the phone's cache and rendering -- is reused
by addressing, not by copying.
Claude reports a subagent's messages on the parent's own stream-json output,
each carrying `parent_tool_use_id` = the id of the Task `tool_use` that started
it. Before this the translator dropped those lines
(`subagent_events_are_not_duplicated_into_the_transcript`); now it routes
them to that subagent's own translator and transcript. The parent's
transcript still shows only the Task call itself.
Codex app-server multiplexes every thread in the session tree onto the root
process's stdout. Its notifications carry `threadId`; `subAgentActivity`
items name the child thread and its lifecycle, and `collabAgentToolCall`
items carry the spawn prompt. The Codex translator routes a non-root
`threadId` exactly as Claude routes a `parent_tool_use_id`. The child thread
id is the subagent id on disk. An asynchronously delivered `agentMessage` is
a `PeerMessage`, not assistant text from the recipient. Its delta notification
does not repeat the completed item's `delivery` field, so the translator
remembers that field from `item/started` and suppresses those deltas. Letting
one into the recipient's provisional assistant row makes its next completed
message replace the combined row, visibly erasing text that Codex still has.
The parent draws the initial `spawnAgent` as its ordinary `Task` card and
closes it when the matching `subAgentActivity.started` arrives. The remaining
collaboration calls remain visible as coordination -- waiting, messaging,
listing and lifecycle controls -- rather than being mistaken for generic task
output. Null optional fields and a bare `completed` status carry no information
and are omitted; their useful result is the child transcript, status or peer
message beside them.
## Storage
Under the session directory:
```
<session>/subagents/<subagent_id>/meta.json {title, created}
<session>/subagents/<subagent_id>/transcript.jsonl same SeqEvent lines as the session's
```
The id is Claude's Task tool_use id (`toolu_…`) or Codex's child thread id.
Both are unique, stable across a backend restart, and already the key their
parent-side lifecycle uses.
Only ids matching `[A-Za-z0-9_-]+` are ever created or looked up, since the
id becomes a path.
The transcript's sequence numbers are its own, starting at 1. `Transcript`,
`read_window`, `catch_up` and `read_after` work on it unchanged.
Its path out: deleting the session deletes its directory, subagents included,
and `POST /sessions/{id}/subagents/delete` removes finished ones on their own
-- all or nothing, and refused while any named one is still running, since its
transcript is still being written to and its process is the session's to stop.
## Lifecycle, as events in the subagent's transcript
1. Created when the parent Task/Agent call is seen. A current Claude CLI's
`task_started` with `task_type: local_agent` is a recovery source when an
adopted stream begins after that call. A bare `parent_tool_use_id` is not
enough: other operations can also parent nested lines, and treating one as
proof created false subagents named after their first subcommand.
First lines written:
`Status Running`, then `UserMessage { text: <the Task's prompt> }` when
the prompt is known -- it genuinely is the subagent's first user turn.
2. Every child line is translated by that subagent's own `Translator`
(one per subagent: tool ids are unique but streaming deltas are by
content-block index, and parallel subagents interleave).
3. **What ends a subagent is the CLI's own task lifecycle**, on top-level
`system` lines that carry no `parent_tool_use_id`: `task_started`
(`task_id`, `tool_use_id`, `task_type`, `is_backgrounded`, the prompt),
`task_progress` repeatedly, then `task_updated` (`patch.status`, naming the
*task* only) and `task_notification` (`tool_use_id`, `status`, and `summary`
-- the agent's own report). `translate_task` keeps the
`task_id -> tool_use_id` mapping from the first so the update can be
attributed, records the summary as the subagent's closing text, and writes
`Status Exited`. A
`completed` update is deliberately not the end: its notification carries
the summary and would otherwise land after the ending. Any other terminal
status ends it from the update, since the failure to avoid is a subagent
nothing ever finishes.
Since Claude Code 2.1.261, `background_tasks_changed { tasks: [...] }` is
the authoritative level beside those edges: its set replaces the previous
set, so a missed terminal edge cannot leave a subagent running forever. Its
ids are deliberately not correlated with the edge stream; what is read off
each entry is its own description and kind, and what is read off the set is
whether it is empty and how large. The session API and stream expose that
size as `backgroundTasks`, which the phone draws beside the status, and
`GET /sessions/{id}/background` serves the entries themselves -- listed in
the session's panel *above* the subagents and never as subagent cards. An
`ambient` entry is excluded from both, on the CLI's own instruction: a
live-update watcher is not activity. A backgrounded subagent is legitimately
in both lists, since it is both running and a transcript. The edges still
carry mapping, outcome and closing summary. On adoption the driver sends a repeated `initialize`,
which makes a current CLI send the full set; an older CLI accepts it and sends no level,
leaving the edge-based path unchanged. A snapshot is reconciled immediately
when the persisted parent status proves it is between turns, and otherwise
at the next `result` boundary -- while a turn is open, a foreground agent is
legitimately absent from the background set. Reconciliation writes
`Status Exited`, which is also what makes a formerly stale row deletable;
a task notification ordered after the level can still add its summary.
The two rules this replaces were both wrong, in opposite directions. The
parent's `tool_result` is not it: a backgrounded Task's arrives at launch
("Async agent launched..."), so ending there truncated a running agent's
transcript at the moment it started. Nor is the subagent's own
`end_turn`: measured against 2.1.237 on 2026-09-06, **a subagent's lines
carry no `stream_event` at all** -- they are whole `user`/`assistant`
lines with a null `stop_reason`, no `result` line is sent for one, and the
sub's final report never appears as a child line -- so that rule could
never fire and every subagent stayed `running` for ever. `ends_a_turn` is
kept as a second detector for a dialect that does say either, and must
never be the only one again.
`Status Exited` either way; the subagent's vocabulary has no `Idle` or
`Waiting`, so the end-of-turn status `dispatch` produces for an ordinary
session is dropped rather than written.
**The ending reaches the parent's transcript as nothing at all**
(2026-09-06). It was tried, and a row per finished subagent is a screenful
of dividers about work the reader was not asking after; the closing report
is *this* transcript's last line and here is where somebody reads it. What
the parent gets a row for is a message a subagent genuinely sends it, which
arrives by the peer path. A backgrounded *command* is the other half of
this and goes the other way: it has no transcript of its own, so its report
updates the tool card that launched it, which was still saying the command
was running. The two lifecycle shapes are still handled once:
whichever gets there first is the one that finds the task still open, and
`finish` below closes it. See PLAN.md's "Two turns must never be drawn as
one".
**While any task is outstanding the session's turn ends in
`Status Waiting` rather than `Idle`.** `Idle` means "waiting for a person",
and a session with a backgrounded subagent is not doing that. The edge
fallback has two sources: the translator's `open_tasks`, and
`Subagents::any_open` -- which covers a subagent launched before a backend
restart adopted the session, whose `task_started` is behind the durable
stdout offset. On current Claude versions the replace-semantics level above
reconciles both at a safe turn boundary.
**A limit the account hits inside a subagent is hoisted to the session**
as well as recorded here, because `resume.rs` can only schedule against a
session, and a background subagent outliving its parent's turn is the
ordinary case -- see PLAN.md's "A limit a subagent hits is the session's".
4. **A child line for a subagent that already finished reopens it**
(`Status Running`) rather than being dropped: a background Task can be
sent another message long after its first turn ended, and that is
exactly what a further line for it means. Same transcript, same child
`Translator`, just picking back up.
5. When the parent session's process exits (`Status Exited` on the
session), every subagent still `Running` gets `Status Exited` too: its
process was the parent's. Read from the directory rather than from the
live map, because one left `Running` by a previous run of the server is
precisely the one nothing in this process has touched -- and it would
otherwise read `running` again every time its session was started.
For Codex the same lifecycle is expressed by app-server rather than Claude's
task notices: `subAgentActivity.started` creates the child,
`subAgentActivity.interacted` reopens it, and `completed` or `interrupted`
finishes it. A child's own `turn/completed` is not its end; it remains running
until that activity edge. The root's `turn/completed` reports `waiting` while
the registry contains an open child, and the last activity completion reports
`idle` if the root is between turns. Because the child thread id is also the
on-disk id, an adopted driver can route and finish a child whose spawn record
is already behind the durable stdout offset. The registry's open count is also
Codex's `backgroundTasks` measurement: lifecycle changes send it through the
same event and session-summary fields as Claude's provider snapshot. The other
part of that measurement is app-server's runtime
`thread/backgroundTerminals/list` set. Its process ids are held only in memory
and added to the open-child count; the driver refreshes the set at terminal
boundaries and while it remains nonempty, rather than decrementing for an
unmatched ending edge.
A subagent that was mid-flight when the backend restarted keeps working:
the registry reopens the existing transcript on the next child line, and
the file continues its sequence -- the same reopening #4 describes, whether
what closed it was a restart or its own `end_turn`. If its turn ended while
the backend was down nothing recorded that until the next line arrives, so
its last status stays `Running`, which the list reports as **unknown**
rather than as running (see the wire shape) until then.
Title: for Claude, the Task call's `description` input, then
` (<subagent_type>)` when one is given; falling back to `Task` when the
description is absent. An adopted current CLI can recover the same fields from
its `local_agent` lifecycle record. For Codex, the first lifecycle record uses
the spawned thread's name or the last segment of `agentPath`, with underscores
shown as spaces, then falls back to `subagent`.
## Server layout
- `session/subagent.rs` -- the registry: `Subagents` (per session, in
`Shared`), `Subagent` (its `Transcript` behind a mutex plus a
`broadcast::Sender<SeqEvent>`), `record(id, event)`, `start(id, title,
prompt)`, `finish(id)`, `reopen(id)`, `finish_all()`, `list()` from disk, and
`delete(ids)` -- its path out. Drivers get an
`Arc<Subagents>` beside their `EventSink`; llama ignores it.
- `session/claude/translate.rs` -- routes child lines by parent id, holds
one child `Translator` per subagent, remembers pending Task calls'
description/prompt/subagent_type.
- `session/codex/translate.rs` -- routes multiplexed app-server notifications
by thread id, remembers collaboration prompts, and translates activity
edges into the same registry lifecycle.
- `session/echo.rs` -- `/subagent [n]`: the test rig. Starts *n* (default 1)
subagents at once, each named "helper k". Each writes the prompt as its
user message, streams a few words of text, runs one `Bash` tool call, then
finishes about three seconds after starting, and the parent's Task calls
end when their subagent does. Three seconds so the running state can be
seen on the phone.
- `routes.rs` -- four routes, in the doc table.
## Wire shape
```
GET /sessions/{id} SessionInfo gains `subagents: N` (count, 0 when none)
GET /sessions same field on each row
GET /sessions/{id}/subagents [{id, title, status, created, lastActivity}], oldest first
GET /sessions/{id}/subagents/{sub}/transcript exactly the session transcript's query and answer
GET /sessions/{id}/subagents/{sub}/events?after=N exactly the session events stream
POST /sessions/{id}/subagents/delete {subagents} -> 204; refused whole if one is running
```
The delete is a batch rather than a `DELETE` per id for the reason the import
list's is: the phone deletes what a reader selected, and one request per row
means a batch can half-arrive, leaving the rows that were missed looking
exactly like rows nobody picked. Unlike an import delete it is local file
removal, so it is done by the time the reply is sent and there is no per-row
state to follow afterwards. What decides "running" is
`Subagents::list`'s own rule, shared through `routes::has_a_process` so the
list and the delete cannot disagree about it.
`status` is the transcript's last `Status` event, serialised like a session's
(`running`, `exited`), except that a subagent whose session is not itself
running cannot be running: the list answers `unknown` for that one. A
subagent never reports `waiting`: that is a session's word for having
outstanding work of its own, and a subagent has none. The
phone words these as *running*, *finished* and *unknown* on the subcard.
The count on `SessionInfo` is a directory listing, so the list stays cheap.
The per-subagent status is only read when the list route is asked for.
## Phone
- The subcards are ordered **still running first, then most recently
active** -- a display decision made on the phone (`subagentOrder`), over the
server's stable oldest-first answer. Two keys rather than activity alone
because a subagent that is thinking reports nothing meanwhile and would sink
below one that just finished.
- **Holding a subcard selects it, and several at a time**, exactly as the
import list works, with the selection bar drawn inside the panel rather than
at the bottom of the session: this selection belongs to the subagent list,
and a bar under the composer would read as acting on the conversation.
Delete is
*disabled*, with the reason in words, while anything selected is still
running. Deleting confirms first, dims the rows it is acting on
(`BusyItem`), and on success takes them out of the panel without refetching
anything else. The phone's cached copy of
a deleted subagent's transcript is purged with it.
- The main session list does not expand or count subagents. Swiping left over
an open session pulls an 88%-wide panel in from the right and fetches
`/sessions/{id}/subagents`; it draws one `OutlinedCard` per subagent: title,
then the status word and a relative time. The transcript remains composed
under the panel, so its event stream, draft and scroll position stay live.
Horizontal scrollers inside the transcript win the gesture. Collapsing one,
or starting over any ordinary part of the session, gives the gesture back to
the panel; Android keeps its own edge Back gesture. Swiping right on the
panel, tapping outside it, or Back closes it.
- Tapping a subcard opens a `SessionScreen` layer in **read-only** form: the
same transcript, paging, cache, selection,
images and status row, with the composer, the process button, the model
picker, the files button, the settings cog and the usage bar left out.
The header shows the subagent's title with the session's title beneath it.
It is another layer over the still-composed session and its panel; Back
returns to the panel.
- Addressing: `fetchTranscript`, `EventStream`, `TranscriptSource` and the
cache take a transcript address rather than a session id --
`sessions/{id}` or `sessions/{id}/subagents/{sub}` -- so the cache nests a
subagent's copy under its session's and the same code serves both.
TEXTURES.md
-496
View File
@@ -1,496 +0,0 @@
# How iris should render an unbounded number of images
## Status (2026-09-04)
**Implemented**, on the `rustify` branch of `ai-app-2`, in `iris/core` and
`iris/src/default/render.rs`. See "Implemented, 2026-09-04" at the bottom for
what landed, what differs from the proposal below and why, and what was
verified versus merely reasoned about. The short version: the binding array
is gone, `request_device` asks for no features and no binding-array limits,
and that is now proven on the emulator's software Vulkan
(`rigs/gpu-probe`), not just read from the code. `RUST.md`'s blocking item
is resolved.
Iris (the person) asked whether iris's (the library's) approach to
"draw however many images happen to be on screen" — relevant here because a
transcript can hold an unbounded number of attached screenshots — actually
works on mobile, her recollection being that it does not. Checked rather
than assumed, on 2026-09-04, on the `rustify` branch of `ai-app-2`. This
file is that investigation and the resulting recommendation, written for a
second agent to review before anything in iris's render core changes — no
code has been written against this yet.
## The problem
Every texture iris ever creates — every `Image` widget
(`iris/src/widget/image.rs`) and every glyph atlas page — gets a permanent
slot in one array via `Textures::add` (`iris/core/src/primitive/texture.rs:65`).
Both of iris's texture-sampling primitives (`TEXTURE` and `GLYPH`) read that
array by index: `core/src/render/shader.wgsl:56` declares
`var views: binding_array<texture_2d<f32>>`, sized by
`UiLimits::default()` (`core/src/render/mod.rs:347`) at **100,000 textures,
1,000 samplers**. Getting a device to accept that layout needs three wgpu
features — `TEXTURE_BINDING_ARRAY`,
`SAMPLED_TEXTURE_AND_STORAGE_BUFFER_ARRAY_NON_UNIFORM_INDEXING`,
`PARTIALLY_BOUND_BINDING_ARRAY` — which correspond to Vulkan's
`VK_EXT_descriptor_indexing` ("bindless"), promoted to Vulkan core at 1.2.
A transcript with an unbounded number of image attachments is exactly the
case that grows this array without bound: each attachment becomes its own
`Image` widget, which takes its own permanent array slot until dropped.
## What was measured
**A new rig, `rigs/gpu-probe`**, asks a device for exactly iris's features
and limits with no window and no APK — a plain executable pushed with
`adb push` and run from `/data/local/tmp`. It has two parts:
`wgpu::Adapter::request_device` with iris's exact `Features`/`Limits`
(`src/main.rs`), and a raw Vulkan query bypassing wgpu entirely via `ash`
(`src/vk.rs`), to tell "the driver doesn't have it" apart from "wgpu didn't
detect it."
- **On this VM's own GPU** (Vulkan via Venus onto an RX 7900 XT):
`IRIS DEVICE: ok`. Not the case that matters — nobody's phone is a
discrete desktop GPU — but it is why the design was never checked before
now: it always worked in the one place it was tried.
- **On the Android emulator's guest Vulkan**, both ICDs it ships
(`vk_swiftshader_icd.json` and, cold-booted, `lvp_icd.json`/lavapipe):
`request_device` **fails** —
`Unsupported features were requested: TEXTURE_BINDING_ARRAY |
SAMPLED_TEXTURE_AND_STORAGE_BUFFER_ARRAY_NON_UNIFORM_INDEXING |
PARTIALLY_BOUND_BINDING_ARRAY`. The raw `ash` query on lavapipe shows the
driver itself reporting all seven descriptor-indexing sub-features as
`true` at device API version 1.3 — so wgpu-hal's own feature detection is
being more conservative than the driver here, for a reason not chased
further (a likely instance-version negotiation gap, since the extension
only promoted to core at 1.2). That part is a wgpu-hal/emulator question,
not the finding that matters, and is **not** why this design is rejected.
**The finding that matters is about real phones, sourced rather than
recalled:**
- The **Android Vulkan Profile 2025** — Google and Khronos's current
baseline, covering **80.1% of active Vulkan-capable Android devices** as
of October 2025
([developer.android.com/ndk/guides/graphics/android-vulkan-profile](https://developer.android.com/ndk/guides/graphics/android-vulkan-profile)) —
does **not** require `VK_EXT_descriptor_indexing` or any descriptor-
indexing feature. It requires `shaderSampledImageArrayDynamicIndexing`
(indexing by a value uniform across the invocation — Vulkan 1.0 baseline,
unrelated to bindless) and stops there; true of the 2021 and 2022
profiles as well.
- Arm's own developer documentation states **"`VK_EXT_descriptor_indexing`
is supported on all Valhall and 5th Gen GPUs"**
([developer.arm.com/mobile-graphics-and-gaming/vulkan-api-best-practices-on-arm-gpus](https://developer.arm.com/mobile-graphics-and-gaming/vulkan-api-best-practices-on-arm-gpus)) —
Mali generations from roughly 2019 (Mali-G77) onward, with no claim made
for Bifrost, Midgard or Utgard, which are still common in budget and
older Android phones that are still in daily use.
- A search engine's summarized claim of "1% support on Android" for this
extension was checked against its cited source (an Arm blog post from
2021) and **was not actually there** — that number does not appear in
any primary source found and should not be repeated. The baseline-
profile finding above is the one with an attributable source; use it
instead.
So this is not a software-renderer artifact. A real, currently-shipping
share of the Android fleet lacks the feature iris's texture pipeline asks
for unconditionally, and neither the emulator's failure nor the current
official hardware baseline gives any reason to expect that to change soon.
## What growth already costs today, before any redesign
Checked directly in `core/src/render/mod.rs` and `core/src/render/texture.rs`,
because "does this redesign make things worse" needs the current baseline
first:
- The `RenderPipeline` (`UiRenderNode::new`) is created **once** and never
rebuilt for any reason related to texture count — its bind group
*layouts* declare fixed slot counts (`limits.max_textures`,
`limits.max_samplers`) up front and that never changes at runtime. Growth
was never at risk of recreating the pipeline, in the current design or
any redesign discussed below.
- What **does** get rebuilt: `UiRenderNode::update` calls
`self.textures.update(&mut ui.textures)`, and if that reports any change,
rebuilds `self.rsc_group` — one `BindGroup` whose entries are
`BindingResource::TextureViewArray(&tex_manager.views())`, collected
fresh over **every currently-live texture**, plus the sampler array and
the mask buffer. This happens on every texture `Push`, `Set`, or `Free`
— an image added anywhere in the whole app rebuilds one shared structure
referencing every other image too.
- The one path already excluded from this, on purpose, is a `Patch` —
writing into an existing texture's pixels without changing which
textures exist. The code says why directly
(`core/src/render/texture.rs`, in `GpuTextures::update`): *"A patch
changes texture contents, not the binding array, so it must not report
`changed` — rebuilding the bind group per glyph is the cost this exists
to avoid."* This is exactly the mechanism I1 built for the glyph atlas:
growing an existing atlas page costs a `write_texture` into a sub-rect,
nothing else.
So today, growth that stays inside an existing texture (glyphs added to an
atlas page) is already free. Growth that adds a *new* texture — a new atlas
page, or any standalone image — already rebuilds the one shared array
regardless of how the array is populated, before any change discussed
below. That existing cost is O(live texture count) in CPU work to collect
the view list and in however expensive the driver finds a
descriptor-set-sized-for-N-descriptors to be.
## Prior art, checked rather than assumed
Two independent projects were checked to see whether "atlas for images"
is actually how this is normally done, rather than a guess:
- **egui_wgpu** (`crates/egui-wgpu/src/renderer.rs` in emilk/egui), the
closest prior art to iris — an immediate-mode wgpu-backed UI library that
ships on Android. It keeps a `HashMap<TextureId, Texture>` and gives
**each texture its own ordinary `BindGroup`** — one texture, one sampler,
no array, no descriptor indexing of any kind. Draw calls are batched by
texture id and the bind group is switched between batches within the
render pass.
- **Vello** — the renderer Masonry (E1/E2's Linebender stack) draws
through — hit the identical problem and wrote down why in their own
roadmap document
([github.com/linebender/vello/blob/main/doc/roadmap_2023.md](https://github.com/linebender/vello/blob/main/doc/roadmap_2023.md)):
*"The number of images that may appear in a scene is not bounded, which
is not a good fit for the basic descriptor binding model... Until then,
we'll do a workaround of having a single atlas image containing all the
images in the scene."* Their reason is broader than Android — WebGPU 1.0
has no descriptor indexing at all — but it reaches the same conclusion
for the same shape of problem: atlas, not a bigger bindless array.
**This is also a live hazard, not a solved one.** Vello's own changelog
(Sparse Strips v0.2.0) lists a fix titled *"WebGL image-atlas allocation
and growth on Mali-G52 GPUs, avoiding application-not-responding errors"*
— an actual ANR, from atlas growth, on an actual mid-range Android GPU,
in the renderer Masonry is built on. The same release added
`AtlasSpaceDiagnostics`/`AtlasLayerDiagnostics` (per-layer free-space,
utilization, fragmentation) because growth needed instrumenting in
production, not because it turned out to be free.
## Recommendation (not yet implemented)
1. **Small, plentiful textures** — glyphs (already done, I1), thumbnails,
downscaled attachment previews, icons — go through a shared atlas, the
same technique as `core/src/render/atlas.rs` generalized beyond glyphs.
Adding one to an existing page is a `Patch`, already free per the
section above.
2. **Large or one-off images** — a photo attachment opened at full
resolution, anything that would fragment a shared page — get their
**own ordinary, non-array bind group**, the egui_wgpu way. Creating one
is O(1): it references only itself, and does not touch any other
texture's binding, unlike today's shared array where every push
rebuilds a structure listing everything.
3. **Opening a new atlas page** is the one case that still resembles
today's rebuild — infrequent (bounded by how many *pages* are needed,
not by how many images have ever been attached) but not free, and
Vello's Mali-G52 fix says this specifically deserves care: it should
never be allowed to block a frame, and it is worth having the
equivalent of Vello's atlas diagnostics before trusting it under load.
4. **Net effect**: dropping `TEXTURE_BINDING_ARRAY`,
`SAMPLED_TEXTURE_AND_STORAGE_BUFFER_ARRAY_NON_UNIFORM_INDEXING`, and
`PARTIALLY_BOUND_BINDING_ARRAY` from iris's device request entirely.
Every path above is plain Vulkan 1.0 / GLES-level texture sampling.
This is also what fixes the emulator failure measured above, regardless
of the unresolved wgpu-hal question: a device that never asks for the
feature cannot be refused for lacking it.
## What this touches, and what is still open
Implementing this reworks iris's rendering core: the shader's binding
group layout (`shader.wgsl`), `Textures` and `GpuTextures`
(`core/src/primitive/texture.rs`, `core/src/render/texture.rs`), both
texture-sampling primitives, and `core/src/ui/painter.rs`'s draw-call
batching (today one draw call can reference any texture by index; the
per-texture-bind-group path needs draws grouped by which bind group they
use). Nothing has been started.
Open questions a reviewer should weigh in on:
- **The size threshold** between "goes in an atlas page" and "gets its own
bind group." Too low and ordinary attachment thumbnails end up as
one-off bind groups, losing the batching benefit the atlas exists for;
too high and a page fragments on a handful of medium images.
- **Eviction policy** for atlas pages once the working set does not fit —
today's `GlyphAtlas` never evicts, because a font's glyph set is small
and bounded; images are not. An LRU at the page level, or at the
individual-image level within a page, has not been designed.
- **Whether iris should keep any binding array at all**, even a small
fixed one (say, capped at a few dozen slots) for atlas pages themselves,
or whether every atlas page should also be its own ordinary bind group
like standalone images — the array's only remaining justification would
be avoiding a bind-group-per-draw-call switch cost that has not been
measured on this project's actual target hardware.
- **How this interacts with I2/E2's virtualised list** (I3): a
bottom-anchored transcript composes only visible rows, so the live
texture set should already be bounded by what is on screen rather than
by the whole conversation — worth confirming that invariant holds before
relying on it to keep atlas/bind-group churn small.
## Review, 2026-09-04
A second pass over the file above against the code, done before anything
is implemented. Iris's worry going in: a bind group per texture means a
draw call per image, and she wants this as efficient as it can be.
### What checked out
Every code reference above is accurate as of this commit: the 100,000 /
1,000 limits, the one-time pipeline, the `rsc_group` rebuild on every
`Push`/`Set`/`Free`, and the `Patch` exclusion. The device request that
asks for the three features is `iris/src/default/render.rs:96`, which the
text above does not name. egui-wgpu and Vello are described correctly.
### The emulator refusal is a wgpu-hal gap, now located
The file guessed "a likely instance-version negotiation gap." It is
narrower than that and it is in wgpu-hal, not the emulator. wgpu-hal
28.0.0 (`src/vulkan/adapter.rs:1618`) only queries
`PhysicalDeviceDescriptorIndexingFeaturesEXT` **when the device advertises
the `VK_EXT_descriptor_indexing` extension string**. A Vulkan 1.2+ driver
that has descriptor indexing as core need not list the extension, and
lavapipe at 1.3 evidently does not, so wgpu never asks and reports the
features absent, which is why `ash` sees seven `true`s and wgpu sees none.
The properties query beside it (line 1486) correctly accepts
`device_api_version >= 1.2 || extension`; the features query does not.
wgpu-hal 30.0.1 in the local registry has the same asymmetry (lines
1872 and 2036). Worth an upstream issue, but not a reason to keep the
design: on real phones the gate that matters is stricter still.
**wgpu's `TEXTURE_BINDING_ARRAY` needs six sub-features, not one**
(`adapter.rs:160-177`): non-uniform indexing *and* update-after-bind for
sampled images, storage images and storage buffers, all together, because
wgpu marks every array-bearing descriptor set update-after-bind. So Arm's
"the extension is supported on Valhall" is necessary but not sufficient;
a driver with sampled-image indexing and without storage-buffer
update-after-bind is refused too. That widens the excluded set beyond
what the Arm quote suggests and strengthens the conclusion.
### A live bug in the current code, found on the way
`GpuTextures::update` (`core/src/render/texture.rs:33`) implements
"a patch must not report changed" as `changed = false`, unconditionally,
which also **cancels a `Push` earlier in the same batch**. That ordering is
exactly what opening a new atlas page produces: `GlyphAtlas::allocate`
pushes the page and `insert` patches it in the same frame, so the bind
group is not rebuilt and the new page's view is not bound until some
unrelated texture change happens to rebuild it. It is hidden today only
because the masks path also sets `changed`. The fix is one line
(`changed |= !matches!(update, Patch)` in spirit); it should go in with
the redesign since that code is being replaced, and it is recorded here
so it is not rediscovered.
### In-layer draw order is already undefined
Relevant to any batching redesign: `Primitives::apply_free`
(`core/src/render/primitive.rs:147`) uses `swap_remove`, so the instance
order within a layer is permuted whenever anything is freed. Overlap order
inside one layer is therefore not something the renderer promises today;
ordering is done with layers. That means grouping a layer's draws by
texture, or drawing a layer's images after its rects and glyphs, loses
nothing that currently exists. It should be written down as an invariant
when the redesign lands, because the new code will depend on it.
### On "a draw call per image"
Two corrections to the worry. First, it is a draw per *distinct texture per
layer*, not per image primitive: every glyph quad in a layer shares the
atlas and stays one instanced draw, and a thumbnail atlas would do the same
for previews. Second, the count is bounded by what is on screen, which I3's
virtualised transcript already bounds, and a mobile GPU is not draw-call
bound at tens of draws per frame; egui ships exactly this on Android. What
does cost is per-frame *bind group creation* and per-frame *sorting*, and
the current code already creates a `primitive_group` bind group every time
a layer updates (`render/mod.rs:103`), so one more per new image is not a
regression in kind.
### Recommended shape (proposal, for Iris to accept or change)
Aimed at the fewest moving parts that need no feature beyond Vulkan 1.0:
1. **Atlas pages become layers of one `texture_2d_array`**, not separate
textures. Every page is already `PAGE`x`PAGE` RGBA8, which is the one
constraint an array texture imposes. A layer index is an ordinary
sampling operand in WGSL and needs no indexing feature, so `GLYPH`
(and any future atlased-image primitive) carries a layer instead of a
`view_idx` and all of a layer's text stays **one draw**. This answers
the open question above about keeping a small binding array: no. Cost
of opening a page: recreate the array with one more layer and
`copy_texture_to_texture` the old ones, GPU-side, no readback; grow
with headroom (double) so it is rare. wgpu's default
`max_texture_array_layers` is 256, at 4 MB each, so the cap is memory
rather than the API.
2. **Every standalone image is its own texture with its own bind group**,
and its instances live in a **separate per-layer instance list**, not
the main one. Then the main instance buffer never contains an image,
there is nothing to sort, no handle remapping beyond what
`apply_free` already does, and each image is `draw(0..4, k..k+1)` with
its bind group set first. Group 2's layout becomes `{atlas array,
one image texture, sampler, masks}`; the main draw binds a 1x1 null
image in the image slot, each image draw binds its own. One pipeline,
one shader, one layout.
3. **No thumbnail atlas in the first version.** With images on their own
textures, the threshold and eviction questions above disappear: an
image is freed when the row that owns its `TextureHandle` scrolls out.
Add an image atlas only if a measured screen shows enough small images
to matter, which a transcript rarely does.
4. **Drop the three features and the two `max_binding_array_*` limits from
`src/default/render.rs`**, and the `UiLimits` counts with them.
5. **Sampling is `NonFiltering` today** (`render/mod.rs:290,299`), so a
downscaled attachment will alias. Either request a filtering sampler
for the image slot or downscale on the CPU before upload; decide when
the image widget is touched, not as part of this.
What this costs against the file's original recommendation: `Textures`
needs to know an image from a page (two kinds of handle, or a kind on
`TextureHandle`), and `Primitives` gets a second instance list per layer.
What it saves: the sort, the size threshold, the eviction policy, and any
per-page bind group switch.
## Implemented, 2026-09-04
The shape above, built as proposed with one structural addition the proposal
didn't need to spell out and one bug it predicted made moot rather than
literally fixed. Files: `core/src/primitive/texture.rs` (`Textures`,
`TextureHandle`), `core/src/render/texture.rs` (`GpuTextures`),
`core/src/render/primitive.rs` (`Primitives`, `GlyphPrimitive`),
`core/src/render/atlas.rs`, `core/src/ui/painter.rs`,
`core/src/render/mod.rs` (`UiRenderNode`, `UiLimits` removed),
`core/src/render/shader.wgsl`, `src/default/render.rs`, and
`rigs/gpu-probe/src/main.rs`.
**1. Atlas pages as array layers.** `GpuTextures` owns one
`texture_2d_array` (`array_texture`/`array_view`), grown by doubling
(`grow_array`): a new texture is created at twice the layer capacity, the
old layers are copied across with `copy_texture_to_texture` (GPU-side, no
readback), and every bind group that referenced the old view — the main
one and every live standalone image's — is rebuilt, since the view's
identity changed. `GlyphPrimitive` carries `layer: u32` instead of
`view_idx`/`sampler_idx`; the layer number is assigned synchronously in
`Textures::add_page` (a plain counter, `next_page_layer`), not by the
renderer, because `GlyphAtlas::insert` needs it in the same call, before
any GPU sync happens — the renderer only finds out later, when it
processes the queued `Push`.
**2. Standalone images, one bind group each.** `TextureKind` on
`TextureHandle`/`Textures` distinguishes `Image` (a plain bind-group index,
`slot`) from `Page { layer }`. `Primitives` gained a second per-layer list
— `images: Vec<PrimitiveInstance>`, tagged `IMAGE_BINDING` — separate from
`instances` (rects and glyphs), written by `Painter::write_image` rather
than through the generic `Primitive` trait, since an image has nowhere in
`PrimitiveData` to put a per-instance entry once the bind group already
picks the texture. `UiRenderNode::draw` draws a layer's `instance` buffer
once as before, then walks `image_instance` one entry at a time, binding
that texture's `BindGroup` (`GpuTextures::image_bind_group`) and issuing
`draw(0..4, k..k+1)` per image. Group 2's layout is exactly the proposed
`{atlas array, one image texture, sampler, masks}`; the main draw binds a
1x1 null view in the image slot.
**The one addition beyond the proposal**: the masks storage buffer lives
in every per-image bind group (group 2, binding 3), and `ArrBuf<Mask>`
recreates its buffer whenever the mask count changes size
(`render/util/mod.rs`'s `ArrBuf::update` now returns whether it resized).
A resize invalidates every bind group holding the old buffer, not just the
main one, so `GpuTextures::update` takes a `masks_resized: bool` and calls
`rebuild_image_bind_groups` when it's set, alongside the same rebuild the
array-growth path already needed. This wasn't a design question the
proposal had to answer (it treated bind-group construction as a given),
but it's exactly the shape of trap layer growth already had, so it uses
the same fix.
**3. No thumbnail atlas.** Not built, as proposed.
**4. Removed**: `TEXTURE_BINDING_ARRAY`, `PARTIALLY_BOUND_BINDING_ARRAY`,
`SAMPLED_TEXTURE_AND_STORAGE_BUFFER_ARRAY_NON_UNIFORM_INDEXING` from
`src/default/render.rs`'s `request_device`, and `UiLimits` (the type
itself, not just its binding-array methods — once its two fields were
gone there was nothing left in it, and `UiRenderNode::new` no longer takes
a limits parameter). `binding_array` no longer appears anywhere in
`shader.wgsl`.
**5. Sampling** is still `NonFiltering`, unchanged, per the proposal's own
note that this is a separate decision for whenever the image widget itself
is touched.
**The `changed = false` bug is structurally gone, not patched.** The old
`GpuTextures::update` held one `changed: bool` that a `Patch` reset
unconditionally, which could erase an earlier `Push` in the same batch (a
new atlas page's `Push` immediately followed by `GlyphAtlas::insert`'s
`Patch`, both queued before the renderer ever runs). The new `update`
computes the rebuild signal by OR-ing each event's own answer
(`rebuild_main |= self.push(...)`), and `Patch`'s arm simply never
contributes to it — there is no shared mutable flag left for a `Patch` to
stomp on. Documented at the call site
(`core/src/render/texture.rs`, `GpuTextures::update`'s doc comment and the
`Patch` match arm's comment) rather than fixed as a one-line diff, since
the mechanism that could go wrong no longer exists.
**In-layer draw order is an explicit invariant now, not just a fact about
`swap_remove`.** `UiRenderNode::draw` draws every layer's images after its
rects and glyphs, and `Primitives::apply_free`'s doc comment states
directly that both of a layer's lists (`instances` and `images`) free with
`swap_remove` and that nothing may assume adjacency survives a free —
recorded there because `apply_free` is the one place a change to either
list's ordering would have to be reconciled.
**Verified:**
- `cargo fmt --all -- --check`, `cargo build --workspace --all-targets`,
`cargo clippy --all-targets`, `cargo test --workspace` all clean in
`iris/`, on the pinned `nightly-2026-09-03` toolchain. 14 tests pass
(unchanged from I1; nothing here is pure-logic enough to add a unit
test to — it's all GPU resource wiring).
- `iris/run-headless.sh minimal --shot /tmp/minimal.png` and
`iris/run-headless.sh tabs --shot /tmp/tabs.png`: both render correctly
on this VM's GPU (Venus) — `tabs`'s glyph-atlas text renders in every
panel, confirming `GlyphPrimitive.layer` addresses the array correctly.
- The standalone-image path specifically: a throwaway example (not
committed) with an `image(...)` widget as part of the root, run the same
way, rendered the image next to glyph-atlas text in one frame —
confirming a live `BindGroup` built by `GpuTextures::create_image` and
bound per-`draw()` call actually samples the right texture. `tabs`'s own
"image span" tab exercises the same widget but needs a click to reach,
which the headless compositor can't deliver (no seat devices, per I1's
own note on this file) — the throwaway example is what stood in for it.
- **Exercised, 2026-09-04: `grow_array` under real load, on `tabs`.**
Rather than building a purpose-made glyph flood, `PAGE`
(`core/src/render/atlas.rs`) was temporarily dropped from 1024 to 64 —
small enough that `tabs`'s ordinary mix of sizes and families (nothing
exotic: a handful of `Text` widgets at a few sizes, one at
`Family::Monospace`) already exceeds one page's worth of distinct
glyphs. A one-line `eprintln!` in `grow_array` confirmed two real grows
in a single run (`GROW_ARRAY: 1 -> 2` then `GROW_ARRAY: 2 -> 4`, i.e.
glyphs landed on at least a third layer), and
`iris/run-headless.sh tabs --shot` showed every tab's text rendering
correctly with no corruption or missing glyphs — confirming the
`copy_texture_to_texture` grow-and-relocate path and cross-layer
sampling (`GlyphPrimitive.layer` addressing a layer beyond the first)
both work. Command:
`sed -i 's/PAGE: u32 = 1024/PAGE: u32 = 64/' core/src/render/atlas.rs`,
rebuild, `./run-headless.sh tabs --shot /tmp/x.png`, then
`git checkout -- core/src/render/atlas.rs` to revert — this is a
throwaway diagnostic value, never a committed change, since a real
1024px page holding only a handful of glyphs at a time would be mostly
wasted space in normal use. Confirmed the revert left `tabs` and
`minimal` byte-identical to the pre-check screenshots afterward.
- **The decisive check**, `rigs/gpu-probe` rewritten to request iris's new
(empty) feature/limit set and run on this checkout's own emulator
(`ai-app-2`, via `emu`), booted with `EMU_GPU=software` so the guest gets
a real Vulkan device (SwiftShader) rather than the `-gpu host` default,
which disables Vulkan in this VM entirely (`-feature -Vulkan`, because
gfxstream can't pair Venus with the real GPU here — worth remembering,
since the *default* `emu up` gives a device with **no** Vulkan adapter
at all, which reads exactly like the old bindless failure if you don't
know to ask for `EMU_GPU=software`):
cd rigs/gpu-probe
ANDROID_NDK_HOME=$HOME/Android/Sdk/ndk/29.0.14206865 \
cargo ndk -t arm64-v8a -P 26 build --release
EMU_GPU=software emu up # from ~/repos/emulator-tools
adb push target/aarch64-linux-android/release/gpu-probe /data/local/tmp/
adb shell chmod 755 /data/local/tmp/gpu-probe
adb shell /data/local/tmp/gpu-probe
Output: `adapters: 1 — Vulkan SwiftShader Device (Subzero) (Cpu)`,
`features iris requires:` (none listed — the set is empty),
`max_buffer_size … ok`, and **`IRIS DEVICE: ok`**. This is the fix
measured working, on the exact rig that first measured it failing.
Emulator stopped afterward (`emu down`); nothing was left running.
TODO.md
+22
View File
@@ -5,6 +5,10 @@ one in place when it turns out to need a decision.
## App — transcript
- [ ] Decide how running background tasks can be inspected. For now the session
status shows only the provider-reported count; command details stay in
their existing tool cards and must not become subagent cards.
- [ ] Messages received from other agents are inconsistent — sometimes they
appear, sometimes they don't. **Needs a rig.** Read the code rather than
measured: a live Claude session only learns of a peer message from the
@@ -33,3 +37,21 @@ one in place when it turns out to need a decision.
that would work today, for Claude sessions, and it is the option that was
not chosen.
## A session the server could not load
- [ ] **A session whose transcript will not parse is skipped with nothing but a
log line, and from the phone it looks exactly like an idle unresponsive
one.** `SessionManager::new` catches a failing `launch` and logs
"couldn't relaunch session <id>", so the session has no pump and no
driver: no status, no history, nothing sendable. That is what the
`taskNote` incident (fd71d87) looked like from Bryan's phone, and why it
needed a report from him rather than being visible in the app.
`Event::Unreadable` removes the cause that time, but not the class — an
unreadable `process.json`, a provider edited away and an unreachable host
all reach the same place.
This is the "design the unknown state first" rule: a session the server
could not load is not a session with nothing to say, and only the phone
can show the difference. It needs a status the wire can carry for it —
the failure with its reason, reported on the session itself — rather than
the reader having to tell it apart from silence.
app/androidApp/src/main/kotlin/com/example/aiapp/Api.kt
+806 -108
View File
File diff suppressed because it is too large. Load diff
app/androidApp/src/main/kotlin/com/example/aiapp/AppRoot.kt
+141 -18
View File
@@ -1,7 +1,9 @@
package com.example.aiapp
import androidx.activity.compose.BackHandler
import androidx.compose.foundation.background
import androidx.compose.foundation.layout.Box
import androidx.compose.foundation.layout.fillMaxSize
import androidx.compose.foundation.layout.imePadding
import androidx.compose.foundation.layout.padding
import androidx.compose.material3.AlertDialog
@@ -9,6 +11,7 @@ import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.CompositionLocalProvider
import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.getValue
import androidx.compose.runtime.key
@@ -19,6 +22,7 @@ import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.setValue
import androidx.compose.ui.Modifier
import androidx.compose.ui.platform.LocalContext
import androidx.compose.ui.semantics.clearAndSetSemantics
import androidx.compose.ui.unit.dp
import com.example.wgapplink.localNetworkAllowed
import kotlinx.coroutines.Dispatchers
@@ -29,25 +33,39 @@ import kotlinx.coroutines.withContext
* One `when` rather than a navigation library: a handful of screens, with [Screen.Main] as the root
* and the back button the only other way between them.
*
* Import, models and setups are tabs inside [MainScreen] -- four views of the same backend, none of
* them a step down from another -- and what is left here is only what genuinely is a step down: one
* session, spawning one, and settings.
* Import, models and machines are tabs inside [MainScreen] -- four views of the same backend, none
* of them a step down from another -- and what is left here is only what genuinely is a step down:
* one session, spawning one, and settings.
*/
private sealed class Screen {
data object Main : Screen()
/**
* One session, with the file explorer over it when [files] is set.
* One session, with the file explorer or a subagent transcript over it when set.
*
* The explorer is a layer on this screen rather than a screen of its own, so the session under
* it stays composed: its event stream keeps flowing, its scroll position and draft stay put,
* and coming back from a file costs nothing. As a sibling `Screen` it would be disposed and re-
* created on every return, refetching the transcript over the tunnel.
* Both are layers on this screen rather than screens of their own, so the session under them
* stays composed: its event stream keeps flowing, its scroll position and draft stay put, and
* coming back costs nothing. As sibling `Screen`s they would dispose and recreate it on every
* return, refetching the transcript over the tunnel.
*/
data class Session(val summary: SessionSummary, val files: FilesTarget? = null) : Screen()
data class Session(
val summary: SessionSummary,
val files: FilesTarget? = null,
val subagent: SubagentSummary? = null,
) : Screen()
data object Spawn : Screen()
/**
* One provider on one machine: its settings, and what its shared server is holding.
*
* A step down from the machines tab rather than a tab of its own, because it is about one
* machine rather than about the backend. Addressed by ids and names rather than by the
* [Provider] it was tapped from: what it shows is fetched, and a stale copy of a card would be
* a second version of the same truth.
*/
data class ProviderSettings(val machineId: String, val provider: String) : Screen()
data object Settings : Screen()
}
@@ -178,7 +196,7 @@ fun AppRoot(
// deliberately does not: resizing a whole screen on every frame of the keyboard animation is
// the cost that made it lag, so it moves only its composer and transcript.
when (val here = screen) {
is Screen.Main ->
Screen.Main ->
Box(Modifier.imePadding()) {
MainScreen(
settings = current,
@@ -191,6 +209,9 @@ fun AppRoot(
screen = Screen.Session(imported)
},
onSettings = { screen = Screen.Settings },
onProvider = { machineId, provider ->
screen = Screen.ProviderSettings(machineId, provider)
},
)
}
is Screen.Session ->
@@ -204,14 +225,107 @@ fun AppRoot(
// A Box so the explorer can be drawn *over* the session rather than instead of it.
// No imePadding here, for the reason above -- the explorer adds its own.
Box {
SessionScreen(
settings = current,
summary = here.summary,
onBack = goToMain,
onFiles = { screen = here.copy(files = it) },
share = share,
onShareTaken = { share = null },
)
val fileLinkHandler = rememberFileLinkHandler { path ->
screen = here.copy(files = here.summary.filesTarget(path))
}
Box(
Modifier.then(
if (here.subagent != null || here.files != null)
Modifier.clearAndSetSemantics {}
else Modifier
)
) {
// How much background work the session has, from the one subscription
// to its events the screen below holds. Here because the panel and that
// screen both draw it, and must draw the same number.
var backgroundTasks by
remember(here.summary.id) {
mutableIntStateOf(here.summary.backgroundTasks)
}
// Where the panel has asked the session screen to put the reader: the
// call a background task was started by. Held here rather than inside
// either, because the two are siblings -- the panel is the one being
// tapped and the transcript is the one that can travel.
var goTo by remember(here.summary.id) { mutableStateOf<CallSite?>(null) }
// The two panels this session can be pulled aside for: its subagents
// from the right, and the whole main screen from the left. Both are here
// rather than screens of their own for the same reason the explorer is --
// the session under them stays composed. The main panel exists only
// inside a session, which is what makes it unswipeable until one has been
// opened.
SidePanels(
left = { active, close ->
MainPanel(
settings = current,
sessionId = here.summary.id,
active = active,
onOpen = { screen = Screen.Session(it) },
onSpawn = { screen = Screen.Spawn },
onImported = { imported ->
reloadToken++
screen = Screen.Session(imported)
},
onSettings = { screen = Screen.Settings },
onProvider = { machineId, provider ->
screen = Screen.ProviderSettings(machineId, provider)
},
onClose = close,
onGone = goToMain,
)
},
// The whole width: it stands in for the screen Back would have shown,
// rather than sitting over the session the way the subagents do.
leftFraction = 1f,
right = { active, close ->
SubagentPanel(
settings = current,
summary = here.summary,
active = active,
backgroundTasks = backgroundTasks,
onOpenSubagent = { screen = here.copy(subagent = it) },
// Closed with it: what the reader asked to see is under this
// panel, and a panel left open over the answer is the one
// thing the tap cannot have meant.
onOpenCall = {
goTo = it
close()
},
)
},
) {
CompositionLocalProvider(
LocalFileLinkHandler provides fileLinkHandler
) {
SessionScreen(
settings = current,
summary = here.summary,
onBack = goToMain,
onFiles = { screen = here.copy(files = it) },
share = share,
onShareTaken = { share = null },
onBackgroundTasks = { backgroundTasks = it },
goTo = goTo,
onGoToTaken = { goTo = null },
)
}
}
}
here.subagent?.let { subagent ->
BackHandler { screen = here.copy(subagent = null) }
Box(
Modifier.fillMaxSize().background(MaterialTheme.colorScheme.background)
) {
key(subagent.id) {
SessionScreen(
settings = current,
summary = here.summary,
onBack = { screen = here.copy(subagent = null) },
onFiles = {},
subagent = subagent,
)
}
}
}
// Its own back handler is registered after this screen's, so it is the one the
// platform asks first, and it steps back inside itself before closing.
here.files?.let { target ->
@@ -223,6 +337,15 @@ fun AppRoot(
}
}
}
is Screen.ProviderSettings ->
Box(Modifier.imePadding()) {
ProviderScreen(
settings = current,
machineId = here.machineId,
provider = here.provider,
onBack = goToMain,
)
}
is Screen.Spawn ->
Box(Modifier.imePadding()) {
SpawnScreen(
app/androidApp/src/main/kotlin/com/example/aiapp/AskQuestion.kt
+3 -5
View File
@@ -20,7 +20,6 @@ import androidx.compose.material3.LocalContentColor
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.OutlinedButton
import androidx.compose.material3.OutlinedCard
import androidx.compose.material3.OutlinedTextField
import androidx.compose.material3.Surface
import androidx.compose.material3.Text
import androidx.compose.runtime.Composable
@@ -335,12 +334,11 @@ private fun OtherAnswer(text: String, onText: (String) -> Unit) {
// No Send of its own: this is one more way to answer the question, and the card's Submit is
// what sends it. A second send button beside the field made the shorter half of the card look
// like the one that finishes it.
OutlinedTextField(
LabelledField(
label = "Other",
value = text,
onValueChange = onText,
label = { Text("Other") },
singleLine = true,
modifier = Modifier.fillMaxWidth().padding(top = 8.dp),
modifier = Modifier.padding(top = 8.dp),
)
}
app/androidApp/src/main/kotlin/com/example/aiapp/Attachment.kt
+24
View File
@@ -11,6 +11,30 @@ import androidx.compose.ui.text.font.FontFamily
import androidx.compose.ui.text.style.TextOverflow
import androidx.compose.ui.unit.dp
/**
* Whether a session can be sent a picture, as the server answers it.
*
* Three states rather than a switch, because for a local model the answer belongs to the server
* that loaded it: one still coming off disk genuinely has not said. [UNKNOWN] is offered -- a
* control withheld because nobody could ask is a photo button missing from a session that would
* have read the photo perfectly well, and the send path says so if the guess was wrong.
*/
enum class ImageSupport {
ACCEPTED,
REFUSED,
UNKNOWN,
}
/**
* What the server called it; anything else -- an older server, a newer word -- is not an answer.
*/
fun imageSupport(word: String): ImageSupport =
when (word) {
"accepted" -> ImageSupport.ACCEPTED
"refused" -> ImageSupport.REFUSED
else -> ImageSupport.UNKNOWN
}
/**
* Whether [ref] names an image the server stored as one -- `<hex>.<extension>`, with an extension
* from the list it writes -- rather than a file kept under its own name. Mirrors the server's
app/androidApp/src/main/kotlin/com/example/aiapp/Attachments.kt
+11
View File
@@ -38,6 +38,11 @@ suspend fun uploadPickedImage(
* Uploads whatever [uri] names, the way its kind needs. An image goes through [uploadPickedImage]
* and is shrunk; anything else goes whole, under the name the other app or the file chooser gave
* it, because the session is told that name rather than shown the bytes.
*
* A picture is refused here, before anything is read or sent, when [images] says this session's
* model cannot read one. Here rather than beside the photo button because this is where every way
* of attaching meets: the picker, the file chooser, and another app's share sheet -- and only the
* first of those has a button to disable.
*/
suspend fun uploadPicked(
context: Context,
@@ -45,10 +50,16 @@ suspend fun uploadPicked(
sessionId: String,
uri: Uri,
maxEdge: Int?,
images: ImageSupport,
): String {
val resolver = context.contentResolver
val mime = resolver.getType(uri)
if (mime != null && mime.startsWith("image/")) {
if (images == ImageSupport.REFUSED) {
throw ApiException(
"this session's model can't read pictures, so that one wasn't attached"
)
}
return uploadPickedImage(context, settings, sessionId, uri, maxEdge)
}
// Opened before the request starts, so a provider that refuses says so here and not from inside
app/androidApp/src/main/kotlin/com/example/aiapp/BackgroundTasks.kt
+244
View File
@@ -0,0 +1,244 @@
package com.example.aiapp
import androidx.compose.foundation.background
import androidx.compose.foundation.clickable
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.height
import androidx.compose.foundation.layout.heightIn
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.layout.width
import androidx.compose.foundation.lazy.LazyListScope
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.LocalContentColor
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.OutlinedCard
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.remember
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.draw.clip
import androidx.compose.ui.semantics.contentDescription
import androidx.compose.ui.semantics.semantics
import androidx.compose.ui.text.font.FontFamily
import androidx.compose.ui.text.style.TextOverflow
import androidx.compose.ui.unit.dp
/**
* The background work a session has going, above its subagents in the panel [SidePanels] slides
* over it from the right.
*
* Collapsed to its one-line count by default, the way everything else this app adds to a screen
* arrives: what a reader came to the panel for is the subagents, and a run of cards about work
* nobody asked after would push them off it. Expanding pushes them down instead of covering them,
* so the two are read together.
*
* The count is drawn even when it is zero, in the same words. A section that appeared only once
* something was running made its own presence the answer, and no heading at all draws "nothing is
* running" and "nobody has asked yet" identically. There is then nothing to expand, so the heading
* carries no chevron either: it is a statement rather than a control.
*/
fun LazyListScope.backgroundTaskSection(
count: Int,
tasks: LoadState<List<BackgroundTaskSummary>?>,
expanded: Boolean,
onToggle: () -> Unit,
onRetry: () -> Unit,
onOpenCall: (CallSite) -> Unit,
) {
item(key = "background-heading") {
Row(
verticalAlignment = Alignment.CenterVertically,
modifier =
Modifier.fillMaxWidth()
.heightIn(min = 48.dp)
.then(if (count == 0) Modifier else Modifier.clickable(onClick = onToggle)),
) {
Text(
"${backgroundTaskLabel(count)} running",
style = MaterialTheme.typography.titleMedium,
modifier = Modifier.weight(1f),
)
if (count > 0) Chevron(if (expanded) Pointing.Up else Pointing.Down)
}
}
// The count as well as the switch: [tasks] is the last answer anybody got, so a section left
// expanded as the work finished would draw cards for tasks that have ended.
if (count == 0 || !expanded) return
when (tasks) {
is LoadState.Loading ->
item(key = "background-loading") {
CircularProgressIndicator(modifier = Modifier.width(24.dp).height(24.dp))
}
is LoadState.Error ->
item(key = "background-error") {
Column {
Text(
tasks.message,
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
)
TextButton(onClick = onRetry) { Text("Try again") }
}
}
// Null is the provider declining to say, which a session whose process has gone answers.
// Said in words: the count above came from somewhere, and an empty space under it would
// read as the tasks having finished rather than as nobody being left to ask.
is LoadState.Loaded ->
when (val rows = tasks.value) {
null ->
item(key = "background-unknown") {
Text(
"This session isn't saying what these are.",
color = MaterialTheme.colorScheme.onSurfaceVariant,
style = MaterialTheme.typography.bodyMedium,
)
}
else ->
uniqueItems(rows, key = { "background-${it.id}" }) { task ->
BackgroundTaskCard(
task,
onOpen = task.call?.let { call -> { onOpenCall(call) } },
)
}
}
}
}
/**
* One background task: what it is doing, drawn as one line that says what kind it is by how it
* looks.
*
* The kind used to be a second line under the words, which on a list of backgrounded commands was
* "background command" repeated down the panel -- and for a provider that names a task by a process
* id it was the *whole* card, so every row said the same two words. A mark carries the same
* difference in a width the text does not have to make room for, and it is the [Glyph]'s
* description that keeps the words for anybody who cannot see it.
*
* A command needs no mark: drawn the way every other verbatim thing here is -- highlighted,
* monospace, on [rawSurface] -- it says "this is a command" in the same appearance the tool card it
* came from uses, and a mark beside that would be the same fact twice.
*
* [onOpen] is where the call that started this is in the transcript, for the readers who tap it:
* null where the provider never said which call it was, or where that call is no longer in the
* transcript, and the card is then a statement rather than a control. The chevron is what says
* which of the two this is, since a card that quietly does nothing when pressed is worse than one
* that never invited the press.
*/
@Composable
private fun BackgroundTaskCard(task: BackgroundTaskSummary, onOpen: (() -> Unit)?) {
val look = backgroundTaskLook(task.kind)
// Null where a provider named the task by a process id and nothing resolved a command out of
// it: there is no code to draw, so the row takes the mark and the words instead.
val command = task.description?.takeIf { look.code }
OutlinedCard(Modifier.fillMaxWidth()) {
Row(
verticalAlignment = Alignment.CenterVertically,
modifier =
Modifier.fillMaxWidth()
.then(
if (onOpen == null) Modifier
else
Modifier.clickable(
onClickLabel = "Show where this started",
onClick = onOpen,
)
)
.padding(horizontal = 12.dp, vertical = 10.dp),
) {
if (command == null) {
Glyph(
look.glyph,
colour = MaterialTheme.colorScheme.onSurfaceVariant,
modifier = Modifier.semantics { contentDescription = look.words },
)
Spacer(Modifier.width(10.dp))
// The kind stands in as the words where the provider gave no description, rather
// than the id it named the task by: Codex reports a process number, which says
// nothing to the person reading and would look like a name somebody chose.
Text(
task.description ?: look.words,
style = MaterialTheme.typography.bodyMedium,
color =
if (task.description == null) MaterialTheme.colorScheme.onSurfaceVariant
else LocalContentColor.current,
maxLines = 2,
overflow = TextOverflow.Ellipsis,
modifier = Modifier.weight(1f),
)
} else {
// Cut at its tail: what identifies a command is the program at its head, and the
// long ones are exactly the ones being read closely.
Text(
// Not cached: one command line lexes in microseconds -- the cache exists for a
// fence with two hundred lines in it.
remember(command) { highlight(command, Language.SHELL) },
style =
MaterialTheme.typography.bodyMedium.copy(fontFamily = FontFamily.Monospace),
maxLines = 2,
overflow = TextOverflow.Ellipsis,
modifier =
Modifier.weight(1f)
// Smaller than the card's own radius, for the reason [RawBlock] rounds
// its corners that way: this sits inside one.
.clip(MaterialTheme.shapes.extraSmall)
.background(rawSurface)
.padding(horizontal = 6.dp, vertical = 4.dp)
// The fill says "command" to everybody else; this says it to a reader
// who cannot see the fill.
.semantics { contentDescription = "${look.words} $command" },
)
}
if (onOpen != null) {
Spacer(Modifier.width(8.dp))
Chevron(Pointing.Right)
}
}
}
}
/**
* How one kind of background task is drawn: see [backgroundTaskLook].
*
* [code] is the kind whose description is verbatim text rather than prose, which is drawn as code
* and takes no [glyph]; the glyph is still what a task of that kind falls back to when nothing said
* what it ran.
*/
private data class TaskLook(val glyph: String, val words: String, val code: Boolean)
/**
* Everything a [BackgroundTaskSummary.kind] decides, answered by one `when`.
*
* One rather than three, which is the rule this screen already learned once with the status word
* and its colour: three `when`s over one set is two of them waiting to miss a member.
*
* A kind this build has not heard of takes the question mark and is named by what every one of them
* has in common. The nearest word or mark we do know -- a robot, a terminal -- would be this screen
* deciding what the server meant by a word it invented after this build shipped.
*/
private fun backgroundTaskLook(kind: String) =
when (kind) {
// A command is drawn in the face a command is drawn in everywhere else here.
"command" -> TaskLook(COMMAND_GLYPH, "background command", code = true)
"agent" -> TaskLook(AGENT_GLYPH, "subagent", code = false)
"workflow" -> TaskLook(WORKFLOW_GLYPH, "workflow", code = false)
else -> TaskLook(UNKNOWN_GLYPH, "background task", code = false)
}
/**
* The heading over one group in the panel, so neither list is a run of cards with no name.
*
* The same band as the background section's own heading row above, rather than a gap chosen to look
* right here: what separates a heading from the cards above it is that both headings sit in a row
* of one height.
*/
@Composable
fun PanelSectionHeading(text: String) {
Row(verticalAlignment = Alignment.CenterVertically, modifier = Modifier.heightIn(min = 48.dp)) {
Text(text, style = MaterialTheme.typography.titleMedium)
}
}
app/androidApp/src/main/kotlin/com/example/aiapp/Bubble.kt
+51
View File
@@ -1,10 +1,15 @@
package com.example.aiapp
import androidx.compose.foundation.layout.PaddingValues
import androidx.compose.foundation.layout.size
import androidx.compose.foundation.shape.CircleShape
import androidx.compose.foundation.shape.RoundedCornerShape
import androidx.compose.material3.Button
import androidx.compose.material3.ButtonDefaults
import androidx.compose.material3.OutlinedButton
import androidx.compose.runtime.Composable
import androidx.compose.ui.Modifier
import androidx.compose.ui.graphics.Color
import androidx.compose.ui.graphics.Shape
import androidx.compose.ui.unit.dp
@@ -49,3 +54,49 @@ val BubbleShape: Shape = RoundedCornerShape(percent = 50)
* ends that tall would bow its sides.
*/
val BubbleMenuShape: Shape = RoundedCornerShape(20.dp)
/**
* A round button sized to the mark it draws.
*
* The composer's three actions -- attach, stop, send -- are single glyphs, and a pill's word-shaped
* padding around one glyph was width taken from the pickers beside it: with a long model name on
* the row, the permission mode ended up too small to hit. One diameter for all three, and it is the
* platform's minimum touch target rather than a button's shorter default height.
*
* [fill] null draws the outlined form, for the one of the three that does not act on the session.
*/
@Composable
fun CircleButton(
onClick: () -> Unit,
modifier: Modifier = Modifier,
fill: Color? = null,
enabled: Boolean = true,
content: @Composable () -> Unit,
) {
val sized = modifier.size(CircleButtonSize)
if (fill == null) {
OutlinedButton(
onClick = onClick,
enabled = enabled,
shape = CircleShape,
contentPadding = PaddingValues(0.dp),
modifier = sized,
) {
content()
}
} else {
Button(
onClick = onClick,
enabled = enabled,
shape = CircleShape,
colors = actionButtonColors(fill),
contentPadding = PaddingValues(0.dp),
modifier = sized,
) {
content()
}
}
}
/** How wide and tall one of those is; see [CircleButton]. */
val CircleButtonSize = 48.dp
app/androidApp/src/main/kotlin/com/example/aiapp/CodeFence.kt
+1
View File
@@ -160,6 +160,7 @@ private val FENCE_LANGUAGES: Map<String, Language> =
"shell" to Language.SHELL,
"zsh" to Language.SHELL,
"console" to Language.SHELL,
"diff" to Language.DIFF,
"python" to Language.PYTHON,
"py" to Language.PYTHON,
"javascript" to Language.JAVASCRIPT,
app/androidApp/src/main/kotlin/com/example/aiapp/Compaction.kt
+20
View File
@@ -60,3 +60,23 @@ fun compactingLabel(seconds: Long?): String =
seconds < 60 -> "compacting ${seconds}s"
else -> "compacting ${seconds / 60}m ${seconds % 60}s"
}
/**
* How full the session is, as the status row says it.
*
* Three states, not two, and the third is the one that needed the words: a session whose occupancy
* is known and whose ceiling is not. That one keeps the bare figure, and a session with a ceiling
* gets both — the reader can see which they are looking at. What must not happen is a missing
* ceiling drawn as a number, or as a proportion of some assumed window, which would be this screen
* inventing the very fact it does not have.
*
* A llama.cpp session always has one, since the window is a flag its own server was started with. A
* coding CLI's is the vendor's business and neither control protocol states it, so those keep the
* bare figure they have always had.
*/
fun contextLabel(held: Long?, limit: Long?): String =
when {
held == null -> "context unknown"
limit == null -> "context ${tokens(held)}"
else -> "context ${tokens(held)} / ${tokens(limit)}"
}
app/androidApp/src/main/kotlin/com/example/aiapp/Dividers.kt
+63
View File
@@ -12,6 +12,10 @@ import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.graphics.Color
import androidx.compose.ui.unit.dp
import java.time.Instant
import java.time.ZoneId
import java.time.format.DateTimeFormatter
import java.time.format.FormatStyle
/**
* A line across the transcript saying what left the session's context.
@@ -36,6 +40,28 @@ fun TranscriptDivider(text: String, color: Color, modifier: Modifier = Modifier)
}
}
/**
* The rule between two replies that met with nothing said in between -- see
* [TranscriptItem.TurnBreak].
*
* No words and no colour. Every other divider here reports something that happened and is worth
* finding by scanning; this one only says "these are two", and it appears once per turn that
* started without anybody typing. Saying more was a screenful of announcements about background
* work the reader was not asking after -- one of them a whole shell command, drawn as centred prose
* because the words came from somewhere that had no reason to keep them short.
*
* The outline colour is the scheme's one for structure rather than for meaning, which is what this
* is. Inset from both edges so it reads as a separator between two rows rather than as the top edge
* of the one under it.
*/
@Composable
fun TurnBreakRow(modifier: Modifier = Modifier) {
HorizontalDivider(
modifier.fillMaxWidth().padding(horizontal = 48.dp, vertical = 6.dp),
color = MaterialTheme.colorScheme.outlineVariant,
)
}
/**
* The mark a clear leaves.
*
@@ -47,3 +73,40 @@ fun TranscriptDivider(text: String, color: Color, modifier: Modifier = Modifier)
fun ClearedRow(modifier: Modifier = Modifier) {
TranscriptDivider("Context cleared", clearedColor, modifier)
}
/**
* The mark running out of quota leaves.
*
* The same red the usage bar takes when a window is spent, because it is the same fact in a second
* place: colour by consequence, so "there is nothing left to spend" is learned once.
*
* A time rather than a countdown. The row is folded once and never re-measured, so a span would go
* stale on screen the moment it was drawn; and this is when the *account* said it would reset,
* which is not a promise about when the session picks back up. A limit the session was told no
* reset time for says nothing about one -- that state has its own words rather than a plausible
* number.
*/
@Composable
fun LimitRow(item: TranscriptItem.LimitNote, modifier: Modifier = Modifier) {
TranscriptDivider(limitSummary(item.resetsAt, ZoneId.systemDefault()), overLimitColor, modifier)
}
/**
* What the row says. Split out so the wording is testable without a screen, since the two states it
* has to keep apart -- a reset time that arrived and one that never did -- are exactly the pair
* that reads the same when it goes wrong.
*
* [zone] is a parameter rather than read here so a test says the same thing wherever it runs.
*/
fun limitSummary(resetsAt: Double?, zone: ZoneId): String {
val at = resetsAt?.let {
try {
DateTimeFormatter.ofLocalizedTime(FormatStyle.SHORT)
.withZone(zone)
.format(Instant.ofEpochSecond(it.toLong()))
} catch (_: Exception) {
null
}
}
return if (at == null) "Usage limit reached" else "Usage limit reached • resets $at"
}
app/androidApp/src/main/kotlin/com/example/aiapp/EventStream.kt
+2 -2
View File
@@ -14,7 +14,7 @@ private const val RESET_EVENT = "reset"
* mean. [close] from any thread ends it, and the caller owns reconnecting -- with the last seq it
* saw as the new cursor.
*/
class EventStream(settings: ServerSettings, private val sessionId: String) {
class EventStream(settings: ServerSettings, private val address: TranscriptAddress) {
private val stream = Sse(settings)
fun close() = stream.close()
@@ -35,7 +35,7 @@ class EventStream(settings: ServerSettings, private val sessionId: String) {
// one and the screen folds the other, and they have to be the same line.
onEvent: (raw: String, event: SeqEvent) -> Unit,
) {
stream.run("/sessions/$sessionId/events?after=$after", onOpen) { name, data ->
stream.run("/${address.urlPath}/events?after=$after", onOpen) { name, data ->
// A named frame carries no payload and a data frame has no name.
if (name == RESET_EVENT) onReset()
else if (data.isNotEmpty()) onEvent(data, parseSeqEvent(data))
app/androidApp/src/main/kotlin/com/example/aiapp/Events.kt
+130 -2
View File
@@ -61,6 +61,24 @@ sealed class SessionEvent {
data class AssistantText(val delta: String) : SessionEvent()
/** The durable value of the open assistant message, replacing its provisional deltas. */
data class AssistantTextFinal(val text: String) : SessionEvent()
/**
* The model's working, streamed the way its reply is: its own card, and deliberately not part
* of what the session said. Only a provider that actually streams its reasoning sends it.
*/
data class Thinking(val delta: String) : SessionEvent()
/**
* The thinking above this finished, having taken [ms].
*
* Measured by the driver, because only it can see when the model stopped: this app knows when
* an event *arrived*, and the last fragment of a block followed by a slow tool call looks
* exactly like thinking that went on that long.
*/
data class ThinkingDone(val ms: Long) : SessionEvent()
data class ToolStart(val id: String, val tool: String, val input: String) : SessionEvent()
data class ToolUpdate(val id: String, val output: String) : SessionEvent()
@@ -108,6 +126,26 @@ sealed class SessionEvent {
val turnStart: Long? = null,
) : SessionEvent()
/**
* A line in the transcript this build cannot read: a kind a newer server wrote, or one an older
* server wrote that has since been dropped.
*
* [kind] is the word the line called itself, so the row can say what is missing rather than
* that something is. The server makes these when reading; no driver sends one.
*/
data class Unreadable(val kind: String) : SessionEvent()
/**
* Retired on 2026-09-06, hours after it was added: a background task finishing, which turned
* out to be a screenful of notices about work nobody was asking after.
*
* Kept because a transcript is append-only -- the sessions that ran a background task in that
* window have these lines for ever. It draws no row, which is the whole reason it is still
* named here rather than left to fall through to [Unknown]: that would draw a placeholder per
* background task, which is the same wall the row was removed for.
*/
object RetiredTaskNote : SessionEvent()
/**
* A command the session was asked to run on itself and cannot run yet. Resolved by
* [CommandSent] with the same id; a command that ran straight away has only that one.
@@ -117,6 +155,9 @@ sealed class SessionEvent {
/** The same command, handed to the session. */
data class CommandSent(val id: String, val text: String) : SessionEvent()
/** Provider-reported number of background tasks alive now. */
data class BackgroundTasks(val count: Int) : SessionEvent()
data class Status(val state: String) : SessionEvent()
/**
@@ -127,6 +168,16 @@ sealed class SessionEvent {
*/
data class Settings(val model: String?, val permissionMode: String?) : SessionEvent()
/**
* Whether a picture can be sent to this session now, as the thing serving its model answered.
*
* Only a llama.cpp session says this, and it says it twice per model: unknown the moment the
* old one is left, then the loaded server's answer. It carries no row -- it is what the
* composer's photo button is drawn from, and a line in the transcript about a control is not
* something anybody asked after.
*/
data class Images(val images: ImageSupport) : SessionEvent()
/**
* What a turn cost, and how much the model was holding when it ended.
*
@@ -135,7 +186,25 @@ sealed class SessionEvent {
* so adding turns up would report a figure the session stopped being true of. Null where the
* dialect did not say, which leaves the context unmeasured rather than unchanged.
*/
data class UsageDelta(val tokens: Long, val context: Long?) : SessionEvent()
data class UsageDelta(
val tokens: Long,
val context: Long?,
/**
* How fast the reply came out, where the provider measured it -- null everywhere else,
* which is most of them. Never worked out here: the time this app watched a reply arrive
* over includes the network and whatever the server was doing between tokens.
*/
val tokensPerSecond: Double? = null,
/**
* How long the provider spent reading the prompt before it began answering; null where
* nothing measured it. The same rule as [tokensPerSecond]: the provider's own figure, or
* nothing at all.
*/
val prefillMs: Long? = null,
) : SessionEvent()
/** How much context this session's model has, which is what [UsageDelta.context] is out of. */
data class ContextWindow(val tokens: Long) : SessionEvent()
/**
* A compaction that finished, and how much context it recovered.
@@ -157,6 +226,20 @@ sealed class SessionEvent {
*/
data object Cleared : SessionEvent()
/**
* The session stopped because its account's usage limit was reached.
*
* Its own event rather than an [Error] carrying the CLI's sentence, because it is a state
* rather than something that went wrong -- and because the raw sentence is `Claude AI usage
* limit reached|1788546972`, which is not readable by the person it is shown to.
*
* [resetsAt] is epoch seconds and null where the session was told nothing. Only the server acts
* on it; what this draws it as is a time, not a countdown, because nothing here re-measures it.
*/
data class LimitReached(val resetsAt: Double?) : SessionEvent()
data class AuthenticationRequired(val message: String) : SessionEvent()
data class Error(val message: String) : SessionEvent()
/**
@@ -193,6 +276,9 @@ fun parseSeqEvent(json: String): SeqEvent {
)
"messageDropped" -> SessionEvent.MessageDropped(body.getString("id"))
"assistantText" -> SessionEvent.AssistantText(body.getString("delta"))
"assistantTextFinal" -> SessionEvent.AssistantTextFinal(body.getString("text"))
"thinking" -> SessionEvent.Thinking(body.getString("delta"))
"thinkingDone" -> SessionEvent.ThinkingDone(body.getLong("ms"))
"toolStart" ->
SessionEvent.ToolStart(
id = body.getString("id"),
@@ -240,19 +326,26 @@ fun parseSeqEvent(json: String): SeqEvent {
body.getString("text"),
if (body.has("turnStart")) body.getLong("turnStart") else null,
)
"unreadable" -> SessionEvent.Unreadable(body.getString("kind"))
"taskNote" -> SessionEvent.RetiredTaskNote
"commandQueued" ->
SessionEvent.CommandQueued(body.getString("id"), body.getString("text"))
"commandSent" -> SessionEvent.CommandSent(body.getString("id"), body.getString("text"))
"backgroundTasks" -> SessionEvent.BackgroundTasks(body.getInt("count"))
"status" -> SessionEvent.Status(body.getString("state"))
"settings" ->
SessionEvent.Settings(
model = body.optString("model").ifEmpty { null },
permissionMode = body.optString("permissionMode").ifEmpty { null },
)
"images" -> SessionEvent.Images(imageSupport(body.optString("images")))
"contextWindow" -> SessionEvent.ContextWindow(body.getLong("tokens"))
"usageDelta" ->
SessionEvent.UsageDelta(
body.getLong("tokens"),
if (body.has("context")) body.getLong("context") else null,
if (body.has("tokensPerSecond")) body.getDouble("tokensPerSecond") else null,
if (body.has("prefillMs")) body.getLong("prefillMs") else null,
)
"compacted" ->
SessionEvent.Compacted(
@@ -261,6 +354,12 @@ fun parseSeqEvent(json: String): SeqEvent {
trigger = body.optString("trigger").ifEmpty { null },
)
"cleared" -> SessionEvent.Cleared
"limitReached" ->
SessionEvent.LimitReached(
if (body.has("resetsAt")) body.getDouble("resetsAt") else null
)
"authenticationRequired" ->
SessionEvent.AuthenticationRequired(body.getString("message"))
"error" -> SessionEvent.Error(body.getString("message"))
else -> SessionEvent.Unknown(type)
}
@@ -275,7 +374,21 @@ fun parseSeqEvent(json: String): SeqEvent {
* first time the server grows a state, and the drift would be a reply that never splits or one
* split mid-stream.
*/
fun sessionWorking(state: String): Boolean = state == "running" || state == "compacting"
fun sessionWorking(state: String): Boolean =
state == "running" || state == "compacting" || state == "loading" || state == "reading"
/** Whether the latest events still say this session needs an explicit provider login. */
internal fun authenticationPromptAfter(open: Boolean, event: SessionEvent): Boolean =
when (event) {
is SessionEvent.AuthenticationRequired -> true
// A later provider response proves an older authentication failure in a replayed page is
// no longer current. Without this, one old failure reopened sign-in after every later
// successful turn.
is SessionEvent.AssistantText,
is SessionEvent.AssistantTextFinal,
is SessionEvent.ToolStart -> false
else -> open
}
/**
* The context after [event], given what it was before.
@@ -299,3 +412,18 @@ fun contextAfter(current: Long?, event: SessionEvent): Long? =
is SessionEvent.Cleared -> null
else -> current
}
/**
* The context window after [event], mirroring the server's `context_limit_after` for the same
* reason [contextAfter] mirrors its neighbour: the screen has to keep up between page loads.
*
* A window belongs to the process, so a session whose process has exited has none — left standing,
* a session restarted on a different model would draw its occupancy against the old model's
* ceiling.
*/
fun contextLimitAfter(current: Long?, event: SessionEvent): Long? =
when (event) {
is SessionEvent.ContextWindow -> event.tokens
is SessionEvent.Status -> if (event.state == "exited") null else current
else -> current
}
app/androidApp/src/main/kotlin/com/example/aiapp/Field.kt
+127
View File
@@ -0,0 +1,127 @@
package com.example.aiapp
import androidx.compose.foundation.background
import androidx.compose.foundation.border
import androidx.compose.foundation.interaction.MutableInteractionSource
import androidx.compose.foundation.interaction.collectIsFocusedAsState
import androidx.compose.foundation.layout.Box
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.shape.RoundedCornerShape
import androidx.compose.foundation.text.BasicTextField
import androidx.compose.foundation.text.KeyboardActions
import androidx.compose.foundation.text.KeyboardOptions
import androidx.compose.material3.LocalTextStyle
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Text
import androidx.compose.runtime.Composable
import androidx.compose.runtime.getValue
import androidx.compose.runtime.remember
import androidx.compose.ui.Modifier
import androidx.compose.ui.graphics.SolidColor
import androidx.compose.ui.text.TextStyle
import androidx.compose.ui.unit.dp
/**
* A text field whose label is a line above it rather than a thing floating inside it.
*
* Every field in this app goes through here, and the reason is vertical space. Material's outlined
* field reserves room for a label that animates into its own border and pads the value by half a
* line top and bottom, so one setting costs the height of three lines of text to hold one. A form
* of ten settings is then a screen and a half of scrolling to read ten short answers.
*
* What is *not* shrunk is the value itself: it stays at body size, because what is expensive here
* is the framing rather than the text, and a field whose contents are smaller than the text beside
* it is a field the reader has to lean in to check. See UI_RULES on never shrinking text to fit.
*
* [hint] is what leaving it blank means, drawn inside the empty box. It had a grey line of its own
* above the box until 2026-09-21: a form of a dozen settings was then mostly explanation, and a
* setting should be a title and a box to type in. Inside, it costs no height and is gone the moment
* anybody types -- which is the trade, since that is also when somebody might look back at it.
*/
@Composable
fun LabelledField(
label: String,
value: String,
onValueChange: (String) -> Unit,
modifier: Modifier = Modifier,
hint: String? = null,
enabled: Boolean = true,
/**
* How many lines the box is, at rest. One for a value; several for prose, where the reader is
* writing rather than filling in -- see `ParamKind::Prose`.
*/
lines: Int = 1,
keyboardOptions: KeyboardOptions = KeyboardOptions.Default,
/** What the keyboard's own action key does, which is usually what the button beside it does. */
keyboardActions: KeyboardActions = KeyboardActions.Default,
) {
Column(modifier.fillMaxWidth()) {
// Body size in the ordinary text colour, which is what a setting's label is where the
// control beside it is a switch or a picker. Smaller and greyer on the ones that are
// fields reads as two ranks of setting where there is one.
Text(label, modifier = Modifier.padding(bottom = 2.dp))
FieldBox(value, onValueChange, enabled, lines, keyboardOptions, keyboardActions, hint)
}
}
/** The box itself: the border, the padding, and the text. Shared so the two fields agree. */
@Composable
private fun FieldBox(
value: String,
onValueChange: (String) -> Unit,
enabled: Boolean,
lines: Int,
keyboardOptions: KeyboardOptions,
keyboardActions: KeyboardActions,
hint: String?,
) {
val interactions = remember { MutableInteractionSource() }
val focused by interactions.collectIsFocusedAsState()
// The focused border is the accent at the same width as the resting one. Growing it instead
// would move the text inside by a pixel on every focus, which is a whole form twitching as the
// reader moves down it.
val edge =
when {
!enabled -> MaterialTheme.colorScheme.outlineVariant
focused -> MaterialTheme.colorScheme.primary
else -> MaterialTheme.colorScheme.outline
}
val shape = RoundedCornerShape(8.dp)
val style =
LocalTextStyle.current.merge(
TextStyle(
color =
if (enabled) MaterialTheme.colorScheme.onSurface
else MaterialTheme.colorScheme.onSurfaceVariant
)
)
BasicTextField(
value = value,
onValueChange = onValueChange,
enabled = enabled,
singleLine = lines == 1,
minLines = lines,
textStyle = style,
keyboardOptions = keyboardOptions,
keyboardActions = keyboardActions,
interactionSource = interactions,
cursorBrush = SolidColor(MaterialTheme.colorScheme.primary),
modifier =
Modifier.fillMaxWidth()
.background(MaterialTheme.colorScheme.surfaceContainerHighest, shape)
.border(1.dp, edge, shape)
.padding(horizontal = 10.dp, vertical = 8.dp),
decorationBox = { field ->
Box {
// Under the text rather than beside it: the value is what the box is for, and a
// hint that pushed it sideways would move every character as somebody typed.
if (value.isEmpty() && hint != null) {
Text(hint, color = MaterialTheme.colorScheme.onSurfaceVariant)
}
field()
}
},
)
}
app/androidApp/src/main/kotlin/com/example/aiapp/FilesScreen.kt
+166 -56
View File
@@ -20,7 +20,6 @@ import androidx.compose.foundation.verticalScroll
import androidx.compose.material3.AlertDialog
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.OutlinedTextField
import androidx.compose.material3.Switch
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
@@ -44,26 +43,60 @@ import kotlinx.coroutines.withContext
/**
* Which machine's files to show, and where to start.
*
* A **setup**, not a session: a filesystem is a property of a machine, and a session only says
* where it was working. That is what makes a second way in -- from the setups tab -- one more
* A **machine**, not a session: a filesystem is a property of a machine, and a session only says
* where it was working. That is what makes a second way in -- from the machines tab -- one more
* caller rather than any new code here.
*/
data class FilesTarget(val setup: String, val setupName: String, val start: String)
data class FilesTarget(
val machine: String,
val machineName: String,
val start: String,
/** A document to open immediately; [start] remains the fallback directory. */
val file: String? = null,
)
/**
* The explorer opened on one file, wherever that file is.
*
* Its directory is what the reader lands in on the way back, which for a session's transcript is
* that session's own directory -- the log, the process record and the rest of what it wrote.
*/
fun fileTarget(file: FileOnMachine) =
FilesTarget(
machine = file.machine,
machineName = file.machineName,
start = parentOf(file.path) ?: "/",
file = file.path,
)
/** The explorer target for this session's machine, optionally opened on [file]. */
fun SessionSummary.filesTarget(file: String? = null) =
FilesTarget(
machine = machine,
machineName = machineName,
start = cwd?.takeIf { it.isNotBlank() } ?: "~",
file = file,
)
/** Where the explorer is: in a directory, or in one file. */
private sealed class Spot(val path: String) {
class Dir(path: String) : Spot(path)
class Doc(path: String) : Spot(path)
class Doc(path: String, val directory: Dir) : Spot(path)
}
private enum class UnsavedDestination {
Directory,
Session,
}
/**
* The files on the machine a session runs on: browse them, read one, change one.
*
* Drawn **over** the session rather than instead of it (see [AppRoot]), so its event stream keeps
* flowing and coming back from a file costs nothing. Back steps one level inside here -- editor to
* viewer, viewer to the directory it came from, directory to the one above -- and only closes from
* where it opened.
* flowing and coming back from a file costs nothing. Both back controls return from a file to its
* directory. In a directory, Android back walks toward the session's project directory and closes
* the explorer once it gets there; the header's back button closes it immediately.
*
* Every directory that has been visited is kept for as long as this is open; the refresh glyph is
* how one gets asked again on purpose, and creating something refetches the directory it was
@@ -72,51 +105,82 @@ private sealed class Spot(val path: String) {
@Composable
fun FilesScreen(settings: ServerSettings, target: FilesTarget, onClose: () -> Unit) {
val scope = rememberCoroutineScope()
var stack by remember { mutableStateOf(listOf<Spot>(Spot.Dir(target.start))) }
val initialDirectory =
target.file?.let(::parentOf)?.let { Spot.Dir(it) } ?: Spot.Dir(target.start)
var here by
remember(target) {
mutableStateOf<Spot>(
target.file?.let { Spot.Doc(it, initialDirectory) } ?: initialDirectory
)
}
val listings = remember { mutableStateMapOf<String, LoadState<Listing>>() }
var creating by remember { mutableStateOf(false) }
// Edit mode and whether anything has been typed live here rather than in the pane below,
// because they are what back has to know about -- and back arrives from two places, the arrow
// and the platform's own gesture, which must mean the same thing.
// because both ways out have to ask before discarding it.
var editing by remember { mutableStateOf(false) }
var dirty by remember { mutableStateOf(false) }
var askUnsaved by remember { mutableStateOf(false) }
val here = stack.last()
var unsavedDestination by remember { mutableStateOf<UnsavedDestination?>(null) }
fun go(spot: Spot) {
editing = false
dirty = false
stack = stack + spot
here = spot
}
fun back() {
when {
editing && dirty -> askUnsaved = true
editing -> editing = false
stack.size > 1 -> {
stack = stack.dropLast(1)
editing = false
dirty = false
}
else -> onClose()
fun leave(destination: UnsavedDestination) {
if (editing && dirty) {
unsavedDestination = destination
} else if (destination == UnsavedDestination.Directory) {
go((here as Spot.Doc).directory)
} else {
onClose()
}
}
suspend fun load(path: String, again: Boolean) {
if (!again && listings[path] is LoadState.Loaded) return
val existing = listings[path]
if (!again && (existing is LoadState.Loaded || existing is LoadState.Loading)) return
listings[path] = LoadState.Loading
listings[path] =
try {
withContext(Dispatchers.IO) {
LoadState.Loaded(fetchDir(settings, target.setup, path))
LoadState.Loaded(fetchDir(settings, target.machine, path))
}
} catch (e: ApiException) {
LoadState.failed(e)
}
}
BackHandler(onBack = ::back)
val projectDirectory = (listings[target.start] as? LoadState.Loaded)?.value?.path
val homeDirectory =
if (target.start == "~") projectDirectory
else (listings["~"] as? LoadState.Loaded)?.value?.path
fun systemBack() {
when (val spot = here) {
is Spot.Doc -> leave(UnsavedDestination.Directory)
is Spot.Dir -> {
val path = (listings[spot.path] as? LoadState.Loaded)?.value?.path ?: spot.path
when {
path == projectDirectory || path == target.start -> onClose()
projectDirectory != null ->
nextDirectoryToward(path, projectDirectory)?.let { go(Spot.Dir(it)) }
?: onClose()
else -> parentOf(path)?.let { go(Spot.Dir(it)) } ?: onClose()
}
}
}
}
// A file link can open without visiting the project first, but Back still needs to know where
// the project is. Home is likewise resolved by the machine rather than guessed on the phone;
// it is what lets every path beneath it be displayed with `~`, including over ssh.
LaunchedEffect(target.machine, target.start) {
if (target.file != null) load(target.start, again = false)
if (target.start != "~") load("~", again = false)
}
BackHandler(onBack = ::systemBack)
Box(
Modifier.fillMaxSize()
@@ -129,14 +193,15 @@ fun FilesScreen(settings: ServerSettings, target: FilesTarget, onClose: () -> Un
when (val spot = here) {
is Spot.Dir -> {
val state = listings[spot.path] ?: LoadState.Loading
// The resolved path once there is one: a directory opened as `~` is called what
// it turned out to be, not what it was asked for.
// Navigate with the resolved path, but name anything under the machine's home
// the way somebody working there would write it.
val at = (state as? LoadState.Loaded)?.value?.path ?: spot.path
val shownAt = tildePath(at, homeDirectory)
FilesHeader(
title = baseName(at),
path = at,
machine = target.setupName,
onBack = ::back,
title = baseName(shownAt),
path = shownAt,
machine = target.machineName,
onBack = { leave(UnsavedDestination.Session) },
) {
GlyphButton(
REFRESH_GLYPH,
@@ -152,7 +217,7 @@ fun FilesScreen(settings: ServerSettings, target: FilesTarget, onClose: () -> Un
)
}
LaunchedEffect(spot.path) { load(spot.path, again = false) }
DirectoryBody(state, onOpen = ::go)
DirectoryBody(state, directory = spot, onOpen = ::go)
}
is Spot.Doc ->
DocPane(
@@ -161,22 +226,26 @@ fun FilesScreen(settings: ServerSettings, target: FilesTarget, onClose: () -> Un
path = spot.path,
name = baseName(spot.path),
editing = editing,
homeDirectory = homeDirectory,
onEditing = { editing = it },
onDirty = { dirty = it },
onBack = ::back,
onBack = { leave(UnsavedDestination.Directory) },
)
}
}
}
if (askUnsaved) {
unsavedDestination?.let { destination ->
UnsavedDialog(
onDiscard = {
askUnsaved = false
editing = false
dirty = false
unsavedDestination = null
if (destination == UnsavedDestination.Directory) {
go((here as Spot.Doc).directory)
} else {
onClose()
}
},
onCancel = { askUnsaved = false },
onCancel = { unsavedDestination = null },
)
}
@@ -185,7 +254,7 @@ fun FilesScreen(settings: ServerSettings, target: FilesTarget, onClose: () -> Un
if (creating && dir != null && listing != null) {
CreateDialog(
settings = settings,
setup = target.setup,
machine = target.machine,
directory = listing.path,
onDismiss = { creating = false },
onCreated = { path, isDirectory ->
@@ -196,7 +265,7 @@ fun FilesScreen(settings: ServerSettings, target: FilesTarget, onClose: () -> Un
load(dir.path, again = true)
// A new file has nothing to look at, so it opens where it can be filled in.
if (!isDirectory) {
go(Spot.Doc(path))
go(Spot.Doc(path, dir))
editing = true
}
}
@@ -248,7 +317,11 @@ private fun FilesHeader(
* looks like a right one.
*/
@Composable
private fun ColumnScope.DirectoryBody(state: LoadState<Listing>, onOpen: (Spot) -> Unit) {
private fun ColumnScope.DirectoryBody(
state: LoadState<Listing>,
directory: Spot.Dir,
onOpen: (Spot) -> Unit,
) {
when (state) {
is LoadState.Loading -> CircularProgressIndicator(Modifier.padding(16.dp))
is LoadState.Error ->
@@ -289,7 +362,9 @@ private fun ColumnScope.DirectoryBody(state: LoadState<Listing>, onOpen: (Spot)
name = entry.name,
trailing = trailingOf(entry),
onClick = {
onOpen(if (entry.isDirectory) Spot.Dir(path) else Spot.Doc(path))
onOpen(
if (entry.isDirectory) Spot.Dir(path) else Spot.Doc(path, directory)
)
},
)
}
@@ -357,6 +432,7 @@ private fun ColumnScope.DocPane(
path: String,
name: String,
editing: Boolean,
homeDirectory: String?,
onEditing: (Boolean) -> Unit,
onDirty: (Boolean) -> Unit,
onBack: () -> Unit,
@@ -381,7 +457,7 @@ private fun ColumnScope.DocPane(
state = LoadState.Loading
state =
try {
val got = withContext(Dispatchers.IO) { fetchFile(settings, target.setup, path) }
val got = withContext(Dispatchers.IO) { fetchFile(settings, target.machine, path) }
if (got is FileContent.Text) draft = TextFieldValue(got.content)
LoadState.Loaded(got)
} catch (e: ApiException) {
@@ -404,7 +480,7 @@ private fun ColumnScope.DocPane(
try {
val written =
withContext(Dispatchers.IO) {
writeFile(settings, target.setup, path, draft.text, against)
writeFile(settings, target.machine, path, draft.text, against)
}
state =
LoadState.Loaded(
@@ -430,7 +506,12 @@ private fun ColumnScope.DocPane(
}
}
FilesHeader(title = name, path = path, machine = target.setupName, onBack = onBack) {
FilesHeader(
title = name,
path = tildePath(path, homeDirectory),
machine = target.machineName,
onBack = onBack,
) {
if (editing) {
if (saving) {
GlyphSpinner("Saving")
@@ -527,7 +608,9 @@ private fun ColumnScope.DocPane(
scope.launch {
val fresh =
try {
withContext(Dispatchers.IO) { fetchFile(settings, target.setup, path) }
withContext(Dispatchers.IO) {
fetchFile(settings, target.machine, path)
}
} catch (e: ApiException) {
saveError = e.message
conflict = null
@@ -571,7 +654,7 @@ private fun Note(text: String) {
@Composable
private fun CreateDialog(
settings: ServerSettings,
setup: String,
machine: String,
directory: String,
onDismiss: () -> Unit,
onCreated: (String, Boolean) -> Unit,
@@ -591,8 +674,8 @@ private fun CreateDialog(
scope.launch {
try {
withContext(Dispatchers.IO) {
if (isDirectory) createDir(settings, setup, path)
else createFile(settings, setup, path)
if (isDirectory) createDir(settings, machine, path)
else createFile(settings, machine, path)
}
onCreated(path, isDirectory)
} catch (e: ApiException) {
@@ -609,13 +692,11 @@ private fun CreateDialog(
title = { Text("Create in ${baseName(directory)}") },
text = {
Column {
OutlinedTextField(
LabelledField(
label = "Name",
value = name,
onValueChange = { name = it },
label = { Text("Name") },
singleLine = true,
enabled = !busy,
modifier = Modifier.fillMaxWidth(),
)
Spacer(Modifier.height(8.dp))
Row(verticalAlignment = Alignment.CenterVertically) {
@@ -678,6 +759,35 @@ internal fun parentOf(path: String): String? {
}
}
/**
* The next directory on the filesystem path from [current] to [destination], or null when there.
*
* Moving between two branches first walks upward to their common ancestor. Once [current] is that
* ancestor, the next press walks one segment down toward [destination]. Both paths are answers from
* the machine, so they are absolute and have no symlinks or `..` left to resolve here.
*/
internal fun nextDirectoryToward(current: String, destination: String): String? {
val here = current.trimEnd('/').ifEmpty { "/" }
val there = destination.trimEnd('/').ifEmpty { "/" }
if (here == there) return null
val beneathHere = if (here == "/") there.startsWith('/') else there.startsWith("$here/")
if (!beneathHere) return parentOf(here)
val next = there.removePrefix(here).trimStart('/').substringBefore('/')
return join(here, next)
}
/** A path as somebody on [home] writes it, leaving paths outside that home unchanged. */
internal fun tildePath(path: String, home: String?): String {
val at = path.trimEnd('/').ifEmpty { "/" }
val resolvedHome = home?.trimEnd('/')?.ifEmpty { "/" } ?: return at
return when {
at == resolvedHome -> "~"
resolvedHome != "/" && at.startsWith("$resolvedHome/") ->
"~${at.removePrefix(resolvedHome)}"
else -> at
}
}
/** What a path names: its last segment, with `/` naming itself. */
internal fun baseName(path: String): String {
val trimmed = path.trimEnd('/')
app/androidApp/src/main/kotlin/com/example/aiapp/Highlighter.kt
+26
View File
@@ -7,6 +7,8 @@ import androidx.compose.ui.text.buildAnnotatedString
/** What a span of code is, in the terms the palette has a colour for. */
enum class Kind {
ADDITION,
DELETION,
KEYWORD,
STRING,
LITERAL,
@@ -24,6 +26,8 @@ data class Span(val start: Int, val end: Int, val kind: Kind)
* one instance and lives with the rest of the palette.
*/
data class SyntaxPalette(
val addition: Color,
val deletion: Color,
val keyword: Color,
val string: Color,
val literal: Color,
@@ -34,6 +38,8 @@ data class SyntaxPalette(
) {
fun of(kind: Kind): Color =
when (kind) {
Kind.ADDITION -> addition
Kind.DELETION -> deletion
Kind.KEYWORD -> keyword
Kind.STRING -> string
Kind.LITERAL -> literal
@@ -44,6 +50,26 @@ data class SyntaxPalette(
}
}
/** A unified diff is line-oriented: colour the changed lines and leave context untouched. */
fun scanDiff(code: String): List<Span> {
val spans = ArrayList<Span>()
var start = 0
while (start < code.length) {
val end = code.indexOf('\n', start).let { if (it == -1) code.length else it }
val kind =
when {
code.startsWith("+++", start) || code.startsWith("---", start) -> Kind.METADATA
code.startsWith("+", start) -> Kind.ADDITION
code.startsWith("-", start) -> Kind.DELETION
code.startsWith("@@", start) -> Kind.METADATA
else -> null
}
if (kind != null) spans.add(Span(start, end, kind))
start = if (end == code.length) end else end + 1
}
return spans
}
/**
* [code] with its keywords, strings and comments coloured, or plain if there is no language for it.
*
app/androidApp/src/main/kotlin/com/example/aiapp/ImportScreen.kt
+31 -29
View File
@@ -82,8 +82,8 @@ private const val SETTLE_MS = 500L
@Composable
fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (SessionSummary) -> Unit) {
val scope = rememberCoroutineScope()
var setups by remember { mutableStateOf<LoadState<List<Setup>>>(LoadState.Loading) }
var chosen by remember { mutableStateOf<Setup?>(null) }
var machines by remember { mutableStateOf<LoadState<List<Machine>>>(LoadState.Loading) }
var chosen by remember { mutableStateOf<Machine?>(null) }
var sessions by remember { mutableStateOf<LoadState<List<Importable>>>(LoadState.Loading) }
// What is happening to each row right now, as the word the row shows. A map keyed by id rather
@@ -100,9 +100,8 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
// Deleting a transcript cannot be undone, so it is asked rather than done. Held as the rows
// themselves, not a flag, so the dialog can say what it is about.
var confirming by remember { mutableStateOf<List<Importable>?>(null) }
// Same default as the spawn screen: a phone is the wrong place to answer "allow Bash?" forty
// times.
var permissionMode by remember { mutableStateOf("auto") }
// Set from the selected Claude provider rather than repeated in the app.
var permissionMode by remember { mutableStateOf("") }
// When each row last slid upwards, as a plain map rather than state: nothing is drawn from it,
// so a tap reading it needs no recomposition.
val movedAt = remember { mutableMapOf<String, Long>() }
@@ -114,9 +113,9 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
* Taken from the answer rather than kept across the load: the server is what knows what is
* running, and this screen may be opening on work another phone started.
*/
suspend fun fetchInto(setup: Setup): LoadState<List<Importable>> =
suspend fun fetchInto(machine: Machine): LoadState<List<Importable>> =
try {
val rows = withContext(Dispatchers.IO) { fetchImportable(settings, setup.id) }
val rows = withContext(Dispatchers.IO) { fetchImportable(settings, machine.id) }
running = rows.mapNotNull { row -> row.pending?.let { row.id to it } }.toMap()
rowErrors = rows.mapNotNull { row -> row.error?.let { row.id to it } }.toMap()
LoadState.Loaded(rows)
@@ -124,10 +123,10 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
LoadState.Error(err.message ?: "Couldn't list sessions")
}
fun loadSessions(setup: Setup) {
fun loadSessions(machine: Machine) {
sessions = LoadState.Loading
selected = emptySet()
scope.launch { sessions = fetchInto(setup) }
scope.launch { sessions = fetchInto(machine) }
}
/** Takes a row out of the list, once the machine no longer has it to offer. */
@@ -141,9 +140,9 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
}
LaunchedEffect(reloadToken) {
setups =
machines =
try {
val found = withContext(Dispatchers.IO) { fetchSetups(settings) }
val found = withContext(Dispatchers.IO) { fetchMachines(settings) }
found.firstOrNull()?.let {
chosen = it
loadSessions(it)
@@ -171,7 +170,7 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
selected = emptySet()
running = running + targets.associate { it.id to WAITING }
rowErrors = rowErrors - targets.map { it.id }.toSet()
val setup = chosen
val machine = chosen
val ids = targets.map { it.id }
scope.launch {
// One request for the whole batch, not one per row. Sent row by row, a handover was
@@ -198,25 +197,28 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
// The listing is the repair, because it carries the same state the events do. Only when
// something still looks outstanding, so the ordinary case does not pay for a second
// listing, which is the most expensive call this screen makes.
if (setup != null && targets.any { running.containsKey(it.id) }) {
if (machine != null && targets.any { running.containsKey(it.id) }) {
// Quietly: no Loading, because blanking the list to report on rows that are already
// saying what is happening to them is the flicker this screen avoids everywhere
// else.
sessions = fetchInto(setup)
sessions = fetchInto(machine)
}
}
}
val provider = chosen?.providers?.firstOrNull { it.kind == "claude_cli" }
LaunchedEffect(chosen?.id, provider?.name) {
permissionMode = provider?.defaultPermissionMode.orEmpty()
}
/** Continues [targets] in the background, leaving the screen where it is. */
fun importAll(targets: List<Importable>) {
val setup = chosen ?: return
val machine = chosen ?: return
val useProvider = provider ?: return
handOver(targets) { ids ->
startImport(
settings,
setup = setup.id,
machine = machine.id,
sessionIds = ids,
provider = useProvider.name,
permissionMode = permissionMode,
@@ -232,7 +234,7 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
* it, which is the case where waiting is the right thing anyway.
*/
fun importAndOpen(target: Importable) {
val setup = chosen ?: return
val machine = chosen ?: return
val useProvider = provider ?: return
running = running + (target.id to IMPORTING)
rowErrors = rowErrors - target.id
@@ -242,7 +244,7 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
withContext(Dispatchers.IO) {
spawnSession(
settings,
setup = setup.id,
machine = machine.id,
provider = useProvider.name,
// Nothing to say: the server titles it from the session it continues.
title = "",
@@ -270,10 +272,10 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
java.util.concurrent.atomic.AtomicReference<ImportableStream?>(null)
}
LaunchedEffect(chosen?.id) {
val setup = chosen?.id ?: return@LaunchedEffect
val machine = chosen?.id ?: return@LaunchedEffect
try {
while (true) {
val stream = ImportableStream(settings, setup)
val stream = ImportableStream(settings, machine)
liveChanges.set(stream)
try {
withContext(Dispatchers.IO) {
@@ -341,24 +343,24 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
)
Spacer(Modifier.height(12.dp))
when (val loaded = setups) {
when (val loaded = machines) {
is LoadState.Loading -> CircularProgressIndicator()
is LoadState.Error -> Text(loaded.message, color = MaterialTheme.colorScheme.error)
is LoadState.Loaded -> {
// Only worth choosing when there is a choice.
if (loaded.value.size > 1) {
Row(Modifier.fillMaxWidth()) {
loaded.value.forEach { setup ->
loaded.value.forEach { machine ->
TextButton(
onClick = {
chosen = setup
loadSessions(setup)
chosen = machine
loadSessions(machine)
}
) {
Text(
setup.name,
machine.name,
color =
if (setup.id == chosen?.id)
if (machine.id == chosen?.id)
MaterialTheme.colorScheme.primary
else MaterialTheme.colorScheme.onSurfaceVariant,
)
@@ -375,7 +377,7 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
} else {
ChipGroup(
label = "Permissions",
options = PERMISSION_MODES,
options = provider?.permissionModes.orEmpty(),
selected = permissionMode,
onSelect = { permissionMode = it },
)
@@ -439,9 +441,9 @@ fun ImportScreen(settings: ServerSettings, reloadToken: Int, onImported: (Sessio
confirmButton = {
TextButton(
onClick = {
val setup = chosen ?: return@TextButton
val machine = chosen ?: return@TextButton
confirming = null
handOver(targets) { ids -> deleteImportable(settings, setup.id, ids) }
handOver(targets) { ids -> deleteImportable(settings, machine.id, ids) }
}
) {
// Coloured by consequence: this takes something away, wherever it appears.
app/androidApp/src/main/kotlin/com/example/aiapp/ImportableStream.kt
+2 -2
View File
@@ -12,13 +12,13 @@ package com.example.aiapp
* the caller owns reconnecting -- there is no cursor to resume from, because anything missed is in
* the next listing.
*/
class ImportableStream(settings: ServerSettings, private val setup: String) {
class ImportableStream(settings: ServerSettings, private val machine: String) {
private val stream = Sse(settings)
fun close() = stream.close()
fun run(onOpen: () -> Unit, onChange: (ImportableChange) -> Unit) {
stream.run("/setups/$setup/importable/events", onOpen) { _, data ->
stream.run("/machines/$machine/importable/events", onOpen) { _, data ->
if (data.isNotEmpty()) parseImportableChange(data)?.let(onChange)
}
}
app/androidApp/src/main/kotlin/com/example/aiapp/Languages.kt
+2 -1
View File
@@ -16,6 +16,7 @@ enum class Language {
CPP,
CSHARP,
DART,
DIFF,
FISH,
GO,
JAVA,
@@ -100,7 +101,7 @@ fun spansOf(code: String, language: Language): List<Span> = SCANNERS.getValue(la
// Lazy for the same reason [RULES] is, since it reads it.
private val SCANNERS: Map<Language, (String) -> List<Span>> by lazy {
RULES.mapValues { (_, rules) -> { code: String -> scan(code, rules) } } +
mapOf(Language.MARKDOWN to ::scanMarkdown)
mapOf(Language.DIFF to ::scanDiff, Language.MARKDOWN to ::scanMarkdown)
}
private val C_STYLE = BlockComment("/*", "*/", nests = false)
app/androidApp/src/main/kotlin/com/example/aiapp/MachineModels.kt
+402
View File
@@ -0,0 +1,402 @@
package com.example.aiapp
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.height
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.lazy.LazyListScope
import androidx.compose.foundation.text.KeyboardActions
import androidx.compose.foundation.text.KeyboardOptions
import androidx.compose.material3.Card
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.LinearProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.Stable
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.platform.LocalSoftwareKeyboardController
import androidx.compose.ui.text.input.ImeAction
import androidx.compose.ui.text.style.TextOverflow
import androidx.compose.ui.unit.dp
import kotlinx.coroutines.CoroutineScope
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.delay
import kotlinx.coroutines.launch
import kotlinx.coroutines.withContext
/**
* The models on one machine, the downloads putting more there, and HuggingFace to find them in.
*
* This was a tab of its own, about the backend's own disk. It moved under the machine's llama.cpp
* provider on 2026-09-19, when a download came to run on the machine that will serve the file:
* there is no such thing as "the models", only this machine's, and the screen that decides how a
* model is loaded is the screen that should be able to fetch one.
*
* Everything here is the machine's state rather than this screen's. A download is a process on that
* machine with its progress written beside the partial file, so closing the app, locking the phone
* or restarting the backend does not touch it, and a second device watching sees the same numbers.
*/
@Stable
class MachineModelsState(
private val settings: ServerSettings,
private val machineId: String,
private val scope: CoroutineScope,
) {
var state by mutableStateOf<LoadState<Models>>(LoadState.Loading)
private set
var query by mutableStateOf("")
var results by mutableStateOf<LoadState<List<RemoteRepo>>?>(null)
private set
var openRepo by mutableStateOf<String?>(null)
private set
var repoFiles by mutableStateOf<LoadState<List<RemoteFile>>?>(null)
private set
/** What the last action said went wrong, shown above the list that action was taken in. */
var actionError by mutableStateOf<String?>(null)
private set
val models: Models?
get() = (state as? LoadState.Loaded)?.value
val downloads: List<Download>
get() = models?.downloads.orEmpty()
/** How big each downloaded model is, by key, for the cards the provider screen draws. */
val sizes: Map<String, Long>
get() = models?.local.orEmpty().associate { it.key to it.bytes }
suspend fun reload() {
state =
try {
withContext(Dispatchers.IO) {
LoadState.Loaded(fetchMachineModels(settings, machineId))
}
} catch (e: ApiException) {
LoadState.failed(e)
}
}
/** Runs [action], says what it said if it failed, and asks the machine again either way. */
private fun act(action: suspend () -> Unit) {
scope.launch {
actionError =
runCatching { withContext(Dispatchers.IO) { action() } }.exceptionOrNull()?.message
reload()
}
}
fun search() {
openRepo = null
results = LoadState.Loading
scope.launch {
results =
try {
withContext(Dispatchers.IO) { LoadState.Loaded(searchModels(settings, query)) }
} catch (e: ApiException) {
LoadState.failed(e)
}
}
}
fun toggleRepo(repo: String) {
if (openRepo == repo) {
openRepo = null
return
}
openRepo = repo
repoFiles = LoadState.Loading
scope.launch {
repoFiles =
try {
withContext(Dispatchers.IO) {
LoadState.Loaded(fetchRepoFiles(settings, machineId, repo))
}
} catch (e: ApiException) {
LoadState.failed(e)
}
}
}
fun download(repo: String, file: String) = act {
startDownload(settings, machineId, repo, file)
}
fun cancel(key: String) = act { cancelDownload(settings, machineId, key) }
fun remove(key: String) = act { deleteModel(settings, machineId, key) }
}
/**
* One machine's models, asked for again while this screen is open.
*
* Polled rather than pushed: a download belongs to a machine, not to any session, so it has no
* event stream of its own. Faster while something is downloading, because that is the only thing
* here that changes by itself -- each ask is a round trip to that machine, and once a minute would
* be a progress bar that moved in jumps.
*
* [onLocalChange] fires when the set of models on the machine changes, which is how the screen
* around this learns that a download has become a model it must now draw settings for.
*
* [enabled] is false for a provider that holds no files of its own -- the Claude CLI names its
* models rather than storing them -- and then nothing is asked of the machine at all. Taken as a
* parameter rather than decided by the caller's `if`, so that this is composed unconditionally and
* keeps its search results across the moment the provider's kind arrives.
*/
@Composable
fun rememberMachineModels(
settings: ServerSettings,
machineId: String,
enabled: Boolean,
onLocalChange: () -> Unit,
): MachineModelsState {
val scope = rememberCoroutineScope()
val state = remember(settings, machineId) { MachineModelsState(settings, machineId, scope) }
LaunchedEffect(state, enabled) {
if (!enabled) return@LaunchedEffect
var known: List<String>? = null
while (true) {
state.reload()
val local = state.models?.local?.map { it.key }
if (local != null) {
if (known != null && known != local) onLocalChange()
known = local
}
delay(if (state.downloads.any { it.state == "running" }) 1500 else 5000)
}
}
return state
}
/** What is being fetched onto this machine, above the models it already has. */
fun LazyListScope.downloadCards(state: MachineModelsState) {
uniqueItems(state.downloads, key = { "download:" + it.key }) { download ->
DownloadCard(
download = download,
onCancel = { state.cancel(download.key) },
onResume = { state.download(download.repo, download.file) },
onRemove = { state.remove(download.key) },
)
}
}
/**
* Finding a model to fetch: a search, and what it found.
*
* Below the models this machine has rather than above them, because what is here is what the reader
* came for and getting another is the rarer errand.
*/
fun LazyListScope.modelSearch(state: MachineModelsState) {
item("search") {
Spacer(Modifier.height(16.dp))
Text("Get another model", style = MaterialTheme.typography.titleSmall)
Text(
"Downloaded onto this machine, which is where llama.cpp reads it from.",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
Spacer(Modifier.height(8.dp))
val keyboard = LocalSoftwareKeyboardController.current
LabelledField(
label = "Search HuggingFace",
value = state.query,
onValueChange = { state.query = it },
// The keyboard's own key searches, and puts itself away to show what it found. The
// button below this is under the keyboard while it is up, so without this the only
// way to press it is to dismiss the keyboard first -- which nothing on screen says.
keyboardOptions = KeyboardOptions(imeAction = ImeAction.Search),
keyboardActions =
KeyboardActions(
onSearch = {
keyboard?.hide()
state.search()
}
),
modifier = Modifier.fillMaxWidth(),
)
TextButton(
enabled = state.query.isNotBlank(),
onClick = {
keyboard?.hide()
state.search()
},
) {
Text("Search")
}
}
when (val found = state.results) {
null -> {}
is LoadState.Loading -> item("searching") { CircularProgressIndicator() }
is LoadState.Error ->
item("search-failed") { Text(found.message, color = MaterialTheme.colorScheme.error) }
is LoadState.Loaded ->
uniqueItems(found.value, key = { "repo:" + it.id }) { repo ->
val open = state.openRepo == repo.id
RepoRow(repo, expanded = open) { state.toggleRepo(repo.id) }
// Inside the expanded repository's own item rather than as a section after the
// list: drawn after every card, a repository's files read as belonging to
// whichever card happened to be last.
if (open) {
when (val files = state.repoFiles) {
null -> {}
is LoadState.Loading -> CircularProgressIndicator()
is LoadState.Error ->
Text(files.message, color = MaterialTheme.colorScheme.error)
is LoadState.Loaded ->
Column {
val busy = state.downloads.map { it.key }.toSet()
files.value.forEach { file ->
RepoFileRow(
file,
downloading = "${repo.id}/${file.path}" in busy,
) {
state.download(repo.id, file.path)
}
}
}
}
}
}
}
}
@Composable
private fun DownloadCard(
download: Download,
onCancel: () -> Unit,
onResume: () -> Unit,
onRemove: () -> Unit,
) {
val running = download.state == "running" || download.state == "verifying"
Card(Modifier.fillMaxWidth().padding(vertical = 4.dp)) {
Column(Modifier.padding(12.dp)) {
Text(download.file, style = MaterialTheme.typography.titleSmall)
Text(
download.repo,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
Spacer(Modifier.height(8.dp))
// A determinate bar only when the size is known. HuggingFace sends no size when it
// was never told one, and a bar drawn from a guess is worse than one that admits it
// is counting.
if (download.total != null && download.total > 0) {
LinearProgressIndicator(
progress = { download.done.toFloat() / download.total.toFloat() },
// Blue at every value, unlike a quota bar: a download nearing its end is
// nearing success, and colouring it like a limit being approached would say
// the opposite.
color = progressColor,
modifier = Modifier.fillMaxWidth(),
)
Text(
"${gigabytes(download.done)} of ${gigabytes(download.total)}",
style = MaterialTheme.typography.bodySmall,
)
} else if (running) {
LinearProgressIndicator(color = progressColor, modifier = Modifier.fillMaxWidth())
Text(
"${gigabytes(download.done)} so far, total size unknown",
style = MaterialTheme.typography.bodySmall,
)
}
download.error?.let {
Text(
it,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.error,
)
}
Row(verticalAlignment = Alignment.CenterVertically) {
Text(
download.state,
style = MaterialTheme.typography.bodySmall,
modifier = Modifier.weight(1f),
)
if (running) {
TextButton(onClick = onCancel) { Text("Cancel") }
} else {
// A stopped download kept its partial file, so carrying on is the cheap
// answer and starting again is not the only one offered.
TextButton(onClick = onResume) { Text("Resume") }
TextButton(onClick = onRemove) { Text("Remove") }
}
}
}
}
}
@Composable
private fun RepoRow(repo: RemoteRepo, expanded: Boolean, onToggle: () -> Unit) {
Card(Modifier.fillMaxWidth().padding(vertical = 4.dp)) {
Row(Modifier.padding(12.dp), verticalAlignment = Alignment.CenterVertically) {
Column(Modifier.weight(1f)) {
Text(
repo.id,
style = MaterialTheme.typography.titleSmall,
maxLines = 1,
// The owner is the part that repeats; the model name at the end is what tells
// two entries apart.
overflow = TextOverflow.StartEllipsis,
)
Text(
"${repo.downloads} downloads · ${repo.likes} likes",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
TextButton(onClick = onToggle) { Text(if (expanded) "Hide" else "Files") }
}
}
}
@Composable
private fun RepoFileRow(file: RemoteFile, downloading: Boolean, onDownload: () -> Unit) {
Row(
Modifier.fillMaxWidth().padding(start = 16.dp, top = 4.dp, bottom = 4.dp),
verticalAlignment = Alignment.CenterVertically,
) {
Column(Modifier.weight(1f)) {
Text(file.path, style = MaterialTheme.typography.bodyMedium)
Text(
gigabytes(file.bytes),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
// Disabled rather than absent, so the row reads the same whether this one is absent,
// already here, or on its way. Offering "Download" for a file that is downloading would be
// a button that does nothing anyone can see.
TextButton(enabled = !file.have && !downloading, onClick = onDownload) {
Text(
when {
file.have -> "Downloaded"
downloading -> "Downloading"
else -> "Download"
}
)
}
}
}
fun gigabytes(bytes: Long): String =
if (bytes >= 1_000_000_000) {
"%.2f GB".format(bytes / 1_000_000_000.0)
} else {
"%.0f MB".format(bytes / 1_000_000.0)
}
app/androidApp/src/main/kotlin/com/example/aiapp/SetupsScreen.kt → app/androidApp/src/main/kotlin/com/example/aiapp/MachinesScreen.kt
+146 -64
View File
@@ -1,5 +1,7 @@
package com.example.aiapp
import androidx.compose.foundation.BorderStroke
import androidx.compose.foundation.clickable
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
@@ -10,9 +12,9 @@ import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.lazy.LazyColumn
import androidx.compose.material3.AlertDialog
import androidx.compose.material3.Card
import androidx.compose.material3.CardDefaults
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.OutlinedTextField
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
@@ -24,6 +26,11 @@ import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.graphics.Color
import androidx.compose.ui.semantics.contentDescription
import androidx.compose.ui.semantics.semantics
import androidx.compose.ui.text.font.FontFamily
import androidx.compose.ui.text.style.TextOverflow
import androidx.compose.ui.unit.dp
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.launch
@@ -37,19 +44,25 @@ import kotlinx.coroutines.withContext
* which is what keeps the enrolled token from being able to introduce commands.
*/
@Composable
fun SetupsScreen(settings: ServerSettings, reloadToken: Int) {
fun MachinesScreen(
settings: ServerSettings,
reloadToken: Int,
/** Opens one provider on one machine -- its settings, and what its server is holding. */
onProvider: (String, String) -> Unit,
) {
val scope = rememberCoroutineScope()
var state by remember { mutableStateOf<LoadState<List<Setup>>>(LoadState.Loading) }
var state by remember { mutableStateOf<LoadState<List<Machine>>>(LoadState.Loading) }
var adding by remember { mutableStateOf(false) }
var renaming by remember { mutableStateOf<Setup?>(null) }
var confirmingDelete by remember { mutableStateOf<Setup?>(null) }
var renaming by remember { mutableStateOf<Machine?>(null) }
var confirmingDelete by remember { mutableStateOf<Machine?>(null) }
var signingIn by remember { mutableStateOf<Pair<Machine, Provider>?>(null) }
var busy by remember { mutableStateOf<String?>(null) }
var actionError by remember { mutableStateOf<String?>(null) }
suspend fun reload() {
state =
try {
withContext(Dispatchers.IO) { LoadState.Loaded(fetchSetups(settings)) }
withContext(Dispatchers.IO) { LoadState.Loaded(fetchMachines(settings)) }
} catch (e: ApiException) {
LoadState.failed(e)
}
@@ -82,19 +95,19 @@ fun SetupsScreen(settings: ServerSettings, reloadToken: Int) {
is LoadState.Error -> Text(current.message, color = MaterialTheme.colorScheme.error)
is LoadState.Loaded ->
LazyColumn(Modifier.fillMaxSize()) {
uniqueItems(current.value, key = { it.id }) { setup ->
SetupCard(
setup = setup,
onRename = { renaming = setup },
uniqueItems(current.value, key = { it.id }) { machine ->
MachineCard(
machine = machine,
onRename = { renaming = machine },
onRediscover = {
scope.launch {
busy = "Asking ${setup.name} what it has…"
busy = "Asking ${machine.name} what it has…"
actionError =
runCatching {
withContext(Dispatchers.IO) {
updateSetup(
updateMachine(
settings,
setup.id,
machine.id,
rediscover = true,
)
}
@@ -105,7 +118,9 @@ fun SetupsScreen(settings: ServerSettings, reloadToken: Int) {
reload()
}
},
onDelete = { confirmingDelete = setup },
onDelete = { confirmingDelete = machine },
onSignIn = { provider -> signingIn = machine to provider },
onProvider = { provider -> onProvider(machine.id, provider.name) },
)
}
}
@@ -113,7 +128,7 @@ fun SetupsScreen(settings: ServerSettings, reloadToken: Int) {
}
if (adding) {
AddSetupDialog(
AddMachineDialog(
onDismiss = { adding = false },
onAdd = { name, ssh ->
adding = false
@@ -121,7 +136,7 @@ fun SetupsScreen(settings: ServerSettings, reloadToken: Int) {
busy = "Asking $name what it has…"
actionError =
runCatching {
withContext(Dispatchers.IO) { addSetup(settings, name, ssh) }
withContext(Dispatchers.IO) { addMachine(settings, name, ssh) }
}
.exceptionOrNull()
?.message
@@ -129,13 +144,13 @@ fun SetupsScreen(settings: ServerSettings, reloadToken: Int) {
reload()
}
},
onTest = { ssh -> withContext(Dispatchers.IO) { probeSetup(settings, ssh) } },
onTest = { ssh -> withContext(Dispatchers.IO) { probeMachine(settings, ssh) } },
)
}
renaming?.let { setup ->
renaming?.let { machine ->
RenameDialog(
setup = setup,
machine = machine,
onDismiss = { renaming = null },
onRename = { name ->
renaming = null
@@ -143,7 +158,7 @@ fun SetupsScreen(settings: ServerSettings, reloadToken: Int) {
actionError =
runCatching {
withContext(Dispatchers.IO) {
updateSetup(settings, setup.id, name = name)
updateMachine(settings, machine.id, name = name)
}
}
.exceptionOrNull()
@@ -154,10 +169,10 @@ fun SetupsScreen(settings: ServerSettings, reloadToken: Int) {
)
}
confirmingDelete?.let { setup ->
confirmingDelete?.let { machine ->
AlertDialog(
onDismissRequest = { confirmingDelete = null },
title = { Text("Remove \"${setup.name}\"?") },
title = { Text("Remove \"${machine.name}\"?") },
text = {
Text(
"The machine is left alone -- this only stops this app offering it. " +
@@ -171,7 +186,9 @@ fun SetupsScreen(settings: ServerSettings, reloadToken: Int) {
scope.launch {
actionError =
runCatching {
withContext(Dispatchers.IO) { deleteSetup(settings, setup.id) }
withContext(Dispatchers.IO) {
deleteMachine(settings, machine.id)
}
}
.exceptionOrNull()
?.message
@@ -187,34 +204,99 @@ fun SetupsScreen(settings: ServerSettings, reloadToken: Int) {
},
)
}
signingIn?.let { (machine, provider) ->
ProviderLoginDialog(
settings = settings,
machineId = machine.id,
machineName = machine.name,
provider = provider.name,
onDismiss = { signingIn = null },
onSignedIn = {
signingIn = null
scope.launch { reload() }
},
)
}
}
@Composable
private fun SetupCard(
setup: Setup,
private fun MachineCard(
machine: Machine,
onRename: () -> Unit,
onRediscover: () -> Unit,
onDelete: () -> Unit,
onSignIn: (Provider) -> Unit,
onProvider: (Provider) -> Unit,
) {
Card(Modifier.fillMaxWidth().padding(vertical = 4.dp)) {
Column(Modifier.padding(12.dp)) {
Text(setup.name, style = MaterialTheme.typography.titleSmall)
Text(machine.name, style = MaterialTheme.typography.titleSmall)
Text(
// Not "this machine": the seeded setup is *called* that, and the card read "this
// Not "this machine": the seeded machine is *called* that, and the card read "this
// machine / this machine".
setup.address ?: "runs where the backend does",
machine.address ?: "runs where the backend does",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
Spacer(Modifier.height(4.dp))
Text(
if (setup.providers.isEmpty()) {
"Nothing found on it. Install something and rediscover."
} else {
setup.providers.joinToString(" · ") { it.name }
},
style = MaterialTheme.typography.bodySmall,
)
if (machine.providers.isEmpty()) {
Text(
"Nothing found on it. Install something and rediscover.",
style = MaterialTheme.typography.bodySmall,
)
} else {
machine.providers.forEach { provider ->
// A card of its own rather than a line of text: a provider is where the
// settings that belong to *this machine* live -- how each of its models is
// loaded, the models themselves, and the server holding them -- and those had
// nowhere to be until one llama-server came to serve every session on a
// machine. Sized by its own padding rather than by whatever control happened
// to be on its row, like the tool call cards it is built after.
Card(
Modifier.fillMaxWidth()
.padding(vertical = 4.dp)
.clickable { onProvider(provider) }
.semantics { contentDescription = "Open ${provider.name}" },
// A border, and the machine card's own surface kept underneath it.
// The tint that was here before is one step along the surface ladder
// from the card it sits in, and two adjacent surfaces render as one flat
// block: these read as lines of text in a box rather than as things to
// open. One cue, and a visible one.
colors = CardDefaults.cardColors(containerColor = Color.Transparent),
border = BorderStroke(1.dp, MaterialTheme.colorScheme.outlineVariant),
) {
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth().padding(12.dp),
) {
Column(Modifier.weight(1f)) {
Text(provider.name, style = MaterialTheme.typography.titleSmall)
// What was actually found, which is the honest second line and
// the one thing here nobody can change. No arrow: a card that
// lifts off the one behind it already reads as something to open,
// and the chevron was the only thing making these look like rows
// of a list.
provider.command?.let {
Text(
it,
style = MaterialTheme.typography.bodySmall,
fontFamily = FontFamily.Monospace,
color = MaterialTheme.colorScheme.onSurfaceVariant,
maxLines = 1,
// A program is identified by its name, which is the tail
// of its path.
overflow = TextOverflow.StartEllipsis,
)
}
}
if (provider.kind == "claude_cli") {
TextButton(onClick = { onSignIn(provider) }) { Text("Sign in") }
}
}
}
}
}
Row(verticalAlignment = Alignment.CenterVertically) {
TextButton(onClick = onRename) { Text("Rename") }
TextButton(onClick = onRediscover) { Text("Rediscover") }
@@ -226,7 +308,7 @@ private fun SetupCard(
}
@Composable
private fun AddSetupDialog(
private fun AddMachineDialog(
onDismiss: () -> Unit,
onAdd: (String, SshDetails?) -> Unit,
onTest: suspend (SshDetails?) -> List<Provider>,
@@ -236,6 +318,7 @@ private fun AddSetupDialog(
var address by remember { mutableStateOf("") }
var identity by remember { mutableStateOf("") }
var attachmentsDir by remember { mutableStateOf("") }
var modelsDir by remember { mutableStateOf("") }
var tested by remember { mutableStateOf<String?>(null) }
var testing by remember { mutableStateOf(false) }
@@ -250,6 +333,7 @@ private fun AddSetupDialog(
port = typedPort,
identityFile = identity.trim().ifEmpty { null },
attachmentsDir = attachmentsDir.trim().ifEmpty { null },
modelsDir = modelsDir.trim().ifEmpty { null },
)
}
@@ -265,33 +349,36 @@ private fun AddSetupDialog(
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
Spacer(Modifier.height(8.dp))
OutlinedTextField(
value = name,
onValueChange = { name = it },
label = { Text("Name") },
singleLine = true,
)
OutlinedTextField(
value = address,
onValueChange = { address = it },
LabelledField(label = "Name", value = name, onValueChange = { name = it })
Spacer(Modifier.height(8.dp))
LabelledField(
// Just the shape. What a blank one means is said once, in the text above this
// form -- repeating it here wrapped the label onto a second line.
label = { Text("user@host[:port]") },
singleLine = true,
label = "user@host[:port]",
value = address,
onValueChange = { address = it },
)
OutlinedTextField(
Spacer(Modifier.height(8.dp))
LabelledField(
label = "Key path on the backend",
value = identity,
onValueChange = { identity = it },
label = { Text("Key path on the backend") },
singleLine = true,
)
// Where a file attached from the phone lands on that machine. Blank means the
// session's own directory, which is what most people want.
OutlinedTextField(
Spacer(Modifier.height(8.dp))
LabelledField(
// Where a file attached from the phone lands on that machine.
label = "Folder for attached files",
value = attachmentsDir,
onValueChange = { attachmentsDir = it },
label = { Text("Folder for attached files (optional)") },
singleLine = true,
hint = "the session's own directory",
)
Spacer(Modifier.height(8.dp))
LabelledField(
// Where that machine's GGUFs are, for a llama.cpp session on it.
label = "Folder for models",
value = modelsDir,
onValueChange = { modelsDir = it },
hint = "the same place this backend keeps its own downloads",
)
tested?.let {
Spacer(Modifier.height(8.dp))
@@ -339,19 +426,14 @@ private fun AddSetupDialog(
}
@Composable
private fun RenameDialog(setup: Setup, onDismiss: () -> Unit, onRename: (String) -> Unit) {
var name by remember { mutableStateOf(setup.name) }
private fun RenameDialog(machine: Machine, onDismiss: () -> Unit, onRename: (String) -> Unit) {
var name by remember { mutableStateOf(machine.name) }
AlertDialog(
onDismissRequest = onDismiss,
title = { Text("Rename") },
text = {
Column {
OutlinedTextField(
value = name,
onValueChange = { name = it },
label = { Text("Name") },
singleLine = true,
)
LabelledField(label = "Name", value = name, onValueChange = { name = it })
Spacer(Modifier.height(8.dp))
Text(
"Sessions already running on it keep working -- they refer to the machine, " +
app/androidApp/src/main/kotlin/com/example/aiapp/MainPanel.kt
+53
View File
@@ -0,0 +1,53 @@
package com.example.aiapp
import androidx.compose.runtime.Composable
import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableIntStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.setValue
/**
* The app's root screen, in the full-width panel [SidePanels] slides over a session from the left.
*
* Not a list of its own but [MainScreen] itself, and the whole width of the screen: what a right
* swipe gets is the screen Back would have got, moved over the session instead of replacing it. The
* session stays composed underneath, with its stream open and its draft and scroll position where
* they were, so swiping the panel back off returns to it for nothing -- where Back and a tap costs
* the whole transcript over the tunnel again.
*
* Tapping the session already open is that same swipe back rather than a fresh screen: reopening it
* would hand [SessionScreen] a new summary for the conversation it is already showing.
*
* [onGone] is the one thing the list can do that this panel cannot survive -- deleting the very
* session it is drawn over. There is nothing left to swipe back into, so that closes the screen.
*/
@Composable
fun MainPanel(
settings: ServerSettings,
sessionId: String,
active: Boolean,
onOpen: (SessionSummary) -> Unit,
onSpawn: () -> Unit,
onImported: (SessionSummary) -> Unit,
onSettings: () -> Unit,
onProvider: (String, String) -> Unit,
onClose: () -> Unit,
onGone: () -> Unit,
) {
// Asked again each time the panel opens: who is working and who is waiting on an answer is
// exactly what changed while the session underneath was being read.
var reloadToken by remember(sessionId) { mutableIntStateOf(0) }
LaunchedEffect(active) { if (active) reloadToken++ }
MainScreen(
settings = settings,
reloadToken = reloadToken,
onOpen = { if (it.id == sessionId) onClose() else onOpen(it) },
onSpawn = onSpawn,
onImported = onImported,
onSettings = onSettings,
onProvider = onProvider,
onDeleted = { if (it == sessionId) onGone() },
)
}
app/androidApp/src/main/kotlin/com/example/aiapp/MainScreen.kt
+19 -10
View File
@@ -26,19 +26,23 @@ import androidx.lifecycle.compose.LocalLifecycleOwner
import androidx.lifecycle.repeatOnLifecycle
/**
* The app's root: one title, and four views of the backend behind it.
* The app's root: one title, and three views of the backend behind it.
*
* These were four screens reached by four words in a row under the title, and the row was already
* full. Tabs say the same thing in less space and say one more thing besides: that these are places
* to be rather than errands to run. Sessions, the machine's importable history, the models on it
* and the machines themselves are all *the same backend*, looked at four ways, and none is a step
* down from another. Settings still is, which is why it stays a pushed screen with its own Back.
* These were screens reached by words in a row under the title, and the row was already full. Tabs
* say the same thing in less space and say one more thing besides: that these are places to be
* rather than errands to run. Sessions, the machine's importable history and the machines
* themselves are all *the same backend*, looked at three ways, and none is a step down from
* another. Settings still is, which is why it stays a pushed screen with its own Back.
*
* Models were a fourth tab until 2026-09-19. They are a machine's models now -- downloaded onto the
* machine that has to serve them -- so they live under that machine's llama.cpp provider, beside
* the settings deciding how each one is loaded. A tab about "the models" was a claim that there is
* one such set, and there is one per machine.
*/
private enum class MainTab(val label: String) {
Sessions("Sessions"),
Import("Import"),
Models("Models"),
Setups("Setups"),
Machines("Machines"),
}
@Composable
@@ -51,6 +55,10 @@ fun MainScreen(
onSpawn: () -> Unit,
onImported: (SessionSummary) -> Unit,
onSettings: () -> Unit,
/** One machine's provider, opened from the machines tab. */
onProvider: (String, String) -> Unit,
/** A session the list has just deleted; see [SessionListScreen]. */
onDeleted: (String) -> Unit = {},
) {
var tab by remember { mutableStateOf(MainTab.Sessions) }
var refreshToken by remember { mutableIntStateOf(0) }
@@ -140,11 +148,12 @@ fun MainScreen(
reloadToken = token,
onOpen = onOpen,
onSpawn = onSpawn,
onDeleted = onDeleted,
)
MainTab.Import ->
ImportScreen(settings = settings, reloadToken = token, onImported = onImported)
MainTab.Models -> ModelsScreen(settings = settings, reloadToken = token)
MainTab.Setups -> SetupsScreen(settings = settings, reloadToken = token)
MainTab.Machines ->
MachinesScreen(settings = settings, reloadToken = token, onProvider = onProvider)
}
}
}
app/androidApp/src/main/kotlin/com/example/aiapp/MarkdownLinks.kt
+52 -1
View File
@@ -32,6 +32,7 @@ import com.mikepenz.markdown.model.markdownAnnotator
import com.mikepenz.markdown.utils.getUnescapedTextInNode
import com.mikepenz.markdown.utils.resolveImageAlt
import com.mikepenz.markdown.utils.resolveImageLink
import java.net.URI
import org.intellij.markdown.MarkdownElementTypes
import org.intellij.markdown.MarkdownTokenTypes
import org.intellij.markdown.ast.ASTNode
@@ -89,6 +90,7 @@ fun LinkedText(content: String, node: ASTNode, style: TextStyle, modifier: Modif
content.buildMarkdownAnnotatedString(node, style, settings)
}
val uriHandler = LocalUriHandler.current
val fileLinkHandler = LocalFileLinkHandler.current
val onPlainTap = LocalMarkdownTap.current
val layout = remember { Ref<TextLayoutResult>() }
// The renderer's own rule for a style that names no colour: the theme's text colour.
@@ -127,7 +129,7 @@ fun LinkedText(content: String, node: ASTNode, style: TextStyle, modifier: Modif
when {
url != null -> {
up.consume()
uriHandler.openUri(url)
if (fileLinkHandler?.invoke(url) != true) uriHandler.openUri(url)
}
onPlainTap != null -> {
up.consume()
@@ -164,6 +166,55 @@ fun LinkedText(content: String, node: ASTNode, style: TextStyle, modifier: Modif
*/
val LocalMarkdownTap = compositionLocalOf<(() -> Unit)?> { null }
/**
* Opens a markdown destination inside the current session when it names a file on that session's
* machine. Null outside a session, where every link keeps its ordinary URI behaviour.
*/
val LocalFileLinkHandler = compositionLocalOf<((String) -> Boolean)?> { null }
/**
* A stable markdown link handler whose behaviour follows the latest [onFile]. Keeping its identity
* stable matters: every visible markdown paragraph reads it, and a session recomposes on every
* streamed event.
*/
@Composable
fun rememberFileLinkHandler(onFile: (String) -> Unit): (String) -> Boolean {
val latest = rememberUpdatedState(onFile)
return remember {
{ destination ->
val path = filePathOf(destination)
if (path == null) false
else {
latest.value(path)
true
}
}
}
}
/**
* The path named by a local-file markdown destination.
*
* Only absolute paths and local `file:` URIs are claimed. A relative destination might be a web
* link, and sending one to a machine's filesystem would silently give an ordinary link a different
* meaning. Editors commonly append a line and optional column; the current viewer opens the file
* itself, so those coordinates are removed here.
*/
internal fun filePathOf(destination: String): String? {
val uri = runCatching { URI(destination) }.getOrNull()
val path =
when {
destination.startsWith("/") && !destination.startsWith("//") ->
uri?.path ?: destination.substringBefore('#').substringBefore('?')
uri != null &&
uri.scheme.equals("file", ignoreCase = true) &&
(uri.host.isNullOrEmpty() || uri.host == "localhost") -> uri.path
else -> null
}
if (path.isNullOrEmpty() || !path.startsWith('/')) return null
return path.replace(Regex(":\\d+(?::\\d+)?$"), "")
}
/**
* [onTap] as a stable value to provide for [LocalMarkdownTap]. The identity stays put while the
* behaviour follows the latest [onTap], which is what keeps providing it from invalidating the text
app/androidApp/src/main/kotlin/com/example/aiapp/ModelName.kt
+13
View File
@@ -21,12 +21,25 @@ const val DEFAULT_MODEL = "default"
* one model rather than one model from another. Anything that does not look like that is returned
* untouched.
*
* A llama.cpp session's model is not an identifier at all -- it is `owner/repo/file.gguf`, where
* the file was downloaded from -- so what is kept is the file, which is the part that tells two
* models apart, and the extension goes with the directories. The model's *own* name is better still
* and is not derivable here: it is inside the file, and only the server has ever opened it. Where a
* screen has the server's answer it should prefer it; this is the floor under every screen that
* does not.
*
* A display decision, not a correction: the full name is what the session reports.
*/
fun modelLabel(model: String?): String {
val name = model?.takeIf { it.isNotBlank() } ?: return DEFAULT_MODEL
if (name.endsWith(GGUF)) {
return name.substringAfterLast('/').removeSuffix(GGUF)
}
return name.removePrefix("claude-").replace(DATED_SUFFIX, "")
}
/** A trailing `-YYYYMMDD`, which is how these identifiers carry their release date. */
private val DATED_SUFFIX = Regex("""-\d{8}$""")
/** What every model a llama.cpp session can run is stored as. */
private const val GGUF = ".gguf"
app/androidApp/src/main/kotlin/com/example/aiapp/ModelsScreen.kt
-374
View File
@@ -1,374 +0,0 @@
package com.example.aiapp
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.fillMaxSize
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.height
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.lazy.LazyColumn
import androidx.compose.material3.Card
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.LinearProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.OutlinedTextField
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.text.style.TextOverflow
import androidx.compose.ui.unit.dp
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.delay
import kotlinx.coroutines.launch
import kotlinx.coroutines.withContext
/**
* Models on the backend, and HuggingFace to get more from.
*
* Everything here is the server's state rather than this screen's: what is downloaded, and what is
* downloading, are the same answers on every enrolled device, and a download started here keeps
* going when this screen closes.
*/
@Composable
fun ModelsScreen(settings: ServerSettings, reloadToken: Int) {
val scope = rememberCoroutineScope()
var state by remember { mutableStateOf<LoadState<Models>>(LoadState.Loading) }
var query by remember { mutableStateOf("") }
var results by remember { mutableStateOf<LoadState<List<RemoteRepo>>?>(null) }
var openRepo by remember { mutableStateOf<String?>(null) }
var repoFiles by remember { mutableStateOf<LoadState<List<RemoteFile>>?>(null) }
var actionError by remember { mutableStateOf<String?>(null) }
suspend fun reload() {
state =
try {
withContext(Dispatchers.IO) { LoadState.Loaded(fetchModels(settings)) }
} catch (e: ApiException) {
LoadState.failed(e)
}
}
// Polled rather than pushed: a download belongs to the machine, not to any session, so it has
// no event stream of its own. Keyed on the token as well, so the header's Refresh restarts the
// loop with a read now rather than leaving the reader watching for a second and a half.
LaunchedEffect(reloadToken) {
while (true) {
reload()
delay(1500)
}
}
Column(Modifier.fillMaxSize().padding(16.dp)) {
actionError?.let {
Text(it, color = MaterialTheme.colorScheme.error)
Spacer(Modifier.height(8.dp))
}
OutlinedTextField(
value = query,
onValueChange = { query = it },
label = { Text("Search HuggingFace") },
singleLine = true,
modifier = Modifier.fillMaxWidth(),
)
Spacer(Modifier.height(8.dp))
TextButton(
enabled = query.isNotBlank(),
onClick = {
openRepo = null
results = LoadState.Loading
scope.launch {
results =
try {
withContext(Dispatchers.IO) {
LoadState.Loaded(searchModels(settings, query))
}
} catch (e: ApiException) {
LoadState.failed(e)
}
}
},
) {
Text("Search")
}
Spacer(Modifier.height(8.dp))
LazyColumn(Modifier.fillMaxSize()) {
when (val current = state) {
is LoadState.Loading -> item { CircularProgressIndicator() }
is LoadState.Error ->
item { Text(current.message, color = MaterialTheme.colorScheme.error) }
is LoadState.Loaded -> {
if (current.value.downloads.isNotEmpty()) {
item { SectionLabel("Downloading") }
uniqueItems(current.value.downloads, key = { it.key + it.run }) { download
->
DownloadCard(download) {
scope.launch {
actionError =
runCatching {
withContext(Dispatchers.IO) {
cancelDownload(settings, download.key)
}
}
.exceptionOrNull()
?.message
}
}
}
}
item { SectionLabel("On the backend") }
if (current.value.local.isEmpty()) {
item {
Text(
"None yet. Search above to find one.",
style = MaterialTheme.typography.bodyMedium,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
}
uniqueItems(current.value.local, key = { it.key }) { model ->
LocalModelCard(model) {
scope.launch {
actionError =
runCatching {
withContext(Dispatchers.IO) {
deleteModel(settings, model.key)
}
}
.exceptionOrNull()
?.message
reload()
}
}
}
}
}
results?.let { found ->
item { SectionLabel("HuggingFace") }
when (found) {
is LoadState.Loading -> item { CircularProgressIndicator() }
is LoadState.Error ->
item { Text(found.message, color = MaterialTheme.colorScheme.error) }
is LoadState.Loaded ->
uniqueItems(found.value, key = { it.id }) { repo ->
val open = openRepo == repo.id
RepoRow(repo, expanded = open) {
if (open) {
openRepo = null
} else {
openRepo = repo.id
repoFiles = LoadState.Loading
scope.launch {
repoFiles =
try {
withContext(Dispatchers.IO) {
LoadState.Loaded(
fetchRepoFiles(settings, repo.id)
)
}
} catch (e: ApiException) {
LoadState.failed(e)
}
}
}
}
// Inside the expanded repository's own item rather than as a section
// after the list: drawn after every card, a repository's files read as
// belonging to whichever card happened to be last.
if (open) {
when (val files = repoFiles) {
null -> {}
is LoadState.Loading -> CircularProgressIndicator()
is LoadState.Error ->
Text(files.message, color = MaterialTheme.colorScheme.error)
is LoadState.Loaded ->
Column {
val busy =
(state as? LoadState.Loaded)
?.value
?.downloads
.orEmpty()
.filter { it.state == "running" }
.map { it.key }
.toSet()
files.value.forEach { file ->
RepoFileRow(
file,
downloading = "${repo.id}/${file.path}" in busy,
) {
scope.launch {
actionError =
runCatching {
withContext(Dispatchers.IO) {
startDownload(
settings,
repo.id,
file.path,
)
}
}
.exceptionOrNull()
?.message
reload()
}
}
}
}
}
}
}
}
}
}
}
}
@Composable
private fun SectionLabel(text: String) {
Spacer(Modifier.height(12.dp))
Text(text, style = MaterialTheme.typography.titleSmall)
Spacer(Modifier.height(4.dp))
}
@Composable
private fun DownloadCard(download: Download, onCancel: () -> Unit) {
Card(Modifier.fillMaxWidth().padding(vertical = 4.dp)) {
Column(Modifier.padding(12.dp)) {
Text(download.file, style = MaterialTheme.typography.titleSmall)
Text(
download.repo,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
Spacer(Modifier.height(8.dp))
// A determinate bar only when the size is known. The server sends no total when it was
// never told one, and a bar drawn from a guess is worse than one that admits it is
// counting.
if (download.total != null && download.total > 0) {
LinearProgressIndicator(
progress = { download.done.toFloat() / download.total.toFloat() },
// Blue at every value, unlike a quota bar: a download nearing its end is
// nearing success, and colouring it like a limit being approached would say the
// opposite.
color = progressColor,
modifier = Modifier.fillMaxWidth(),
)
Text(
"${gigabytes(download.done)} of ${gigabytes(download.total)}",
style = MaterialTheme.typography.bodySmall,
)
} else {
LinearProgressIndicator(color = progressColor, modifier = Modifier.fillMaxWidth())
Text(
"${gigabytes(download.done)} so far, total size unknown",
style = MaterialTheme.typography.bodySmall,
)
}
download.error?.let {
Text(
it,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.error,
)
}
Row {
Text(
download.state,
style = MaterialTheme.typography.bodySmall,
modifier = Modifier.weight(1f),
)
if (download.state == "running") {
TextButton(onClick = onCancel) { Text("Cancel") }
}
}
}
}
}
@Composable
private fun LocalModelCard(model: LocalModel, onDelete: () -> Unit) {
Card(Modifier.fillMaxWidth().padding(vertical = 4.dp)) {
Row(Modifier.padding(12.dp), verticalAlignment = Alignment.CenterVertically) {
Column(Modifier.weight(1f)) {
Text(model.file, style = MaterialTheme.typography.titleSmall)
Text(
"${model.repo} · ${gigabytes(model.bytes)}",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
TextButton(onClick = onDelete) { Text("Delete") }
}
}
}
@Composable
private fun RepoRow(repo: RemoteRepo, expanded: Boolean, onToggle: () -> Unit) {
Card(Modifier.fillMaxWidth().padding(vertical = 4.dp)) {
Row(Modifier.padding(12.dp), verticalAlignment = Alignment.CenterVertically) {
Column(Modifier.weight(1f)) {
Text(
repo.id,
style = MaterialTheme.typography.titleSmall,
maxLines = 1,
// The owner is the part that repeats; the model name at the end is what tells
// two entries apart.
overflow = TextOverflow.StartEllipsis,
)
Text(
"${repo.downloads} downloads · ${repo.likes} likes",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
TextButton(onClick = onToggle) { Text(if (expanded) "Hide" else "Files") }
}
}
}
@Composable
private fun RepoFileRow(file: RemoteFile, downloading: Boolean, onDownload: () -> Unit) {
Row(
Modifier.fillMaxWidth().padding(start = 16.dp, top = 4.dp, bottom = 4.dp),
verticalAlignment = Alignment.CenterVertically,
) {
Column(Modifier.weight(1f)) {
Text(file.path, style = MaterialTheme.typography.bodyMedium)
Text(
gigabytes(file.bytes),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
// Disabled rather than absent, so the row reads the same whether this one is absent,
// already here, or on its way. Offering "Download" for a file that is downloading would be
// a button that does nothing anyone can see.
TextButton(enabled = !file.have && !downloading, onClick = onDownload) {
Text(
when {
file.have -> "Downloaded"
downloading -> "Downloading"
else -> "Download"
}
)
}
}
}
private fun gigabytes(bytes: Long): String =
if (bytes >= 1_000_000_000) {
"%.2f GB".format(bytes / 1_000_000_000.0)
} else {
"%.0f MB".format(bytes / 1_000_000.0)
}
app/androidApp/src/main/kotlin/com/example/aiapp/NerdIcons.kt
+30 -2
View File
@@ -28,7 +28,7 @@ import androidx.compose.ui.unit.sp
* This replaced a hand-drawn canvas gear, whose doc comment argued against icon fonts on the
* grounds that a system font may not have the glyph. That objection is about *relying* on a system
* font, and it is exactly right: the answer is not to avoid glyphs but to ship them. The font here
* is `app/build-icon-font.sh`'s output -- seventeen glyphs, 2.8 KB, subset out of the 3 MB symbols
* is `app/build-icon-font.sh`'s output -- eighteen glyphs, 2.9 KB, subset out of the 3 MB symbols
* font and committed. Adding one means adding its codepoint in *both* places; a codepoint here that
* the script did not subset is a glyph that silently isn't there.
*
@@ -136,6 +136,34 @@ val EDIT_GLYPH = glyph(0xF03EB)
*/
val SAVE_GLYPH = glyph(0xF0193)
/**
* `md-menu` -- the burger: three stacked rules, drawn as the handle a row is dragged by.
*
* The mark for "take hold of this and move it" rather than for a menu, which is what it means on a
* row that has one: three rules look like the rows of a list, and the only thing here that draws
* them is a list being rearranged. Nothing else in this app opens a menu from a burger, so the two
* senses cannot be confused.
*/
val DRAG_GLYPH = glyph(0xF035C)
/**
* `md-console_line` -- a shell prompt: a backgrounded command, in the panel beside the turn.
*
* The four marks here are one set, drawn by `backgroundTaskLook`: they exist because the kind of a
* background task used to be a word on a line of its own, which on a list of commands was the same
* two words down the whole panel. Each keeps its words as the description a screen reader is given.
*/
val COMMAND_GLYPH = glyph(0xF07B7)
/** `md-robot` -- a subagent: something running that is doing its own reasoning. */
val AGENT_GLYPH = glyph(0xF06A9)
/** `md-sitemap` -- a workflow: steps arranged by something other than the agent itself. */
val WORKFLOW_GLYPH = glyph(0xF04AA)
/** `md-help_circle_outline` -- a background task of a kind this build has not heard of. */
val UNKNOWN_GLYPH = glyph(0xF0625)
/**
* The size an icon draws at beside a line of text.
*
@@ -165,7 +193,7 @@ private val GLYPH_EXTENT = GLYPH_SIZE.value.dp
* own corners and beside a title it arrived at the first letter. And it is taller than any header's
* text, which is what lets the button fill a header row rather than sit in the middle of one.
*/
private val GLYPH_BUTTON_SIZE = 48.dp
val GLYPH_BUTTON_SIZE = 48.dp
/**
* The ring itself, for putting something that is *not* a glyph button next to one -- a title beside
app/androidApp/src/main/kotlin/com/example/aiapp/Notifications.kt
+15 -6
View File
@@ -34,6 +34,8 @@ import org.json.JSONObject
* gets a push from Google's servers, which would mean this backend talking to Google about
* somebody's coding sessions, and the whole point of the tunnel is that it does not.
*
* Every moment it hears about goes to the drawer; [show] decides what else is done with it.
*
* The cost Android charges is a notification of its own that cannot be dismissed. That is made as
* quiet as the platform allows: [ONGOING_CHANNEL] is `IMPORTANCE_MIN`, so it makes no sound, shows
* no status-bar icon, and sits at the bottom of the shade. It is not hidden outright, because it
@@ -138,10 +140,11 @@ class NotificationService : Service() {
// Nothing to tell somebody about the session they are reading. The transcript in front of
// them is already saying it.
if (isOnScreen(notification.sessionId)) return
// The app is up: it says this itself, as a banner over whatever screen they are on. Never
// both -- one thing happened, and a drawer filling up behind an app that already showed you
// each one is a drawer nobody reads.
if (handOver(notification)) return
// The app is up, so it says this itself as a banner over whatever screen they are on --
// which interrupts, where the drawer's row records: a banner lasts seconds and reaches only
// somebody already looking. Both go up, and the banner having done the interrupting is what
// makes the row a silent one.
val banner = handOver(notification)
val manager = NotificationManagerCompat.from(this)
// Two different noes, and both are answers rather than faults: the runtime permission
// refused, and notifications switched off for the app in Android's own settings.
@@ -172,6 +175,7 @@ class NotificationService : Service() {
.setAutoCancel(true)
.setWhen((notification.at * 1000).toLong())
.setShowWhen(true)
.setSilent(banner)
.build()
manager.notify(notification.sessionId, ALERT_ID, built)
}
@@ -262,7 +266,8 @@ class NotificationService : Service() {
* Whether there is an app to reach is the subscriber count rather than a flag of its own:
* [SessionAlerts] collects this exactly while it is on screen. `tryEmit` neither suspends
* nor blocks the thread reading the stream, and the buffer is there so a handful of
* sessions finishing together all land rather than the last one winning.
* sessions finishing together all land rather than the last one winning. Reaching the app
* does not stop the drawer's row; it makes it a silent one.
*/
private val toApp = MutableSharedFlow<SessionNotification>(extraBufferCapacity = 8)
@@ -272,7 +277,11 @@ class NotificationService : Service() {
private fun handOver(notification: SessionNotification) =
toApp.subscriptionCount.value > 0 && toApp.tryEmit(notification)
/** Somebody is looking at [sessionId]; nothing is posted about it until they stop. */
/**
* Somebody is looking at [sessionId]; nothing is posted about it until they stop, and
* whatever the drawer is already holding about it goes now rather than waiting to be swiped
* away. Opening the session *is* reading the notification, whichever way they got here.
*/
fun showing(context: Context, sessionId: String) {
onScreen = sessionId
// Whatever was posted about it before is about to be read, so it has nothing left to
app/androidApp/src/main/kotlin/com/example/aiapp/PendingMessages.kt
+121
View File
@@ -0,0 +1,121 @@
package com.example.aiapp
import android.content.Context
import androidx.core.content.edit
import java.util.UUID
import org.json.JSONArray
import org.json.JSONObject
private const val PENDING_MESSAGES = "pending-messages"
/** A quiet user bubble below the durable transcript. */
internal data class QueuedMessage(
val id: String,
val text: String,
val attachments: List<String>,
val refusal: String? = null,
/** This phone is still waiting for any durable event that says the server accepted it. */
val local: Boolean = false,
/** The HTTP request returned successfully; the provider event is still outstanding. */
val serverAccepted: Boolean = false,
)
internal fun localPendingMessage(text: String, attachments: List<String>) =
QueuedMessage("local-${UUID.randomUUID()}", text, attachments, local = true)
private fun QueuedMessage.matches(text: String, attachments: List<String>) =
this.text == text && this.attachments == attachments
/** Replaces the local bridge with the server's durable waiting message, without drawing both. */
internal fun reconcileQueuedMessage(
queued: List<QueuedMessage>,
event: SessionEvent.MessageQueued,
): List<QueuedMessage> {
if (queued.any { !it.local && it.id == event.id }) return queued
val at = queued.indexOfFirst { it.local && it.matches(event.text, event.attachments) }
if (at < 0) return queued + QueuedMessage(event.id, event.text, event.attachments)
return queued.mapIndexed { index, message ->
if (index == at) QueuedMessage(event.id, event.text, event.attachments) else message
}
}
/** Removes exactly the pending bubble that became a provider-received user message. */
internal fun reconcileUserMessage(
queued: List<QueuedMessage>,
event: SessionEvent.UserMessage,
): List<QueuedMessage> {
val at =
event.id?.let { id -> queued.indexOfFirst { !it.local && it.id == id }.takeIf { it >= 0 } }
?: queued.indexOfFirst { it.local && it.matches(event.text, event.attachments) }
return if (at < 0) queued else queued.filterIndexed { index, _ -> index != at }
}
/** Keeps a failed send in place and puts its actionable failure in that message's bubble. */
internal fun markPendingFailure(
queued: List<QueuedMessage>,
id: String,
failure: String,
): List<QueuedMessage> = queued.map { message ->
if (message.local && message.id == id) message.copy(refusal = failure) else message
}
/** Stops persisting a send once the server owns it, while its bubble awaits the provider event. */
internal fun markPendingAccepted(queued: List<QueuedMessage>, id: String): List<QueuedMessage> =
queued.map { message ->
if (message.local && message.id == id) message.copy(serverAccepted = true) else message
}
internal fun discardPendingMessage(
queued: List<QueuedMessage>,
id: String,
): List<QueuedMessage> = queued.filterNot { it.local && it.id == id }
/** Restores sends for which this phone has not yet seen a durable server event. */
internal fun loadPendingMessages(context: Context, key: String): List<QueuedMessage> {
val encoded =
context.getSharedPreferences(PENDING_MESSAGES, Context.MODE_PRIVATE).getString(key, null)
?: return emptyList()
return try {
val messages = JSONArray(encoded)
List(messages.length()) { index ->
val message = messages.getJSONObject(index)
val attachments = message.optJSONArray("attachments") ?: JSONArray()
QueuedMessage(
id = message.getString("id"),
text = message.getString("text"),
attachments = List(attachments.length()) { attachments.getString(it) },
refusal = message.optString("refusal").takeIf { it.isNotEmpty() },
local = true,
)
}
} catch (_: org.json.JSONException) {
// A corrupt local outbox is not useful on the next open either. Remove it rather than
// repeatedly pretending it decoded to an intentionally empty one.
context.getSharedPreferences(PENDING_MESSAGES, Context.MODE_PRIVATE).edit { remove(key) }
emptyList()
}
}
/** Stores only sends the server has not confirmed; everything accepted is the server's to keep. */
internal fun savePendingMessages(context: Context, key: String, queued: List<QueuedMessage>) {
val local = queued.filter { it.local && !it.serverAccepted }
context.getSharedPreferences(PENDING_MESSAGES, Context.MODE_PRIVATE).edit {
if (local.isEmpty()) {
remove(key)
} else {
putString(
key,
JSONArray(
local.map { message ->
JSONObject()
.put("id", message.id)
.put("text", message.text)
.put("attachments", JSONArray(message.attachments))
.put("refusal", message.refusal ?: "")
}
)
.toString(),
)
}
}
}
app/androidApp/src/main/kotlin/com/example/aiapp/ProviderLoginDialog.kt
+211
View File
@@ -0,0 +1,211 @@
package com.example.aiapp
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.height
import androidx.compose.material3.AlertDialog
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableIntStateOf
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.platform.LocalUriHandler
import androidx.compose.ui.unit.dp
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.delay
import kotlinx.coroutines.launch
import kotlinx.coroutines.withContext
/**
* Relays a provider CLI's headless browser login without ever owning its credentials.
*
* The URL and code live only in this composition. The CLI process on [machineId] remains the one
* OAuth client and the only writer of its credential file.
*/
@Composable
fun ProviderLoginDialog(
settings: ServerSettings,
machineId: String,
machineName: String,
provider: String,
onDismiss: () -> Unit,
onSignedIn: () -> Unit,
) {
val scope = rememberCoroutineScope()
val uriHandler = LocalUriHandler.current
var login by remember(machineId, provider) { mutableStateOf<ProviderLogin?>(null) }
var code by remember(machineId, provider) { mutableStateOf("") }
var error by remember(machineId, provider) { mutableStateOf<String?>(null) }
var retry by remember(machineId, provider) { mutableIntStateOf(0) }
suspend fun follow(initial: ProviderLogin): ProviderLogin {
var current = initial
val wasSubmitting = initial.state == "submitting"
while (current.state == "starting" || current.state == "submitting") {
delay(400)
current =
withContext(Dispatchers.IO) {
fetchProviderLogin(
settings,
machineId,
provider,
current.attempt,
)
}
login = current
}
if (wasSubmitting && current.state == "waitingForCode" && current.detail == null) {
current =
current.copy(
detail = "That code was not accepted. Copy the complete code and try again."
)
login = current
}
return current
}
LaunchedEffect(machineId, provider, retry) {
error = null
code = ""
login = null
try {
val started =
withContext(Dispatchers.IO) { startProviderLogin(settings, machineId, provider) }
login = started
if (follow(started).state == "succeeded") {
onSignedIn()
}
} catch (e: ApiException) {
error = e.message
}
}
fun dismiss() {
login
?.takeUnless { it.state in setOf("succeeded", "failed", "cancelled") }
?.let {
scope.launch(Dispatchers.IO) {
runCatching { cancelProviderLogin(settings, machineId, provider, it.attempt) }
}
}
onDismiss()
}
AlertDialog(
onDismissRequest = ::dismiss,
title = { Text("Sign in to Claude") },
text = {
Column {
Text(
"Claude will sign in on $machineName. Open the authorization page, then " +
"paste the code it gives you here."
)
Spacer(Modifier.height(12.dp))
when (val current = login) {
null ->
if (error == null) {
Row(verticalAlignment = Alignment.CenterVertically) {
CircularProgressIndicator()
Text("Starting sign-in…")
}
}
else ->
when (current.state) {
"starting",
"submitting" ->
Row(verticalAlignment = Alignment.CenterVertically) {
CircularProgressIndicator()
Text(
if (current.state == "submitting") "Checking code…"
else "Starting sign-in…"
)
}
"waitingForCode" -> {
TextButton(
onClick = {
runCatching {
current.authorizationUrl?.let(uriHandler::openUri)
}
.onFailure {
error = "Couldn't open the authorization page."
}
},
enabled = current.authorizationUrl != null,
) {
Text("Open authorization page")
}
LabelledField(
label = "Authorization code",
value = code,
onValueChange = { code = it },
)
current.detail?.let {
Text(it, color = MaterialTheme.colorScheme.error)
}
}
"succeeded" -> Text("Signed in on $machineName.")
"cancelled" -> Text("Sign-in was cancelled.")
else ->
Text(
current.detail ?: "Sign-in failed.",
color = MaterialTheme.colorScheme.error,
)
}
}
error?.let { Text(it, color = MaterialTheme.colorScheme.error) }
}
},
confirmButton = {
val current = login
when {
current?.state == "waitingForCode" ->
TextButton(
onClick = {
scope.launch {
error = null
try {
val submitted =
withContext(Dispatchers.IO) {
submitProviderLoginCode(
settings,
machineId,
provider,
current.attempt,
code,
)
}
login = submitted
if (follow(submitted).state == "succeeded") {
onSignedIn()
}
} catch (e: ApiException) {
error = e.message
}
}
},
enabled = code.isNotBlank(),
) {
Text("Continue")
}
error != null || current?.state == "failed" || current?.state == "cancelled" ->
TextButton(onClick = { retry++ }) { Text("Try again") }
current?.state == "succeeded" -> TextButton(onClick = onDismiss) { Text("Done") }
}
},
dismissButton = {
if (login?.state != "succeeded") {
TextButton(onClick = ::dismiss) { Text("Cancel") }
}
},
)
}
app/androidApp/src/main/kotlin/com/example/aiapp/ProviderParams.kt
+111
View File
@@ -0,0 +1,111 @@
package com.example.aiapp
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.height
import androidx.compose.foundation.text.KeyboardOptions
import androidx.compose.material3.Text
import androidx.compose.runtime.Composable
import androidx.compose.ui.Modifier
import androidx.compose.ui.text.input.KeyboardType
import androidx.compose.ui.unit.dp
/**
* The controls for whatever settings a provider says it takes.
*
* One composable for both screens that offer them — the spawn form and the session settings dialog
* — and for every provider, because the server declares the list (see `DriverKind::params`) rather
* than this file knowing it. A driver that grows a setting gets a control here with no change to
* the app, which is the whole point: the values that suit one machine ship as defaults, and every
* one of them stays reachable from a phone.
*
* [values] is the whole map and [onChange] hands back the whole map. A key absent from it means the
* setting is unset, which is what every [ParamSpec.unset] describes — so clearing a field and never
* touching it are deliberately the same state.
*/
@Composable
fun ProviderParamFields(
specs: List<ParamSpec>,
values: Map<String, String>,
onChange: (Map<String, String>) -> Unit,
/**
* How a choice is drawn here, which is the screen's to decide rather than the setting's: chips
* on a form somebody is filling in, a picker row in a list of settings. A provider's choices
* have to look like the choices beside them, whichever screen that is.
*/
choices: ChoiceStyle,
modifier: Modifier = Modifier,
) {
if (specs.isEmpty()) return
Column(modifier.fillMaxWidth()) {
specs.forEach { spec ->
val set = { value: String ->
onChange(
// Blank clears rather than storing an empty string: the server reads an absent
// key as "use the default", and an empty one would be a value it then failed
// to parse.
if (value.isBlank()) values - spec.key else values + (spec.key to value)
)
}
when (spec.kind) {
"choice" -> {
// The first option is what unset means, so selecting it clears the key — see
// `ParamKind::Choice`. Without that the picker could show a default it could
// not return to.
val default = spec.options.firstOrNull().orEmpty()
val selected = values[spec.key] ?: default
val pick = { chosen: String -> set(if (chosen == default) "" else chosen) }
when (choices) {
ChoiceStyle.Chips ->
ChipGroup(
label = spec.label,
options = spec.options,
selected = selected,
onSelect = pick,
)
ChoiceStyle.Picker -> PickerRow(spec.label, selected, spec.options, pick)
}
}
else ->
LabelledField(
label = spec.label,
value = values[spec.key].orEmpty(),
onValueChange = set,
hint = spec.unset,
// Prose is written rather than filled in, so it gets the room to be read
// back -- see `ParamKind::Prose`.
lines = if (spec.kind == "prose") 4 else 1,
keyboardOptions = KeyboardOptions(keyboardType = keyboardFor(spec.kind)),
)
}
Spacer(Modifier.height(12.dp))
}
}
}
/** Which control a [ParamKind.Choice] gets -- see [ProviderParamFields]'s `choices`. */
enum class ChoiceStyle {
Chips,
Picker,
}
/**
* The keyboard for a value's shape. A number field that opens the letter keyboard is one every
* entry is made harder by, and these are nearly all numbers.
*/
private fun keyboardFor(kind: String): KeyboardType =
when (kind) {
"integer" -> KeyboardType.Number
"decimal" -> KeyboardType.Decimal
else -> KeyboardType.Text
}
/**
* How long typing has to stop before edited settings are sent.
*
* Long enough that a number is one request rather than one per digit, short enough that closing the
* dialog straight after typing still saves — the save runs on the screen behind it, which outlives
* the dialog, so this delay is not a window the value can be lost in.
*/
const val PARAM_SAVE_DELAY_MS = 700L
app/androidApp/src/main/kotlin/com/example/aiapp/ProviderScreen.kt
+509
View File
@@ -0,0 +1,509 @@
package com.example.aiapp
import androidx.compose.foundation.clickable
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.fillMaxSize
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.height
import androidx.compose.foundation.layout.imePadding
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.lazy.LazyColumn
import androidx.compose.foundation.rememberScrollState
import androidx.compose.foundation.text.KeyboardOptions
import androidx.compose.foundation.verticalScroll
import androidx.compose.material3.AlertDialog
import androidx.compose.material3.Card
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableIntStateOf
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.text.font.FontFamily
import androidx.compose.ui.text.input.KeyboardType
import androidx.compose.ui.unit.dp
import androidx.compose.ui.window.DialogProperties
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.launch
import kotlinx.coroutines.withContext
/**
* One provider on one machine: what it is, what its shared server is holding, and how each of its
* models is loaded.
*
* This is where a setting that belongs to a *machine* lives, as opposed to one that belongs to a
* session. The two were one list until llama.cpp sessions came to share one server per machine: how
* a model is loaded stopped being anything a single session could decide, because one copy of it in
* memory is what several sessions are talking to.
*
* It is also the only place a loaded model is taken out of memory. Nothing does that on its own —
* closing a session leaves the model loaded on purpose, since the next one to want it would
* otherwise pay the load again — so the memory is freed here, where what it costs everybody is
* visible.
*/
@Composable
fun ProviderScreen(
settings: ServerSettings,
machineId: String,
provider: String,
/**
* The way back, or null where this is drawn inside something that has one of its own -- the
* session settings screen's second tab. Two ways out stacked above each other is a reader
* asking which of them goes where.
*/
onBack: (() -> Unit)?,
) {
val scope = rememberCoroutineScope()
var state by remember { mutableStateOf<LoadState<ProviderView>>(LoadState.Loading) }
var reload by remember { mutableIntStateOf(0) }
var editing by remember { mutableStateOf<ProviderModel?>(null) }
var confirmingStop by remember { mutableStateOf(false) }
// What is being done to the server or to one of its models, in a word, and what went wrong
// when it did. Both here rather than per row: these act on the whole machine.
var busy by remember { mutableStateOf<String?>(null) }
var actionError by remember { mutableStateOf<String?>(null) }
var confirmingDelete by remember { mutableStateOf<ProviderModel?>(null) }
// The machine's own models and what is being fetched onto it. Only for a provider that serves
// files off that machine's disk -- everything else names its models rather than holding them,
// and a search for a GGUF under the Claude CLI would be an offer that leads nowhere.
val kind = (state as? LoadState.Loaded)?.value?.kind
val machineModels =
rememberMachineModels(
settings = settings,
machineId = machineId,
enabled = kind == "llama_cpp",
// A download that became a model is a model this screen has no settings for yet, so
// the view it is drawing is now one model short of the truth.
onLocalChange = { reload++ },
)
LaunchedEffect(reload) {
state =
try {
withContext(Dispatchers.IO) {
LoadState.Loaded(fetchProvider(settings, machineId, provider))
}
} catch (e: ApiException) {
LoadState.failed(e)
}
}
// Say what is happening, do it, say what went wrong, refetch: every action on this screen
// changes what it is showing.
val act = { what: String, action: suspend () -> Unit ->
scope.launch {
busy = what
actionError =
runCatching { withContext(Dispatchers.IO) { action() } }.exceptionOrNull()?.message
busy = null
reload++
}
Unit
}
// The models search at the bottom takes the keyboard, and everything below the field it is
// typed in -- the Search button, the results -- is behind it without this.
Column(Modifier.fillMaxSize().imePadding().padding(16.dp)) {
onBack?.let {
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
TextButton(onClick = it) { Text("Back") }
}
}
when (val current = state) {
is LoadState.Loading -> CircularProgressIndicator()
is LoadState.Error -> Text(current.message, color = MaterialTheme.colorScheme.error)
is LoadState.Loaded -> {
val view = current.value
Text(view.name, style = MaterialTheme.typography.titleMedium)
Text(
"on ${view.machine}",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
view.command?.let {
Text(
it,
style = MaterialTheme.typography.bodySmall,
fontFamily = FontFamily.Monospace,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
Spacer(Modifier.height(12.dp))
actionError?.let {
Text(it, color = MaterialTheme.colorScheme.error)
Spacer(Modifier.height(8.dp))
}
busy?.let {
Row(verticalAlignment = Alignment.CenterVertically) {
CircularProgressIndicator(Modifier.height(16.dp).padding(end = 8.dp))
Text(it, style = MaterialTheme.typography.bodySmall)
}
Spacer(Modifier.height(8.dp))
}
LazyColumn(Modifier.fillMaxSize()) {
view.server?.let { server ->
item("server") {
ServerCard(
server = server,
maxLoaded = view.maxLoaded,
enabled = busy == null,
onStop = { confirmingStop = true },
onMaxLoaded = { chosen ->
act("Saving…") {
setProviderSettings(
settings,
machineId,
provider,
chosen,
)
}
},
)
Spacer(Modifier.height(12.dp))
}
}
if (view.models.isNotEmpty() && view.modelParams.isNotEmpty()) {
item("models-heading") {
Text("Models", style = MaterialTheme.typography.titleSmall)
Spacer(Modifier.height(8.dp))
}
}
machineModels.actionError?.let { failure ->
item("models-error") {
Text(failure, color = MaterialTheme.colorScheme.error)
}
}
// Above the models: this is what is about to be one of them.
downloadCards(machineModels)
val sizes = machineModels.sizes
uniqueItems(view.models, key = { it.id }) { model ->
ModelCard(
model = model,
specs = view.modelParams,
bytes = sizes[model.id],
onDelete =
if (model.id in sizes) ({ confirmingDelete = model }) else null,
// Tapping opens the settings; a provider whose models take none has
// nothing to open, so the row is not a control.
onEdit =
if (view.modelParams.isEmpty()) null else ({ editing = model }),
onUnload =
if (model.status == "loaded" || model.status == "sleeping") {
{
act("Unloading ${model.label}…") {
unloadProviderModel(
settings,
machineId,
provider,
model.id,
)
}
}
} else null,
enabled = busy == null,
)
}
if (kind == "llama_cpp") modelSearch(machineModels)
if (view.mcpServers.isNotEmpty()) {
item("mcp") {
Spacer(Modifier.height(12.dp))
Text("Tool servers", style = MaterialTheme.typography.titleSmall)
Text(
view.mcpServers.joinToString(", ") +
" — configured on the backend, in its config file.",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
}
}
}
}
}
editing?.let { model ->
val view = (state as? LoadState.Loaded)?.value
ModelSettingsDialog(
model = model,
specs = view?.modelParams.orEmpty(),
onDismiss = { editing = null },
onSave = { params ->
editing = null
act("Saving ${model.label}…") {
setModelSettings(settings, machineId, provider, model.id, params)
}
},
)
}
confirmingDelete?.let { model ->
AlertDialog(
onDismissRequest = { confirmingDelete = null },
title = { Text("Delete ${model.label}?") },
text = {
Text(
"The file is removed from ${(state as? LoadState.Loaded)?.value?.machine ?: "this machine"}. " +
"Nothing here can get it back -- downloading it again is the whole file again. " +
"Sessions using it keep their conversations and cannot start it."
)
},
confirmButton = {
TextButton(
onClick = {
confirmingDelete = null
machineModels.remove(model.id)
}
) {
Text("Delete")
}
},
dismissButton = {
TextButton(onClick = { confirmingDelete = null }) { Text("Cancel") }
},
)
}
if (confirmingStop) {
AlertDialog(
onDismissRequest = { confirmingStop = false },
title = { Text("Stop this server?") },
text = {
// Said plainly rather than hidden: this is the only thing that frees the memory,
// and what it costs is that every session on this machine reloads its model.
Text(
"Every model it is holding is unloaded. Sessions using it will show as " +
"exited, and the next message to one loads its model again — which is " +
"the slow part, not the sending."
)
},
confirmButton = {
TextButton(
onClick = {
confirmingStop = false
act("Stopping…") { stopProviderServer(settings, machineId, provider) }
}
) {
Text("Stop")
}
},
dismissButton = { TextButton(onClick = { confirmingStop = false }) { Text("Cancel") } },
)
}
}
@Composable
private fun ServerCard(
server: ServerState,
maxLoaded: Int?,
enabled: Boolean,
onStop: () -> Unit,
onMaxLoaded: (Int?) -> Unit,
) {
// The saved value is what this starts at and what Save is compared against, so a field left
// half-typed is visibly not saved rather than quietly either way.
val saved = maxLoaded?.toString().orEmpty()
var typed by remember(saved) { mutableStateOf(saved) }
var confirming by remember { mutableStateOf(false) }
Card(Modifier.fillMaxWidth()) {
Column(Modifier.padding(12.dp)) {
Text("Model server", style = MaterialTheme.typography.titleSmall)
Text(
if (server.running) {
"Running" + (server.port?.let { ", reached on port $it" } ?: "")
} else {
// Not a fault: nothing is loaded because nothing has asked. Saying it in
// words rather than colouring the row, since "stopped" and "we could not
// ask" would otherwise look the same.
"Not running. A session starts it when it needs a model."
},
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
Spacer(Modifier.height(8.dp))
LabelledField(
label = "Models loaded at once",
value = typed,
onValueChange = { typed = it.filter(Char::isDigit) },
hint = "one -- a second model replaces the first",
keyboardOptions = KeyboardOptions(keyboardType = KeyboardType.Number),
)
Row(verticalAlignment = Alignment.CenterVertically) {
// Shown whether or not it is running, and disabled when there is nothing to stop:
// a button that comes and goes makes its own absence the message.
TextButton(enabled = enabled && server.running, onClick = onStop) { Text("Stop") }
Spacer(Modifier.weight(1f))
TextButton(
enabled = enabled && typed != saved,
// Saving this while the server is up changes nothing until it comes down
// again, which is asked rather than written underneath -- see [RestartDialog].
onClick = {
if (server.running) confirming = true else onMaxLoaded(typed.toIntOrNull())
},
) {
Text("Save")
}
}
}
}
if (confirming) {
RestartDialog(
title = "Save for the next start?",
text =
"This server is running, and how many models it keeps loaded was decided when it " +
"started. Saving now changes what it does the next time it starts -- stop it " +
"here to have that be now.",
onConfirm = {
confirming = false
onMaxLoaded(typed.toIntOrNull())
},
onDismiss = { confirming = false },
)
}
}
@Composable
private fun ModelCard(
model: ProviderModel,
specs: List<ParamSpec>,
/** How big the file is on the machine, for a provider whose models are files. */
bytes: Long?,
onEdit: (() -> Unit)?,
onUnload: (() -> Unit)?,
onDelete: (() -> Unit)?,
enabled: Boolean,
) {
Card(
Modifier.fillMaxWidth()
.padding(vertical = 4.dp)
.then(if (onEdit != null && enabled) Modifier.clickable(onClick = onEdit) else Modifier)
) {
Column(Modifier.padding(12.dp)) {
Row(verticalAlignment = Alignment.CenterVertically) {
Text(
model.label,
style = MaterialTheme.typography.bodyMedium,
modifier = Modifier.weight(1f),
)
bytes?.let {
Text(
gigabytes(it),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
}
// What the server is doing with it, in its own word. Absent means nobody could ask --
// the server is not running -- and the line is left out rather than guessed at.
model.status?.let {
Text(
it,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
if (model.settings.isNotEmpty()) {
Text(
// In the words the dialog uses, and in the order it draws them: a summary
// naming `contextSize` is a summary of a different screen than the one it
// sits under.
specs
.mapNotNull { spec ->
model.settings[spec.key]?.let { "${spec.label} $it" }
}
.joinToString(", "),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
if (onUnload != null || onDelete != null) {
Row(verticalAlignment = Alignment.CenterVertically) {
// Both shown whenever this kind of model has them, disabled rather than
// absent: unloading frees memory and deleting frees disk, and a button that
// comes and goes makes its own absence the message.
onUnload?.let { TextButton(enabled = enabled, onClick = it) { Text("Unload") } }
Spacer(Modifier.weight(1f))
onDelete?.let { TextButton(enabled = enabled, onClick = it) { Text("Delete") } }
}
}
}
}
}
/**
* How one model is loaded.
*
* Saved on Save rather than as it is typed, unlike the session settings dialog: writing this
* unloads the model for everybody using it, which is not something to do once per keystroke.
*/
@Composable
private fun ModelSettingsDialog(
model: ProviderModel,
specs: List<ParamSpec>,
onDismiss: () -> Unit,
onSave: (Map<String, String>) -> Unit,
) {
var params by remember(model.id) { mutableStateOf(model.settings) }
// Asked over this dialog rather than instead of it, so Cancel comes back to the edits rather
// than throwing them away.
var confirming by remember(model.id) { mutableStateOf(false) }
// Whether saving costs anything worth asking about: something has to be in memory, and at
// least one of these settings has to be one it read on the way in.
val reloads =
(model.status == "loaded" || model.status == "sleeping") && specs.any { it.restart }
AlertDialog(
onDismissRequest = onDismiss,
// Every control here is a number, so the keyboard is up for most of this dialog's life --
// and a dialog that keeps its own size under the keyboard puts Save off the bottom of the
// screen, where nothing on screen says it is there. Taking the insets ourselves is what
// lets `imePadding` shrink it instead.
properties = DialogProperties(decorFitsSystemWindows = false),
modifier = Modifier.imePadding(),
title = { Text(model.label) },
text = {
Column(Modifier.verticalScroll(rememberScrollState())) {
ProviderParamFields(
specs = specs,
values = params,
onChange = { params = it },
// Chips: this is a form of its own rather than a row in a list of settings.
choices = ChoiceStyle.Chips,
)
}
},
confirmButton = {
TextButton(onClick = { if (reloads) confirming = true else onSave(params) }) {
Text("Save")
}
},
dismissButton = { TextButton(onClick = onDismiss) { Text("Cancel") } },
)
if (confirming) {
RestartDialog(
title = "Unload ${model.label}?",
text =
"It is in memory now, and these are read when it is loaded. Saving takes it out " +
"of memory; the sessions using it load it again with these settings on their " +
"next message.",
onConfirm = {
confirming = false
onSave(params)
},
onDismiss = { confirming = false },
)
}
}
app/androidApp/src/main/kotlin/com/example/aiapp/RawBlock.kt
+14
View File
@@ -1,10 +1,12 @@
package com.example.aiapp
import androidx.compose.foundation.background
import androidx.compose.foundation.horizontalScroll
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.ColumnScope
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.rememberScrollState
import androidx.compose.material3.MaterialTheme
import androidx.compose.runtime.Composable
import androidx.compose.ui.Modifier
@@ -18,6 +20,14 @@ import androidx.compose.ui.unit.dp
* monospace text drawn hard against the edge of a tinted block reads as a clipping fault, and three
* copies of "clip, fill, pad" drift apart the first time one is adjusted.
*
* **Nothing in here wraps; it scrolls sideways instead.** This is column-aligned far more often
* than it is prose -- a diff, a table, a test run, a command and its arguments -- and wrapping
* destroys exactly the alignment that was carrying the meaning, while turning one line into four
* and a run of them into a wall. The scroll belongs to the block rather than to each line so that
* the lines stay aligned with each other as it moves: one offset for the whole column is what makes
* a shifted diff still read as a diff. Every [Text] inside is therefore drawn with `softWrap =
* false`, which is the half of this a caller has to remember.
*
* The colour is [rawSurface], which is also what a code block inside a reply is given.
*/
@Composable
@@ -29,6 +39,10 @@ fun RawBlock(modifier: Modifier = Modifier, content: @Composable ColumnScope.()
// rectangle drawn at the same radius as the one behind it reads as a misprint.
.clip(MaterialTheme.shapes.extraSmall)
.background(rawSurface)
// Clipped and filled before this, so the tint is the viewport and does not scroll away
// from under the text; padded after it, so the inset travels with the content and the
// last column does not end flush against the edge.
.horizontalScroll(rememberScrollState())
.padding(horizontal = 8.dp, vertical = 6.dp),
content = content,
)
app/androidApp/src/main/kotlin/com/example/aiapp/Reorder.kt
+284
View File
@@ -0,0 +1,284 @@
package com.example.aiapp
import androidx.compose.foundation.gestures.detectDragGestures
import androidx.compose.foundation.gestures.scrollBy
import androidx.compose.foundation.layout.Box
import androidx.compose.foundation.layout.size
import androidx.compose.foundation.lazy.LazyListItemInfo
import androidx.compose.foundation.lazy.LazyListState
import androidx.compose.material3.MaterialTheme
import androidx.compose.runtime.Composable
import androidx.compose.runtime.State
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableFloatStateOf
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.rememberUpdatedState
import androidx.compose.runtime.setValue
import androidx.compose.runtime.withFrameNanos
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.hapticfeedback.HapticFeedback
import androidx.compose.ui.hapticfeedback.HapticFeedbackType
import androidx.compose.ui.input.pointer.pointerInput
import androidx.compose.ui.platform.LocalDensity
import androidx.compose.ui.platform.LocalHapticFeedback
import androidx.compose.ui.semantics.contentDescription
import androidx.compose.ui.semantics.semantics
import androidx.compose.ui.unit.Density
import androidx.compose.ui.unit.dp
import androidx.compose.ui.unit.sp
import kotlin.math.abs
import kotlinx.coroutines.CoroutineScope
import kotlinx.coroutines.launch
/**
* Dragging a row of a [androidx.compose.foundation.lazy.LazyColumn] into a different place in it.
*
* Generic rather than the session list's own, because "hold this and move it" is one gesture
* wherever it appears and the arithmetic below is the whole of it. The list itself is left alone:
* this reports a move and the caller decides what a move means -- it is the caller that holds the
* rows and the caller that tells a server about the new order.
*
* The drag is on a [ReorderHandle] rather than on the row, which is what keeps it out of the way of
* the scroll. A whole row that can be dragged sideways-ish is a row that sometimes eats a fling,
* and a list is scrolled far more often than it is rearranged.
*/
class Reorder
internal constructor(
private val listState: LazyListState,
private val scope: CoroutineScope,
private val haptics: HapticFeedback,
/** What the [EDGE] band is in pixels here; a band in raw pixels is one screen's answer. */
private val density: Density,
/**
* The caller's own lists are what move; these are [State] so that the gesture, which outlives a
* recomposition, is never holding the first composition's copy of them.
*/
private val onMove: State<(from: Int, to: Int) -> Unit>,
private val onSettled: State<() -> Unit>,
) {
/** The key of the row in hand, or null when nothing is being dragged. */
var held by mutableStateOf<Any?>(null)
private set
/** Where the list had laid the row out when it was taken hold of, in viewport pixels. */
private var grabbedAt = 0
/** How far the finger has moved since, which is what the row is drawn following. */
private var dragged by mutableFloatStateOf(0f)
/** How far the list has scrolled under it since -- see [follow]. */
private var scrolled = 0f
/** The index the row has been moved to so far, which is what the next move counts from. */
private var at = 0
/** Where it started, so that a handle merely pressed is not reported as a rearrangement. */
private var from = 0
/**
* How much of the travel below the moves so far have accounted for.
*
* The travel is what decides a crossing, rather than where the row is drawn *now*: a lazy list
* animates an item into its new place, so for a few frames after a move `offset` still reports
* roughly the old one. Deciding from that offset re-decided the same crossing on every frame
* until the animation caught up, and a drag of two rows arrived six rows down.
*/
private var settled = 0f
private fun info(key: Any): LazyListItemInfo? =
listState.layoutInfo.visibleItemsInfo.firstOrNull { it.key == key }
private fun itemAt(index: Int): LazyListItemInfo? =
listState.layoutInfo.visibleItemsInfo.firstOrNull { it.index == index }
/**
* How far from where the list laid it out this row should be drawn -- zero for every row but
* the one in hand.
*
* Measured against where the row is laid out *now* rather than accumulated, which is what makes
* it self-correcting: a move, or a scroll under the finger, puts the row somewhere new, and the
* same subtraction cancels that out so the row stays under the finger instead of jumping by its
* own height.
*/
fun offsetOf(key: Any): Float {
if (key != held) return 0f
val now = info(key) ?: return 0f
return grabbedAt + dragged - now.offset
}
internal fun grab(key: Any) {
val from = info(key) ?: return
held = key
grabbedAt = from.offset
at = from.index
this.from = from.index
dragged = 0f
scrolled = 0f
settled = 0f
// The platform's "you have picked this up", the same feedback a long press gives, because
// the gesture it confirms is the same kind of commitment.
haptics.performHapticFeedback(HapticFeedbackType.LongPress)
}
internal fun drag(by: Float) {
if (held == null) return
dragged += by
cross()
}
/**
* Trades places with as many neighbours as the travel so far has earned.
*
* Half a neighbour's height each way, so the row changes place when it covers most of the one
* it is passing -- and a full height of hysteresis before it can come back, since the move has
* already paid that half in the other direction. A loop rather than one step: a fast drag, or a
* list scrolling under a parked finger, crosses several rows between two events.
*/
private fun cross() {
while (true) {
val slack = dragged + scrolled - settled
val next = itemAt(if (slack > 0) at + 1 else at - 1) ?: return
if (abs(slack) < next.size / 2f) return
// Where the list is looking, taken before the move and put back after it. A lazy list
// keeps its place by the *key* of the item at the top, so moving that item takes the
// viewport with it -- drag the top row down two places and the list scrolls two rows
// to follow it, which reads as the row never having moved. The correction is by index,
// which is the thing that did not change.
val anchor = listState.firstVisibleItemIndex
val within = listState.firstVisibleItemScrollOffset
onMove.value(at, next.index)
// Requested rather than scrolled to: this has to take effect in the *same* measurement
// as the move, and a scroll launched beside it lands before the list has taken the new
// order and is then undone by it.
listState.requestScrollToItem(anchor, within)
settled += if (slack > 0) next.size.toFloat() else -next.size.toFloat()
at = next.index
// Loud on purpose: the row is under a finger that is covering it, so the tick is how
// the reader knows a place was taken rather than that they are still between two.
haptics.performHapticFeedback(HapticFeedbackType.SegmentTick)
}
}
internal fun release() {
// Only where the row actually went somewhere: a handle pressed and let go has rearranged
// nothing, and reporting one would have the server rewrite the order it already has.
val moved = held != null && at != from
held = null
dragged = 0f
scrolled = 0f
settled = 0f
if (moved) onSettled.value()
}
/**
* Scrolls the list while the row in hand is held against one end of it, so a row can be moved
* further than one screenful. A frame loop rather than a response to the drag, because a finger
* parked at the bottom edge sends no more events and is exactly the case this exists for.
*/
internal fun follow() {
val key = held ?: return
scope.launch {
while (held == key) {
withFrameNanos {}
val moving = info(key) ?: continue
val viewport = listState.layoutInfo.viewportEndOffset
val edge = with(density) { EDGE.toPx() }
val top = grabbedAt + dragged
val bottom = top + moving.size
val step =
when {
top < edge -> -(edge - top).coerceAtMost(edge)
bottom > viewport - edge -> (bottom - (viewport - edge)).coerceAtMost(edge)
else -> 0f
}
if (step == 0f) continue
// Counted as travel of its own: the finger has not moved, but the rows have moved
// under it, which is the same thing to everything above. Nothing is added to the
// drag, because where the row is *drawn* is measured against the list's own
// offsets and those have already moved.
scrolled += listState.scrollBy(step * SPEED)
cross()
}
}
}
private companion object {
/** How close to an end of the list a held row has to be before the list follows it. */
val EDGE = 36.dp
/** A fraction of the overshoot per frame, so the scroll eases in rather than lurching. */
const val SPEED = 0.12f
}
}
@Composable
fun rememberReorder(
listState: LazyListState,
/** Two indices into the lazy list, which is the caller's own order to rearrange. */
onMove: (from: Int, to: Int) -> Unit,
/** The drag is over: the order on screen is the one to keep. */
onSettled: () -> Unit,
): Reorder {
val move = rememberUpdatedState(onMove)
val settled = rememberUpdatedState(onSettled)
val haptics = LocalHapticFeedback.current
val density = LocalDensity.current
val scope = rememberCoroutineScope()
return remember(listState) { Reorder(listState, scope, haptics, density, move, settled) }
}
/**
* The handle a row is dragged by: the burger, at about the size of a heading.
*
* Bigger than an icon beside a line of text -- this is what a row is taken hold of by, and at
* [GLYPH_SIZE] it read as decoration on the end of the row. Not as big as the row either: a mark
* scaled to the card's whole inner height came out heavier than anything else on screen, since
* these rules thicken with the glyph.
*
* The touch square around it is [GLYPH_BUTTON_SIZE], the same as every other icon control here, so
* the mark and the area that answers to a finger are two different sizes -- which is why the caller
* subtracts [HANDLE_MARGIN] from the gap it wants: what has to line up with the text on the other
* side is the mark, not the box around it.
*
* [key] is the row's own key in the list, which is how a gesture that started here finds the row it
* belongs to -- an index would be stale the moment the first move landed.
*/
@Composable
fun ReorderHandle(state: Reorder, key: Any, modifier: Modifier = Modifier) {
Box(
contentAlignment = Alignment.Center,
modifier =
modifier
.size(GLYPH_BUTTON_SIZE)
// Nothing here draws a word, and a handle is the kind of control somebody using a
// screen reader has no other way to find.
.semantics { contentDescription = "Drag to reorder" }
.pointerInput(key) {
detectDragGestures(
onDragStart = {
state.grab(key)
state.follow()
},
onDrag = { _, amount -> state.drag(amount.y) },
onDragEnd = { state.release() },
onDragCancel = { state.release() },
)
},
) {
Glyph(DRAG_GLYPH, colour = MaterialTheme.colorScheme.onSurfaceVariant, size = HANDLE_MARK)
}
}
/** How big the mark itself is: a heading's size, which is what the font is asked for in `sp`. */
private val HANDLE_MARK = 24.sp
/**
* How much of the touch square lies outside the mark on each side.
*
* A caller that wants the *mark* a given distance from something takes this off that distance --
* see the rule about aligning the mark rather than the box it is centred in.
*/
val HANDLE_MARGIN = (GLYPH_BUTTON_SIZE - HANDLE_MARK.value.dp) / 2
app/androidApp/src/main/kotlin/com/example/aiapp/ReplyFooter.kt
+89
View File
@@ -0,0 +1,89 @@
package com.example.aiapp
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Text
import androidx.compose.runtime.Composable
import androidx.compose.ui.Modifier
import androidx.compose.ui.text.style.TextAlign
import java.time.Instant
import java.time.ZoneId
import java.time.format.DateTimeFormatter
import java.time.format.FormatStyle
import java.util.Locale
/**
* The line under a finished reply: what it cost to produce, and when it was sent.
*
* Small and set back, in the tone the session's own subtitle takes: it is about the message rather
* than part of it, and at the reply's own size it would read as the last thing the model said.
*
* Right-aligned because it closes the message rather than opening one -- a reader scanning down the
* left edge is reading what was said, and this is where that ends.
*/
@Composable
fun ReplyFooter(
ts: Double,
tokensPerSecond: Double?,
prefillMs: Long?,
modifier: Modifier = Modifier,
) {
val text = replyFooterText(ts, tokensPerSecond, prefillMs, ZoneId.systemDefault()) ?: return
Text(
text,
style = MaterialTheme.typography.labelSmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
textAlign = TextAlign.End,
modifier = modifier.fillMaxWidth(),
)
}
/**
* What the footer says, or null when there is nothing to say: "read 9.5s · 50.3 tok/s · 3:00 PM".
*
* Split out so the wording is testable without a screen, and [zone] is a parameter for the same
* reason [limitSummary] takes one: a test has to say the same thing wherever it runs.
*
* **The time is last, and so sits against the right edge whatever else is on the line.** The
* measurements in front of it are the provider's, so a session on another provider has fewer of
* them or none -- and a reader who has learned where the clock is should not have to find it again
* because the model changed. The costs grow leftwards into the space instead.
*
* Those measurements are drawn only where the provider made them. Most do not -- a coding CLI
* reports what a turn cost and never how long the model spent on it -- and the time this app
* watched a reply arrive over is a different quantity: it counts the network, the pauses between
* tokens and whatever else the machine was doing. So the line is the clock alone rather than a
* plausible figure beside it.
*/
fun replyFooterText(
ts: Double,
tokensPerSecond: Double?,
prefillMs: Long?,
zone: ZoneId,
): String? {
val at =
if (ts <= 0.0) null
else
try {
DateTimeFormatter.ofLocalizedTime(FormatStyle.SHORT)
.withZone(zone)
.format(Instant.ofEpochMilli((ts * 1000).toLong()))
} catch (_: Exception) {
null
}
// A tenth up to three digits, where the difference between 18 and 18.4 tok/s is something a
// reader comparing two models can use; past that the tenth is noise on a figure that moves by
// more than that between turns.
val rate =
tokensPerSecond
?.takeIf { it > 0.0 }
?.let {
if (it >= 100) String.format(Locale.getDefault(), "%.0f tok/s", it)
else String.format(Locale.getDefault(), "%.1f tok/s", it)
}
// Named "read" rather than given a unit alone, because a second figure in seconds beside a
// rate is unreadable otherwise -- and it is the same word the status row uses while it is
// happening, so the wait and the figure for it are one vocabulary.
val read = prefillMs?.takeIf { it > 0 }?.let { "read ${formatMillis(it)}" }
return listOfNotNull(read, rate, at).joinToString(" · ").ifEmpty { null }
}
app/androidApp/src/main/kotlin/com/example/aiapp/RestartDialog.kt
+35
View File
@@ -0,0 +1,35 @@
package com.example.aiapp
import androidx.compose.material3.AlertDialog
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
/**
* What a setting costs, asked before it is written.
*
* Every setting in this app whose consequence is worth saying says it here rather than in a
* paragraph beside the control: a sentence under a switch is read after the decision if it is read
* at all, and a form of a dozen settings each carrying its own explanation is mostly explanation. A
* modal interrupts at the moment the consequence becomes real, and it is also a way out.
*
* What they have in common, and why it is one dialog rather than four: each of them ends something
* that is running — a process, a loaded model, a server — and says what starting it again costs.
*/
@Composable
fun RestartDialog(
title: String,
text: String,
onConfirm: () -> Unit,
onDismiss: () -> Unit,
/** The word on the button, which is the action rather than a bare "OK". */
confirm: String = "Save",
) {
AlertDialog(
onDismissRequest = onDismiss,
title = { Text(title) },
text = { Text(text) },
confirmButton = { TextButton(onClick = onConfirm) { Text(confirm) } },
dismissButton = { TextButton(onClick = onDismiss) { Text("Cancel") } },
)
}
app/androidApp/src/main/kotlin/com/example/aiapp/SessionAlerts.kt
+6 -6
View File
@@ -31,13 +31,13 @@ import androidx.lifecycle.compose.LocalLifecycleOwner
import androidx.lifecycle.repeatOnLifecycle
/**
* A session wanting attention, said over the app rather than through Android's drawer.
* A session wanting attention, said over the app as well as in Android's drawer.
*
* Two places can carry the same fact and only one is right at a time. A row in the shade is for
* somebody looking at something else: it makes a sound, it waits however long it has to, and acting
* on it means leaving whatever they were doing. Somebody with this app open needs none of that. So
* while these are on screen the stream is delivered here instead, which is arranged by the
* collection below and nothing else.
* Two places carry the same fact and they are doing different jobs: a row in the shade waits
* however long it has to, which makes it the record, and a banner is read now or not at all, which
* makes it the interruption. So somebody with the app open gets both -- this, and a silent row
* behind it that is still there when they go looking and goes by itself when they open the session.
* Whether the app is open at all is this collection and nothing else.
*
* A banner can go three ways, each somebody deciding something different: tapped, which opens the
* session; pushed off either side; or left alone, in which case it goes when the bar runs out.
app/androidApp/src/main/kotlin/com/example/aiapp/SessionImage.kt
+220 -15
View File
@@ -5,24 +5,32 @@ import androidx.compose.foundation.Image
import androidx.compose.foundation.background
import androidx.compose.foundation.clickable
import androidx.compose.foundation.gestures.detectTransformGestures
import androidx.compose.foundation.interaction.MutableInteractionSource
import androidx.compose.foundation.layout.Box
import androidx.compose.foundation.layout.fillMaxSize
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.height
import androidx.compose.foundation.layout.navigationBarsPadding
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.layout.size
import androidx.compose.material3.Button
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Text
import androidx.compose.runtime.Composable
import androidx.compose.runtime.DisposableEffect
import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.SideEffect
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableFloatStateOf
import androidx.compose.runtime.mutableIntStateOf
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.draw.clip
import androidx.compose.ui.geometry.Offset
import androidx.compose.ui.graphics.Color
import androidx.compose.ui.graphics.FilterQuality
import androidx.compose.ui.graphics.ImageBitmap
@@ -30,12 +38,20 @@ import androidx.compose.ui.graphics.asImageBitmap
import androidx.compose.ui.graphics.graphicsLayer
import androidx.compose.ui.input.pointer.pointerInput
import androidx.compose.ui.layout.ContentScale
import androidx.compose.ui.layout.onSizeChanged
import androidx.compose.ui.platform.LocalDensity
import androidx.compose.ui.platform.LocalView
import androidx.compose.ui.unit.Dp
import androidx.compose.ui.unit.IntSize
import androidx.compose.ui.unit.dp
import androidx.compose.ui.unit.isSpecified
import androidx.compose.ui.window.Dialog
import androidx.compose.ui.window.DialogProperties
import androidx.compose.ui.window.DialogWindowProvider
import androidx.core.view.ViewCompat
import androidx.core.view.WindowCompat
import androidx.core.view.WindowInsetsCompat
import androidx.core.view.WindowInsetsControllerCompat
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.withContext
@@ -140,12 +156,23 @@ fun SessionImageViewer(
onClose: () -> Unit,
) {
val (bitmap, failed) = rememberSessionBitmap(settings, sessionId, ref)
val view = LocalView.current
var hiddenBars by remember(ref) { mutableStateOf(ViewerBars()) }
var barInsets by remember(ref) { mutableStateOf(ViewerBarInsets()) }
Dialog(
onDismissRequest = onClose,
properties = DialogProperties(usePlatformDefaultWidth = false),
properties =
DialogProperties(usePlatformDefaultWidth = false, decorFitsSystemWindows = false),
) {
ViewerSystemBars(hiddenBars)
Box(
Modifier.fillMaxSize().background(Color.Black).clickable(onClick = onClose),
Modifier.fillMaxSize()
.background(Color.Black)
.clickable(
interactionSource = remember { MutableInteractionSource() },
indication = null,
onClick = onClose,
),
contentAlignment = Alignment.Center,
) {
when (val image = bitmap) {
@@ -165,7 +192,46 @@ fun SessionImageViewer(
// beside it are.
CircularProgressIndicator(color = Color.White)
}
else -> ZoomableImage(image)
else -> {
var viewport by remember { mutableStateOf(IntSize.Zero) }
var nativeSizeRequest by remember { mutableIntStateOf(0) }
ZoomableImage(
image,
nativeSizeRequest = nativeSizeRequest,
onViewportChanged = {
viewport = it
ViewCompat.getRootWindowInsets(view)?.let { insets ->
barInsets =
ViewerBarInsets(
status =
insets
.getInsetsIgnoringVisibility(
WindowInsetsCompat.Type.statusBars()
)
.top,
navigation =
insets
.getInsetsIgnoringVisibility(
WindowInsetsCompat.Type.navigationBars()
)
.bottom,
)
}
},
onBarsChanged = { hiddenBars = it },
barInsets = barInsets,
viewport = viewport,
)
Button(
onClick = { nativeSizeRequest++ },
modifier =
Modifier.align(Alignment.BottomEnd)
.navigationBarsPadding()
.padding(16.dp),
) {
Text("100%")
}
}
}
}
}
@@ -225,34 +291,72 @@ private fun enlargingFilter(sourceHeight: Int, drawnHeight: Int): FilterQuality
* The image on its own, as large as it fits, with pinch to zoom.
*
* Inside a dialog rather than a screen -- see [SessionImageViewer] -- so the platform's back
* gesture returns to the transcript instead of leaving the app. It opens fitted, the whole image
* visible.
* gesture returns to the transcript instead of leaving the app. It opens fitted, with the whole
* image visible without enlarging a smaller one; the 100% control changes to one bitmap pixel per
* screen pixel and recenters it.
*/
@Composable
private fun ZoomableImage(image: ImageBitmap) {
private fun ZoomableImage(
image: ImageBitmap,
nativeSizeRequest: Int,
onViewportChanged: (IntSize) -> Unit,
onBarsChanged: (ViewerBars) -> Unit,
barInsets: ViewerBarInsets,
viewport: IntSize,
) {
var scale by remember { mutableFloatStateOf(1f) }
var offsetX by remember { mutableFloatStateOf(0f) }
var offsetY by remember { mutableFloatStateOf(0f) }
val nativeScale = nativeScale(image.width, image.height, viewport.width, viewport.height)
LaunchedEffect(nativeSizeRequest, nativeScale) {
if (nativeSizeRequest > 0) {
scale = nativeScale
offsetX = 0f
offsetY = 0f
}
}
val bars =
viewerBars(
image.width,
image.height,
viewport.width,
viewport.height,
scale,
Offset(offsetX, offsetY),
barInsets,
)
SideEffect { onBarsChanged(bars) }
Image(
bitmap = image,
contentDescription = "Attached image",
contentScale = ContentScale.Fit,
contentScale = ContentScale.Inside,
// Zoomed in, the reader is looking at pixels on purpose.
filterQuality = FilterQuality.None,
modifier =
Modifier.fillMaxSize()
.pointerInput(Unit) {
detectTransformGestures { _, pan, zoom, _ ->
// Floor of 1 so the image cannot be pinched smaller than fitted, which is
// already the whole of it; a ceiling so it cannot be lost off-screen.
scale = (scale * zoom).coerceIn(1f, 8f)
if (scale > 1f) {
offsetX += pan.x
offsetY += pan.y
.onSizeChanged(onViewportChanged)
.pointerInput(nativeScale) {
detectTransformGestures { centroid, pan, zoom, _ ->
val oldScale = scale
val maximumScale = maxOf(8f, nativeScale)
val newScale = (oldScale * zoom).coerceIn(1f, maximumScale)
if (newScale > 1f) {
val offset =
zoomOffset(
Offset(offsetX, offsetY),
centroid,
pan,
oldScale,
newScale,
Offset(size.width / 2f, size.height / 2f),
)
offsetX = offset.x
offsetY = offset.y
} else {
offsetX = 0f
offsetY = 0f
}
scale = newScale
}
}
.graphicsLayer {
@@ -263,3 +367,104 @@ private fun ZoomableImage(image: ImageBitmap) {
},
)
}
/** Lets the picture use the whole display, hiding only the system bars it actually reaches. */
@Composable
private fun ViewerSystemBars(hidden: ViewerBars) {
val view = LocalView.current
val window = (view.parent as? DialogWindowProvider)?.window
val controller = window?.let { WindowCompat.getInsetsController(it, view) }
SideEffect {
controller?.systemBarsBehavior =
WindowInsetsControllerCompat.BEHAVIOR_SHOW_TRANSIENT_BARS_BY_SWIPE
if (hidden.status) {
controller?.hide(WindowInsetsCompat.Type.statusBars())
} else {
controller?.show(WindowInsetsCompat.Type.statusBars())
}
if (hidden.navigation) {
controller?.hide(WindowInsetsCompat.Type.navigationBars())
} else {
controller?.show(WindowInsetsCompat.Type.navigationBars())
}
}
DisposableEffect(view) {
onDispose {
controller?.show(
WindowInsetsCompat.Type.statusBars() or WindowInsetsCompat.Type.navigationBars()
)
}
}
}
internal data class ViewerBars(val status: Boolean = false, val navigation: Boolean = false)
internal data class ViewerBarInsets(val status: Int = 0, val navigation: Int = 0)
/** Which full-screen system-bar regions the fitted, zoomed and panned image intersects. */
internal fun viewerBars(
imageWidth: Int,
imageHeight: Int,
viewportWidth: Int,
viewportHeight: Int,
scale: Float,
offset: Offset,
insets: ViewerBarInsets,
): ViewerBars {
if (imageWidth <= 0 || imageHeight <= 0 || viewportWidth <= 0 || viewportHeight <= 0) {
return ViewerBars()
}
val fittedScale = insideScale(imageWidth, imageHeight, viewportWidth, viewportHeight)
val width = imageWidth * fittedScale * scale
val height = imageHeight * fittedScale * scale
val left = viewportWidth / 2f + offset.x - width / 2f
val right = left + width
val top = viewportHeight / 2f + offset.y - height / 2f
val bottom = top + height
val crossesScreen = right > 0f && left < viewportWidth
return ViewerBars(
status = crossesScreen && insets.status > 0 && bottom > 0f && top < insets.status,
navigation =
crossesScreen &&
insets.navigation > 0 &&
bottom > viewportHeight - insets.navigation &&
top < viewportHeight,
)
}
/** Scale relative to [ContentScale.Inside] at which bitmap and screen pixels are one-to-one. */
internal fun nativeScale(
imageWidth: Int,
imageHeight: Int,
viewportWidth: Int,
viewportHeight: Int,
): Float {
if (imageWidth <= 0 || imageHeight <= 0 || viewportWidth <= 0 || viewportHeight <= 0) return 1f
return 1f / insideScale(imageWidth, imageHeight, viewportWidth, viewportHeight)
}
/** The downscale-only factor used by [ContentScale.Inside]. */
private fun insideScale(
imageWidth: Int,
imageHeight: Int,
viewportWidth: Int,
viewportHeight: Int,
): Float =
minOf(
1f,
viewportWidth.toFloat() / imageWidth,
viewportHeight.toFloat() / imageHeight,
)
/** Keeps the image point beneath [centroid] beneath the fingers as its scale changes. */
internal fun zoomOffset(
offset: Offset,
centroid: Offset,
pan: Offset,
oldScale: Float,
newScale: Float,
viewportCenter: Offset,
): Offset {
val scaleChange = newScale / oldScale
return offset * scaleChange + (centroid - viewportCenter) * (1f - scaleChange) + pan
}
app/androidApp/src/main/kotlin/com/example/aiapp/SessionListScreen.kt
+401 -137
View File
@@ -1,9 +1,11 @@
package com.example.aiapp
import androidx.activity.compose.BackHandler
import androidx.compose.foundation.ExperimentalFoundationApi
import androidx.compose.foundation.combinedClickable
import androidx.compose.foundation.layout.Box
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.PaddingValues
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.fillMaxSize
@@ -12,11 +14,15 @@ import androidx.compose.foundation.layout.height
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.layout.width
import androidx.compose.foundation.lazy.LazyColumn
import androidx.compose.foundation.lazy.rememberLazyListState
import androidx.compose.material3.AlertDialog
import androidx.compose.material3.Card
import androidx.compose.material3.CardDefaults
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.FloatingActionButton
import androidx.compose.material3.LinearProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Surface
import androidx.compose.material3.Switch
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
@@ -29,14 +35,30 @@ import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.draw.alpha
import androidx.compose.ui.graphics.graphicsLayer
import androidx.compose.ui.layout.onSizeChanged
import androidx.compose.ui.platform.LocalContext
import androidx.compose.ui.platform.LocalDensity
import androidx.compose.ui.unit.dp
import androidx.compose.ui.zIndex
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.launch
import kotlinx.coroutines.withContext
/**
* The sessions tab: sessions awaiting an answer sort to the top, which is the "your turn" inbox.
* The sessions tab: every session, in the order the reader has put them in.
*
* Nothing here sorts. The order is the server's `sessions` list and the reader's own -- see
* [reorderSessions] -- which is the one arrangement a row cannot be moved out of by something the
* session does. It replaced sorting by activity, and then sorting by when each agent was turned on:
* both meant a list that rearranged itself under whoever was reading it, and the status word and
* its colour already say which session wants something without the row having to move to say it.
*
* Holding a row puts the screen in selection mode, the same gesture and the same bottom bar as the
* import tab, so the two lists are learned once. Rearranging is deliberately *not* part of a
* selection -- the handle moves the row it is on, whether or not that row is picked out -- because
* "which rows am I acting on" and "where does this one go" are two questions.
*
* No title and no Back of its own -- [MainScreen] owns the header and the tab that names this one.
* What stays here is the button that adds a session, because that acts on this list and nothing
@@ -48,10 +70,23 @@ fun SessionListScreen(
reloadToken: Int,
onOpen: (SessionSummary) -> Unit,
onSpawn: () -> Unit,
/** A session this list has just deleted, for whoever is showing it elsewhere. */
onDeleted: (String) -> Unit = {},
) {
val scope = rememberCoroutineScope()
var listState by remember { mutableStateOf<LoadState<List<SessionSummary>>>(LoadState.Loading) }
var confirmingDelete by remember { mutableStateOf<SessionSummary?>(null) }
// Which rows the reader has picked out. Empty means selection mode is off, as on the import
// tab: a selection mode with nothing in it has no controls and no way out but Back.
var selected by remember { mutableStateOf<Set<String>>(emptySet()) }
// The sessions a delete has been confirmed for, or none. A list rather than one session,
// because a selection is what the bar below acts on.
var confirmingDelete by remember { mutableStateOf<List<SessionSummary>>(emptyList()) }
// Whether an answer is outstanding, which is a different question from whether there is
// anything to draw: see [refresh].
var reloading by remember { mutableStateOf(false) }
// Failures that belong to one session rather than to the list, keyed by its id and shown on its
// own card. The two scopes are decided by whether the server answered: it answered and refused,
@@ -60,6 +95,10 @@ fun SessionListScreen(
// Cleared on the next successful load below -- an entry outlives its session otherwise.
var deleteErrors by remember { mutableStateOf<Map<String, String>>(emptyMap()) }
// Why the order on screen is not the order that was saved, when saving one failed. The list is
// what failed, so it is reported over the list rather than on any row.
var orderError by remember { mutableStateOf<String?>(null) }
// Which sessions have a delete in flight. A set of ids rather than a flag on the row, because
// the rows are rebuilt from whatever the server last said and this belongs to the request.
var deleting by remember { mutableStateOf<Set<String>>(emptySet()) }
@@ -70,31 +109,142 @@ fun SessionListScreen(
val transcriptCache = remember(settings) { TranscriptCache(cacheRoot(context, settings)) }
fun refresh() {
listState = LoadState.Loading
// The rows stay while the answer is on its way, with the bar below saying one is: this
// list is asked again every time the panel over a session is opened, and blanking it each
// time hands the reader an empty screen to report on something that was never in doubt.
// A first load has nothing to keep, and says so with the spinner instead.
if (listState !is LoadState.Loaded) listState = LoadState.Loading
reloading = true
scope.launch {
listState =
try {
val loaded =
withContext(Dispatchers.IO) { LoadState.Loaded(fetchSessions(settings)) }
deleteErrors = emptyMap()
val alive = loaded.value.map { it.id }.toSet()
// A selection is of sessions, so one deleted somewhere else leaves it. Only
// that one: the other rows the reader picked out are still there.
selected = selected.intersect(alive)
// The path out for a cached transcript whose session was deleted somewhere
// else. This list is the only place that ever learns the full set. On the
// answer rather than in `finally`: a list that failed to arrive says nothing
// about which sessions exist.
withContext(Dispatchers.IO) {
transcriptCache.retainOnly(loaded.value.map { it.id }.toSet())
}
withContext(Dispatchers.IO) { transcriptCache.retainOnly(alive) }
loaded
} catch (e: ApiException) {
LoadState.failed(e)
}
reloading = false
}
}
/**
* Deletes every session in [targets], one after another.
*
* One at a time and in the order they are drawn: the server has no batch delete for sessions,
* and each one ends a process. Each row says what is happening to it from the moment the work
* is handed over, which is also when the selection goes -- a bar still naming sessions being
* deleted is a set nobody can act on.
*/
fun deleteChosen(targets: List<SessionSummary>, alsoDeleteForeign: Boolean) {
selected = emptySet()
// Marked here rather than after the request returns: a row has to say something is
// happening to it from the moment it is asked for.
deleting = deleting + targets.map { it.id }
deleteErrors = deleteErrors - targets.map { it.id }.toSet()
scope.launch {
for (session in targets) {
try {
withContext(Dispatchers.IO) {
deleteSession(settings, session.id, alsoDeleteForeign)
// After it succeeded, not before: a refused delete leaves the session
// exactly as it was, and its transcript with it.
transcriptCache.session(TranscriptAddress(session.id)).purge()
}
// Only this row, and only what changed. Refetching the list instead put every
// other session back through loading and handed the reader an empty screen, to
// report on something never in doubt.
val loaded = listState
if (loaded is LoadState.Loaded) {
listState = LoadState.Loaded(loaded.value.filterNot { it.id == session.id })
}
onDeleted(session.id)
} catch (e: ApiException) {
// Kept, because it is still there: the server refused, so the session it
// refused about is exactly as it was.
deleteErrors = deleteErrors + (session.id to (e.message ?: "Delete failed"))
} finally {
deleting = deleting - session.id
}
}
}
}
LaunchedEffect(reloadToken) { refresh() }
val rows = rememberLazyListState()
val reorder =
rememberReorder(
listState = rows,
onMove = { from, to ->
// Moved here and now, because the row is under a finger: waiting for the server to
// agree would drag the handle away from the card it is on. What the server thinks
// is asked for when the finger comes up, and a refusal puts the list back.
val loaded = listState
if (loaded is LoadState.Loaded) {
val moved = loaded.value.toMutableList()
moved.add(to, moved.removeAt(from))
listState = LoadState.Loaded(moved)
}
},
onSettled = {
val loaded = listState
if (loaded is LoadState.Loaded) {
val order = loaded.value.map { it.id }
scope.launch {
try {
withContext(Dispatchers.IO) { reorderSessions(settings, order) }
orderError = null
} catch (e: ApiException) {
orderError = e.message ?: "The new order couldn't be saved"
// The screen must not go on showing an arrangement nothing kept, so
// the server's own order comes back -- which is also the only way to
// see what it does think.
refresh()
}
}
}
},
)
// Back leaves selection mode rather than the tab, which is the level it is one step above.
// Nested inside MainScreen's own handler, so it wins while there is a selection.
BackHandler(enabled = selected.isNotEmpty()) { selected = emptySet() }
// Measured rather than assumed: the list reserves exactly what the bar covers, so the last row
// can still be scrolled to while it is up.
var barHeight by remember { mutableStateOf(0.dp) }
val density = LocalDensity.current
// What the bar covers *now*: its measurement is kept while it is away, but nothing is
// reserved for a bar that is not up.
val covered = if (selected.isEmpty()) 0.dp else barHeight
// The spawn button floats over the list, so the list ends above it -- measured, for the
// reason the bar is. Without this the last row sat under the button, which was survivable
// while every part of a row did the same thing and is not now that corner is a handle.
var buttonHeight by remember { mutableStateOf(0.dp) }
Box(Modifier.fillMaxSize()) {
Column(Modifier.fillMaxSize().padding(16.dp)) {
orderError?.let { message ->
// The server's own words, unprefixed, the way every other failure is shown.
Text(
message,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.error,
)
Spacer(Modifier.height(8.dp))
}
when (val state = listState) {
is LoadState.Loading -> CircularProgressIndicator()
// The message as Api.kt wrote it, with nothing added: it is already a whole
@@ -113,20 +263,32 @@ fun SessionListScreen(
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
// Awaiting-answer first (the point of the screen), then most recently active.
val ordered =
state.value.sortedWith(
compareByDescending<SessionSummary> { it.status == "awaitingInput" }
.thenByDescending { it.lastActivity }
)
LazyColumn {
uniqueItems(ordered, key = { it.id }) { session ->
LazyColumn(
state = rows,
contentPadding =
PaddingValues(bottom = covered + buttonHeight + BUTTON_RING * 2),
) {
uniqueItems(state.value, key = { it.id }) { session ->
SessionCard(
session = session,
error = deleteErrors[session.id],
deleting = session.id in deleting,
onOpen = { onOpen(session) },
onLongPress = { confirmingDelete = session },
picked = session.id in selected,
// The handle is a selection-mode control, so it is absent rather
// than disabled outside one: this is not a capability being
// withheld, it is a mode the list is not in.
reorder = reorder.takeIf { selected.isNotEmpty() },
onClick = {
// In selection mode a tap is a selection, so the reader is
// never one mis-tap away from opening a session they were only
// picking rows for.
if (selected.isEmpty()) onOpen(session)
else
selected =
if (session.id in selected) selected - session.id
else selected + session.id
},
onLongPress = { selected = selected + session.id },
)
Spacer(Modifier.height(12.dp))
}
@@ -135,70 +297,114 @@ fun SessionListScreen(
}
}
// Over the list rather than above it: a bar that appears in the flow moves every row down
// by its own height at the moment the reader is looking at them.
if (reloading) {
LinearProgressIndicator(Modifier.align(Alignment.TopCenter).fillMaxWidth())
}
// Beside nothing in particular, because a selection is not one row: the options that act on
// it belong to the screen, and the bottom is where a thumb already is.
if (selected.isNotEmpty()) {
val picked =
(listState as? LoadState.Loaded)?.value?.filter { it.id in selected }.orEmpty()
SessionSelectionBar(
count = picked.size,
modifier =
Modifier.align(Alignment.BottomCenter).onSizeChanged {
barHeight = with(density) { it.height.toDp() }
},
onDelete = { confirmingDelete = picked },
)
}
// Above the bar when there is one, by what that bar measured: the button stays rather than
// coming and going, since an absent control cannot say whether there was nothing to do.
FloatingActionButton(
onClick = onSpawn,
modifier = Modifier.align(Alignment.BottomEnd).padding(24.dp),
modifier =
Modifier.align(Alignment.BottomEnd)
.padding(end = BUTTON_RING, bottom = BUTTON_RING + covered)
.onSizeChanged { buttonHeight = with(density) { it.height.toDp() } },
) {
Text("+", style = MaterialTheme.typography.headlineMedium)
}
}
confirmingDelete?.let { session ->
// Reset per session, so a toggle turned on for one conversation is not still on for the
// next. Off to begin with: see [deleteSession].
var alsoDeleteForeign by remember(session.id) { mutableStateOf(false) }
val targets = confirmingDelete
if (targets.isNotEmpty()) {
// Reset per selection, so a toggle turned on for one set of conversations is not still
// on for the next. Off to begin with: see [deleteSession].
var alsoDeleteForeign by remember(targets) { mutableStateOf(false) }
// Whichever of these keep a transcript of their own decide what the sentences below say,
// and whether the switch is offered at all. Old servers reported only the capability, when
// Claude Code was its sole owner.
val owned = targets.filter { it.keepsOwnTranscript }
val transcriptOwner = owned.firstOrNull()?.ownTranscriptName ?: "Claude Code"
AlertDialog(
onDismissRequest = { confirmingDelete = null },
title = { Text("Delete \"${session.title}\"?") },
onDismissRequest = { confirmingDelete = emptyList() },
title = {
Text(
if (targets.size == 1) "Delete \"${targets.first().title}\"?"
else "Delete ${targets.size} sessions?"
)
},
text = {
// Two different acts behind one button, so it says which one this is. What
// separates them is whether the *driver* keeps its own record of the conversation
// -- the Claude Code CLI does, whether this app spawned the session or imported it;
// separates them is whether the *driver* keeps its own record of the
// conversation
// -- the coding CLIs do, whether this app spawned the session or imported it;
// echo and llama.cpp do not.
//
// This used to branch on `imported`, above a comment asserting that "a session
// started here has no copy anywhere". That was false for every claude-cli session
// started here has no copy anywhere". That was false for every coding-CLI session
// this app spawned, and getting it wrong in that direction is the expensive one:
// "this can't be undone", said of something that can, spends the credibility the
// "this can't be undone", said of something that can, spends the credibility that
// sentence needs.
//
// Neither branch promises a restore. The recoverable one says what is known -- the
// driver keeps its own record -- rather than that the file is still there, and it
// names what goes either way, because this app's transcript holds images, peer
// messages and commands the CLI's own record never had.
// Neither branch promises a restore. The recoverable one says what is known,
// that the driver keeps its own record, rather than that the file is still there,
// and it names what goes either way, because this app's transcript holds images,
// peer messages and commands the CLI's own record never had.
//
// A selection takes the sentence that covers all of it: "some of these" is what
// makes the mixed case true without either half of it being read as a promise
// about every row.
Column {
Text(
when {
!session.keepsOwnTranscript ->
owned.isEmpty() ->
"Kills the process and deletes the conversation. Nothing else " +
"keeps a copy, so this can't be undone."
// The sentence below is the one the toggle makes false, which is why it
// is written twice rather than appended to: leaving "should still be
// there to import again" on screen beside a switch that removes it is
// the reassurance being read at the moment it stops being true.
// The sentence below is the one the toggle makes false, which is
// why it is written twice rather than appended to: "should still be
// there to import again", left on screen beside a switch that removes
// it, is the reassurance being read as it stops being true.
alsoDeleteForeign ->
"Kills the process and deletes both copies of the conversation: " +
"this app's, and Claude Code's own transcript on the " +
"this app's, and $transcriptOwner's own transcript on the " +
"machine. Nothing keeps another, so this can't be undone."
else ->
"Stops the process and deletes this app's copy of the " +
"conversation, including any images, peer messages and " +
"commands recorded only here. Claude Code keeps its own " +
"transcript on the machine, so the conversation itself " +
"should still be there to import again."
"commands recorded only here. $transcriptOwner keeps its own " +
"transcript on the machine" +
(if (owned.size < targets.size) " for some of these" else "") +
", so the conversation itself should still be there to " +
"import again."
}
)
// Only where there is a second copy to decide about. Absent rather than
// disabled, because this is not a capability being withheld: for echo and
// llama.cpp there is no other transcript, and a switch offering to delete one
// would be asking about something that does not exist.
if (session.keepsOwnTranscript) {
if (owned.isNotEmpty()) {
Spacer(Modifier.height(16.dp))
// Its own row rather than beside the paragraph: a switch is taller than a
// line of text and re-centres whatever shares a row with it.
// Its own row rather than beside the paragraph: a switch is taller
// than a line of text and re-centres whatever shares a row with it.
Row(verticalAlignment = Alignment.CenterVertically) {
Text(
"Delete Claude Code's transcript too",
"Delete $transcriptOwner's transcript too",
style = MaterialTheme.typography.bodyMedium,
modifier = Modifier.weight(1f),
)
@@ -214,38 +420,8 @@ fun SessionListScreen(
confirmButton = {
TextButton(
onClick = {
confirmingDelete = null
// Marked here rather than after the request returns: the row has to say
// something is happening to it from the moment it is asked for.
deleting = deleting + session.id
deleteErrors = deleteErrors - session.id
scope.launch {
try {
withContext(Dispatchers.IO) {
deleteSession(settings, session.id, alsoDeleteForeign)
// After it succeeded, not before: a refused delete leaves the
// session exactly as it was, and its transcript with it.
transcriptCache.session(session.id).purge()
}
// Only this row, and only what changed. Refetching the list instead
// put every other session back through loading and handed the
// reader an empty screen, to report on something never in doubt.
val loaded = listState
if (loaded is LoadState.Loaded) {
listState =
LoadState.Loaded(
loaded.value.filterNot { it.id == session.id }
)
}
} catch (e: ApiException) {
// Kept, because it is still there: the server refused, so the
// session it refused about is exactly as it was.
deleteErrors =
deleteErrors + (session.id to (e.message ?: "Delete failed"))
} finally {
deleting = deleting - session.id
}
}
confirmingDelete = emptyList()
deleteChosen(targets, alsoDeleteForeign)
}
) {
// Coloured by consequence: this takes something away, and does so wherever it
@@ -254,12 +430,54 @@ fun SessionListScreen(
}
},
dismissButton = {
TextButton(onClick = { confirmingDelete = null }) { Text("Cancel") }
TextButton(onClick = { confirmingDelete = emptyList() }) { Text("Cancel") }
},
)
}
}
/**
* The ring of space inside a session's card, which is also what its handle leaves around itself.
*/
private val CARD_PADDING = 16.dp
/** The gap the spawn button keeps from the edges it floats over, and from the list above it. */
private val BUTTON_RING = 24.dp
/**
* What can be done to the sessions that are selected.
*
* Delete only, for now, which is the one thing this screen has ever done to a session from the list
* rather than from inside it. The same bar as the import tab's, down to the wording of the count.
*/
@Composable
private fun SessionSelectionBar(
count: Int,
modifier: Modifier = Modifier,
onDelete: () -> Unit,
) {
Surface(
modifier = modifier.fillMaxWidth(),
color = MaterialTheme.colorScheme.surfaceContainerHigh,
tonalElevation = 3.dp,
) {
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth().padding(horizontal = 16.dp, vertical = 8.dp),
) {
Text(
"$count selected",
style = MaterialTheme.typography.bodyMedium,
color = MaterialTheme.colorScheme.onSurfaceVariant,
modifier = Modifier.weight(1f),
)
TextButton(onClick = onDelete) {
Text("Delete", color = MaterialTheme.colorScheme.error)
}
}
}
}
@OptIn(ExperimentalFoundationApi::class)
@Composable
private fun SessionCard(
@@ -274,62 +492,118 @@ private fun SessionCard(
* about a request that has not been answered yet.
*/
deleting: Boolean,
onOpen: () -> Unit,
/** Whether this row is one of the selection the bottom bar acts on. */
picked: Boolean,
/** The drag this row can be moved by, or null where the list is not in selection mode. */
reorder: Reorder?,
onClick: () -> Unit,
onLongPress: () -> Unit,
) {
val held = reorder?.held == session.id
BusyItem(label = if (deleting) "deleting" else null) {
Card(
// Off while the delete is in flight: a card that still opens a session it is deleting
// is a race the reader can start by tapping. On the card rather than in [BusyItem],
// which leaves gestures alone so the list still scrolls.
Modifier.fillMaxWidth()
.combinedClickable(
enabled = !deleting,
onClick = onOpen,
onLongClick = onLongPress,
)
colors =
if (picked)
CardDefaults.cardColors(
containerColor = MaterialTheme.colorScheme.secondaryContainer,
contentColor = MaterialTheme.colorScheme.onSecondaryContainer,
)
else CardDefaults.cardColors(),
// Lifted while it is in hand, which is the one cue that says this row is being carried
// rather than sitting where it belongs.
elevation = CardDefaults.cardElevation(defaultElevation = if (held) 8.dp else 0.dp),
modifier =
Modifier.fillMaxWidth()
// Drawn where the finger has taken it, above the rows it is passing over. Both
// in the layer rather than in the layout, so nothing around it moves and the
// list does not remeasure per frame of a drag.
.zIndex(if (held) 1f else 0f)
.graphicsLayer { translationY = reorder?.offsetOf(session.id) ?: 0f },
) {
Column(Modifier.padding(16.dp)) {
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
Row(verticalAlignment = Alignment.CenterVertically) {
Column(
// Everything but the handle, which is what makes the two gestures separate
// rather than competing: a press that lands on the handle never reaches this,
// so holding it cannot select the row it is about to move. The card had the
// click while the handle was the only thing inside it that did not want one,
// and a hold on the handle then both selected the row and ate the drag.
//
// Off while the delete is in flight: a card that still opens a session it is
// deleting is a race the reader can start by tapping. Here rather than in
// [BusyItem], which leaves gestures alone so the list still scrolls.
Modifier.weight(1f)
.combinedClickable(
enabled = !deleting,
onClick = onClick,
onLongClick = onLongPress,
)
.padding(CARD_PADDING)
) {
Text(
session.title,
style = MaterialTheme.typography.titleMedium,
modifier = Modifier.weight(1f),
)
StatusText(session.status)
}
Spacer(Modifier.height(4.dp))
Row(modifier = Modifier.fillMaxWidth()) {
Text(
// Machine, then what runs on it, then what it is set to: the same order and
// separator as the session screen's header and the usage dialog, so one
// pair of facts is not written three ways.
listOfNotNull(
session.setupName,
session.provider,
session.model?.let { modelLabel(it) },
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
Text(
session.title,
style = MaterialTheme.typography.titleMedium,
modifier = Modifier.weight(1f),
)
StatusText(session.status)
if (session.backgroundTasks > 0) {
Text(
backgroundTaskLabel(session.backgroundTasks),
style = MaterialTheme.typography.labelLarge,
color = MaterialTheme.colorScheme.onSurfaceVariant,
modifier = Modifier.padding(start = 8.dp),
)
.joinToString(" · "),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
modifier = Modifier.weight(1f),
)
Text(
relativeTime(session.lastActivity),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
}
Spacer(Modifier.height(4.dp))
Row(modifier = Modifier.fillMaxWidth()) {
Text(
// Machine, then what runs on it, then what it is set to: the same order
// and separator as the session screen's header and the usage dialog, so
// one pair of facts is not written three ways.
listOfNotNull(
session.machineName,
session.provider,
session.model?.let { modelLabel(it) },
)
.joinToString(" · "),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
modifier = Modifier.weight(1f),
)
Text(
relativeTime(session.lastActivity),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
error?.let {
Spacer(Modifier.height(8.dp))
// The server's own words, unprefixed, the way every other failure is shown.
Text(
it,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.error,
)
}
}
error?.let {
Spacer(Modifier.height(8.dp))
// The server's own words, unprefixed, the way every other failure is shown.
Text(
it,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.error,
// Inside the card, so what it moves is the thing it is drawn on. Nothing is held
// open for it outside selection mode: the row is then the row it always was.
if (reorder != null) {
ReorderHandle(
reorder,
session.id,
// Dimmed with the rest of the row while something is happening to it, since
// a row on its way out is not one to rearrange -- see [BusyItem], whose
// appearance this matches rather than repeating its dimming rule.
// The mark lines up with the text on the other side of the card,
// which means taking the square it is centred in off the gap: see
// [HANDLE_MARGIN].
Modifier.alpha(if (deleting) 0.4f else 1f)
.padding(end = CARD_PADDING - HANDLE_MARGIN),
)
}
}
@@ -339,18 +613,8 @@ private fun SessionCard(
@Composable
fun StatusText(status: String) {
val (label, color) =
when (status) {
"awaitingInput" -> "your turn" to awaitingColor
"running" -> "running" to runningColor
"compacting" -> "compacting" to commandColor
"exited" -> "exited" to MaterialTheme.colorScheme.onSurfaceVariant
// Said in words, because it differs in kind from the others rather than in degree: the
// session is not idle and has not exited, nobody has been able to find out which. A
// muted colour alone would read as one of the quiet states.
"unknown" -> "can't tell" to MaterialTheme.colorScheme.onSurfaceVariant
else -> status to MaterialTheme.colorScheme.onSurfaceVariant
}
val label = sessionStatusWord(status)
val color = sessionStatusColour(status)
Row(verticalAlignment = Alignment.CenterVertically) {
if (sessionWorking(status)) {
// The same colour as the word beside it: the two are one signal, and a spinner in the
app/androidApp/src/main/kotlin/com/example/aiapp/SessionScreen.kt
+1417 -578
View File
File diff suppressed because it is too large. Load diff
app/androidApp/src/main/kotlin/com/example/aiapp/SessionSettingsDialog.kt
-331
View File
@@ -1,331 +0,0 @@
package com.example.aiapp
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.height
import androidx.compose.foundation.layout.width
import androidx.compose.foundation.text.KeyboardActions
import androidx.compose.foundation.text.KeyboardOptions
import androidx.compose.material3.AlertDialog
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.OutlinedTextField
import androidx.compose.material3.Switch
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.text.input.ImeAction
import androidx.compose.ui.unit.dp
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.launch
import kotlinx.coroutines.withContext
/**
* What can be changed about one session, as opposed to about this app.
*
* Over the session rather than a step down from it: everything here is about the conversation
* behind it, and a dialog keeps that conversation on screen while it is being adjusted. It was a
* screen of its own until 2026-08-30, which put a page transition and a back stack around two
* controls and hid the thing they act on.
*
* The model and the permission mode are deliberately still on the session's own bar, because those
* are changed *while* reading a turn -- "not this model, try that one".
*
* Captions are for what a control costs rather than for what it is. A paragraph under every control
* made the dialog longer than the conversation it covers -- so Notifications has none, while Move
* and Reload do, because what those two take away is not visible from here.
*/
@Composable
fun SessionSettingsDialog(
settings: ServerSettings,
sessionId: String,
/**
* What the session is called now, as the screen behind this knows it -- see the rename below.
*/
title: String,
onRenamed: (String) -> Unit,
/**
* What this phone is holding of the conversation, or null while that is being measured -- see
* the Reload row below, which is what would discard it.
*/
cachedBytes: Long?,
onReload: () -> Unit,
onDismiss: () -> Unit,
/**
* Copies what this session costs to draw. Built by the session screen, because everything it
* measures is that screen's own state.
*/
onCopyRenderReport: () -> Unit,
) {
val scope = rememberCoroutineScope()
var name by remember(sessionId) { mutableStateOf(title) }
var saving by remember { mutableStateOf(false) }
var error by remember { mutableStateOf<String?>(null) }
// Null until the server has been asked. The row this dialog was opened over is a snapshot of
// whenever the list was last fetched, so drawing the switch straight from it would show a
// position that may have been changed since. Until the answer arrives the switch is disabled
// and a spinner sits beside it, which is what not knowing looks like.
var notify by remember(sessionId) { mutableStateOf<Boolean?>(null) }
var notifyError by remember { mutableStateOf<String?>(null) }
// Where the session works. Null until the server has been asked, for the same reason the switch
// above is. An empty answer is a session that was never given a directory, which is not the
// same as one whose directory is unknown -- the field is only enabled once one of those is
// settled.
var cwd by remember(sessionId) { mutableStateOf<String?>(null) }
var typedCwd by remember(sessionId) { mutableStateOf("") }
var cwdError by remember { mutableStateOf<String?>(null) }
var movingCwd by remember { mutableStateOf(false) }
LaunchedEffect(sessionId) {
try {
val fresh = withContext(Dispatchers.IO) { fetchSession(settings, sessionId) }
notify = fresh.notify
cwd = fresh.cwd.orEmpty()
typedCwd = fresh.cwd.orEmpty()
} catch (e: ApiException) {
// Left unknown rather than falling back to the stale row: the switch stays disabled,
// instead of offering a position nothing confirmed.
notifyError = e.message
notify = null
}
}
/**
* Moves the session, which ends the process that is in the old directory.
*
* Said plainly beside the field rather than confirmed in a second dialog: what it costs is a
* process, and a stopped session is a state this app already has a word and a button for.
*/
fun moveCwd() {
val chosen = typedCwd.trim()
if (movingCwd || chosen.isEmpty() || chosen == cwd) return
movingCwd = true
cwdError = null
scope.launch {
try {
withContext(Dispatchers.IO) { setSessionCwd(settings, sessionId, chosen) }
cwd = chosen
} catch (e: ApiException) {
// Where it happened: this field is the only thing on screen that knows a move was
// asked for, and the reason is usually the path itself.
cwdError = e.message
} finally {
movingCwd = false
}
}
}
// Moved optimistically so the switch answers the finger that moved it, and put back if the
// request is refused -- a switch that waits for a round trip reads as broken on a slow tunnel,
// and one that stays moved after a refusal lies.
fun setNotify(wanted: Boolean) {
val was = notify
notify = wanted
notifyError = null
scope.launch {
try {
withContext(Dispatchers.IO) { setSessionNotify(settings, sessionId, wanted) }
} catch (e: ApiException) {
notify = was
notifyError = e.message
}
}
}
// Nothing to do when the name has not changed, so the button says so rather than sending a
// request whose success would look exactly like the failure of having typed nothing.
val changed = name.trim().isNotEmpty() && name.trim() != title
fun save() {
if (!changed || saving) return
val chosen = name.trim()
saving = true
error = null
scope.launch {
try {
withContext(Dispatchers.IO) { renameSession(settings, sessionId, chosen) }
onRenamed(chosen)
} catch (e: ApiException) {
// Reported here, where it happened, because this dialog is the only place that
// knows a rename was attempted.
error = e.message
saving = false
}
}
}
AlertDialog(
onDismissRequest = onDismiss,
title = { Text("Session settings") },
text = {
Column {
OutlinedTextField(
value = name,
onValueChange = { name = it },
label = { Text("Name") },
singleLine = true,
enabled = !saving,
modifier = Modifier.fillMaxWidth(),
// The keyboard's own action does what the button does: a one-field form where
// the return key does nothing is a form people press return at anyway.
keyboardOptions = KeyboardOptions(imeAction = ImeAction.Done),
keyboardActions = KeyboardActions(onDone = { save() }),
)
Spacer(Modifier.height(8.dp))
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
Glyph(BELL_GLYPH, colour = MaterialTheme.colorScheme.onSurface)
Spacer(Modifier.width(8.dp))
Text("Notifications", modifier = Modifier.weight(1f))
if (notify == null && notifyError == null) {
CircularProgressIndicator(
modifier = Modifier.width(16.dp).height(16.dp),
strokeWidth = 2.dp,
)
Spacer(Modifier.width(8.dp))
}
Switch(
checked = notify == true,
onCheckedChange = { setNotify(it) },
enabled = notify != null,
)
}
// Beside the switch that failed, not with the rename's error: they are two requests
// and a reader has to be able to tell which one the server refused.
notifyError?.let {
Text(
it,
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
)
}
Spacer(Modifier.height(8.dp))
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
OutlinedTextField(
value = typedCwd,
onValueChange = { typedCwd = it },
label = { Text("Working directory") },
// What the field cannot say by being empty: a session that was never given
// one starts wherever its launcher does, and this names that rather than
// showing a path nobody chose.
placeholder = { Text("wherever the session was started") },
singleLine = true,
enabled = cwd != null && !movingCwd,
modifier = Modifier.weight(1f),
keyboardOptions = KeyboardOptions(imeAction = ImeAction.Done),
keyboardActions = KeyboardActions(onDone = { moveCwd() }),
)
TextButton(
onClick = { moveCwd() },
enabled =
cwd != null &&
!movingCwd &&
typedCwd.trim().isNotEmpty() &&
typedCwd.trim() != cwd,
) {
Text(if (movingCwd) "Moving..." else "Move")
}
}
// The whole of what pressing Move does, where it is about to be pressed. A
// directory is settled when the process is spawned, so it is ended and the next
// thing said to the session starts it in the new place.
Text(
"Moving stops the session's process. It starts again in the new directory " +
"with the next message, or with Start.",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
cwdError?.let {
Text(
it,
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
)
}
Spacer(Modifier.height(8.dp))
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
Text("Transcript", modifier = Modifier.weight(1f))
// The size is what the button discards, and the unknown state is drawn rather
// than guessed: a spinner while the directory is being measured, and words when
// there is nothing there, because "nothing cached" and "0 B" read as different
// claims.
when {
cachedBytes == null ->
CircularProgressIndicator(
modifier = Modifier.width(16.dp).height(16.dp),
strokeWidth = 2.dp,
)
else ->
Text(
humanSize(cachedBytes)?.let { "$it cached" } ?: "nothing cached",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
Spacer(Modifier.width(12.dp))
// Enabled whether or not anything is cached: "what I see disagrees with the
// machine" is a state an empty cache can be in too, and a control that comes
// and goes makes its own presence the signal.
TextButton(onClick = onReload) { Text("Reload") }
}
// Captioned, unlike the controls above it, for the same reason Move is: what it
// costs is not visible, and neither is the case it exists for.
Text(
"Reload throws away this phone's copy and fetches the transcript from the " +
"server again. Use it when what is shown here disagrees with the file " +
"on the machine.",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
error?.let {
Spacer(Modifier.height(8.dp))
Text(
it,
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
)
}
Spacer(Modifier.height(8.dp))
// About this session, which is what everything in here is -- and it was on the
// header until 2026-09-03, where the folder button now is. It copies rather than
// opening anything, so it says so and then says it happened: a row that looks like
// a control and gives no sign of having run is one people press twice.
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
Glyph(SPEED_GLYPH, colour = MaterialTheme.colorScheme.onSurface)
Spacer(Modifier.width(8.dp))
Text("Render timings", modifier = Modifier.weight(1f))
TextButton(onClick = onCopyRenderReport) { Text("Copy") }
}
}
},
// Disabled rather than absent while there is nothing to save: a button that comes and goes
// makes its own presence the signal, and its absence cannot say why.
confirmButton = {
TextButton(onClick = { save() }, enabled = changed && !saving) {
Text(if (saving) "Saving..." else "Save")
}
},
dismissButton = { TextButton(onClick = onDismiss) { Text("Close") } },
)
}
app/androidApp/src/main/kotlin/com/example/aiapp/SessionSettingsScreen.kt
+685
View File
@@ -0,0 +1,685 @@
package com.example.aiapp
import androidx.activity.compose.BackHandler
import androidx.compose.foundation.layout.Arrangement
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.fillMaxSize
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.height
import androidx.compose.foundation.layout.imePadding
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.layout.width
import androidx.compose.foundation.rememberScrollState
import androidx.compose.foundation.text.KeyboardActions
import androidx.compose.foundation.text.KeyboardOptions
import androidx.compose.foundation.verticalScroll
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Surface
import androidx.compose.material3.Switch
import androidx.compose.material3.Tab
import androidx.compose.material3.TabRow
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableIntStateOf
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.text.input.ImeAction
import androidx.compose.ui.unit.dp
import java.time.Instant
import java.time.ZoneId
import java.time.format.DateTimeFormatter
import java.time.format.FormatStyle
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.launch
import kotlinx.coroutines.withContext
/**
* What can be changed about one session, and about the provider serving it.
*
* A screen rather than a dialog, again, and for the reason the dialog was chosen in the first place
* turned around: it has outgrown one. A Material dialog constrains its own height and scrolls
* inside itself, so a form of a dozen settings is read through a letterbox that also covers the
* conversation it is about -- and there is nowhere in it to put a second tab. Drawn over the
* session rather than as a `Screen` of its own, so the session under it stays composed and its
* stream keeps flowing; the back gesture closes it.
*
* **Two tabs, and the second is not a copy.** It is [ProviderScreen] -- the same composable the
* machines tab opens, for this session's machine and provider. A session's settings and its
* provider's are different things with different owners (one rides on a request, one decides how a
* model is loaded for everybody), and this is the second way in rather than a second version of
* them.
*
* The model and the permission mode are on the session's own bar as well, because those are changed
* *while* reading a turn -- "not this model, try that one". They are here too because that bar is
* one row shared with three actions: a long model name leaves the other picker a few pixels wide,
* and this is where somebody goes looking for a setting anyway.
*
* Captions are for what a control costs rather than for what it is. A paragraph under every control
* made the dialog longer than the conversation it covers -- so Notifications has none, while Move
* and Reload do, because what those two take away is not visible from here.
*/
@Composable
fun SessionSettingsScreen(
settings: ServerSettings,
sessionId: String,
/** Which machine and provider the second tab is about. */
machineId: String,
provider: String,
/**
* What the session is called now, as the screen behind this knows it -- see the rename below.
*/
title: String,
onRenamed: (String) -> Unit,
/**
* How hard the model thinks, or null for the CLI's own default.
*
* Owned by the screen behind this rather than held here, like [title]: this dialog is what
* changes it, and a level kept only for as long as the dialog is open is the old one again the
* next time it is opened.
*
* Not fetched, because unlike the notification switch there is nothing else that changes it:
* the level is this app's to set and the server does not resolve it into something else.
*/
effort: String?,
onEffortChanged: (String?) -> Unit,
/** Whether a level does anything here; the row is left out entirely where it does not. */
takesEffort: Boolean,
/**
* The settings that are one of a list -- the model and the permission mode.
*
* Owned by the screen behind this, like [title] and [effort]: it is what asked the machine what
* the provider offers. Whichever of them this one has no answer for is not in the list, and
* draws no row.
*/
choices: List<SessionChoice>,
/**
* The settings this session's provider takes, and what they are set to.
*
* Declared by the server rather than listed here -- see [ProviderParamFields]. Empty for a
* provider with none, which draws no section at all.
*/
paramSpecs: List<ParamSpec>,
params: Map<String, String>,
onParamsChanged: (Map<String, String>) -> Unit,
/**
* What this phone is holding of the conversation, or null while that is being measured -- see
* the Reload row below, which is what would discard it.
*/
cachedBytes: Long?,
/**
* How big the record on the server is, or null where it did not say. The other half of the pair
* beside it: what the conversation costs there, against what this phone is holding of it.
*/
transcriptBytes: Long?,
onReload: () -> Unit,
/**
* Opens the transcript file itself in the explorer. Null from a server that does not say where
* it is, which draws no button rather than one that cannot work.
*/
onViewRaw: (() -> Unit)?,
onDismiss: () -> Unit,
/**
* Copies what this session costs to draw. Built by the session screen, because everything it
* measures is that screen's own state.
*/
onCopyRenderReport: () -> Unit,
) {
val scope = rememberCoroutineScope()
var name by remember(sessionId) { mutableStateOf(title) }
var effortError by remember { mutableStateOf<String?>(null) }
var saving by remember { mutableStateOf(false) }
var error by remember { mutableStateOf<String?>(null) }
// Null until the server has been asked. The row this dialog was opened over is a snapshot of
// whenever the list was last fetched, so drawing the switch straight from it would show a
// position that may have been changed since. Until the answer arrives the switch is disabled
// and a spinner sits beside it, which is what not knowing looks like.
var notify by remember(sessionId) { mutableStateOf<Boolean?>(null) }
var notifyError by remember { mutableStateOf<String?>(null) }
// The same three-state shape the notification switch has, for the same reason: until the
// server has answered, the switch is disabled rather than showing a position nothing confirmed.
var autoResume by remember(sessionId) { mutableStateOf<Boolean?>(null) }
var resumeMessage by remember(sessionId) { mutableStateOf(DEFAULT_RESUME_MESSAGE) }
// When the server next intends to ask whether the limit has lifted, or null when nothing is
// waiting. Read once with everything else: it moves on the server's schedule, not this
// screen's, and a figure that redrew itself here would be this app re-measuring what it was
// told.
var resumeAt by remember(sessionId) { mutableStateOf<Double?>(null) }
var resumeError by remember { mutableStateOf<String?>(null) }
// Where the session works. Null until the server has been asked, for the same reason the switch
// above is. An empty answer is a session that was never given a directory, which is not the
// same as one whose directory is unknown -- the field is only enabled once one of those is
// settled.
var cwd by remember(sessionId) { mutableStateOf<String?>(null) }
var typedCwd by remember(sessionId) { mutableStateOf("") }
var cwdError by remember { mutableStateOf<String?>(null) }
var movingCwd by remember { mutableStateOf(false) }
// The two settings on this screen that end the session's process, held while the reader is
// asked whether that is what they meant. Null is nobody being asked.
var askedCwd by remember(sessionId) { mutableStateOf<String?>(null) }
var askedEffort by remember(sessionId) { mutableStateOf<String?>(null) }
LaunchedEffect(sessionId) {
try {
val fresh = withContext(Dispatchers.IO) { fetchSession(settings, sessionId) }
notify = fresh.notify
autoResume = fresh.autoResume
resumeMessage = fresh.autoResumeMessage
resumeAt = fresh.resumeAt
cwd = fresh.cwd.orEmpty()
typedCwd = fresh.cwd.orEmpty()
} catch (e: ApiException) {
// Left unknown rather than falling back to the stale row: the switch stays disabled,
// instead of offering a position nothing confirmed.
notifyError = e.message
notify = null
resumeError = e.message
autoResume = null
}
}
/**
* Moves the session, which ends the process that is in the old directory.
*
* Only ever reached through [RestartDialog], which is where what it costs is said -- see
* `askToMove`.
*/
fun askToMove() {
val chosen = typedCwd.trim()
if (movingCwd || chosen.isEmpty() || chosen == cwd) return
askedCwd = chosen
}
fun moveCwd() {
val chosen = typedCwd.trim()
if (movingCwd || chosen.isEmpty() || chosen == cwd) return
movingCwd = true
cwdError = null
scope.launch {
try {
withContext(Dispatchers.IO) { setSessionCwd(settings, sessionId, chosen) }
cwd = chosen
} catch (e: ApiException) {
// Where it happened: this field is the only thing on screen that knows a move was
// asked for, and the reason is usually the path itself.
cwdError = e.message
} finally {
movingCwd = false
}
}
}
/**
* Chooses a thinking level, which ends the process the old level was launched with. Asked for
* first, the same way a move is.
*
* Put back if the request is refused, for the reason the notification switch below gives: a
* control that stays where it was put after a refusal is stating something untrue.
*/
fun setEffort(chosen: String?) {
val was = effort
onEffortChanged(chosen)
effortError = null
scope.launch {
try {
withContext(Dispatchers.IO) { setSessionEffort(settings, sessionId, chosen) }
} catch (e: ApiException) {
onEffortChanged(was)
effortError = e.message
}
}
}
// Moved optimistically so the switch answers the finger that moved it, and put back if the
// request is refused -- a switch that waits for a round trip reads as broken on a slow tunnel,
// and one that stays moved after a refusal lies.
fun setNotify(wanted: Boolean) {
val was = notify
notify = wanted
notifyError = null
scope.launch {
try {
withContext(Dispatchers.IO) { setSessionNotify(settings, sessionId, wanted) }
} catch (e: ApiException) {
notify = was
notifyError = e.message
}
}
}
/**
* Turns auto-resume on or off, or changes what it would say.
*
* One request for both, because the server takes one: switching it on and typing the message
* are two halves of the same decision, and sending them separately would leave a moment where
* the session is armed with the old words.
*
* Put back if refused, like the notification switch. Turning it off also clears what was
* scheduled -- said here rather than only on the server, or the row would go on naming a time
* that no longer exists.
*/
fun setAutoResume(on: Boolean, message: String) {
val wasOn = autoResume
val wasMessage = resumeMessage
val wasAt = resumeAt
autoResume = on
resumeMessage = message
if (!on) resumeAt = null
resumeError = null
scope.launch {
try {
withContext(Dispatchers.IO) {
setSessionAutoResume(settings, sessionId, on, message)
}
} catch (e: ApiException) {
autoResume = wasOn
resumeMessage = wasMessage
resumeAt = wasAt
resumeError = e.message
}
}
}
// Nothing to do when the name has not changed, so the button says so rather than sending a
// request whose success would look exactly like the failure of having typed nothing.
val changed = name.trim().isNotEmpty() && name.trim() != title
fun save() {
if (!changed || saving) return
val chosen = name.trim()
saving = true
error = null
scope.launch {
try {
withContext(Dispatchers.IO) { renameSession(settings, sessionId, chosen) }
onRenamed(chosen)
} catch (e: ApiException) {
// Reported here, where it happened, because this dialog is the only place that
// knows a rename was attempted.
error = e.message
saving = false
}
}
}
// The platform's own way back out of a layer: without it, back falls through to whatever is
// under this and closes the session -- which reads as a crash to somebody who meant to return
// to what they were reading.
BackHandler(onBack = onDismiss)
var tab by remember(sessionId) { mutableIntStateOf(0) }
Surface(Modifier.fillMaxSize()) {
// The keyboard covers the lower half of a form of fields, and this is a screen rather
// than a dialog now -- nothing else is going to move it out of the way.
Column(Modifier.fillMaxSize().imePadding()) {
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth().padding(horizontal = 16.dp),
) {
GlyphButton(BACK_GLYPH, "Back", onDismiss)
Spacer(Modifier.width(GLYPH_BUTTON_MARGIN))
Text(
"Settings",
style = MaterialTheme.typography.titleMedium,
modifier = Modifier.weight(1f),
)
// Disabled rather than absent while there is nothing to save: a button that comes
// and goes makes its own presence the signal, and its absence cannot say why.
TextButton(onClick = { save() }, enabled = changed && !saving) {
Text(if (saving) "Saving..." else "Save")
}
}
// The same two-tab shape the main screen uses for its three, so a reader who has
// learned one has learned the other.
TabRow(selectedTabIndex = tab) {
Tab(selected = tab == 0, onClick = { tab = 0 }, text = { Text("Session") })
Tab(selected = tab == 1, onClick = { tab = 1 }, text = { Text(provider) })
}
if (tab == 1) {
// The machines tab's own screen, with its back control left off: this one has a
// header of its own, and two ways out stacked above each other is a reader asking
// which of them goes where.
ProviderScreen(
settings = settings,
machineId = machineId,
provider = provider,
onBack = null,
)
return@Column
}
Column(
Modifier.verticalScroll(rememberScrollState())
.padding(horizontal = 16.dp)
.padding(top = 12.dp, bottom = 16.dp)
) {
LabelledField(
label = "Name",
value = name,
onValueChange = { name = it },
enabled = !saving,
keyboardOptions = KeyboardOptions(imeAction = ImeAction.Done),
// The keyboard's own action does what the button does: a one-field form where
// the return key does nothing is a form people press return at anyway.
keyboardActions = KeyboardActions(onDone = { save() }),
)
Spacer(Modifier.height(8.dp))
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
Glyph(BELL_GLYPH, colour = MaterialTheme.colorScheme.onSurface)
Spacer(Modifier.width(8.dp))
Text("Notifications", modifier = Modifier.weight(1f))
if (notify == null && notifyError == null) {
CircularProgressIndicator(
modifier = Modifier.width(16.dp).height(16.dp),
strokeWidth = 2.dp,
)
Spacer(Modifier.width(8.dp))
}
Switch(
checked = notify == true,
onCheckedChange = { setNotify(it) },
enabled = notify != null,
)
}
// Beside the switch that failed, not with the rename's error: they are two requests
// and a reader has to be able to tell which one the server refused.
notifyError?.let {
Text(
it,
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
)
}
Spacer(Modifier.height(8.dp))
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
Text("Resume after a usage limit", modifier = Modifier.weight(1f))
if (autoResume == null && resumeError == null) {
CircularProgressIndicator(
modifier = Modifier.width(16.dp).height(16.dp),
strokeWidth = 2.dp,
)
Spacer(Modifier.width(8.dp))
}
Switch(
checked = autoResume == true,
onCheckedChange = { setAutoResume(it, resumeMessage) },
enabled = autoResume != null,
)
}
// Disabled rather than hidden while the switch is off: a field that comes and goes
// makes its own presence the signal, and a visible one teaches what the switch will
// do. Committed on the keyboard's Done rather than on every keystroke, so typing a
// sentence is one request instead of one per letter.
LabelledField(
label = "Message to send",
value = resumeMessage,
onValueChange = { resumeMessage = it },
// What an empty field means: the server's own word rather than a session
// poked with nothing to read.
hint = DEFAULT_RESUME_MESSAGE,
enabled = autoResume == true,
keyboardOptions = KeyboardOptions(imeAction = ImeAction.Done),
keyboardActions =
KeyboardActions(onDone = { setAutoResume(true, resumeMessage) }),
)
// Only where something is actually waiting. Absent is not a state worth a row: a
// session that has not hit a limit has nothing scheduled, which the reader can see
// from the switch.
resumeAt?.let { at ->
Text(
"Waiting now -- next check ${formatCheckTime(at)}.",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
resumeError?.let {
Text(
it,
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
)
}
Spacer(Modifier.height(8.dp))
// The button sits at the bottom of the row rather than centred on it: the field
// beside it is a label above a box, and a control centred against the pair lands
// beside the label rather than beside the thing it acts on.
Row(
verticalAlignment = Alignment.Bottom,
modifier = Modifier.fillMaxWidth(),
) {
LabelledField(
label = "Working directory",
value = typedCwd,
onValueChange = { typedCwd = it },
// What the field cannot say by being empty: a session that was never
// given one starts wherever its launcher does.
hint = "wherever the session was started",
enabled = cwd != null && !movingCwd,
keyboardOptions = KeyboardOptions(imeAction = ImeAction.Done),
keyboardActions = KeyboardActions(onDone = { askToMove() }),
modifier = Modifier.weight(1f),
)
TextButton(
onClick = { askToMove() },
enabled =
cwd != null &&
!movingCwd &&
typedCwd.trim().isNotEmpty() &&
typedCwd.trim() != cwd,
) {
Text(if (movingCwd) "Moving..." else "Move")
}
}
cwdError?.let {
Text(
it,
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
)
}
choices.forEach { choice ->
Spacer(Modifier.height(8.dp))
PickerRow(choice.label, choice.current, choice.options, choice.onPick)
}
// Left out rather than disabled, the one place this dialog does that: a disabled
// control teaches what the thing can do, and a llama session cannot do this at all
// -- the row would be teaching something false about it.
if (takesEffort) {
Spacer(Modifier.height(8.dp))
PickerRow(
"Thinking",
effort ?: DEFAULT_EFFORT,
// The level the CLI picks for itself is in the list as well as in the
// button, so leaving a level is not a one-way trip -- the same correction
// the model picker carries.
listOf(DEFAULT_EFFORT) + EFFORT_LEVELS,
) { chosen ->
askedEffort = chosen
}
effortError?.let {
Text(
it,
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
)
}
}
if (paramSpecs.isNotEmpty()) {
Spacer(Modifier.height(16.dp))
Text(
"Model settings",
style = MaterialTheme.typography.titleSmall,
)
Spacer(Modifier.height(8.dp))
// Edited here and saved by the screen behind this, which is what makes
// typing in a text field affordable: the save is debounced, and a dialog
// dismissed mid-edit would take an unsaved value with it.
ProviderParamFields(
specs = paramSpecs,
values = params,
onChange = onParamsChanged,
// The same picker the model and permission rows above use: how hard a
// session thinks is one kind of setting, whichever provider declares it.
choices = ChoiceStyle.Picker,
)
}
Spacer(Modifier.height(8.dp))
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
Text("Transcript", modifier = Modifier.weight(1f))
// What the conversation costs on the server, and then what Reload would
// discard here -- one line, so the two sizes read as a pair. A server that
// did not measure its file leaves its half out rather than saying zero.
val onServer = transcriptBytes?.let { humanSize(it) ?: "0 B" }
// The unknown state is drawn rather than guessed: a spinner while the cache
// is being measured, and words when there is nothing in it, because "nothing
// cached" and "0 B" read as different claims.
when {
cachedBytes == null -> {
onServer?.let {
Text(
it,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
Spacer(Modifier.width(8.dp))
}
CircularProgressIndicator(
modifier = Modifier.width(16.dp).height(16.dp),
strokeWidth = 2.dp,
)
}
else -> {
val cached =
humanSize(cachedBytes)?.let { "$it cached" } ?: "nothing cached"
Text(
onServer?.let { "$it · $cached" } ?: cached,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
}
}
// Both on a line of their own under what they act on, rather than crowded against
// the size on the line above: two buttons and a measurement do not fit the width
// of a phone, and the one that would lose is the number.
Row(
horizontalArrangement = Arrangement.End,
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
// The record as it is on disk, for the question the drawn conversation cannot
// answer -- which is most of what anybody opens this dialog to debug.
onViewRaw?.let { TextButton(onClick = it) { Text("View raw") } }
Spacer(Modifier.width(8.dp))
// Enabled whether or not anything is cached: "what I see disagrees with the
// machine" is a state an empty cache can be in too, and a control that comes
// and goes makes its own presence the signal.
TextButton(onClick = onReload) { Text("Reload") }
}
error?.let {
Spacer(Modifier.height(8.dp))
Text(
it,
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
)
}
Spacer(Modifier.height(8.dp))
// About this session, which is what everything in here is -- and it was on the
// header until 2026-09-03, where the folder button now is. It copies rather than
// opening anything, so it says so and then says it happened: a row that looks like
// a control and gives no sign of having run is one people press twice.
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
Glyph(SPEED_GLYPH, colour = MaterialTheme.colorScheme.onSurface)
Spacer(Modifier.width(8.dp))
Text("Render timings", modifier = Modifier.weight(1f))
TextButton(onClick = onCopyRenderReport) { Text("Copy") }
}
}
}
}
// A directory is settled when the process is spawned, so moving means ending it. Said here
// rather than under the field, which is the rule the whole screen follows -- see
// [RestartDialog].
askedCwd?.let { chosen ->
RestartDialog(
title = "Move to $chosen?",
text =
"This stops the session's process. It starts again in the new directory with " +
"the next message, or with Start.",
confirm = "Move",
onConfirm = {
askedCwd = null
moveCwd()
},
onDismiss = { askedCwd = null },
)
}
// The same cost for the same reason: the CLI reads the level when it launches, and has no
// control request for changing one.
askedEffort?.let { chosen ->
RestartDialog(
title = "Think $chosen?",
text =
"This stops the session's process. It starts again with the next message, or " +
"with Start.",
confirm = "Change",
onConfirm = {
askedEffort = null
setEffort(chosen.takeIf { it != DEFAULT_EFFORT })
},
onDismiss = { askedEffort = null },
)
}
}
/**
* One session setting that is a choice from a list, as this dialog draws it.
*
* A shape rather than a pair of parameters each, because a provider may offer either of them, both
* or neither, and they are otherwise the same control.
*/
data class SessionChoice(
val label: String,
val current: String,
val options: List<String>,
val onPick: (String) -> Unit,
)
/**
* When the server will next look, as a local time.
*
* A time rather than a countdown, for the reason the transcript's own limit row gives: this screen
* reads the figure once, and a span drawn from a value nothing refreshes goes stale while somebody
* is looking at it.
*/
private fun formatCheckTime(epochSeconds: Double): String =
try {
DateTimeFormatter.ofLocalizedTime(FormatStyle.SHORT)
.withZone(ZoneId.systemDefault())
.format(Instant.ofEpochSecond(epochSeconds.toLong()))
} catch (_: Exception) {
// A time that cannot be read is not a time to show: the sentence above still says a check
// is coming, which is the part the reader can act on.
"soon"
}
app/androidApp/src/main/kotlin/com/example/aiapp/SessionStatusWords.kt
+74
View File
@@ -0,0 +1,74 @@
package com.example.aiapp
import androidx.compose.material3.MaterialTheme
import androidx.compose.runtime.Composable
import androidx.compose.ui.graphics.Color
/**
* What a session's status is called on screen, and what colour it is drawn in.
*
* One pair of functions rather than a branch on each screen that shows a status. There were two,
* and the second silently fell short the moment the server grew a state: `waiting` arrived and the
* session list learned the word and the colour while the session screen's status row printed the
* wire's own word in the muted grey every quiet state uses. That comment already said the words
* were "the session list's own"; this is what makes that true rather than a promise.
*
* A subagent's own three states are deliberately not here -- see `subagentStatusLabel`, which
* collapses everything it does not recognise rather than passing it through, because a subagent has
* fewer states than a session and reporting one it cannot have is worse than reporting none.
*/
fun sessionStatusWord(status: String, subagent: Boolean = false): String =
when (status) {
"idle" -> "idle"
"running" -> "running"
"compacting" -> "compacting"
// Not "running": a model coming off disk is not a model answering, and the difference is
// minutes. Said in its own word so a first message that waits is explained rather than
// looking like a session that has stopped responding. See `SessionStatus::Loading`.
//
// "model" rather than "loading" alone, because there are two waits before an answer and
// the reader is entitled to know which one they are in: this one happens once, and
// "reading prompt" below happens on every turn.
"loading" -> "loading model"
// The model has the prompt and has not started answering. Its own word for the same
// reason: a long conversation spends real time here, and reported as "running" it looked
// like a model thinking. See `SessionStatus::Reading`.
"reading" -> "reading prompt"
// Its own word, because the state it is easily mistaken for means the opposite: "idle"
// invites the reader to type something, and a waiting session is going to carry on without
// them. See `SessionStatus::Waiting`.
"waiting" -> "waiting"
"awaitingInput" -> "your turn"
// A subagent's process was always its parent's, so it had none of its own to merely stop.
"exited" -> if (subagent) "finished" else "exited"
// Said in words, because it differs in kind from the others rather than in degree: the
// session is not idle and has not exited, nobody has been able to find out which. A muted
// colour alone would read as one of the quiet states.
"unknown" -> "can't tell"
// A state this build has never heard of, said as itself. The nearest word we do know would
// read as a fact somebody established.
else -> status
}
fun backgroundTaskLabel(count: Int): String = "$count bg ${if (count == 1) "task" else "tasks"}"
/**
* The colour that goes with [sessionStatusWord]: the accent is spent on the states that are about
* to do something or want something, and every quiet one shares the muted colour.
*
* Stated beside whatever draws it rather than inherited -- a colour that carries meaning has to
* carry its own contrast, since the surface under it will not change to rescue it.
*/
@Composable
fun sessionStatusColour(status: String): Color =
when (status) {
"awaitingInput" -> awaitingColor
"running" -> runningColor
"compacting" -> commandColor
// The same accent as the other states that are busy on their own account, because that is
// what this is: something is happening and nothing is wanted from the reader.
"loading",
"reading" -> commandColor
"waiting" -> waitingColor
else -> MaterialTheme.colorScheme.onSurfaceVariant
}
app/androidApp/src/main/kotlin/com/example/aiapp/SessionUsageBar.kt
+158 -42
View File
@@ -15,6 +15,8 @@ import androidx.compose.runtime.remember
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.draw.drawWithContent
import androidx.compose.ui.geometry.Offset
import androidx.compose.ui.graphics.Color
import androidx.compose.ui.unit.dp
import java.time.Duration
@@ -70,13 +72,22 @@ class UsageFeed(
/** Ask the backend again now. The dialog's refresh button; the poll does it on its own. */
val refresh: () -> Unit,
) {
/** What [setup]'s own limits came back as. See [usageFor] for why the states are these. */
fun forSetup(setup: String): SessionUsage =
when (val state = snapshots) {
/**
* What meters [session], and what that meter came back as. See [usageFor] for the states.
*
* A session rather than a machine, because a machine is not what is metered: one machine runs
* the Claude CLI and an echo session side by side, and only the first of them spends anything.
*/
fun forSession(session: SessionSummary): SessionUsage {
// Settled without asking anybody: a session nothing meters has nothing to check, and
// "checking" is what the fetch's own states would say about it for as long as one is out.
val provider = session.usageProvider ?: return SessionUsage.NotMetered
return when (val state = snapshots) {
is LoadState.Loading -> SessionUsage.Waiting
is LoadState.Error -> SessionUsage.Unavailable(state.message)
is LoadState.Loaded -> usageFor(state.value, setup)
is LoadState.Loaded -> usageFor(state.value, session.machine, provider, session.model)
}
}
}
/**
@@ -116,8 +127,8 @@ fun rememberUsageFeed(settings: ServerSettings): UsageFeed {
*
* Worst rather than the five-hour one, because the button it colours opens *all* of them, and a
* blue icon over a weekly quota at 97% would be the interface answering a question nobody asked.
* Taken over however many windows came back rather than the three Claude sends today -- the backend
* passes windows it does not recognise straight through.
* Taken over however many windows this session's provider returned rather than the three Claude
* sends today -- the backend passes windows it does not recognise straight through.
*
* Every state that is not a measurement takes the ordinary control colour instead. That is the
* point where colour stops being able to help: blue is the low end of a scale here, so colouring an
@@ -133,7 +144,7 @@ fun usageGlyphColour(usage: SessionUsage): Color =
}
/**
* The five-hour window for the machine this session runs on, under the session's own header.
* The shortest usage window for the pool this session uses, under the session's own header.
*
* Here rather than only in the usage dialog because it is the number that decides whether to keep
* going, and it was a screen away from the place that decision gets made. It reports on this
@@ -150,17 +161,17 @@ fun SessionUsageBar(usage: SessionUsage, modifier: Modifier = Modifier) {
// rather than recomputed at draw time: a percentage that comes back unchanged is an equal
// value, Compose skips the recomposition, and a "left" that only ticked when the quota moved
// would sit at a stale figure for hours.
var now by remember { mutableStateOf(OffsetDateTime.now()) }
LaunchedEffect(Unit) {
while (true) {
delay(REFRESH_MS)
now = OffsetDateTime.now()
}
}
val now = rememberUsageNow()
// Nothing at all for a machine that meters nothing: a row saying "unknown" there would report a
// problem about a setup somebody chose, on every screen, forever.
if (usage is SessionUsage.NotMetered) {
// Nothing at all for a session that meters nothing: a row saying "unknown" there would report
// a problem about a machine somebody chose, on every screen, forever.
//
// And nothing while the first fetch is out, which is a different silence. A request in flight
// is not a state to report -- and the session that meters nothing is exactly the one this
// cannot yet tell apart, so "5-hour usage: checking" appeared under an echo session for half a
// second and was then taken away. A row that has to be withdrawn is worse than one that
// arrives late.
if (usage is SessionUsage.NotMetered || usage is SessionUsage.Waiting) {
return
}
@@ -171,24 +182,18 @@ fun SessionUsageBar(usage: SessionUsage, modifier: Modifier = Modifier) {
// Words, not a colour and not an empty bar: every one of these is a different kind of
// answer from "this much is used", and only words carry a difference in kind.
when (val state = usage) {
SessionUsage.NotMetered -> Unit
is SessionUsage.Unavailable -> UsageNote("5-hour usage unknown -- ${state.why}")
SessionUsage.Waiting -> UsageNote("5-hour usage: checking")
// Both handled above, before the row exists at all.
SessionUsage.NotMetered,
SessionUsage.Waiting -> Unit
is SessionUsage.Unavailable -> UsageNote("Usage unknown -- ${state.why}")
is SessionUsage.Known -> {
val window = state.windows.firstOrNull { it.kind == "session" }
val window = shortestUsageWindow(state.windows)
if (window == null) {
UsageNote("5-hour usage unknown -- no five-hour window reported")
UsageNote("Usage unknown -- no window duration was reported")
} else {
LinearProgressIndicator(
progress = { (window.percent / 100.0).toFloat().coerceIn(0f, 1f) },
// The same step at the same percentages as the dialog's bars: this is the
// same measurement, and a reader who learned the colour there has to be
// able to read it here without checking which screen they are on.
color = quotaColor(window.percent),
modifier = Modifier.weight(1f),
)
UsageProgressIndicator(window, now, Modifier.weight(1f))
Text(
fiveHourLabel(window, now),
usageWindowLabel(window, now),
style = MaterialTheme.typography.labelSmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
modifier = Modifier.padding(start = 8.dp),
@@ -199,6 +204,56 @@ fun SessionUsageBar(usage: SessionUsage, modifier: Modifier = Modifier) {
}
}
/** A clock shared by each usage surface, advanced independently of changes to the quota. */
@Composable
internal fun rememberUsageNow(): OffsetDateTime {
var now by remember { mutableStateOf(OffsetDateTime.now()) }
LaunchedEffect(Unit) {
while (true) {
delay(REFRESH_MS)
now = OffsetDateTime.now()
}
}
return now
}
/** The quota fill with a white tick showing how far the current time window has progressed. */
@Composable
internal fun UsageProgressIndicator(
window: UsageWindow,
now: OffsetDateTime,
modifier: Modifier = Modifier,
) {
val elapsed = usageWindowElapsedFraction(window, now)
LinearProgressIndicator(
progress = { (window.percent / 100.0).toFloat().coerceIn(0f, 1f) },
// The same step at the same percentages everywhere: this is the same measurement, and a
// reader who learned the colour on one surface should not have to relearn it on another.
color = quotaColor(window.percent),
modifier =
modifier.drawWithContent {
drawContent()
elapsed?.let { fraction ->
drawLine(
color = Color.White,
start = Offset(size.width * fraction, 0f),
end = Offset(size.width * fraction, size.height),
strokeWidth = 2.dp.toPx(),
)
}
},
)
}
/** Elapsed time divided by the reported window duration, or null when either value is unknown. */
internal fun usageWindowElapsedFraction(window: UsageWindow, now: OffsetDateTime): Float? {
val durationMinutes = window.durationMinutes?.takeIf { it > 0 } ?: return null
val end = windowEnd(window.resetsAt, now) as? WindowEnd.Ends ?: return null
val remainingMinutes =
end.until.seconds.toDouble() / 60.0 + end.until.nano.toDouble() / 60_000_000_000.0
return (1.0 - remainingMinutes / durationMinutes).coerceIn(0.0, 1.0).toFloat()
}
/** Anything this row says instead of drawing a bar, so all of them look the same. */
@Composable
private fun UsageNote(text: String) {
@@ -210,42 +265,103 @@ private fun UsageNote(text: String) {
}
/**
* "42% -- 2h 15m left": how much is gone, then how long what is left has to last.
* "42% -- 2h 15m left / 5h": how much is gone, then how long what is left has to last, then how
* long the whole window is.
*
* The percentage on its own does not answer the question it gets asked, which is whether to start
* something now; 80% with twenty minutes to go and 80% with four hours to go are opposite answers.
*
* The window's *length* is what the provider's own name for it used to carry ("5-hour window"), and
* it is worth more beside the time left than in front of the percentage: "3h 42m left / 5h" says in
* one reading both how much of the cycle is to come and which cycle this is. Where the provider
* reported no duration there is simply nothing after the span -- the name it gave is not a
* measurement of one, so nothing is inferred from it.
*
* The window's end has two missing cases, worded differently on purpose; see [WindowEnd]. A window
* that is not running gets the percentage and nothing else.
*/
private fun fiveHourLabel(window: UsageWindow, now: OffsetDateTime): String {
private fun usageWindowLabel(window: UsageWindow, now: OffsetDateTime): String {
val percent = "${window.percent.toInt()}%"
val outOf =
window.durationMinutes?.takeIf { it > 0 }?.let { " / ${formatMillis(it * 60_000)}" } ?: ""
return when (val end = windowEnd(window.resetsAt, now)) {
// Between blocks the five-hour window has no reset time, and saying so is a fact about
// nothing: there is no window to run out. The percentage is the whole answer.
// Between blocks a window can have no reset time, and saying so is a fact about nothing:
// there is no window to run out. The percentage is the whole answer.
WindowEnd.NotRunning -> percent
WindowEnd.Unreadable -> "$percent · reset time unreadable"
is WindowEnd.Ends ->
// Under a minute, including past the end: the number would round to "0m left", which
// reads as a measurement rather than as the window having run out.
if (end.until < Duration.ofMinutes(1)) "$percent · refresh soon"
else "$percent · ${formatSpan(end.until)} left"
else "$percent · ${formatSpan(end.until)} left$outOf"
}
}
/**
* One machine's snapshot, out of every machine's.
* One meter's snapshot, out of every machine's: [machine]'s row for [provider].
*
* Both halves are needed to pick it. A machine can hold more than one meter -- the Claude CLI's
* account and, while a test has one set, an echo session's invented one -- and a snapshot is one
* service on one machine.
*
* Every way of having *failed* to get numbers is [SessionUsage.Unavailable] with the reason in it.
* None of them may look like zero, and none may look like [SessionUsage.NotMetered], which is the
* machine having no quota rather than the question going unanswered.
*/
fun usageFor(snapshots: List<UsageSnapshot>, setup: String): SessionUsage {
// No snapshot at all means the backend never asked, which it only does for a machine with
// nothing metered on it. That is a different answer from having asked and failed.
val mine = snapshots.firstOrNull { it.setup == setup } ?: return SessionUsage.NotMetered
fun usageFor(
snapshots: List<UsageSnapshot>,
machine: String,
provider: String,
model: String?,
): SessionUsage {
// No snapshot at all means the backend never asked, which it only does where there is nothing
// to ask about. That is a different answer from having asked and failed.
val pools = usageSnapshotsFor(snapshots, machine, provider)
if (pools.isEmpty()) return SessionUsage.NotMetered
val mine =
usagePoolFor(pools, model)
?: return SessionUsage.Unavailable("couldn't tell which usage pool this session uses")
if (mine.state != "ok") {
return SessionUsage.Unavailable(mine.detail ?: mine.state)
val why =
mine.detail
?: when (mine.state) {
"notLoggedIn" -> "no Claude account is signed in on this machine"
"authenticating" -> "Claude sign-in is in progress"
"loginRequired" -> "Claude sign-in is required"
else -> mine.state
}
return SessionUsage.Unavailable(why)
}
return SessionUsage.Known(mine.windows)
}
/** Every billing pool reported for one provider on one machine. */
internal fun usageSnapshotsFor(
snapshots: List<UsageSnapshot>,
machine: String,
provider: String?,
): List<UsageSnapshot> =
if (provider == null) emptyList()
else snapshots.filter { it.machine == machine && it.provider == provider }
/** The pool an explicit model names, or the provider's generic pool for every other model. */
internal fun usagePoolFor(pools: List<UsageSnapshot>, model: String?): UsageSnapshot? {
if (pools.size == 1) return pools.first()
val normalizedModel = model?.normalizedPoolName()
val named = normalizedModel?.let { wanted ->
pools.firstOrNull { pool ->
val name = pool.limitName?.normalizedPoolName()
name == wanted || (wanted.contains("luna") && name == "gptreserve")
}
}
return named ?: pools.firstOrNull { it.limitId == "codex" }
}
/** The shortest cycle the selected pool actually reported. */
internal fun shortestUsageWindow(windows: List<UsageWindow>): UsageWindow? =
windows
.mapNotNull { window -> window.durationMinutes?.let { duration -> duration to window } }
.minByOrNull { it.first }
?.second
private fun String.normalizedPoolName(): String = lowercase().filter(Char::isLetterOrDigit)
app/androidApp/src/main/kotlin/com/example/aiapp/SettingsScreen.kt
+4 -19
View File
@@ -15,7 +15,6 @@ import androidx.compose.foundation.layout.width
import androidx.compose.material3.Button
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.OutlinedButton
import androidx.compose.material3.OutlinedTextField
import androidx.compose.material3.Text
import androidx.compose.runtime.Composable
import androidx.compose.runtime.getValue
@@ -125,28 +124,14 @@ fun SettingsScreen(
}
Spacer(Modifier.height(16.dp))
OutlinedTextField(
value = host,
onValueChange = { host = it },
label = { Text("Host") },
singleLine = true,
modifier = Modifier.fillMaxWidth(),
)
LabelledField(label = "Host", value = host, onValueChange = { host = it })
Spacer(Modifier.height(8.dp))
OutlinedTextField(
value = port,
onValueChange = { port = it },
label = { Text("Port") },
singleLine = true,
modifier = Modifier.fillMaxWidth(),
)
LabelledField(label = "Port", value = port, onValueChange = { port = it })
Spacer(Modifier.height(8.dp))
OutlinedTextField(
LabelledField(
label = if (existing != null) "Token (unchanged if left blank)" else "Token",
value = token,
onValueChange = { token = it },
label = { Text(if (existing != null) "Token (unchanged if left blank)" else "Token") },
singleLine = true,
modifier = Modifier.fillMaxWidth(),
)
Spacer(Modifier.height(24.dp))
app/androidApp/src/main/kotlin/com/example/aiapp/SidePanels.kt
+250
View File
@@ -0,0 +1,250 @@
package com.example.aiapp
import androidx.activity.compose.BackHandler
import androidx.compose.animation.core.Animatable
import androidx.compose.foundation.background
import androidx.compose.foundation.clickable
import androidx.compose.foundation.gestures.Orientation
import androidx.compose.foundation.gestures.draggable
import androidx.compose.foundation.gestures.rememberDraggableState
import androidx.compose.foundation.layout.Box
import androidx.compose.foundation.layout.BoxScope
import androidx.compose.foundation.layout.BoxWithConstraints
import androidx.compose.foundation.layout.fillMaxHeight
import androidx.compose.foundation.layout.fillMaxSize
import androidx.compose.foundation.layout.width
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Surface
import androidx.compose.runtime.Composable
import androidx.compose.runtime.derivedStateOf
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableFloatStateOf
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.graphics.graphicsLayer
import androidx.compose.ui.platform.LocalDensity
import androidx.compose.ui.semantics.clearAndSetSemantics
import androidx.compose.ui.semantics.contentDescription
import androidx.compose.ui.semantics.semantics
import androidx.compose.ui.unit.Dp
import androidx.compose.ui.unit.dp
import kotlin.math.absoluteValue
import kotlinx.coroutines.launch
private const val OPEN_THRESHOLD = 0.35f
private val FLING_THRESHOLD = 400.dp
/**
* How much of the screen a panel takes by default, leaving a sliver of what it is over.
*
* A panel given the whole width instead is standing in for the screen rather than sitting over it,
* and then the sliver would be a strip of a screen the reader has just left behind.
*/
const val PANEL_FRACTION = 0.88f
/** How dark the scrim over [SidePanels]' content goes with a panel fully open. */
private const val SCRIM_ALPHA = 0.32f
/**
* Which side of the content a panel comes in from: where it sits, and which way it slides out.
*
* [sign] is also the direction of the reveal this side owns, so the drag arithmetic is written once
* rather than once per side with the minus signs moved around.
*/
enum class PanelSide(val alignment: Alignment, val sign: Float) {
Left(Alignment.CenterStart, -1f),
Right(Alignment.CenterEnd, 1f),
}
/**
* Keeps [content] composed while a panel belonging to it moves over from the left or the right.
*
* One gesture drives both sides rather than one handler each, because two `draggable`s over the
* same content cannot share a horizontal drag: the inner one claims it whichever way the finger
* went, and the outer never sees a thing. So the position is a single signed reveal -- negative is
* the left panel showing, positive the right -- which also makes it impossible to have both open.
*
* A side left null has no panel and no gesture toward it -- the reveal cannot travel that way at
* all -- so one composable serves a screen with one panel and a screen with two.
*
* The root drag handler deliberately sits behind descendants. A horizontal scroller consumes its
* drag first, so code blocks, attachments and tool inputs keep their existing gesture. Collapsing
* that content, or starting over any ordinary part of the session, gives the gesture back to the
* panel; Android's own right-edge Back gesture remains untouched.
*/
@Composable
fun SidePanels(
left: (@Composable (active: Boolean, close: () -> Unit) -> Unit)? = null,
leftFraction: Float = PANEL_FRACTION,
right: (@Composable (active: Boolean, close: () -> Unit) -> Unit)? = null,
rightFraction: Float = PANEL_FRACTION,
content: @Composable () -> Unit,
) {
val scope = rememberCoroutineScope()
// Which panel the gesture settled on, null for neither. The *settled* side rather than the
// current position, so a panel's contents know they are being looked at while the animation
// is still running.
var opened by remember { mutableStateOf<PanelSide?>(null) }
var dragging by remember { mutableStateOf(false) }
var draggedReveal by remember { mutableFloatStateOf(0f) }
val animatedReveal = remember { Animatable(0f) }
// Read from a draw or layout lambda, never from the composable body: where the panel has got
// to changes every frame of a drag, and a body that reads it recomposes this whole subtree --
// the session included -- once per frame. The booleans below are what composition is allowed
// to know, and each of them changes twice per gesture. (Same rule as the keyboard inset in
// SessionScreen, and found the same way.)
fun revealNow() = if (dragging) draggedReveal else animatedReveal.value
val leftShown by remember { derivedStateOf { revealNow() < 0f } }
val rightShown by remember { derivedStateOf { revealNow() > 0f } }
val engaged = leftShown || rightShown
val flingThreshold = with(LocalDensity.current) { FLING_THRESHOLD.toPx() }
suspend fun startDrag() {
animatedReveal.stop()
draggedReveal = animatedReveal.value
dragging = true
}
// Which panel the reveal belongs to, [bias] breaking the tie at rest -- a drag away from
// nothing is toward whichever panel that direction opens.
fun sideOf(bias: Float): PanelSide? =
when {
draggedReveal < 0f -> PanelSide.Left
draggedReveal > 0f -> PanelSide.Right
bias > 0f -> PanelSide.Left
bias < 0f -> PanelSide.Right
else -> null
}
suspend fun finishDrag(velocity: Float) {
val side = sideOf(0f)
// How fast the finger is moving toward that side's open position: the left panel opens
// rightwards and the right panel leftwards, so the sign of a velocity only means something
// once it is read against the side. A fling decides on its own; anything slower is decided
// by how far in the panel already is.
val toward = side?.let { -it.sign * velocity } ?: 0f
val opens =
if (toward.absoluteValue > flingThreshold) toward > 0f
else draggedReveal.absoluteValue >= OPEN_THRESHOLD
val target = side.takeIf { opens }
opened = target
animatedReveal.snapTo(draggedReveal)
dragging = false
animatedReveal.animateTo(target?.sign ?: 0f)
}
fun close() {
opened = null
scope.launch { animatedReveal.animateTo(0f) }
}
BackHandler(enabled = opened != null) { close() }
BoxWithConstraints(Modifier.fillMaxSize()) {
val dragState = rememberDraggableState { delta ->
// Against the width of the panel this drag is moving, since the reveal is a fraction
// of it and the two sides need not be the same width.
val width =
sideOf(delta)?.let {
constraints.maxWidth * if (it == PanelSide.Left) leftFraction else rightFraction
} ?: return@rememberDraggableState
draggedReveal =
(draggedReveal - delta / width.coerceAtLeast(1f)).coerceIn(
if (left == null) 0f else -1f,
if (right == null) 0f else 1f,
)
}
val drag =
Modifier.draggable(
state = dragState,
orientation = Orientation.Horizontal,
onDragStarted = { startDrag() },
onDragStopped = { velocity -> finishDrag(velocity) },
)
Box(
Modifier.fillMaxSize()
.then(drag)
.then(if (engaged) Modifier.clearAndSetSemantics {} else Modifier)
) {
content()
}
if (engaged) {
Box(
Modifier.fillMaxSize()
.graphicsLayer { alpha = revealNow().absoluteValue * SCRIM_ALPHA }
.background(MaterialTheme.colorScheme.scrim)
.semantics { contentDescription = "Dismiss panel" }
.clickable { close() }
)
}
// Both panels stay composed while they are off screen, so opening one costs no
// composition -- but an off-screen panel is cleared from the semantics tree, since nothing
// a reader cannot see should be reachable by swiping through the screen.
left?.let { panel ->
SlidingPanel(
side = PanelSide.Left,
width = maxWidth * leftFraction,
raised = leftFraction < 1f,
shown = { (-revealNow()).coerceAtLeast(0f) },
visible = leftShown,
drag = drag,
) {
panel(opened == PanelSide.Left, ::close)
}
}
right?.let { panel ->
SlidingPanel(
side = PanelSide.Right,
width = maxWidth * rightFraction,
raised = rightFraction < 1f,
shown = { revealNow().coerceAtLeast(0f) },
visible = rightShown,
drag = drag,
) {
panel(opened == PanelSide.Right, ::close)
}
}
}
}
/**
* One panel at [shown] of the way in, sliding out to its own [side].
*
* [raised] is for a panel with some of the screen still beside it, which takes a tonal step to say
* it is above what it has not covered. A panel covering the whole width has nothing to be above,
* and a step there is a screen that is simply the wrong colour.
*
* [visible] says the same thing as `shown() > 0f` and is the form composition may read; see
* [SidePanels].
*/
@Composable
private fun BoxScope.SlidingPanel(
side: PanelSide,
width: Dp,
raised: Boolean,
shown: () -> Float,
visible: Boolean,
drag: Modifier,
contents: @Composable () -> Unit,
) {
Surface(
tonalElevation = if (raised) 3.dp else 0.dp,
shadowElevation = 8.dp,
modifier =
Modifier.align(side.alignment)
.width(width)
.fillMaxHeight()
.graphicsLayer { translationX = side.sign * size.width * (1f - shown()) }
.then(if (visible) Modifier else Modifier.clearAndSetSemantics {})
.then(drag),
) {
contents()
}
}
app/androidApp/src/main/kotlin/com/example/aiapp/SpawnScreen.kt
+189 -123
View File
@@ -16,7 +16,6 @@ import androidx.compose.material3.Button
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.FilterChip
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.OutlinedTextField
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
@@ -48,44 +47,52 @@ fun SpawnScreen(
val scope = rememberCoroutineScope()
// What the form is made of, and whether we have it yet. A failure here is not the same as a
// server with nothing to offer, so it must not reach the pickers as empty lists.
var options by remember { mutableStateOf<LoadState<List<Setup>>>(LoadState.Loading) }
var options by remember { mutableStateOf<LoadState<List<Machine>>>(LoadState.Loading) }
// Setup first, then one of its providers. Choosing a setup can invalidate the provider, so the
// provider is stored by name and resolved against the current setup rather than held as an
// Machine first, then one of its providers. Choosing a machine can invalidate the provider, so
// the
// provider is stored by name and resolved against the current machine rather than held as an
// object that could outlive the list it came from.
var setupName by remember { mutableStateOf<String?>(null) }
var machineName by remember { mutableStateOf<String?>(null) }
var providerName by remember { mutableStateOf<String?>(null) }
var title by remember { mutableStateOf("") }
var model by remember { mutableStateOf("") }
var providerModels by remember { mutableStateOf<List<OfferedModel>>(emptyList()) }
var providerModelsLoading by remember { mutableStateOf(false) }
var providerModelsError by remember { mutableStateOf<String?>(null) }
var cwd by remember { mutableStateOf("") }
// "auto" rather than "manual": on a phone every ask is a round trip to a question card, and
// answering "allow Bash?" dozens of times per task is what this app exists to avoid.
var permissionMode by remember { mutableStateOf("auto") }
// Set only after the selected provider reports its own default. An empty value is not sent.
var permissionMode by remember { mutableStateOf("") }
// Null until the server has been asked, and null again if it answers "no level chosen" -- the
// two are told apart by [defaultsAsked], because a picker that shows a level before the answer
// arrives is one you can spawn at without having chosen it.
var effort by remember { mutableStateOf<String?>(null) }
var defaultsAsked by remember { mutableStateOf(false) }
var busy by remember { mutableStateOf(false) }
// Only the spawn's own failure. The fetch's lives in `options`: this one leaves a filled-in
// form worth keeping, and that one leaves nothing to fill in.
var spawnError by remember { mutableStateOf<String?>(null) }
// Downloaded models, for a llama provider to choose between. Kept separate from the setups: a
// Claude session needs none, so failing to list them must not stop the screen rendering.
var models by remember { mutableStateOf<List<LocalModel>>(emptyList()) }
var modelKey by remember { mutableStateOf<String?>(null) }
var contextSize by remember { mutableStateOf("") }
var temperature by remember { mutableStateOf("") }
// Whatever the chosen provider says it takes, by key. Empty until something is typed: an
// absent key means the server's own default, which is what every field's placeholder says.
var params by remember { mutableStateOf<Map<String, String>>(emptyMap()) }
LaunchedEffect(Unit) {
// Separate from the machines fetch below and deliberately not fatal: failing to learn the
// default must leave a screen you can still spawn from, so the picker stays on "default"
// and says so rather than the whole form refusing to draw.
runCatching { withContext(Dispatchers.IO) { fetchDefaultEffort(settings) } }
.onSuccess { effort = it }
defaultsAsked = true
options =
try {
val fetched = withContext(Dispatchers.IO) { fetchSetups(settings) }
val fetched = withContext(Dispatchers.IO) { fetchMachines(settings) }
val first = fetched.firstOrNull()
setupName = first?.name
machineName = first?.name
providerName = first?.providers?.firstOrNull()?.name
LoadState.Loaded(fetched)
} catch (e: ApiException) {
LoadState.failed(e)
}
models =
runCatching { withContext(Dispatchers.IO) { fetchModels(settings).local } }
.getOrDefault(emptyList())
}
Column(Modifier.fillMaxSize().verticalScroll(rememberScrollState()).padding(16.dp)) {
@@ -102,7 +109,7 @@ fun SpawnScreen(
// Nothing below is fillable until the options are here, and a failure to fetch them leaves
// no form worth showing -- so this reports and stops, rather than offering empty pickers
// under an error message.
val setups =
val machines =
when (val state = options) {
is LoadState.Loading -> {
CircularProgressIndicator()
@@ -114,51 +121,88 @@ fun SpawnScreen(
}
is LoadState.Loaded -> state.value
}
val setup = setups.firstOrNull { it.name == setupName }
val current = setup?.providers?.firstOrNull { it.name == providerName }
// Only the Claude CLI has models, a working directory and permission modes; keying the
// extra fields on the kind rather than the provider name keeps a second Claude provider
val machine = machines.firstOrNull { it.name == machineName }
val current = machine?.providers?.firstOrNull { it.name == providerName }
// Coding CLIs take a working directory, model, permission mode and thinking level. Keying
// the extra fields on the kind rather than the provider name keeps a second installation
// from needing anything here.
val isClaude = current?.kind == "claude_cli"
val isCodex = current?.kind == "codex_cli"
val isCodingCli = isClaude || isCodex
val isLlama = current?.kind == "llama_cpp"
// Echo is the only kind with nothing to choose between.
val offersModels = isCodingCli || isLlama
// Where a session's tools act, which is the only thing a working directory decides.
val takesCwd = isCodingCli || isLlama
// Whichever machine and provider are chosen now, asked again when either changes. The
// previous answer is dropped first rather than left on screen: a model name from another
// machine looks exactly like one from this one.
LaunchedEffect(machine?.id, current?.name) {
model = ""
// A key from the previous provider would be a setting this one does not have, drawn
// by no control and sent at the spawn anyway.
params = emptyMap()
providerModels = emptyList()
providerModelsError = null
permissionMode = current?.defaultPermissionMode.orEmpty()
// Every kind that offers models at all, not only the coding CLIs: a llama provider
// answers with the GGUFs on the machine it runs on, through the same call. One
// question with one answer is what keeps the picker free of a branch on the kind.
if (machine == null || current == null || !offersModels) {
providerModelsLoading = false
return@LaunchedEffect
}
providerModelsLoading = true
try {
providerModels =
withContext(Dispatchers.IO) {
fetchProviderModels(settings, machine.id, current.name)
}
} catch (e: ApiException) {
providerModelsError = e.message
} finally {
providerModelsLoading = false
}
}
// The machine first, because it decides what can be run at all.
ChipGroup(
label = "Setup",
options = setups.map { it.name },
selected = setupName,
label = "Machine",
options = machines.map { it.name },
selected = machineName,
onSelect = { name ->
setupName = name
machineName = name
// The provider list changes with the machine, so a name carried over from the
// previous one would be a selection that isn't in the picker. Take that machine's
// first.
providerName =
setups.firstOrNull { it.name == name }?.providers?.firstOrNull()?.name
machines.firstOrNull { it.name == name }?.providers?.firstOrNull()?.name
},
)
setup?.address?.let {
machine?.address?.let {
Text(
it,
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
// The address belongs to the setup above it, not to the provider label below; without
// The address belongs to the machine above it, not to the provider label below; without
// this they read as one block.
Spacer(Modifier.height(8.dp))
}
// Only what this machine actually has. A setup with none says so rather than showing an
// Only what this machine actually has. A machine with none says so rather than showing an
// empty row that reads as a failure.
if (setup != null && setup.providers.isEmpty()) {
if (machine != null && machine.providers.isEmpty()) {
Text(
"\"${setup.name}\" has no providers configured.",
"\"${machine.name}\" has no providers configured.",
style = MaterialTheme.typography.bodyMedium,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
} else {
ChipGroup(
label = "Provider",
options = setup?.providers?.map { it.name }.orEmpty(),
options = machine?.providers?.map { it.name }.orEmpty(),
selected = providerName,
onSelect = { providerName = it },
)
@@ -166,91 +210,116 @@ fun SpawnScreen(
Spacer(Modifier.height(16.dp))
OutlinedTextField(
value = title,
onValueChange = { title = it },
label = { Text("Title") },
singleLine = true,
modifier = Modifier.fillMaxWidth(),
)
LabelledField(label = "Title", value = title, onValueChange = { title = it })
if (isLlama) {
// A llama session names one of the models this backend has downloaded, so the choice is
// that list rather than free text -- a name that is not on disk is a session that
// cannot start.
if (models.isEmpty()) {
Text(
"No models downloaded yet. Get one from the Models screen first.",
style = MaterialTheme.typography.bodyMedium,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
} else {
ChipGroup(
label = "Model",
// The file, not the whole key: the repository is the same for every
// quantisation of a model, so the file name is what tells two of them apart.
options = models.map { it.file },
selected = models.firstOrNull { it.key == modelKey }?.file,
onSelect = { file -> modelKey = models.first { it.file == file }.key },
)
if (offersModels) {
when {
providerModelsLoading ->
Text(
"Loading model choices…",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
providerModelsError != null ->
Text(
"Model choices unavailable: $providerModelsError",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.error,
)
// A llama session cannot start without one, so this says what to do about it
// rather than only that there is nothing -- the models it needs are on the
// machine that will serve them, which is not always this backend.
providerModels.isEmpty() && isLlama ->
Text(
"No models on ${machine.name}. The Models screen downloads " +
"to the backend; another machine needs the file put there itself.",
style = MaterialTheme.typography.bodyMedium,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
providerModels.isEmpty() ->
Text(
"This machine reported no selectable models.",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
else -> {
Spacer(Modifier.height(16.dp))
ChipGroup(
label = "Model",
// The label, and the id is what is sent: for a llama model those differ,
// since it is chosen by path and named by what is inside the file.
options = providerModels.map { it.label },
selected = providerModels.firstOrNull { it.id == model }?.label,
onSelect = { chosen ->
val id = providerModels.first { it.label == chosen }.id
// A llama session has to have one, so choosing the same chip twice
// must not clear it -- there is nothing to fall back to.
model = if (model == id && !isLlama) "" else id
},
)
}
}
Spacer(Modifier.height(16.dp))
}
OutlinedTextField(
value = contextSize,
onValueChange = { contextSize = it },
label = { Text("Context size (blank = the model's default)") },
singleLine = true,
modifier = Modifier.fillMaxWidth(),
)
Spacer(Modifier.height(16.dp))
OutlinedTextField(
value = temperature,
onValueChange = { temperature = it },
label = { Text("Temperature (blank = llama.cpp's default)") },
singleLine = true,
modifier = Modifier.fillMaxWidth(),
if (isCodingCli) {
// Free text as well as the chips above: the catalog is a shortcut, and a CLI will
// take a name it did not list.
LabelledField(
label = "Model",
value = model,
onValueChange = { model = it },
hint = "the CLI's default",
)
Spacer(Modifier.height(16.dp))
}
if (isClaude) {
if (current.models.isNotEmpty()) {
Spacer(Modifier.height(16.dp))
ChipGroup(
label = "Model",
options = current.models,
selected = model.ifEmpty { null },
onSelect = { chosen -> model = if (model == chosen) "" else chosen },
)
}
Spacer(Modifier.height(8.dp))
OutlinedTextField(
value = model,
onValueChange = { model = it },
label = { Text("Model (blank = the CLI's default)") },
singleLine = true,
modifier = Modifier.fillMaxWidth(),
)
Spacer(Modifier.height(16.dp))
// Nothing is running yet, so nothing here waits for a restart -- every one of these is
// read by the process this form is about to start.
ProviderParamFields(
specs = current?.params.orEmpty(),
values = params,
onChange = { params = it },
// Chips, like every other choice on this form.
choices = ChoiceStyle.Chips,
)
OutlinedTextField(
// Every session whose tools act on files needs one, which is both kinds that have
// tools -- a llama session's built-in tools run in it exactly as a CLI's do.
if (takesCwd) {
LabelledField(
label = "Working directory",
value = cwd,
onValueChange = { cwd = it },
label = { Text("Working directory") },
placeholder = { Text("/home/…") },
singleLine = true,
modifier = Modifier.fillMaxWidth(),
hint = "wherever the session's process starts",
)
Spacer(Modifier.height(16.dp))
}
// Offered wherever the provider has modes, rather than where this screen believes it
// does: the server is what knows, and llama.cpp grew them without this line changing.
if (current != null && current.permissionModes.isNotEmpty()) {
ChipGroup(
label = "Permissions",
options = PERMISSION_MODES,
options = current.permissionModes,
selected = permissionMode,
onSelect = { permissionMode = it },
)
Spacer(Modifier.height(16.dp))
}
if (isCodingCli) {
// Says what it does to *later* spawns as well, because it does: the level chosen here
// is stored as the default, which is the whole way that default is set. A picker that
// quietly changed a global would be the same control with the fact left out.
ChipGroup(
label = "Thinking (kept as the default for new sessions)",
options = listOf(DEFAULT_EFFORT) + EFFORT_LEVELS,
// The CLI's own default is a level in the list, so this cannot be a one-way trip.
// Disabled-looking until the server has answered, for the reason above.
selected = if (defaultsAsked) effort ?: DEFAULT_EFFORT else null,
onSelect = { chosen -> effort = chosen.takeIf { it != DEFAULT_EFFORT } },
)
}
Spacer(Modifier.height(24.dp))
@@ -268,33 +337,29 @@ fun SpawnScreen(
try {
val spawned =
withContext(Dispatchers.IO) {
// Stored before the spawn and not after it: choosing a level is
// an intent about new sessions in general, so a spawn that then
// fails must not also lose the choice. Non-fatal for the same
// reason the fetch above is -- the session is what was asked for.
if (isCodingCli) {
runCatching { setDefaultEffort(settings, effort) }
}
spawnSession(
settings,
// The id, not the label: labels are editable and the server
// resolves by id. Non-null here, since `chosen` came from
// `setup`'s own provider list.
setup = setup.id,
// `machine`'s own provider list.
machine = machine.id,
provider = chosen.name,
title = title.trim(),
model =
if (isLlama) modelKey else model.trim().takeIf { isClaude },
cwd = cwd.trim().takeIf { isClaude },
permissionMode = permissionMode.takeIf { isClaude },
// Sent only when set, so blank means "whatever llama.cpp does
// by default" rather than a zero.
params =
buildMap {
if (isLlama) {
contextSize
.trim()
.takeIf { it.isNotEmpty() }
?.let { put("contextSize", it) }
temperature
.trim()
.takeIf { it.isNotEmpty() }
?.let { put("temperature", it) }
}
},
model = model.trim().takeIf { offersModels },
cwd = cwd.trim().takeIf { takesCwd },
permissionMode = permissionMode.takeIf { it.isNotEmpty() },
effort = effort.takeIf { isCodingCli },
// Already only the keys somebody set: a field left blank
// removes its key rather than sending an empty value, so
// "blank" reaches the server as "your default".
params = params,
)
}
onSpawned(spawned)
@@ -304,7 +369,8 @@ fun SpawnScreen(
}
}
},
enabled = !busy && current != null && !(isLlama && modelKey == null),
// A llama session names the file to load, so there is nothing to spawn without one.
enabled = !busy && current != null && !(isLlama && model.isEmpty()),
) {
Text(if (busy) "Spawning..." else "Spawn")
}
app/androidApp/src/main/kotlin/com/example/aiapp/SubagentPanel.kt
+335
View File
@@ -0,0 +1,335 @@
package com.example.aiapp
import androidx.activity.compose.BackHandler
import androidx.compose.foundation.ExperimentalFoundationApi
import androidx.compose.foundation.combinedClickable
import androidx.compose.foundation.layout.Arrangement
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.fillMaxSize
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.height
import androidx.compose.foundation.layout.heightIn
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.layout.width
import androidx.compose.foundation.lazy.LazyColumn
import androidx.compose.material3.AlertDialog
import androidx.compose.material3.CardDefaults
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.LocalContentColor
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.OutlinedCard
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.LaunchedEffect
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableIntStateOf
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.rememberCoroutineScope
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.platform.LocalContext
import androidx.compose.ui.unit.dp
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.launch
import kotlinx.coroutines.withContext
/**
* What a session has running beside the turn you are reading: its background tasks, then its
* subagents, in the panel [SidePanels] slides over it from the right.
*
* [active] is whether the panel is being looked at: the lists are fetched then rather than on
* composition, since the panel is composed for every session whether or not anybody opens it.
*
* [onOpenCall] takes the reader to where a background task was started, in the transcript under
* this panel -- so the panel is closed with it, which is the caller's to do.
*
* [backgroundTasks] is the live count from the session's own event stream, and is what the
* background list is refetched against: a card for work that has since finished is a stale
* measurement drawn as a current one, which is the one thing a list of what is running now must not
* do.
*
* Both lists are items of one lazy column rather than two stacked scrollers, so expanding the
* background section pushes the subagents down without either being able to run off the panel.
*/
@Composable
fun SubagentPanel(
settings: ServerSettings,
summary: SessionSummary,
active: Boolean,
backgroundTasks: Int,
onOpenSubagent: (SubagentSummary) -> Unit,
onOpenCall: (CallSite) -> Unit,
) {
val scope = rememberCoroutineScope()
val context = LocalContext.current
val transcriptCache = remember(settings) { TranscriptCache(cacheRoot(context, settings)) }
var rows by
remember(summary.id) { mutableStateOf<LoadState<List<SubagentSummary>>>(LoadState.Loading) }
var selected by remember(summary.id) { mutableStateOf(setOf<String>()) }
var deleting by remember(summary.id) { mutableStateOf(setOf<String>()) }
var deleteError by remember(summary.id) { mutableStateOf<String?>(null) }
var confirming by remember(summary.id) { mutableStateOf<List<SubagentSummary>?>(null) }
var refreshToken by remember(summary.id) { mutableIntStateOf(0) }
var background by
remember(summary.id) {
mutableStateOf<LoadState<List<BackgroundTaskSummary>?>>(LoadState.Loading)
}
var backgroundExpanded by remember(summary.id) { mutableStateOf(false) }
LaunchedEffect(active, refreshToken) {
if (!active) return@LaunchedEffect
rows = LoadState.Loading
rows =
try {
val fetched = withContext(Dispatchers.IO) { fetchSubagents(settings, summary.id) }
selected = selected intersect fetched.mapTo(mutableSetOf()) { it.id }
LoadState.Loaded(fetched)
} catch (e: ApiException) {
LoadState.failed(e)
}
}
// No reset to Loading on a refetch: the spinner belongs to the first fetch, and one flashed
// over the list at every start and end would blink precisely when something happened.
LaunchedEffect(active, backgroundTasks, refreshToken) {
if (!active || backgroundTasks == 0) return@LaunchedEffect
background =
try {
LoadState.Loaded(
withContext(Dispatchers.IO) { fetchBackgroundTasks(settings, summary.id) }
)
} catch (e: ApiException) {
LoadState.failed(e)
}
}
BackHandler(enabled = selected.isNotEmpty()) { selected = emptySet() }
val ordered = (rows as? LoadState.Loaded)?.value?.let(::subagentOrder)
Column(Modifier.fillMaxSize()) {
LazyColumn(
verticalArrangement = Arrangement.spacedBy(8.dp),
modifier = Modifier.weight(1f).padding(horizontal = 16.dp),
) {
backgroundTaskSection(
count = backgroundTasks,
tasks = background,
expanded = backgroundExpanded,
onToggle = { backgroundExpanded = !backgroundExpanded },
onRetry = { refreshToken++ },
onOpenCall = onOpenCall,
)
item(key = "subagents-heading") { PanelSectionHeading("Subagents") }
when (val state = rows) {
is LoadState.Loading ->
item(key = "subagents-loading") {
CircularProgressIndicator(modifier = Modifier.width(24.dp).height(24.dp))
}
is LoadState.Error ->
item(key = "subagents-error") {
Column {
Text(
state.message,
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
)
TextButton(onClick = { refreshToken++ }) { Text("Try again") }
}
}
is LoadState.Loaded ->
if (ordered.isNullOrEmpty()) {
item(key = "subagents-empty") {
Text(
"No subagents in this session.",
color = MaterialTheme.colorScheme.onSurfaceVariant,
style = MaterialTheme.typography.bodyMedium,
)
}
} else {
uniqueItems(ordered, key = { it.id }) { subagent ->
SubagentCard(
subagent = subagent,
selected = subagent.id in selected,
selecting = selected.isNotEmpty(),
deleting = subagent.id in deleting,
onClick = { onOpenSubagent(subagent) },
onSelect = {
selected =
if (subagent.id in selected) selected - subagent.id
else selected + subagent.id
},
)
}
}
}
}
if (selected.isNotEmpty()) {
val picked = ordered.orEmpty().filter { it.id in selected }
SubagentSelectionBar(
picked = picked,
onDelete = { confirming = picked },
modifier = Modifier.padding(horizontal = 16.dp),
)
}
deleteError?.let {
Text(
it,
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
modifier = Modifier.padding(horizontal = 16.dp, vertical = 8.dp),
)
}
}
confirming?.let { picked ->
AlertDialog(
onDismissRequest = { confirming = null },
title = {
Text(
if (picked.size == 1) "Delete this subagent?"
else "Delete ${picked.size} subagents?"
)
},
text = {
Text(
(if (picked.size == 1) "\"${picked.first().title}\"\n\n" else "") +
"A subagent's transcript is the only record of what it did: the session " +
"that started it kept just the Task call. Nothing else has a copy, so " +
"this can't be undone. The session itself is untouched."
)
},
confirmButton = {
TextButton(
onClick = {
confirming = null
selected = emptySet()
val ids = picked.map { it.id }
val gone = ids.toSet()
deleting += gone
deleteError = null
scope.launch {
try {
withContext(Dispatchers.IO) {
deleteSubagents(settings, summary.id, ids)
gone.forEach {
transcriptCache
.session(TranscriptAddress(summary.id, it))
.purge()
}
}
val loaded = rows
if (loaded is LoadState.Loaded) {
rows =
LoadState.Loaded(loaded.value.filterNot { it.id in gone })
}
} catch (e: ApiException) {
deleteError = e.message ?: "Delete failed"
} finally {
deleting -= gone
}
}
}
) {
Text("Delete", color = MaterialTheme.colorScheme.error)
}
},
dismissButton = { TextButton(onClick = { confirming = null }) { Text("Cancel") } },
)
}
}
private fun subagentOrder(rows: List<SubagentSummary>): List<SubagentSummary> =
rows.sortedWith(
compareByDescending<SubagentSummary> { it.status == "running" }
.thenByDescending { it.lastActivity }
)
@OptIn(ExperimentalFoundationApi::class)
@Composable
private fun SubagentCard(
subagent: SubagentSummary,
selected: Boolean,
selecting: Boolean,
deleting: Boolean,
onClick: () -> Unit,
onSelect: () -> Unit,
) {
BusyItem(label = if (deleting) "deleting" else null) {
OutlinedCard(
colors =
if (selected)
CardDefaults.outlinedCardColors(
containerColor = MaterialTheme.colorScheme.secondaryContainer,
contentColor = MaterialTheme.colorScheme.onSecondaryContainer,
)
else CardDefaults.outlinedCardColors(),
modifier =
Modifier.fillMaxWidth()
.combinedClickable(
enabled = !deleting,
onClick = { if (selecting) onSelect() else onClick() },
onLongClick = onSelect,
),
) {
Column(Modifier.padding(horizontal = 12.dp, vertical = 8.dp)) {
Text(subagent.title, style = MaterialTheme.typography.titleSmall)
Spacer(Modifier.height(2.dp))
Row(Modifier.fillMaxWidth()) {
Text(
subagentStatusLabel(subagent.status),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
modifier = Modifier.weight(1f),
)
Text(
relativeTime(subagent.lastActivity),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
}
}
}
}
}
@Composable
private fun SubagentSelectionBar(
picked: List<SubagentSummary>,
onDelete: () -> Unit,
modifier: Modifier = Modifier,
) {
val running = picked.count { it.status == "running" }
Row(
verticalAlignment = Alignment.CenterVertically,
modifier = modifier.fillMaxWidth().heightIn(min = 48.dp),
) {
Text(
if (running == 0) "${picked.size} selected" else "$running still running",
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
modifier = Modifier.weight(1f),
)
TextButton(onClick = onDelete, enabled = running == 0) {
Text(
"Delete",
color =
if (running == 0) MaterialTheme.colorScheme.error
else LocalContentColor.current,
)
}
}
}
private fun subagentStatusLabel(status: String) =
when (status) {
"running" -> "running"
"exited" -> "finished"
else -> "unknown"
}
app/androidApp/src/main/kotlin/com/example/aiapp/Theme.kt
+16
View File
@@ -134,6 +134,20 @@ val clearedColor: Color
val awaitingColor: Color
@Composable get() = Mocha.Peach
/**
* Waiting on itself: the session's turn is over, but a subagent or a backgrounded command it
* started is still going, and it will speak again with nobody having typed anything.
*
* Its own colour rather than [awaitingColor], which is the opposite state -- that one means the
* reader has something to do, and this one means they specifically do not. Not [runningColor]
* either: nothing is being written, and a green "running" on a session that will say nothing for
* ten minutes is the wrong promise. Blue for the same reason [commandColor] is blue -- not stuck,
* but not replying to you either -- and a different blue because that one is the session acting on
* itself rather than getting on with what was asked.
*/
val waitingColor: Color
@Composable get() = Mocha.Sky
/** Approaching a limit -- still fine, worth seeing. */
val warningColor: Color
@Composable get() = Mocha.Yellow
@@ -199,6 +213,8 @@ val rawSurface: Color
*/
fun catppuccinSyntax(): SyntaxPalette =
SyntaxPalette(
addition = Mocha.Green,
deletion = Mocha.Red,
keyword = Mocha.Mauve,
string = Mocha.Green,
literal = Mocha.Peach,
app/androidApp/src/main/kotlin/com/example/aiapp/ThinkingCard.kt
+87
View File
@@ -0,0 +1,87 @@
package com.example.aiapp
import androidx.compose.foundation.clickable
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.Row
import androidx.compose.foundation.layout.Spacer
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.height
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.layout.width
import androidx.compose.material3.Card
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Text
import androidx.compose.runtime.Composable
import androidx.compose.runtime.CompositionLocalProvider
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.unit.dp
/**
* A model's working, shut until somebody asks for it.
*
* Shut by default, like a tool call and a memory note and for the same reason: it is not what the
* session said, and left open it puts the reasoning between the question and the answer -- which on
* a small model is most of the conversation.
*
* The heading is the whole of what the reader gets for free, so it carries the one thing worth
* knowing without opening anything: whether this is still going, and if not how long it took. A
* spinner while it runs, because that is the same fact a running command reports and it is drawn
* the same way here.
*/
@Composable
fun ThinkingCard(
item: TranscriptItem.ThinkingRow,
replies: ParsedReplies,
expanded: Boolean,
onToggle: () -> Unit,
modifier: Modifier = Modifier,
) {
Card(modifier.fillMaxWidth().clickable(onClick = onToggle)) {
Column(Modifier.padding(12.dp)) {
Row(verticalAlignment = Alignment.CenterVertically) {
Text(thinkingHeadline(item), style = MaterialTheme.typography.titleSmall)
Spacer(Modifier.width(8.dp))
if (item.open) {
CircularProgressIndicator(
modifier = Modifier.width(16.dp).height(16.dp),
strokeWidth = 2.dp,
)
}
}
// Markdown, like every other thing the model wrote: a model reasons in the same
// lists, headings and fenced code it answers in, and drawn plainly those arrive as
// rows of hashes and asterisks around the working the reader opened the card to read.
// Still arriving means the incremental parse -- see [MarkdownText] -- since an open
// block gains a delta at a time.
//
// The words shut the card too, and have to do it themselves -- see [LocalMarkdownTap].
if (expanded) {
CompositionLocalProvider(LocalMarkdownTap provides rememberMarkdownTap(onToggle)) {
MarkdownText(
item.text,
replies,
Modifier.padding(top = 6.dp),
live = item.open,
)
}
}
}
}
}
/**
* "Thinking", "Thought for 12.4s", or "Thought".
*
* The third is the one worth keeping: a block whose turn ended before the model said anything --
* interrupted, stopped, a process that exited -- was thought about for a length of time nobody
* measured. Naming a span there would be this screen inventing one, and the reader has no way to
* tell an invented one from the rest.
*/
fun thinkingHeadline(item: TranscriptItem.ThinkingRow): String =
when {
item.open -> "Thinking"
item.ms != null -> "Thought for ${formatMillis(item.ms)}"
else -> "Thought"
}
app/androidApp/src/main/kotlin/com/example/aiapp/ToolInput.kt
+44 -9
View File
@@ -1,9 +1,6 @@
package com.example.aiapp
import androidx.compose.foundation.horizontalScroll
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.rememberScrollState
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Text
import androidx.compose.runtime.Composable
@@ -51,24 +48,31 @@ data class ToolInput(
private val SUBJECTS: Map<String, Pair<String, Language?>> =
mapOf(
"Bash" to ("command" to Language.SHELL),
"Shell" to ("command" to Language.SHELL),
"Patch" to ("diff" to Language.DIFF),
"Read" to ("file_path" to null),
"Write" to ("file_path" to null),
"Edit" to ("file_path" to null),
"Glob" to ("pattern" to null),
"Grep" to ("pattern" to null),
"WebFetch" to ("url" to null),
"WebSearch" to ("query" to null),
// Persisted transcripts keep the provider vocabulary they were written with.
"web_search" to ("query" to null),
)
/** Fields that are the tool's own prose about itself rather than input to it. */
private val DESCRIPTIONS = listOf("description", "prompt")
fun parseToolInput(tool: String, input: String): ToolInput {
if (input.trim() == "null") return ToolInput(null, null, null, null, emptyList())
val json =
try {
JSONObject(input)
} catch (_: org.json.JSONException) {
// Not an object: older transcripts and some tools send a bare string. It is still the
// input, so it is still shown.
// input, so it is still shown. JSON null is the one exception: it means the call had
// no input, and drawing the word makes an absent value look like an instruction.
return ToolInput(
null,
null,
@@ -78,15 +82,20 @@ fun parseToolInput(tool: String, input: String): ToolInput {
)
}
val (subjectKey, language) = SUBJECTS[tool] ?: (null to null)
val subject = subjectKey?.let { json.optString(it) }?.takeIf { it.isNotBlank() }
val subject =
subjectKey
?.let { json.text(it) }
?.takeIf { it.isNotBlank() }
?.let { if (tool == "Bash") renderedBashScript(it) ?: it else it }
val description = DESCRIPTIONS.firstNotNullOfOrNull {
json.optString(it).takeIf { v -> v.isNotBlank() }
json.text(it)?.takeIf { value -> value.isNotBlank() }
}
val timeout = json.optString("timeout").takeIf { it.isNotBlank() }?.let { formatMillisText(it) }
val timeout = json.text("timeout")?.takeIf { it.isNotBlank() }?.let { formatMillisText(it) }
val rest =
json
.keys()
.asSequence()
.filterNot(json::isNull)
.filter { it != subjectKey || subject == null }
.filter { it !in DESCRIPTIONS || description == null }
.filter { it != "timeout" || timeout == null }
@@ -96,6 +105,31 @@ fun parseToolInput(tool: String, input: String): ToolInput {
return ToolInput(subject, language, description, timeout, rest)
}
private fun JSONObject.text(key: String): String? =
if (isNull(key)) null else optString(key).takeIf { it.isNotEmpty() }
/**
* Removes Codex's rendered Bash argv from old transcript rows.
*
* New events arrive normalized by the server, but persisted transcripts keep the input originally
* written to them. Only the outer pair are presentation quoting: quotes inside the command belong
* to the command and must not be parsed as an early end delimiter.
*/
internal fun renderedBashScript(command: String): String? {
val prefix =
listOf("/usr/bin/bash -lc ", "/bin/bash -lc ", "bash -lc ").firstOrNull {
command.startsWith(it)
} ?: return null
val quoted = command.removePrefix(prefix)
return quoted
.takeIf {
it.length >= 2 &&
((it.startsWith('\'') && it.endsWith('\'')) ||
(it.startsWith('"') && it.endsWith('"')))
}
?.substring(1, quoted.lastIndex)
}
/**
* A tool call's input: its subject highlighted, then whatever else it carried.
*
@@ -113,7 +147,8 @@ fun ToolInputView(tool: String, input: String, modifier: Modifier = Modifier) {
RawBlock(modifier) {
parsed.subject?.let { subject ->
// Not wrapped: a wrapped command hides where its arguments end, and the long one is the
// one being read closely.
// one being read closely. The sideways scroll that makes that readable is the block's,
// shared with the lines below -- see [RawBlock].
Text(
// Not cached: a tool's subject is one command line, which lexes in microseconds --
// the cache exists for a fence with two hundred lines in it.
@@ -121,7 +156,6 @@ fun ToolInputView(tool: String, input: String, modifier: Modifier = Modifier) {
style = MaterialTheme.typography.bodySmall,
fontFamily = FontFamily.Monospace,
softWrap = false,
modifier = Modifier.fillMaxWidth().horizontalScroll(rememberScrollState()),
)
}
parsed.rest.forEach {
@@ -130,6 +164,7 @@ fun ToolInputView(tool: String, input: String, modifier: Modifier = Modifier) {
style = MaterialTheme.typography.bodySmall,
fontFamily = FontFamily.Monospace,
color = MaterialTheme.colorScheme.onSurfaceVariant,
softWrap = false,
modifier = Modifier.padding(top = 2.dp),
)
}
app/androidApp/src/main/kotlin/com/example/aiapp/ToolRows.kt
+120 -32
View File
@@ -27,6 +27,9 @@ import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.draw.clip
import androidx.compose.ui.graphics.Shape
import androidx.compose.ui.layout.onPlaced
import androidx.compose.ui.layout.onSizeChanged
import androidx.compose.ui.layout.positionInRoot
import androidx.compose.ui.platform.LocalDensity
import androidx.compose.ui.semantics.contentDescription
import androidx.compose.ui.semantics.semantics
@@ -61,7 +64,9 @@ sealed class TranscriptRow {
*
* A tool row therefore keys on [TranscriptItem.ToolRun.runId] rather than on a sequence number,
* and it is the *same* value whether the run is drawn as one card or as a group. Which value
* that is belongs to the item ([TranscriptItem.key]), not to a `when` here.
* that is belongs to the item ([TranscriptItem.key]) everywhere a row is one thing; where
* [groupRuns] cuts a run into several rows it is the one deciding, and it says so by handing
* each piece its key.
*/
abstract val key: Any
@@ -75,23 +80,15 @@ sealed class TranscriptRow {
*/
abstract val startSeq: Long
data class Single(val item: TranscriptItem) : TranscriptRow() {
override val key: Any
get() = item.key
data class Single(val item: TranscriptItem, override val key: Any = item.key) :
TranscriptRow() {
override val startSeq: Long
get() = item.seq
}
/** Two or more calls with nothing between them; drawn as one collapsed card. */
data class Tools(val calls: List<TranscriptItem.ToolRun>) : TranscriptRow() {
/** The run's own name, which every call in it already carries. */
val id: String
get() = calls.first().runId
override val key: Any
get() = id
data class Tools(val calls: List<TranscriptItem.ToolRun>, override val key: String) :
TranscriptRow() {
override val startSeq: Long
get() = calls.first().seq
}
@@ -102,33 +99,75 @@ sealed class TranscriptRow {
*
* A single call is left alone: "Called 1 tool" hides a card to say the same thing in more words,
* and the run this exists for is the burst of five greps nobody wants to scroll past.
*
* The last call is left alone too, and so is one still running wherever in its run it sits. What
* the session is doing, or did last, is the one thing worth seeing without opening anything, and a
* heading counting it hides it. What folds a call back into its run is therefore not finishing but
* being overtaken: anything arriving behind it, a reply included, makes it history.
*
* [heldOut] is the one thing being read can change, and only in that direction: a call standing on
* its own that somebody is reading is not overtaken while they read it. Opening a call *already*
* inside a group does not pull it out (2026-09-16, after it briefly did) -- it is visible where it
* is, and grouping is what gives a row its identity, so a rule that reads the open set both ways
* makes the reader's own tap rebuild the rows around it: three rows became one the moment a call
* was closed, and no anchor survives a row that no longer exists -- the list jumped by 450px and
* took the closed card with it. Which calls are held out is [SessionScreen]'s to say, since being
* inside a group once is what settles it.
*/
fun groupToolRuns(items: List<TranscriptItem>): List<TranscriptRow> =
DebugStats.timed("grouped tool runs") { groupRuns(items) }
fun groupToolRuns(
items: List<TranscriptItem>,
heldOut: Set<String> = emptySet(),
): List<TranscriptRow> = DebugStats.timed("grouped tool runs") { groupRuns(items, heldOut) }
private fun groupRuns(items: List<TranscriptItem>): List<TranscriptRow> {
private fun groupRuns(items: List<TranscriptItem>, heldOut: Set<String>): List<TranscriptRow> {
val rows = mutableListOf<TranscriptRow>()
var run = mutableListOf<TranscriptItem.ToolRun>()
// A run can occupy more than one non-adjacent piece, so claimed keys span the whole transcript
// rather than resetting at each piece.
var runId: String? = null
val claimedKeys = mutableSetOf<String>()
fun flush() {
when (run.size) {
0 -> {}
1 -> rows += TranscriptRow.Single(run.first())
else -> rows += TranscriptRow.Tools(run.toList())
val first = run.firstOrNull() ?: return
// The first piece keeps the run's name, which survives a page landing in front of it
// ([adoptRun]). Later pieces qualify that name with their first call; the suffix is the
// final guard because a duplicate LazyColumn key takes down the whole screen.
var key = first.runId
if (!claimedKeys.add(key)) {
key = "${first.runId}/${first.id}"
var suffix = 2
while (!claimedKeys.add(key)) {
key = "${first.runId}/${first.id}/${suffix++}"
}
}
rows +=
if (run.size == 1) TranscriptRow.Single(first, key)
else TranscriptRow.Tools(run.toList(), key)
run = mutableListOf()
}
items.forEach { item ->
items.forEachIndexed { index, item ->
// Grouped by the run each call says it belongs to, not by adjacency worked out here.
// Adjacency is the same answer most of the time and a worse one at the edges: a call
// arriving next to an existing run, or a page of history arriving in front of one, both
// change which call is *first*.
if (item is TranscriptItem.ToolRun && (run.isEmpty() || run.first().runId == item.runId)) {
run += item
} else {
val call = item as? TranscriptItem.ToolRun
if (call == null || call.runId != runId) {
flush()
if (item is TranscriptItem.ToolRun) run += item else rows += TranscriptRow.Single(item)
runId = call?.runId
}
when {
call == null -> rows += TranscriptRow.Single(item)
// Standing outside the run is the call's place in the list as it is now, not something
// recorded on the call: the same finished call is a row of its own while it is the last
// thing that happened, or open and never yet grouped, and part of its group once a
// reply lands behind it.
call.done && call.id !in heldOut && index != items.lastIndex -> run += call
else -> {
flush()
run += call
flush()
}
}
}
flush()
@@ -160,7 +199,15 @@ fun ToolGroup(
*/
onToggle: () -> Unit,
isToolExpanded: (String) -> Boolean,
onToolToggle: (String) -> Unit,
/**
* Toggles one call, and says where in the group it was drawn: how far down the group's own top
* edge its card begins, and how tall that card is now.
*
* The screen anchors on *rows*, and a call is not one -- but what the reader is opening or
* shutting is the call, and keeping it under their finger needs its place inside the row. Only
* the group knows that, so only the group can say it. See `SessionScreen`'s `toggleAnchored`.
*/
onToolToggle: (id: String, top: Int, height: Int) -> Unit,
onAnswer: (List<QuestionAnswer>, onSettled: () -> Unit) -> Unit,
image: @Composable (String) -> Unit,
) {
@@ -175,8 +222,10 @@ fun ToolGroup(
}
return
}
val placed = remember { Placed() }
Column(
Modifier.fillMaxWidth()
.onPlaced { placed.top = it.positionInRoot().y }
.clip(MaterialTheme.shapes.medium)
.background(MaterialTheme.colorScheme.surfaceContainerLow)
) {
@@ -196,13 +245,19 @@ fun ToolGroup(
verticalArrangement = Arrangement.spacedBy(GROUP_GAP),
) {
group.calls.forEachIndexed { index, call ->
val card = remember(call.id) { Placed() }
ToolCard(
tool = call,
expanded = isToolExpanded(call.id),
onToggle = { onToolToggle(call.id) },
onToggle = {
onToolToggle(call.id, (card.top - placed.top).toInt(), card.height)
},
onAnswer = onAnswer,
image = image,
shape = connectedShape(index, group.calls.size),
modifier =
Modifier.onPlaced { card.top = it.positionInRoot().y }
.onSizeChanged { card.height = it.height },
)
}
}
@@ -212,6 +267,18 @@ fun ToolGroup(
}
}
/**
* Where something was last placed, in the window's coordinates, and how tall it was.
*
* Deliberately not snapshot state: it is written from the layout phase, and a write there that
* composition reads would schedule another recomposition of every group on screen, every frame.
* Nothing reads it except the gesture that follows.
*/
private class Placed {
var top = 0f
var height = 0
}
/**
* The height of a group's heading, and so of the bar at its foot.
*
@@ -293,14 +360,17 @@ fun ToolCard(
image: @Composable (String) -> Unit = {},
/** Square where this card faces another in a group; see [connectedShape]. */
shape: Shape = CardDefaults.shape,
modifier: Modifier = Modifier,
) {
val parsed = remember(tool.tool, tool.input) { parseToolInput(tool.tool, tool.input) }
val name = toolDisplayName(tool.tool)
val output = toolDisplayOutput(tool.tool, tool.output)
val deciding = tool.asks.any { it.answers.isEmpty() }
val open = expanded || deciding
Card(Modifier.fillMaxWidth().clickable(onClick = onToggle), shape = shape) {
Card(modifier.fillMaxWidth().clickable(onClick = onToggle), shape = shape) {
Column(Modifier.padding(GROUP_INSET_LARGE)) {
Row(verticalAlignment = Alignment.CenterVertically) {
Text(tool.tool, style = MaterialTheme.typography.titleSmall)
Text(name, style = MaterialTheme.typography.titleSmall)
if (open) {
Spacer(Modifier.weight(1f))
parsed.timeout?.let {
@@ -355,24 +425,26 @@ fun ToolCard(
if (tool.tool != ASK_USER_QUESTION) {
ToolInputView(tool.tool, tool.input, Modifier.padding(top = 4.dp))
}
if (tool.output.isNotEmpty()) {
if (output.isNotEmpty()) {
Spacer(Modifier.height(8.dp))
Text("Output", style = MaterialTheme.typography.labelSmall)
// What the tool printed, on the surface everything verbatim gets and in the
// face it was written for: this is column-aligned far more often than it is
// prose, and a proportional font silently destroys the alignment that carried
// the meaning.
// the meaning. Unwrapped for the same reason, and scrolled sideways by the
// block around it -- see [RawBlock].
//
// Its terminal styling applied and the rest of the escapes taken out: colour is
// often the whole of what a diff or a test run is saying. Remembered against
// the text, so a card that is open through a scroll parses once.
val palette = remember { ansiPalette() }
val styled = remember(tool.output, palette) { ansiStyled(tool.output, palette) }
val styled = remember(output, palette) { ansiStyled(output, palette) }
RawBlock(Modifier.padding(top = 2.dp)) {
Text(
styled,
style = MaterialTheme.typography.bodySmall,
fontFamily = FontFamily.Monospace,
softWrap = false,
)
}
}
@@ -391,6 +463,22 @@ fun ToolCard(
}
}
private val collaborationToolNames =
mapOf(
"Task" to "Spawn agent",
"TaskOutput" to "Wait for agents",
"SendMessage" to "Message agent",
"CloseAgent" to "Close agent",
"InterruptAgent" to "Interrupt agent",
"ListAgents" to "List agents",
"ResumeAgent" to "Resume agent",
)
internal fun toolDisplayName(tool: String): String = collaborationToolNames[tool] ?: tool
internal fun toolDisplayOutput(tool: String, output: String): String =
if (tool in collaborationToolNames && output == "completed") "" else output
/**
* The permission ask on the call it is about.
*
app/androidApp/src/main/kotlin/com/example/aiapp/TranscriptAddress.kt
+27
View File
@@ -0,0 +1,27 @@
package com.example.aiapp
/**
* Where one transcript lives: a session's own, or one of its subagents'.
*
* The single mechanism [fetchTranscript], [EventStream], [TranscriptSource] and
* [TranscriptCache.session] all take, rather than each growing its own branch between a session and
* a subagent -- see SUBAGENTS.md's "Phone" and "Wire shape". A caller that has only a session id
* builds one with the one-argument constructor; a subagent's screen supplies both ids.
*/
data class TranscriptAddress(val sessionId: String, val subagentId: String? = null) {
/** The URL segment naming this transcript, before `/transcript` or `/events`. */
val urlPath: String
get() =
if (subagentId == null) "sessions/$sessionId"
else "sessions/$sessionId/subagents/$subagentId"
/**
* Where this transcript's cache lives on the phone, relative to the cache root.
*
* A subagent's nests under its session's directory rather than sitting beside it, so deleting a
* session's cache directory takes its subagents' with it -- the same one-way door the server's
* own storage describes.
*/
val cachePath: String
get() = if (subagentId == null) sessionId else "$sessionId/subagents/$subagentId"
}
app/androidApp/src/main/kotlin/com/example/aiapp/TranscriptCache.kt
+9 -2
View File
@@ -35,8 +35,15 @@ class TranscriptCache(
private val root: File,
private val warn: (String) -> Unit = { Log.w("ai-app", it) },
) {
/** The cache for one session, whether or not anything has been stored for it yet. */
fun session(id: String): SessionCache = SessionCache(File(root, id), warn)
/**
* The cache for one transcript, whether or not anything has been stored for it yet.
*
* A subagent's [TranscriptAddress.cachePath] nests it under its session's directory, so
* deleting the session (below) takes its subagents' caches with it -- there is no separate
* purge for one.
*/
fun session(address: TranscriptAddress): SessionCache =
SessionCache(File(root, address.cachePath), warn)
/**
* Deletes every session directory not in [ids], called after a successful list fetch. The path
app/androidApp/src/main/kotlin/com/example/aiapp/TranscriptItems.kt
+222 -23
View File
@@ -63,6 +63,45 @@ sealed class TranscriptItem {
* inside that reply would step the list under them.
*/
val settled: Boolean = false,
/** A final value that supersedes provisional deltas behind a page boundary. */
val replacesPrefix: Boolean = false,
/**
* When the reply was sent, in epoch seconds: the time on its newest delta, which is the
* moment it finished rather than the moment it started.
*
* The transcript's own timestamp rather than a clock read here, so every device draws the
* same time under the same reply and a replayed page agrees with the live stream.
*/
val ts: Double = 0.0,
/**
* How fast it was generated, where the provider measured it; null everywhere else.
*
* Folded on from the turn's usage event rather than carried by the text, because it is not
* known until the reply is over.
*/
val tokensPerSecond: Double? = null,
/** How long the provider spent reading the prompt, where it measured that. */
val prefillMs: Long? = null,
) : TranscriptItem()
/**
* The model's working before -- or between -- the things it said.
*
* Its own row rather than part of the reply, and deliberately not a [ToolRun]: a run of tool
* calls collapses into one card, and folding a model's reasoning into "Called 6 tools" would
* file it as one of them. Shut by default, like every other card that is not what was said.
*
* Three states, because two of them are not the same absence. [open] is a block still being
* thought, which is what the spinner is for. A closed one with an [ms] says how long it took; a
* closed one without is a block whose turn ended before anything said -- an interrupted reply,
* a session stopped mid-thought -- and it says so by not naming a duration rather than by
* naming a wrong one.
*/
data class ThinkingRow(
override val seq: Long,
val text: String,
val ms: Long? = null,
val open: Boolean = true,
) : TranscriptItem()
data class ToolRun(
@@ -143,6 +182,29 @@ sealed class TranscriptItem {
get() = arrived
}
/**
* Where one turn ended and the next began with nothing said in between.
*
* A rule and no words. Two replies meet like this whenever a turn starts without anybody typing
* -- a subagent reporting back, a session the CLI picked up by itself -- and drawn with only
* the ordinary gap between them they read as one answer with a paragraph break through the
* middle of it. What the reader needs is to see that these are two; what started the turn is
* somebody else's transcript's business, and a row per background task is a screenful of
* dividers about work nobody was asking after.
*
* Made by the fold rather than sent by the server, because it is not something that happened:
* it is the boundary between two things that did. See [foldEvent].
*/
data class TurnBreak(override val seq: Long) : TranscriptItem() {
/**
* Its own key, because it shares a [seq] with the reply it sits above -- that reply's first
* delta is the event this was made at, and a keyed list refuses two items with one key by
* taking the app down.
*/
override val key: Any
get() = "break$seq"
}
/**
* A command the session ran on itself -- `/compact`, `/rename`. Kept in the transcript rather
* than only shown while it waits, because it explains what follows: a conversation that
@@ -175,6 +237,18 @@ sealed class TranscriptItem {
val preTokens: Long?,
val postTokens: Long?,
) : TranscriptItem()
/**
* The account ran out of quota, so the turn stopped here.
*
* A divider rather than an error: nothing failed, and what a reader scrolling back needs from
* it is the same thing a clear or a compaction gives them -- why the conversation stops at this
* line.
*
* [resetsAt] is epoch seconds and null where the session was told nothing, which is a state the
* row has words for rather than a time it invents.
*/
data class LimitNote(override val seq: Long, val resetsAt: Double?) : TranscriptItem()
}
/**
@@ -219,7 +293,7 @@ private fun runIdFor(items: List<TranscriptItem>, id: String, tool: String): Str
* with the seam wherever the reader happened to have paged.
*/
fun joinPages(earlier: List<TranscriptItem>, later: List<TranscriptItem>): List<TranscriptItem> {
val (older, newer) = healSplitMessage(earlier, later)
val (older, newer) = healSplitThinking(healSplitMessage(earlier, later))
val startedEarlier =
older.filterIsInstance<TranscriptItem.ToolRun>().mapTo(mutableSetOf()) { it.id }
val endedLater =
@@ -249,9 +323,10 @@ fun joinPages(earlier: List<TranscriptItem>, later: List<TranscriptItem>): List<
/**
* Rejoins a message the page boundary cut, and hands back the two pages to concatenate.
*
* [foldEvent] never leaves two assistant messages next to each other inside one page, so two
* meeting at a join are always the two halves of one reply, and leaving them apart drew a single
* answer as two with a paragraph break through the middle of a sentence.
* [foldEvent] never leaves an *unfinished* assistant message with another behind it inside one
* page, so an unsettled one at a join is always the far half of the reply the boundary cut, and
* leaving the two apart drew a single answer as two with a paragraph break through the middle of a
* sentence. Two settled replies meeting there are two turns and stay two.
*
* The newer half keeps its identity, for the reason [adoptRun] gives. It grows by what the older
* half brings, which is safe here and nowhere else -- the join is at the oldest end of what is
@@ -266,6 +341,36 @@ private fun healSplitMessage(
if (head !is TranscriptItem.AssistantMsg || tail !is TranscriptItem.AssistantMsg) {
return earlier to later
}
// A settled reply is a whole turn, so the two are two answers that happen to meet at the
// boundary rather than one cut in half -- the same distinction the fold makes, and joining them
// here would put back exactly the run-together paragraph it stops.
// The rule between them is put in here too, since the fold that would have made it never saw
// these two side by side.
if (head.settled) return earlier to (listOf(TranscriptItem.TurnBreak(tail.seq)) + later)
if (tail.replacesPrefix) return earlier.dropLast(1) to later
return earlier.dropLast(1) to (listOf(tail.copy(text = head.text + tail.text)) + later.drop(1))
}
/**
* Rejoins a thinking block the page boundary cut, the same way [healSplitMessage] rejoins a reply.
*
* A block streams a fragment at a time exactly as a reply does, so a boundary lands inside one as
* readily. The older half then holds an open block whose [SessionEvent.ThinkingDone] is on the
* newer page -- so it spun for the rest of the conversation, saying the machine was working on a
* thought it finished minutes ago, and the same working was drawn as two blocks.
*
* Only where the older half is still open: a closed one has its own ending and the two are two
* blocks that happen to meet here. The newer half keeps its identity, for the reason [adoptRun]
* gives -- it is the part already on screen.
*/
private fun healSplitThinking(
pages: Pair<List<TranscriptItem>, List<TranscriptItem>>
): Pair<List<TranscriptItem>, List<TranscriptItem>> {
val (earlier, later) = pages
val head = earlier.lastOrNull()
val tail = later.firstOrNull()
if (head !is TranscriptItem.ThinkingRow || tail !is TranscriptItem.ThinkingRow) return pages
if (!head.open) return pages
return earlier.dropLast(1) to (listOf(tail.copy(text = head.text + tail.text)) + later.drop(1))
}
@@ -352,24 +457,75 @@ fun foldEvent(items: List<TranscriptItem>, entry: SeqEvent): List<TranscriptItem
// first of them: a row whose identity changed with every delta would be a new row on
// every frame, and the list would jump for the whole of a streamed answer.
val last = items.lastOrNull()
if (last is TranscriptItem.AssistantMsg) {
// A message growing again is not finished, whatever a status said in between.
items.dropLast(1) + last.copy(text = last.text + event.delta, settled = false)
// Only into a reply that is still arriving. A settled one is a turn that ended, and
// text after it belongs to the next turn -- a separate message, drawn as its own row.
// Growing it instead ran two answers together with not even a space between them,
// which is what happens whenever a turn starts with nothing recorded in front of it:
// a subagent reporting back, or a peer message the CLI only owns up to at the end.
if (last is TranscriptItem.AssistantMsg && !last.settled) {
items.dropLast(1) + last.copy(text = last.text + event.delta, ts = entry.ts)
} else {
items + TranscriptItem.AssistantMsg(entry.seq, event.delta)
// A rule between the two, and only where they actually meet: anything that draws a
// row of its own -- a message, a command, a peer note -- is already the boundary.
val between =
if (last is TranscriptItem.AssistantMsg)
listOf(TranscriptItem.TurnBreak(entry.seq))
else emptyList()
items + between + TranscriptItem.AssistantMsg(entry.seq, event.delta, ts = entry.ts)
}
}
is SessionEvent.AssistantTextFinal -> {
val last = items.lastOrNull()
if (last is TranscriptItem.AssistantMsg && !last.settled) {
items.dropLast(1) +
last.copy(text = event.text, replacesPrefix = true, ts = entry.ts)
} else {
val between =
if (last is TranscriptItem.AssistantMsg)
listOf(TranscriptItem.TurnBreak(entry.seq))
else emptyList()
items +
between +
TranscriptItem.AssistantMsg(
entry.seq,
event.text,
replacesPrefix = true,
ts = entry.ts,
)
}
}
is SessionEvent.Thinking -> {
// Deltas grow the open block, keeping the seq of the first of them, for the same
// reason a reply's do: a row whose identity changed per delta is a new row per frame.
val last = items.lastOrNull()
if (last is TranscriptItem.ThinkingRow && last.open) {
items.dropLast(1) + last.copy(text = last.text + event.delta)
} else {
items + TranscriptItem.ThinkingRow(entry.seq, event.delta)
}
}
// The newest block still open, rather than whatever row happens to be last.
is SessionEvent.ThinkingDone ->
closeThinking(items) { it.copy(ms = event.ms, open = false) }
is SessionEvent.ToolStart ->
items +
TranscriptItem.ToolRun(
entry.seq,
event.id,
runIdFor(items, event.id, event.tool),
event.tool,
event.input,
"",
done = false,
)
// A call id names one call for its whole lifetime. Codex can repeat the start while
// recovering an in-flight item; appending that replay made two rows with one key, and
// Compose aborts the entire LazyColumn when it encounters them. Ignoring the replay
// also repairs transcripts which already contain it when they are folded on reopen.
if (items.any { it is TranscriptItem.ToolRun && it.id == event.id }) {
items
} else {
items +
TranscriptItem.ToolRun(
entry.seq,
event.id,
runIdFor(items, event.id, event.tool),
event.tool,
event.input,
"",
done = false,
)
}
is SessionEvent.ToolUpdate -> updateTool(items, event.id) { it.copy(output = event.output) }
is SessionEvent.ToolEnd ->
// Created when its start is not here, rather than dropped. A fold that only ever
@@ -444,7 +600,13 @@ fun foldEvent(items: List<TranscriptItem>, entry: SeqEvent): List<TranscriptItem
// nothing it belongs above.
is SessionEvent.MessageDropped -> items
is SessionEvent.Settings -> items
// Neither carries a row: both are about what the session can do rather than about anything
// said in it, and the composer is where they are drawn.
is SessionEvent.Images -> items
is SessionEvent.BackgroundTasks -> items
is SessionEvent.Status -> settleReply(items, event.state)
is SessionEvent.AuthenticationRequired ->
items + TranscriptItem.ErrorMsg(entry.seq, event.message)
is SessionEvent.Error -> items + TranscriptItem.ErrorMsg(entry.seq, event.message)
is SessionEvent.Image ->
// Under the call that produced it when there is one, and a row of its own when there is
@@ -458,12 +620,32 @@ fun foldEvent(items: List<TranscriptItem>, entry: SeqEvent): List<TranscriptItem
} else {
items + TranscriptItem.ImageItem(entry.seq, event.ref)
}
is SessionEvent.LimitReached -> items + TranscriptItem.LimitNote(entry.seq, event.resetsAt)
is SessionEvent.Cleared -> items + TranscriptItem.ClearedNote(entry.seq)
is SessionEvent.Compacted ->
items + TranscriptItem.CompactedNote(entry.seq, event.preTokens, event.postTokens)
is SessionEvent.Unknown -> items + TranscriptItem.Note(entry.seq, "[${event.type}]")
// Screen-level state, not transcript rows -- see SessionScreen.
is SessionEvent.UsageDelta -> items
// Said rather than skipped: a line the server could not read is a hole in the conversation,
// and one that draws nothing is a hole nothing on screen ever mentions.
is SessionEvent.Unreadable ->
items + TranscriptItem.Note(entry.seq, "[unreadable: ${event.kind}]")
// No row: see [SessionEvent.RetiredTaskNote].
is SessionEvent.RetiredTaskNote -> items
// No row of its own -- the counts are screen-level state, see SessionScreen -- but the
// generation speed belongs under the reply it measured, and this is where that reply ends.
// Only onto the newest row, and only when that row is a reply: a turn whose usage arrives
// after a tool call has nothing here to put it on, which draws as a footer without it.
is SessionEvent.UsageDelta ->
when (val last = items.lastOrNull()) {
is TranscriptItem.AssistantMsg ->
items.dropLast(1) +
last.copy(
tokensPerSecond = event.tokensPerSecond,
prefillMs = event.prefillMs,
)
else -> items
}
is SessionEvent.ContextWindow -> items
}
/**
@@ -474,9 +656,22 @@ fun foldEvent(items: List<TranscriptItem>, entry: SeqEvent): List<TranscriptItem
*/
private fun settleReply(items: List<TranscriptItem>, state: String): List<TranscriptItem> {
if (sessionWorking(state)) return items
val last = items.lastOrNull() as? TranscriptItem.AssistantMsg ?: return items
if (last.settled) return items
return items.dropLast(1) + last.copy(settled = true)
// A block the turn ended in the middle of is over, however it ended. Left open it spins for
// the rest of the conversation, which says the machine is working when nothing is.
val ended = closeThinking(items) { it.copy(open = false) }
val last = ended.lastOrNull() as? TranscriptItem.AssistantMsg ?: return ended
if (last.settled) return ended
return ended.dropLast(1) + last.copy(settled = true)
}
/** [change] applied to the newest thinking block still open, if there is one. */
private fun closeThinking(
items: List<TranscriptItem>,
change: (TranscriptItem.ThinkingRow) -> TranscriptItem.ThinkingRow,
): List<TranscriptItem> {
val at = items.indexOfLast { it is TranscriptItem.ThinkingRow && it.open }
if (at < 0) return items
return items.toMutableList().apply { this[at] = change(this[at] as TranscriptItem.ThinkingRow) }
}
private fun updateTool(
@@ -523,6 +718,10 @@ suspend fun warm(replies: ParsedReplies, rows: List<TranscriptItem>) {
// transcript often enough that leaving it out was the whole of why one cost a fifth
// of a second to open.
is TranscriptItem.PeerNote -> listOf(row.text)
// A model's working is markdown too, and only once it is settled: an open block
// gains a delta at a time and is drawn by the incremental parse, so warming one
// would hold a parse of every prefix of it.
is TranscriptItem.ThinkingRow -> if (row.open) emptyList() else listOf(row.text)
else -> emptyList()
}
}
app/androidApp/src/main/kotlin/com/example/aiapp/TranscriptList.kt
+26 -6
View File
@@ -1,6 +1,7 @@
package com.example.aiapp
import androidx.compose.foundation.layout.Box
import androidx.compose.foundation.layout.Column
import androidx.compose.foundation.layout.PaddingValues
import androidx.compose.foundation.layout.fillMaxWidth
import androidx.compose.foundation.layout.padding
@@ -10,6 +11,9 @@ import androidx.compose.foundation.lazy.LazyListState
import androidx.compose.foundation.text.selection.SelectionContainer
import androidx.compose.foundation.text.selection.SelectionState
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
@@ -49,6 +53,8 @@ fun TranscriptList(
units: List<TranscriptUnit>,
state: LazyListState,
moreHistory: Boolean,
historyError: String?,
onRetryHistory: () -> Unit,
selection: SelectionState,
modifier: Modifier = Modifier,
below: @Composable () -> Unit,
@@ -94,15 +100,29 @@ fun TranscriptList(
DebugStats.count("unit composed")
Box(Modifier.fillMaxWidth().padding(top = u.gap)) { unit(u) }
}
// Standing in for everything not fetched yet. Only here while there is more -- its
// appearance at the top edge is also roughly when the next page is asked for, so what
// it reports is a fetch in flight rather than an end reached.
// Standing in for everything not fetched yet. A failed fetch stays actionable here:
// when the loaded transcript is too short to scroll, this boundary is the only place
// the reader can be given another way to ask.
if (moreHistory) {
item(key = "history", contentType = "history") {
Box(Modifier.fillMaxWidth().padding(vertical = 24.dp)) {
CircularProgressIndicator(
Modifier.align(Alignment.Center).size(HISTORY_SPINNER)
)
if (historyError == null) {
CircularProgressIndicator(
Modifier.align(Alignment.Center).size(HISTORY_SPINNER)
)
} else {
Column(
Modifier.align(Alignment.Center),
horizontalAlignment = Alignment.CenterHorizontally,
) {
Text(
"Couldn't load earlier messages. $historyError",
color = MaterialTheme.colorScheme.error,
style = MaterialTheme.typography.bodySmall,
)
TextButton(onClick = onRetryHistory) { Text("Try again") }
}
}
}
}
}
app/androidApp/src/main/kotlin/com/example/aiapp/TranscriptSource.kt
+5 -5
View File
@@ -18,7 +18,7 @@ import java.util.concurrent.atomic.AtomicReference
*/
class TranscriptSource(
private val settings: ServerSettings,
private val sessionId: String,
private val address: TranscriptAddress,
val cache: SessionCache,
) {
private val stream = AtomicReference<EventStream?>(null)
@@ -65,7 +65,7 @@ class TranscriptSource(
val tail = cache.tail() ?: return false
// `before = seq + 1` is the newest event with seq <= the cursor, which is the event *at*
// the cursor when the server still has one there.
val answer = fetchTranscript(settings, sessionId, before = tail.seq + 1, limit = 1)
val answer = fetchTranscript(settings, address, before = tail.seq + 1, limit = 1)
val matches =
answer.size == 1 &&
try {
@@ -83,7 +83,7 @@ class TranscriptSource(
*/
suspend fun fetchOpening(): List<SeqEvent> {
DebugStats.count("transcript page from server")
val page = fetchTranscript(settings, sessionId, limit = OPENING_WINDOW)
val page = fetchTranscript(settings, address, limit = OPENING_WINDOW)
page.forEach { (line, entry) -> cache.append(line, entry.seq) }
cache.flush()
return page.map { it.second }
@@ -108,7 +108,7 @@ class TranscriptSource(
val page =
fetchTranscript(
settings,
sessionId,
address,
before = before,
limit = limit,
coalesce = coalesce,
@@ -131,7 +131,7 @@ class TranscriptSource(
* well lose.
*/
fun follow(after: Long, onOpen: () -> Unit, onReset: () -> Unit, onEvent: (SeqEvent) -> Unit) {
val opened = EventStream(settings, sessionId)
val opened = EventStream(settings, address)
stream.getAndSet(opened)?.close()
try {
opened.run(after, onOpen, onReset) { raw, entry ->
app/androidApp/src/main/kotlin/com/example/aiapp/TranscriptUnits.kt
+41
View File
@@ -130,6 +130,25 @@ sealed class TranscriptUnit {
get() = "u$seq:$ordinal"
}
/**
* The line under a finished reply: when it was sent, and how fast it was generated.
*
* A unit of its own rather than something drawn inside the last block, because a settled reply
* *is* its blocks -- there is no row left to hang it on, and the last block is a piece of
* markdown that knows nothing about the message it came from.
*/
data class ReplyFoot(
override val seq: Long,
override val ordinal: Int,
val ts: Double,
val tokensPerSecond: Double?,
val prefillMs: Long?,
override val gap: Dp,
) : TranscriptUnit() {
override val key: Any
get() = "f$seq"
}
/** One memory note of a settled reply; see [MemoryNote]. */
data class Memory(
override val seq: Long,
@@ -237,6 +256,19 @@ fun transcriptUnits(
}
}
}
// Unconditional, because being in this branch is what says the reply is over:
// [splitWanted] is settled-or-overtaken. The case to keep out is a message still
// arriving, whose "sent at" is not yet the one it ends up with, and that is drawn
// whole.
units +=
TranscriptUnit.ReplyFoot(
row.startSeq,
ordinal,
item.ts,
item.tokensPerSecond,
item.prefillMs,
gap(FOOT_SPACING),
)
} else {
units += TranscriptUnit.Whole(row, rowGap)
}
@@ -280,6 +312,14 @@ fun unwarmedReplies(rows: List<TranscriptRow>, replies: ParsedReplies): List<Tra
* length has lines that wrap, so its bubble is at the full width already and the slices match it
* exactly. Below it, one item of at most a few screens is nothing the list minds composing.
*/
/**
* The room between a reply's last block and the line under it.
*
* Tighter than the gap between blocks: the footer belongs to the message above it, and at a block's
* spacing it reads as a row of its own floating between two replies.
*/
private val FOOT_SPACING: Dp = 2.dp
const val USER_SPLIT_CHARS = 4000
/**
@@ -375,6 +415,7 @@ private val TranscriptUnit?.kind: String
is TranscriptUnit.PeerBlock -> "peer block"
is TranscriptUnit.UserChunk -> "user slice"
is TranscriptUnit.Memory -> "memory note"
is TranscriptUnit.ReplyFoot -> "reply footer"
is TranscriptUnit.Whole ->
when (val row = row) {
is TranscriptRow.Tools -> "tool group"
app/androidApp/src/main/kotlin/com/example/aiapp/UsageDialog.kt
+89 -32
View File
@@ -9,12 +9,15 @@ import androidx.compose.foundation.layout.padding
import androidx.compose.foundation.rememberScrollState
import androidx.compose.foundation.verticalScroll
import androidx.compose.material3.CircularProgressIndicator
import androidx.compose.material3.LinearProgressIndicator
import androidx.compose.material3.MaterialTheme
import androidx.compose.material3.Surface
import androidx.compose.material3.Text
import androidx.compose.material3.TextButton
import androidx.compose.runtime.Composable
import androidx.compose.runtime.getValue
import androidx.compose.runtime.mutableStateOf
import androidx.compose.runtime.remember
import androidx.compose.runtime.setValue
import androidx.compose.ui.Alignment
import androidx.compose.ui.Modifier
import androidx.compose.ui.unit.dp
@@ -30,7 +33,14 @@ import java.time.OffsetDateTime
* own, so the only thing its Back could ever have meant was "put this away".
*/
@Composable
fun UsageDialog(feed: UsageFeed, onDismiss: () -> Unit) {
fun UsageDialog(
settings: ServerSettings,
feed: UsageFeed,
session: SessionSummary,
onDismiss: () -> Unit,
) {
var signingIn by remember { mutableStateOf(false) }
val now = rememberUsageNow()
// A plain Dialog rather than an AlertDialog, for the spacing alone. AlertDialog fixes the gaps
// between its title, content and buttons at sizes meant for a sentence of prose and a decision;
// this is a dense read-out, and those gaps left a band of empty dialog above Close that was
@@ -45,10 +55,6 @@ fun UsageDialog(feed: UsageFeed, onDismiss: () -> Unit) {
verticalAlignment = Alignment.CenterVertically,
modifier = Modifier.fillMaxWidth(),
) {
// Deliberately not subtitled with the provider this was opened from. These
// numbers belong to an account on a particular machine -- naming the session's
// provider here made an echo session's screen read "echo" above a line reading
// "claude". Each machine names itself and the service it came from.
Text(
"Usage",
style = MaterialTheme.typography.headlineSmall,
@@ -65,10 +71,24 @@ fun UsageDialog(feed: UsageFeed, onDismiss: () -> Unit) {
}
Spacer(Modifier.height(8.dp))
// Scrolls rather than being trimmed: a machine can report any number of windows and
// there can be any number of machines, and a dialog is the one place where running
// out of room is silent. `fill = false` so a short read-out keeps a short dialog.
// a provider can report several billing pools, and a dialog is the one place where
// running out of room is silent. `fill = false` so a short read-out keeps a short
// dialog.
Column(Modifier.weight(1f, fill = false).verticalScroll(rememberScrollState())) {
UsageBody(feed.snapshots)
val state =
when (val snapshots = feed.snapshots) {
is LoadState.Loading -> LoadState.Loading
is LoadState.Error -> snapshots
is LoadState.Loaded ->
LoadState.Loaded(
usageSnapshotsFor(
snapshots.value,
session.machine,
session.usageProvider,
)
)
}
UsageBody(state, now, onSignIn = { signingIn = true })
}
TextButton(onClick = onDismiss, modifier = Modifier.align(Alignment.End)) {
Text("Close")
@@ -76,29 +96,46 @@ fun UsageDialog(feed: UsageFeed, onDismiss: () -> Unit) {
}
}
}
if (signingIn) {
ProviderLoginDialog(
settings = settings,
machineId = session.machine,
machineName = session.machineName,
provider = session.provider,
onDismiss = { signingIn = false },
onSignedIn = {
signingIn = false
feed.refresh()
},
)
}
}
/** What came back, or why nothing did. Split out so the dialog above reads as its own shape. */
@Composable
private fun UsageBody(state: LoadState<List<UsageSnapshot>>) {
private fun UsageBody(
state: LoadState<List<UsageSnapshot>>,
now: OffsetDateTime,
onSignIn: () -> Unit,
) {
Column {
when (val current = state) {
is LoadState.Loading -> CircularProgressIndicator()
is LoadState.Error -> Text(current.message, color = MaterialTheme.colorScheme.error)
is LoadState.Loaded ->
if (current.value.isEmpty()) {
// Not an error and not a blank screen: no machine offers a paid service, so
// Not an error and not a blank screen: this provider has no paid quota, so
// there is genuinely nothing to report and saying so is the answer.
Text(
"No machine here runs anything with usage limits.",
"This session's provider has no usage limits.",
style = MaterialTheme.typography.bodyMedium,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
} else {
// No card around each machine. A card is a step up the surface ladder, and
// inside a dialog -- itself a raised surface -- the step barely renders while
// costing 16dp on every side. What separates one machine from the next is the
// line naming it.
// No card around each pool. A card is a step up the surface ladder, and inside
// a dialog -- itself a raised surface -- the step barely renders while costing
// 16dp on every side. What separates one pool from the next is the line naming
// it.
current.value.forEachIndexed { index, snapshot ->
if (index > 0) {
Spacer(Modifier.height(20.dp))
@@ -108,18 +145,18 @@ private fun UsageBody(state: LoadState<List<UsageSnapshot>>) {
// read as a section of their own. Small and quiet, because the numbers
// below are what somebody opened this to see.
Text(
"${snapshot.setupName.ifEmpty { snapshot.setup }} · ${snapshot.provider}",
usageSectionTitle(snapshot),
style = MaterialTheme.typography.bodySmall,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
SnapshotState(snapshot)
SnapshotState(snapshot, onSignIn)
snapshot.windows.forEachIndexed { windowIndex, window ->
// Between the bars, not after the last one: a trailing gap here is what
// put a band of empty dialog above the Close button.
if (windowIndex > 0) {
Spacer(Modifier.height(12.dp))
}
WindowBar(window)
WindowBar(window, now)
}
}
}
@@ -127,20 +164,44 @@ private fun UsageBody(state: LoadState<List<UsageSnapshot>>) {
}
}
private fun usageSectionTitle(snapshot: UsageSnapshot): String {
val machine = snapshot.machineName.ifEmpty { snapshot.machine }
val provider = snapshot.provider
val pool =
if (provider == "codex" && snapshot.limitId != "codex") {
when (snapshot.limitName) {
"gpt-reserve" -> "Luna Reserve"
null -> snapshot.limitId
else -> snapshot.limitName
}
} else null
return listOfNotNull(machine, provider, pool).joinToString(" · ")
}
/**
* Anything other than numbers: why this machine has none.
*
* The distinction the old single message could not draw. A machine nobody has logged in on is
* working exactly as somebody set it up, so it reads as a plain statement -- marking it would be
* the interface nagging about a decision already made. Only the two faults are coloured as faults.
* the interface nagging about a decision already made. It still offers the direct sign-in action;
* unreachable and provider failures are the states coloured as faults.
*/
@Composable
private fun SnapshotState(snapshot: UsageSnapshot) {
private fun SnapshotState(snapshot: UsageSnapshot, onSignIn: () -> Unit) {
when (snapshot.state) {
"ok" -> {}
"notLoggedIn" ->
"notLoggedIn",
"loginRequired" -> {
Text(
"No Claude account on this machine.",
snapshot.detail ?: "No Claude account on this machine.",
style = MaterialTheme.typography.bodyMedium,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
TextButton(onClick = onSignIn) { Text("Sign in") }
}
"authenticating" ->
Text(
"Claude sign-in is in progress.",
style = MaterialTheme.typography.bodyMedium,
color = MaterialTheme.colorScheme.onSurfaceVariant,
)
@@ -162,7 +223,7 @@ private fun SnapshotState(snapshot: UsageSnapshot) {
}
@Composable
private fun WindowBar(window: UsageWindow) {
private fun WindowBar(window: UsageWindow, now: OffsetDateTime) {
Column {
Row(modifier = Modifier.fillMaxWidth()) {
Text(
@@ -173,12 +234,8 @@ private fun WindowBar(window: UsageWindow) {
Text("${window.percent.toInt()}%", style = MaterialTheme.typography.bodyMedium)
}
Spacer(Modifier.height(4.dp))
LinearProgressIndicator(
progress = { (window.percent / 100.0).toFloat().coerceIn(0f, 1f) },
color = quotaColor(window.percent),
modifier = Modifier.fillMaxWidth(),
)
resetLine(window)?.let {
UsageProgressIndicator(window, now, Modifier.fillMaxWidth())
resetLine(window, now)?.let {
Spacer(Modifier.height(2.dp))
Text(
it,
@@ -197,8 +254,8 @@ private fun WindowBar(window: UsageWindow) {
* failure appeared as an ISO string in a sentence written for a person. Both are named in
* [WindowEnd], and the session bar words them the same way.
*/
private fun resetLine(window: UsageWindow): String? =
when (val end = windowEnd(window.resetsAt, OffsetDateTime.now())) {
private fun resetLine(window: UsageWindow, now: OffsetDateTime): String? =
when (val end = windowEnd(window.resetsAt, now)) {
WindowEnd.NotRunning -> null
WindowEnd.Unreadable -> "reset time unreadable"
is WindowEnd.Ends ->
app/androidApp/src/main/res/font/nerd_icons.ttf
BIN
View File
Binary file not shown.
app/androidApp/src/test/kotlin/com/example/aiapp/AuthenticationPromptTest.kt
+28
View File
@@ -0,0 +1,28 @@
package com.example.aiapp
import kotlin.test.Test
import kotlin.test.assertFalse
import kotlin.test.assertTrue
class AuthenticationPromptTest {
@Test
fun an_authentication_failure_stays_actionable_through_its_terminal_status() {
val required =
authenticationPromptAfter(
false,
SessionEvent.AuthenticationRequired("sign in again"),
)
assertTrue(authenticationPromptAfter(required, SessionEvent.Status("idle")))
}
@Test
fun a_later_provider_response_makes_an_old_failure_stale() {
assertFalse(
authenticationPromptAfter(
true,
SessionEvent.AssistantText("Working again."),
)
)
}
}
app/androidApp/src/test/kotlin/com/example/aiapp/FilesNavigationTest.kt
+33
View File
@@ -0,0 +1,33 @@
package com.example.aiapp
import kotlin.test.Test
import kotlin.test.assertEquals
class FilesNavigationTest {
@Test
fun `back walks through the common ancestor toward the project`() {
val project = "/home/bob/repos/project"
assertEquals("/", nextDirectoryToward("/etc", project))
assertEquals("/home", nextDirectoryToward("/", project))
assertEquals("/home/bob", nextDirectoryToward("/home", project))
assertEquals("/home/bob/repos", nextDirectoryToward("/home/bob", project))
assertEquals(project, nextDirectoryToward("/home/bob/repos", project))
assertEquals(null, nextDirectoryToward(project, project))
}
@Test
fun `back leaves a project descendant one directory at a time`() {
assertEquals(
"/home/bob/repos/project/src",
nextDirectoryToward("/home/bob/repos/project/src/main", "/home/bob/repos/project"),
)
}
@Test
fun `paths inside the machine home use tilde notation`() {
assertEquals("~", tildePath("/home/bob", "/home/bob"))
assertEquals("~/repos/project", tildePath("/home/bob/repos/project", "/home/bob/"))
assertEquals("/home/bobby/project", tildePath("/home/bobby/project", "/home/bob"))
assertEquals("/etc", tildePath("/etc", "/home/bob"))
}
}
app/androidApp/src/test/kotlin/com/example/aiapp/HighlighterTest.kt
+15
View File
@@ -338,6 +338,21 @@ class HighlighterTest {
assertEquals("+[-]", highlight("+[-]", fenceLanguage("brainfuck")).text)
}
@Test
fun `a diff colours changes and identifies its framing separately`() {
val code = "--- a/file\n+++ b/file\n@@ -1 +1 @@\n-old\n context\n+new"
assertSpans(code, Language.DIFF, Kind.DELETION, "-old")
assertSpans(code, Language.DIFF, Kind.ADDITION, "+new")
assertSpans(
code,
Language.DIFF,
Kind.METADATA,
"--- a/file",
"+++ b/file",
"@@ -1 +1 @@",
)
}
@Test
fun `every language the fence table knows has a scanner`() {
Language.entries.forEach { spansOf("x", it) }
app/androidApp/src/test/kotlin/com/example/aiapp/LimitRowTest.kt
+32
View File
@@ -0,0 +1,32 @@
package com.example.aiapp
import java.time.ZoneId
import kotlin.test.Test
import kotlin.test.assertEquals
import kotlin.test.assertTrue
/**
* What the transcript says where a session ran out of quota.
*
* The pair worth a test is the one that reads the same when it goes wrong: a reset time that
* arrived and one that never did. The second must not turn into a plausible-looking time, because a
* reader has no way of telling an invented one from a reported one.
*/
class LimitRowTest {
private val utc = ZoneId.of("UTC")
@Test
fun `a reported reset time is shown as a time`() {
// 2026-09-05T12:00:00Z. Asserted as a prefix and the clock reading rather than as the
// whole string: the platform's own short-time format is what this asks for, and it
// differs by JDK and locale down to which space character separates the meridiem.
val summary = limitSummary(1_788_609_600.0, utc)
assertTrue(summary.startsWith("Usage limit reached • resets "), summary)
assertTrue(summary.contains("12:00"), summary)
}
@Test
fun `a limit with no reset time says only what is known`() {
assertEquals("Usage limit reached", limitSummary(null, utc))
}
}
app/androidApp/src/test/kotlin/com/example/aiapp/MarkdownLinksTest.kt
+31
View File
@@ -0,0 +1,31 @@
package com.example.aiapp
import kotlin.test.Test
import kotlin.test.assertEquals
import kotlin.test.assertNull
class MarkdownLinksTest {
@Test
fun `absolute file paths are opened on the session machine`() {
assertEquals(
"/home/bob/repos/ai app/Main.kt",
filePathOf("/home/bob/repos/ai%20app/Main.kt"),
)
assertEquals("/home/bob/Main.kt", filePathOf("file:///home/bob/Main.kt"))
assertEquals("/home/bob/Main.kt", filePathOf("file://localhost/home/bob/Main.kt"))
}
@Test
fun `editor coordinates select the file itself`() {
assertEquals("/home/bob/Main.kt", filePathOf("/home/bob/Main.kt:42"))
assertEquals("/home/bob/Main.kt", filePathOf("file:///home/bob/Main.kt:42:7#L42"))
}
@Test
fun `ordinary links keep their external meaning`() {
assertNull(filePathOf("https://example.com/source.kt"))
assertNull(filePathOf("docs/source.kt"))
assertNull(filePathOf("//example.com/source.kt"))
assertNull(filePathOf("file://example.com/source.kt"))
}
}
app/androidApp/src/test/kotlin/com/example/aiapp/PendingMessagesTest.kt
+72
View File
@@ -0,0 +1,72 @@
package com.example.aiapp
import kotlin.test.Test
import kotlin.test.assertEquals
import kotlin.test.assertTrue
class PendingMessagesTest {
private fun local(text: String = "keep this") =
QueuedMessage("local-1", text, emptyList(), local = true)
@Test
fun a_server_queue_replaces_the_local_bridge_instead_of_duplicating_it() {
val queued =
reconcileQueuedMessage(
listOf(local()),
SessionEvent.MessageQueued("server-1", "keep this", emptyList()),
)
assertEquals(1, queued.size)
assertEquals("server-1", queued.single().id)
assertTrue(!queued.single().local)
}
@Test
fun an_immediately_received_message_removes_its_local_bridge() {
val queued =
reconcileUserMessage(
listOf(local()),
SessionEvent.UserMessage("keep this", id = null, attachments = emptyList()),
)
assertTrue(queued.isEmpty())
}
@Test
fun a_transport_failure_stays_on_its_message() {
val queued = markPendingFailure(listOf(local()), "local-1", "Can't reach the server")
assertEquals("Can't reach the server", queued.single().refusal)
assertTrue(queued.single().local)
}
@Test
fun server_acceptance_keeps_the_bubble_until_the_provider_event() {
val queued = markPendingAccepted(listOf(local()), "local-1")
assertEquals(1, queued.size)
assertTrue(queued.single().serverAccepted)
}
@Test
fun discarding_a_failed_send_removes_only_that_local_copy() {
val server = QueuedMessage("server-1", "already accepted", emptyList())
val queued = listOf(local(), local("keep this one").copy(id = "local-2"), server)
val discarded = discardPendingMessage(queued, "local-1")
assertEquals(listOf("local-2", "server-1"), discarded.map { it.id })
}
@Test
fun identical_messages_are_reconciled_one_at_a_time() {
val queued = listOf(local(), local().copy(id = "local-2"))
val afterFirst =
reconcileUserMessage(
queued,
SessionEvent.UserMessage("keep this", id = null, attachments = emptyList()),
)
assertEquals(listOf("local-2"), afterFirst.map { it.id })
}
}
app/androidApp/src/test/kotlin/com/example/aiapp/SessionImageTest.kt
+111
View File
@@ -0,0 +1,111 @@
package com.example.aiapp
import androidx.compose.ui.geometry.Offset
import kotlin.test.Test
import kotlin.test.assertEquals
class SessionImageTest {
@Test
fun `a full-height image hides both bars`() {
assertEquals(
ViewerBars(status = true, navigation = true),
viewerBars(
imageWidth = 1000,
imageHeight = 2000,
viewportWidth = 1000,
viewportHeight = 2000,
scale = 1f,
offset = Offset.Zero,
insets = ViewerBarInsets(status = 100, navigation = 100),
),
)
}
@Test
fun `a letterboxed image leaves both bars visible`() {
assertEquals(
ViewerBars(),
viewerBars(
imageWidth = 1000,
imageHeight = 500,
viewportWidth = 1000,
viewportHeight = 2000,
scale = 1f,
offset = Offset.Zero,
insets = ViewerBarInsets(status = 100, navigation = 100),
),
)
}
@Test
fun `panning into the status bar hides only that bar`() {
assertEquals(
ViewerBars(status = true),
viewerBars(
imageWidth = 1000,
imageHeight = 500,
viewportWidth = 1000,
viewportHeight = 2000,
scale = 2f,
offset = Offset(0f, -500f),
insets = ViewerBarInsets(status = 100, navigation = 100),
),
)
}
@Test
fun `native scale reverses fitting a tall image`() {
assertEquals(
1.25f,
nativeScale(
imageWidth = 1000,
imageHeight = 2000,
viewportWidth = 1000,
viewportHeight = 1600,
),
)
}
@Test
fun `native scale leaves a small image alone`() {
assertEquals(
1f,
nativeScale(
imageWidth = 500,
imageHeight = 500,
viewportWidth = 1000,
viewportHeight = 1000,
),
)
}
@Test
fun `zoom keeps the region panned to in the center`() {
assertEquals(
Offset(240f, -160f),
zoomOffset(
offset = Offset(120f, -80f),
centroid = Offset(500f, 1000f),
pan = Offset.Zero,
oldScale = 2f,
newScale = 4f,
viewportCenter = Offset(500f, 1000f),
),
)
}
@Test
fun `zoom keeps an off-center pinch beneath moving fingers`() {
assertEquals(
Offset(460f, 420f),
zoomOffset(
offset = Offset(100f, -100f),
centroid = Offset(250f, 400f),
pan = Offset(10f, 20f),
oldScale = 2f,
newScale = 4f,
viewportCenter = Offset(500f, 1000f),
),
)
}
}
app/androidApp/src/test/kotlin/com/example/aiapp/SessionUsageTest.kt
+137
View File
@@ -0,0 +1,137 @@
package com.example.aiapp
import java.time.OffsetDateTime
import kotlin.test.Test
import kotlin.test.assertEquals
class SessionUsageTest {
@Test
fun `usage snapshots stay with the session's machine and provider`() {
val claude = snapshot("machine", "claude", null)
val codex = snapshot("machine", "codex", "codex")
val reserve = snapshot("machine", "codex", "gpt-reserve")
val elsewhere = snapshot("other", "codex", "codex")
assertEquals(
listOf(codex, reserve),
usageSnapshotsFor(
listOf(claude, codex, reserve, elsewhere),
machine = "machine",
provider = "codex",
),
)
}
@Test
fun `a session without a meter has no usage snapshots`() {
assertEquals(
emptyList(),
usageSnapshotsFor(
listOf(snapshot("machine", "claude", null)),
machine = "machine",
provider = null,
),
)
}
@Test
fun `the model selects its named pool and other models use the generic pool`() {
val generic = snapshot("machine", "codex", "codex")
val spark =
snapshot(
"machine",
"codex",
"codex_bengalfox",
limitName = "GPT-5.3-Codex-Spark",
)
val reserve =
snapshot("machine", "codex", "base_model_inference", limitName = "gpt-reserve")
val pools = listOf(generic, spark, reserve)
assertEquals(spark, usagePoolFor(pools, "gpt-5.3-codex-spark"))
assertEquals(reserve, usagePoolFor(pools, "gpt-5.6-luna"))
assertEquals(generic, usagePoolFor(pools, "gpt-6-astra"))
}
@Test
fun `the bar uses the shortest reported cycle`() {
val weekly = window("Weekly", 10_080)
val hourly = window("5-hour window", 300)
assertEquals(hourly, shortestUsageWindow(listOf(weekly, hourly)))
assertEquals(null, shortestUsageWindow(listOf(window("unknown", null))))
}
@Test
fun `the time cursor follows elapsed time through the window`() {
val now = OffsetDateTime.parse("2026-09-17T12:00:00Z")
assertEquals(
0.4f,
usageWindowElapsedFraction(
window("5-hour window", 300, "2026-09-17T15:00:00Z"),
now,
),
)
}
@Test
fun `the time cursor clamps at the window ends`() {
val now = OffsetDateTime.parse("2026-09-17T12:00:00Z")
assertEquals(
0f,
usageWindowElapsedFraction(
window("5-hour window", 300, "2026-09-17T18:00:00Z"),
now,
),
)
assertEquals(
1f,
usageWindowElapsedFraction(
window("5-hour window", 300, "2026-09-17T11:00:00Z"),
now,
),
)
}
@Test
fun `the time cursor is absent without a usable duration and reset time`() {
val now = OffsetDateTime.parse("2026-09-17T12:00:00Z")
assertEquals(null, usageWindowElapsedFraction(window("unknown", null), now))
assertEquals(null, usageWindowElapsedFraction(window("not running", 300), now))
assertEquals(
null,
usageWindowElapsedFraction(window("unreadable", 300, "not a timestamp"), now),
)
assertEquals(null, usageWindowElapsedFraction(window("zero", 0), now))
}
private fun snapshot(
machine: String,
provider: String,
limitId: String?,
limitName: String? = null,
) =
UsageSnapshot(
provider = provider,
machine = machine,
machineName = machine,
limitId = limitId,
limitName = limitName,
state = "ok",
detail = null,
windows = emptyList(),
)
private fun window(label: String, durationMinutes: Long?, resetsAt: String? = null) =
UsageWindow(
kind = "test",
label = label,
percent = 12.0,
durationMinutes = durationMinutes,
resetsAt = resetsAt,
active = false,
)
}
app/androidApp/src/test/kotlin/com/example/aiapp/ThinkingTest.kt
+150
View File
@@ -0,0 +1,150 @@
package com.example.aiapp
import java.time.ZoneId
import kotlin.test.Test
import kotlin.test.assertEquals
import kotlin.test.assertNull
import kotlin.test.assertTrue
/**
* The model's working as its own row, and the line under a finished reply.
*
* Both have the same shape of hazard: a state nothing measured must not come out looking like one
* that was. A block interrupted mid-thought has no duration, and a provider that reports no
* generation speed has no figure -- neither may borrow one.
*/
class ThinkingTest {
private val utc = ZoneId.of("UTC")
private var seq = 0L
private fun fold(items: List<TranscriptItem>, event: SessionEvent, ts: Double = 1.0) =
foldEvent(items, SeqEvent(seq = ++seq, ts = ts, event = event))
private fun fold(vararg events: SessionEvent) =
events.fold(emptyList<TranscriptItem>()) { items, event -> fold(items, event) }
private fun thinking(items: List<TranscriptItem>) =
items.filterIsInstance<TranscriptItem.ThinkingRow>()
@Test
fun `deltas accumulate into one block that ends with its duration`() {
val items =
fold(
SessionEvent.Thinking("the user "),
SessionEvent.Thinking("wants a card"),
SessionEvent.ThinkingDone(12_400),
SessionEvent.AssistantText("Here it is."),
)
val block = thinking(items).single()
assertEquals("the user wants a card", block.text)
assertEquals(12_400, block.ms)
assertEquals("Thought for 12.4s", thinkingHeadline(block))
// Its own row, above the reply rather than inside it.
assertEquals(1, items.filterIsInstance<TranscriptItem.AssistantMsg>().size)
}
@Test
fun `a block the turn ended in the middle of stops without naming a span`() {
val items = fold(SessionEvent.Thinking("half a thought"), SessionEvent.Status("idle"))
val block = thinking(items).single()
assertNull(block.ms)
assertTrue(!block.open)
assertEquals("Thought", thinkingHeadline(block))
}
@Test
fun `a block still being thought says so`() {
val block = thinking(fold(SessionEvent.Thinking("hmm"))).single()
assertTrue(block.open)
assertEquals("Thinking", thinkingHeadline(block))
}
@Test
fun `thinking between two replies is two replies and two blocks`() {
val items =
fold(
SessionEvent.Thinking("first"),
SessionEvent.ThinkingDone(1_000),
SessionEvent.AssistantText("One."),
SessionEvent.Thinking("second"),
SessionEvent.ThinkingDone(2_000),
SessionEvent.AssistantText("Two."),
)
assertEquals(listOf("first", "second"), thinking(items).map { it.text })
assertEquals(
listOf("One.", "Two."),
items.filterIsInstance<TranscriptItem.AssistantMsg>().map { it.text },
)
}
@Test
fun `a reply carries when it was sent and what it cost to produce`() {
val items =
fold(emptyList(), SessionEvent.AssistantText("Done."), ts = 1_788_609_600.0).let {
fold(it, SessionEvent.UsageDelta(42, 100, 18.37, 9_489))
}
val reply = items.filterIsInstance<TranscriptItem.AssistantMsg>().single()
assertEquals(1_788_609_600.0, reply.ts)
assertEquals(18.37, reply.tokensPerSecond)
assertEquals(9_489, reply.prefillMs)
val footer = replyFooterText(reply.ts, reply.tokensPerSecond, reply.prefillMs, utc)
// The clock reading rather than the whole string: the platform's own short-time format
// differs by JDK and locale, which is the point of asking it for one.
assertTrue(footer!!.startsWith("read 9.5s · 18.4 tok/s · "), footer)
assertTrue(footer.contains("12:00"), footer)
}
@Test
fun `the clock stays at the end however much the provider measured`() {
// What a provider that measures nothing leaves: the time, and nothing in front of it.
val bare = replyFooterText(1_788_609_600.0, null, null, utc)
assertTrue(bare!!.contains("12:00"), bare)
assertTrue(!bare.contains("tok/s") && !bare.contains("read"), bare)
// Every shape ends with the same thing, which is the whole point of the order: the clock
// does not move because the session is on a provider that measures more or less.
val shapes =
listOf(
bare,
replyFooterText(1_788_609_600.0, 18.37, null, utc)!!,
replyFooterText(1_788_609_600.0, null, 9_489, utc)!!,
replyFooterText(1_788_609_600.0, 18.37, 9_489, utc)!!,
)
assertEquals(1, shapes.map { it.substringAfterLast("· ") }.distinct().size, "$shapes")
// A reply with nothing to say has no line at all rather than an empty one.
assertNull(replyFooterText(0.0, null, null, utc))
}
@Test
fun `a block cut by a page boundary is one block, and it is not still going`() {
// Each page folded on its own, as the app does: the older one holds the fragments before
// the cut and no ending, the newer one the rest and the ending.
val older = fold(SessionEvent.Thinking("half a "))
val newer = fold(SessionEvent.Thinking("thought"), SessionEvent.ThinkingDone(2_000))
val joined = joinPages(older, newer)
val block = thinking(joined).single()
assertEquals("half a thought", block.text)
assertEquals(2_000, block.ms)
assertTrue(!block.open)
}
@Test
fun `two blocks meeting at a page boundary stay two`() {
val older = fold(SessionEvent.Thinking("first"), SessionEvent.ThinkingDone(1_000))
val newer = fold(SessionEvent.Thinking("second"), SessionEvent.ThinkingDone(2_000))
assertEquals(listOf("first", "second"), thinking(joinPages(older, newer)).map { it.text })
}
@Test
fun `usage that lands after a tool call is not folded onto an older reply`() {
val items =
fold(
SessionEvent.AssistantText("Reading it."),
SessionEvent.ToolStart("t1", "Read", "{}"),
SessionEvent.ToolEnd("t1", "done"),
SessionEvent.UsageDelta(42, 100, 18.0, 500),
)
assertNull(items.filterIsInstance<TranscriptItem.AssistantMsg>().single().tokensPerSecond)
}
}
app/androidApp/src/test/kotlin/com/example/aiapp/ToolInputTest.kt
+43
View File
@@ -0,0 +1,43 @@
package com.example.aiapp
import kotlin.test.Test
import kotlin.test.assertEquals
import kotlin.test.assertNull
import kotlin.test.assertTrue
class ToolInputTest {
@Test
fun `bash wrapper ignores double quotes inside its outer pair`() {
assertEquals(
"rg -n \"needle\" server app",
renderedBashScript("/usr/bin/bash -lc \"rg -n \"needle\" server app\""),
)
}
@Test
fun `bash wrapper ignores single quotes inside its outer pair`() {
assertEquals(
"printf 'hello'",
renderedBashScript("/bin/bash -lc 'printf 'hello''"),
)
}
@Test
fun `unquoted or unfamiliar commands stay intact`() {
assertNull(renderedBashScript("/usr/bin/bash -lc echo hello"))
assertNull(renderedBashScript("/usr/bin/fish -lc 'echo hello'"))
}
@Test
fun `missing optional input is not displayed as null`() {
assertTrue(parseToolInput("TaskOutput", "null").rest.isEmpty())
}
@Test
fun `collaboration calls say what they do and omit empty completion`() {
assertEquals("Spawn agent", toolDisplayName("Task"))
assertEquals("Wait for agents", toolDisplayName("TaskOutput"))
assertEquals("", toolDisplayOutput("TaskOutput", "completed"))
assertEquals("failed", toolDisplayOutput("TaskOutput", "failed"))
}
}
app/androidApp/src/test/kotlin/com/example/aiapp/ToolRowsTest.kt
+167
View File
@@ -0,0 +1,167 @@
package com.example.aiapp
import kotlin.test.Test
import kotlin.test.assertEquals
import kotlin.test.assertTrue
/**
* How a run of tool calls is cut into rows: the call still running, the last call in the
* transcript, and one held out because the reader has it open are drawn on their own, and every
* piece the cut leaves behind still has a key of its own -- two rows sharing one key take the app
* down, and a key that moves takes the reader's place with it.
*/
class ToolRowsTest {
private var seq = 0L
private fun call(id: String, runId: String = id, done: Boolean = true) =
TranscriptItem.ToolRun(
seq = ++seq,
id = id,
runId = runId,
tool = "Bash",
input = "{}",
output = if (done) "ok" else "",
done = done,
)
/** Something that is not a tool call, to put behind the run so its last call folds in. */
private fun reply() = TranscriptItem.AssistantMsg(seq = ++seq, text = "done")
private fun shape(rows: List<TranscriptRow>) = rows.map { row ->
when (row) {
is TranscriptRow.Tools -> row.calls.map { it.id }
is TranscriptRow.Single -> listOf((row.item as? TranscriptItem.ToolRun)?.id ?: "reply")
}
}
private fun assertKeysDistinct(rows: List<TranscriptRow>) =
assertEquals(rows.size, rows.map { it.key }.toSet().size, "$rows")
@Test
fun the_call_still_running_is_a_row_of_its_own() {
val rows =
groupToolRuns(
listOf(
call("a"),
call("b", runId = "a"),
call("c", runId = "a", done = false),
call("d", runId = "a"),
reply(),
)
)
assertEquals(
listOf(listOf("a", "b"), listOf("c"), listOf("d"), listOf("reply")),
shape(rows),
)
assertKeysDistinct(rows)
}
@Test
fun a_call_running_in_the_middle_of_its_run_splits_the_group_in_two() {
val rows =
groupToolRuns(
listOf(
call("a"),
call("b", runId = "a", done = false),
call("c", runId = "a"),
call("d", runId = "a"),
reply(),
)
)
assertEquals(
listOf(listOf("a"), listOf("b"), listOf("c", "d"), listOf("reply")),
shape(rows),
)
assertKeysDistinct(rows)
}
/**
* A call held out is one the reader opened while it stood on its own; being overtaken while
* they read it does not fold it away, and closing it hands it back to its run.
*/
@Test
fun a_held_out_call_stays_out_of_its_group() {
val calls =
listOf(
call("a"),
call("b", runId = "a"),
call("c", runId = "a"),
call("d", runId = "a"),
reply(),
)
val whileHeld = groupToolRuns(calls, heldOut = setOf("d"))
val afterItCloses = groupToolRuns(calls)
assertEquals(listOf(listOf("a", "b", "c"), listOf("d"), listOf("reply")), shape(whileHeld))
assertTrue(whileHeld[1] is TranscriptRow.Single, "$whileHeld")
assertKeysDistinct(whileHeld)
assertEquals(listOf(listOf("a", "b", "c", "d"), listOf("reply")), shape(afterItCloses))
assertTrue(afterItCloses.first() is TranscriptRow.Tools, "$afterItCloses")
}
/**
* The one case where the run's name is a call that is not in the run's first row: a page of
* history joined onto a run whose own first call is still going ([joinPages] renames the older
* calls to the newer run's name). Both rows would key on that name.
*/
@Test
fun the_run_keeps_its_name_even_when_the_call_it_is_named_after_is_the_one_running() {
val rows = groupToolRuns(listOf(call("a", runId = "b"), call("b", done = false)))
assertEquals(listOf(listOf("a"), listOf("b")), shape(rows))
assertKeysDistinct(rows)
assertEquals("b", rows.first().key)
}
@Test
fun a_run_that_reappears_after_another_row_keeps_distinct_keys() {
val rows =
groupToolRuns(
listOf(
call("older", runId = "exec-1"),
call("older-2", runId = "exec-1"),
reply(),
call("exec-1", runId = "exec-1"),
call("newer", runId = "exec-1"),
reply(),
)
)
assertTrue(rows[0] is TranscriptRow.Tools, "$rows")
assertTrue(rows[2] is TranscriptRow.Tools, "$rows")
assertKeysDistinct(rows)
assertEquals("exec-1", rows[0].key)
assertEquals("exec-1/exec-1", rows[2].key)
}
/**
* Finishing is not what folds a call back in -- being overtaken is. A session that has run its
* last command and is writing its reply leaves that command standing until the reply starts.
*/
@Test
fun the_last_call_stays_out_when_it_finishes_and_folds_in_when_something_follows() {
val a = call("a")
val running = call("b", runId = "a", done = false)
val finished = running.copy(done = true)
val whileRunning = groupToolRuns(listOf(a, running))
val afterItEnds = groupToolRuns(listOf(a, finished))
val afterTheReply = groupToolRuns(listOf(a, finished, reply()))
assertEquals(listOf(listOf("a"), listOf("b")), shape(whileRunning))
assertEquals(listOf(listOf("a"), listOf("b")), shape(afterItEnds))
assertEquals(listOf(listOf("a", "b"), listOf("reply")), shape(afterTheReply))
// The run keeps the key it was drawn under throughout, so the list rebuilds a row rather
// than losing its anchor.
assertEquals(whileRunning.first().key, afterItEnds.first().key)
assertEquals(whileRunning.first().key, afterTheReply.first().key)
}
@Test
fun a_run_of_finished_calls_is_one_group_once_something_follows_it() {
val rows =
groupToolRuns(
listOf(call("a"), call("b", runId = "a"), call("c", runId = "a"), reply())
)
assertEquals(listOf(listOf("a", "b", "c"), listOf("reply")), shape(rows))
assertTrue(rows.first() is TranscriptRow.Tools, "$rows")
}
}
app/androidApp/src/test/kotlin/com/example/aiapp/TranscriptCacheTest.kt
+1 -1
View File
@@ -23,7 +23,7 @@ class TranscriptCacheTest {
private fun cache() = TranscriptCache(File(temp, "v1/host_8443")) { said += it }
private fun session(id: String = "s") = cache().session(id)
private fun session(id: String = "s") = cache().session(TranscriptAddress(id))
private fun line(seq: Long, type: String = "toolStart") =
"""{"seq":$seq,"ts":1.5,"type":"$type","id":"x"}"""
app/androidApp/src/test/kotlin/com/example/aiapp/TranscriptItemsTest.kt
+185
View File
@@ -0,0 +1,185 @@
package com.example.aiapp
import kotlin.test.Test
import kotlin.test.assertEquals
import kotlin.test.assertTrue
/**
* Where one turn ends and the next begins, which is the part of the fold that had no way of saying
* anything was wrong: two replies run together read as one long answer, and the seam is somewhere
* in the middle of a sentence.
*/
class TranscriptItemsTest {
private var seq = 0L
private fun fold(items: List<TranscriptItem>, event: SessionEvent) =
foldEvent(items, SeqEvent(seq = ++seq, ts = 1.0, event = event))
private fun fold(vararg events: SessionEvent) =
events.fold(emptyList<TranscriptItem>()) { items, event -> fold(items, event) }
private fun texts(items: List<TranscriptItem>) =
items.filterIsInstance<TranscriptItem.AssistantMsg>().map { it.text }
@Test
fun an_authentication_failure_stays_visible_as_an_error_row() {
val entry =
SeqEvent(
seq = 7,
ts = 1.0,
event = SessionEvent.AuthenticationRequired("sign in again"),
)
assertEquals(
TranscriptItem.ErrorMsg(7, "sign in again"),
foldEvent(emptyList(), entry).single(),
)
}
@Test
fun text_after_the_turn_ended_is_a_new_reply_rather_than_more_of_the_last_one() {
val items =
fold(
SessionEvent.AssistantText("You'll get the one-line notice when it lands."),
SessionEvent.Status("idle"),
SessionEvent.AssistantText("Dev Updater fix is pushed."),
)
assertEquals(
listOf("You'll get the one-line notice when it lands.", "Dev Updater fix is pushed."),
texts(items),
)
}
@Test
fun deltas_of_one_reply_still_accumulate_into_it() {
val items =
fold(
SessionEvent.AssistantText("Still "),
SessionEvent.AssistantText("running "),
SessionEvent.Status("running"),
SessionEvent.AssistantText("its tests."),
)
assertEquals(listOf("Still running its tests."), texts(items))
}
@Test
fun completed_text_replaces_provisional_deltas_live() {
val items =
fold(
SessionEvent.AssistantText("I'll inspect the color-c concrete implementation"),
SessionEvent.AssistantTextFinal(
"I’ll inspect the color-correction TODO and the relevant design."
),
)
assertEquals(
listOf("I’ll inspect the color-correction TODO and the relevant design."),
texts(items),
)
}
@Test
fun completed_text_discards_provisional_deltas_across_a_page_boundary() {
val earlier = fold(SessionEvent.AssistantText("1. provisional section\n\n"))
val later =
fold(
SessionEvent.AssistantTextFinal("1. final first section\n\n2. final second section")
)
assertEquals(
listOf("1. final first section\n\n2. final second section"),
texts(joinPages(earlier, later)),
)
}
@Test
fun a_final_value_after_a_settled_reply_is_a_new_message_across_a_page_boundary() {
val earlier =
fold(
SessionEvent.AssistantText("Previous answer."),
SessionEvent.Status("idle"),
)
val later = fold(SessionEvent.AssistantTextFinal("Next answer."))
assertEquals(
listOf("Previous answer.", "Next answer."),
texts(joinPages(earlier, later)),
)
}
/**
* The rule that replaced the wall of reports. A turn that starts with nothing recorded in front
* of it -- a subagent finishing, the CLI picking a conversation back up -- leaves two replies
* abutting, and only the break says they are two.
*/
@Test
fun two_replies_that_meet_are_separated_by_a_rule_and_nothing_else() {
val items =
fold(
SessionEvent.AssistantText("Launched it."),
SessionEvent.Status("waiting"),
SessionEvent.AssistantText("Noted."),
)
assertEquals(3, items.size, "$items")
assertTrue(items[1] is TranscriptItem.TurnBreak, "$items")
assertEquals(listOf("Launched it.", "Noted."), texts(items))
// Distinct keys: the break shares the reply's seq, and two items with one key take the
// app down.
assertEquals(3, items.map { it.key }.toSet().size, "$items")
}
/**
* A reply after anything that draws a row of its own needs no rule: that row is the boundary.
*/
@Test
fun a_reply_after_a_row_of_its_own_gets_no_rule() {
val items =
fold(
SessionEvent.AssistantText("Launched it."),
SessionEvent.Status("idle"),
SessionEvent.UserMessage("carry on", null, emptyList()),
SessionEvent.AssistantText("Noted."),
)
assertTrue(items.none { it is TranscriptItem.TurnBreak }, "$items")
}
@Test
fun a_repeated_tool_start_is_still_one_row() {
val start = SessionEvent.ToolStart("exec-1", "Bash", "{\"command\":\"cargo test\"}")
val items =
fold(
start,
SessionEvent.AssistantText("The test run is still going."),
start,
SessionEvent.ToolEnd("exec-1", "finished"),
)
val tools = items.filterIsInstance<TranscriptItem.ToolRun>()
assertEquals(1, tools.size, "$items")
assertEquals("finished", tools.single().output)
assertTrue(tools.single().done)
}
/**
* The page-join half of the same rule. A boundary that cuts one reply leaves an unfinished half
* to be rejoined; a boundary that lands between two turns must not join anything, or paging
* back puts the run-together paragraph straight back.
*/
@Test
fun paging_back_rejoins_a_cut_reply_and_leaves_two_finished_ones_apart() {
val cut =
joinPages(
listOf(TranscriptItem.AssistantMsg(1, "half a ")),
listOf(TranscriptItem.AssistantMsg(2, "sentence", settled = true)),
)
assertEquals(listOf("half a sentence"), texts(cut))
val whole =
joinPages(
listOf(TranscriptItem.AssistantMsg(1, "One turn.", settled = true)),
listOf(TranscriptItem.AssistantMsg(2, "The next.", settled = true)),
)
assertEquals(listOf("One turn.", "The next."), texts(whole))
// And the rule between them, which the fold that would have made it never got to see.
assertTrue(whole.any { it is TranscriptItem.TurnBreak }, "$whole")
}
}
app/build-icon-font.sh
+5
View File
@@ -45,6 +45,11 @@ GLYPHS=(
U+F0193 # md-content_save
U+F0224 # md-file_outline
U+F201 # fa-line_chart -- Font Awesome's, asked for by name
U+F035C # md-menu -- the burger, as a row's drag handle
U+F07B7 # md-console_line -- a backgrounded command
U+F06A9 # md-robot -- a subagent
U+F04AA # md-sitemap -- a workflow
U+F0625 # md-help_circle_outline -- a background task of a kind this build does not know
)
url=https://github.com/ryanoasis/nerd-fonts/releases/latest/download/NerdFontsSymbolsOnly.zip
app/debug-transcript.sh
+2 -2
View File
@@ -120,10 +120,10 @@ TOKEN=$(grep -o 'token=[A-Za-z0-9_-]*' "$WORK/server.log" | head -1 | cut -d= -f
api() { curl -s --cacert "$CERTS/ca.pem" -H "Authorization: Bearer $TOKEN" "$@"; }
echo "==> Importing"
SETUP=$(api "https://127.0.0.1:$PORT/setups" | sed -n 's/.*"id":"\([^"]*\)".*/\1/p' | head -1)
MACHINE=$(api "https://127.0.0.1:$PORT/machines" | sed -n 's/.*"id":"\([^"]*\)".*/\1/p' | head -1)
SESSION=$(api -H 'Content-Type: application/json' -X POST \
"https://127.0.0.1:$PORT/sessions" \
-d "{\"setup\":\"$SETUP\",\"provider\":\"claude-cli\",\"title\":\"$PROJECT\",\"import\":\"$ID\"}" \
-d "{\"machine\":\"$MACHINE\",\"provider\":\"claude-cli\",\"title\":\"$PROJECT\",\"import\":\"$ID\"}" \
| sed -n 's/.*"id":"\([^"]*\)".*/\1/p' | head -1)
echo " session $SESSION, $(wc -l < "$WORK/sessions/$SESSION/transcript.jsonl") events"
app/run-android.sh
+1 -1
View File
@@ -9,7 +9,7 @@
# emulator" is three places for the memory check that was missing from all of
# them.
#
# Environment setup (SDK location, PATH, ...) lives in ./android-env.sh,
# Environment machine (SDK location, PATH, ...) lives in ./android-env.sh,
# which can also be sourced directly for one-off commands.
set -eu
app/ui-sandbox.sh
+17 -6
View File
@@ -110,7 +110,7 @@ api) # ./ui-sandbox.sh api /path [curl args...]
;;
spawn) # ./ui-sandbox.sh spawn [title] -- an echo session; prints its id
api /sessions -X POST -H 'content-type: application/json' \
-d "{\"setup\":\"local\",\"provider\":\"echo\",\"title\":\"${2:-test}\"}" |
-d "{\"machine\":\"local\",\"provider\":\"echo\",\"title\":\"${2:-test}\"}" |
python3 -c 'import json,sys; print(json.load(sys.stdin)["id"])'
exit 0
;;
@@ -150,7 +150,7 @@ if [ -f "$ROOT/config.ron" ]; then
/^tokens: \[/ { in_tokens = 1; next }
# The server writes the list back compactly, with the last entry
# and the close on one line: " ),],". Reading the close only
# at a line start ran past it into `setups`, and the salvage then
# at a line start ran past it into `machines`, and the salvage then
# carried a second copy of that block into the new config.
in_tokens && /\],/ {
sub(/\],.*/, "")
@@ -213,13 +213,24 @@ while [ "$i" -le 8 ]; do
i=$((i + 1))
done
# A CLI that does nothing, so importing one of these is free and safe.
# Everything the spawn path cares about is here: it holds the fifo open,
# records a real pid, writes nothing, and dies on a signal. A real
# A CLI that does nothing during a session and offers one deterministic login
# during `auth login`, so both paths are free and safe. Everything the spawn
# path cares about is here: it holds the fifo open, records a real pid, writes
# nothing, and dies on a signal. A real
# `claude --resume` against an invented session id would either fail in a
# way that tests nothing or start a turn on somebody's account.
cat >"$ROOT/fake-claude" <<FAKE
#!/bin/sh
if [ "\${1:-}" = auth ] && [ "\${2:-}" = login ]; then
echo 'https://claude.com/cai/oauth/authorize?state=ai-app-sandbox'
while IFS= read -r code; do
if [ "\$code" = sandbox-code ]; then
exit 0
fi
echo 'Invalid code' >&2
done
exit 1
fi
# Slow to start, on purpose. An import against this finishes in
# milliseconds otherwise, so every state on the way -- the row marked
# "importing", the queue behind it, the event that clears them -- is over
@@ -346,7 +357,7 @@ tokens: [
sha256: "$hash",
),
$salvaged],
setups: [
machines: [
(
id: "local",
name: "sandbox",
iris/.gitignore
-2
View File
@@ -1,2 +0,0 @@
/target
perf.data*
iris/Cargo.lock
Generated
-3898
View File
File diff suppressed because it is too large. Load diff
iris/Cargo.toml
-42
View File
@@ -1,42 +0,0 @@
[package]
name = "iris"
version.workspace = true
edition.workspace = true
# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
[dependencies]
iris-core = { workspace = true }
iris-macro = { workspace = true }
parley = { workspace = true }
swash = { workspace = true }
winit = { workspace = true }
arboard = { workspace = true, features = ["wayland-data-control"] }
pollster = { workspace = true }
wgpu = { workspace = true }
image = { workspace = true }
tokio = { workspace = true, features = ["sync", "rt", "rt-multi-thread"] }
[dev-dependencies]
tokio = { workspace = true, features = ["sync", "rt", "rt-multi-thread", "time"] }
[workspace]
members = ["core", "macro"]
[workspace.package]
version = "0.1.0"
edition = "2024"
[workspace.dependencies]
pollster = "0.4.0"
winit = "0.30.12"
wgpu = "28.0.0"
bytemuck = "1.23.1"
image = "0.25.6"
parley = "0.11.1"
swash = "0.2.10"
fxhash = "0.2.1"
arboard = "3.6.1"
iris-core = { path = "core" }
iris-macro = { path = "macro" }
tokio = "1.49.0"
iris/TODO
-40
View File
@@ -1,40 +0,0 @@
images
settings (sampler)
text
figure out ways to speed up / what costs the most
resizing (per frame) is really slow (assuming painter isn't griefing)
j is weird / fix x offset
masks r just made to bare minimum work
scaling
could be just a simple scaling factor that multiplies abs
and need to ensure text uses raw abs and not scaled abs
naming? (pt, px)
want to keep (drawn) regions using px? or should I add another field to UiScalar/Vec
field could be best solution so redrawing stuff isn't needed & you can specify both as user
WidgetRef<W> or smth instead of Id
enum that's either an Id or an actual concrete instance of W
painter takes them in instead of (or in addition to) id
then type wrapper widgets to contain them
allows for compile time optimization if a widget wrapper's inner is known at compile time
and the id of inner is not needed anywhere
maybe introduce InnerWidget trait to allow for editors to expose & modify inner type
maybe could also store a parent widget and keep using InnerWidget trait? unsure if possible
really weird limitation:
I don't think you can currently remove an element from a parent and put it in a child of the same parent
because it removes the unused children after the entire parent redraw
but the child gets drawn during that, so it will think the child is still active !!!
or something like that idk, maybe I need a special enum for parent that includes a undecided state where it may or may not get redrawn by the parent
or just do ref counting and ensure all drawn things == 1 afterwards (seems like best way)
ok so I'm removing the limit for now
don't forget I'm streaming
tags
vecs for each widget type?
POTENTIAL BUG: closures that store IDs will not decrement the id!!! need to not increment id if moved into closure somehow??? wait no, need to decrement ID every time an event fn is added...... only if the id is used in it..??
iris/core/Cargo.toml
-12
View File
@@ -1,12 +0,0 @@
[package]
name = "iris-core"
version.workspace = true
edition.workspace = true
[dependencies]
wgpu = { workspace = true }
bytemuck ={ workspace = true }
image = { workspace = true }
parley = { workspace = true }
swash = { workspace = true }
fxhash = { workspace = true }
iris/core/src/attr.rs
-23
View File
@@ -1,23 +0,0 @@
use crate::{UiRsc, WeakWidget, WidgetIdFn, WidgetLike};
pub trait WidgetAttr<Rsc, W: ?Sized> {
type Input;
fn run(rsc: &mut Rsc, id: WeakWidget<W>, input: Self::Input);
}
pub trait Attrable<Rsc, W: ?Sized, Tag> {
fn attr<A: WidgetAttr<Rsc, W>>(self, input: A::Input) -> impl WidgetIdFn<Rsc, W>;
}
impl<Rsc: UiRsc, WL: WidgetLike<Rsc, Tag>, Tag> Attrable<Rsc, WL::Widget, Tag> for WL {
fn attr<A: WidgetAttr<Rsc, WL::Widget>>(
self,
input: A::Input,
) -> impl WidgetIdFn<Rsc, WL::Widget> {
|rsc| {
let id = self.add(rsc);
A::run(rsc, id, input);
id
}
}
}
iris/core/src/event/ctx.rs
-18
View File
@@ -1,18 +0,0 @@
use crate::{HasEvents, WeakWidget, Widget};
pub struct EventCtx<'a, Rsc: HasEvents, Data> {
pub state: &'a mut Rsc::State,
pub data: Data,
}
pub struct EventIdCtx<'a, Rsc: HasEvents, Data, W: ?Sized> {
pub widget: WeakWidget<W>,
pub state: &'a mut Rsc::State,
pub data: Data,
}
impl<Rsc: HasEvents, Data, W: Widget> EventIdCtx<'_, Rsc, Data, W> {
pub fn widget<'a>(&self, rsc: &'a mut Rsc) -> &'a mut W {
&mut rsc.ui_mut().widgets[self.widget]
}
}
iris/core/src/event/manager.rs
-169
View File
@@ -1,169 +0,0 @@
use crate::{
ActiveData, Event, EventCtx, EventFn, EventIdCtx, EventLike, HasEvents, IdLike, LayerId,
WeakWidget, WidgetEventFn, WidgetId,
util::{HashMap, HashSet, TypeMap},
};
use std::{any::TypeId, rc::Rc};
pub struct EventManager<Rsc> {
widget_to_types: HashMap<WidgetId, HashSet<TypeId>>,
types: TypeMap<dyn EventManagerLike<Rsc>>,
}
impl<Rsc> Default for EventManager<Rsc> {
fn default() -> Self {
Self {
widget_to_types: Default::default(),
types: Default::default(),
}
}
}
impl<Rsc: HasEvents + 'static> EventManager<Rsc> {
pub fn get_type<E: EventLike>(&mut self) -> &mut TypeEventManager<Rsc, E::Event> {
self.types.type_or_default()
}
pub fn register<I: IdLike + 'static, E: EventLike>(
&mut self,
id: I,
event: E,
f: impl for<'a> WidgetEventFn<Rsc, <E::Event as Event>::Data<'a>, I::Widget>,
) {
let i = id.id();
self.get_type::<E>().register(id, event, f);
self.widget_to_types
.entry(i)
.or_default()
.insert(Self::type_key::<E>());
}
pub fn type_key<E: EventLike>() -> TypeId {
TypeId::of::<TypeEventManager<Rsc, E::Event>>()
}
}
pub trait EventsLike {
fn remove(&mut self, id: WidgetId);
fn draw(&mut self, active: &ActiveData);
fn undraw(&mut self, active: &ActiveData);
}
impl<Rsc: HasEvents + 'static> EventsLike for EventManager<Rsc> {
fn remove(&mut self, id: WidgetId) {
for t in self.widget_to_types.get(&id).into_flat_iter() {
self.types.get_mut(t).unwrap().remove(id);
}
}
fn draw(&mut self, active: &ActiveData) {
for t in self.widget_to_types.get(&active.id).into_flat_iter() {
self.types.get_mut(t).unwrap().draw(active);
}
}
fn undraw(&mut self, active: &ActiveData) {
for t in self.widget_to_types.get(&active.id).into_flat_iter() {
self.types.get_mut(t).unwrap().undraw(active);
}
}
}
pub trait EventManagerLike<State> {
fn remove(&mut self, id: WidgetId);
fn draw(&mut self, data: &ActiveData);
fn undraw(&mut self, data: &ActiveData);
}
type EventData<Rsc, E> = (E, Rc<dyn for<'a> EventFn<Rsc, <E as Event>::Data<'a>>>);
pub struct TypeEventManager<Rsc: HasEvents, E: Event> {
// TODO: reduce visiblity!!
pub active: HashMap<LayerId, HashMap<WidgetId, E::State>>,
map: HashMap<WidgetId, Vec<EventData<Rsc, E>>>,
}
impl<Rsc: HasEvents, E: Event> EventManagerLike<Rsc> for TypeEventManager<Rsc, E> {
fn remove(&mut self, id: WidgetId) {
self.map.remove(&id);
for layer in self.active.values_mut() {
layer.remove(&id);
}
}
fn draw(&mut self, data: &ActiveData) {
self.active
.entry(data.layer)
.or_default()
.entry(data.id)
.or_default();
}
fn undraw(&mut self, data: &ActiveData) {
if let Some(layer) = self.active.get_mut(&data.layer) {
layer.remove(&data.id);
}
}
}
impl<Rsc: HasEvents, E: Event> Default for TypeEventManager<Rsc, E> {
fn default() -> Self {
Self {
active: Default::default(),
map: Default::default(),
}
}
}
impl<Rsc: HasEvents + 'static, E: Event> TypeEventManager<Rsc, E> {
fn register<I: IdLike + 'static>(
&mut self,
widget: I,
event: impl EventLike<Event = E>,
f: impl for<'a> WidgetEventFn<Rsc, E::Data<'a>, I::Widget>,
) {
let event = event.into_event();
self.map.entry(widget.id()).or_default().push((
event,
Rc::new(move |ctx, rsc| {
f(
EventIdCtx {
widget: WeakWidget::new(widget.id()),
state: ctx.state,
data: ctx.data,
},
rsc,
);
}),
));
}
/// The event lists this widget was registered with (`register`'s
/// `event` argument, one per call), without running anything. Lets a
/// caller ask "would this widget's registrations match the current
/// state" separately from actually dispatching to it -- used by
/// `sense.rs` to decide whether a widget genuinely consumes a scroll
/// or press this frame (so a lower layer can still receive it if not)
/// without that decision being conflated with "the cursor happens to
/// be over it," which is all `run_fn` running something tells you.
pub fn registered(&self, id: WidgetId) -> impl Iterator<Item = &E> {
self.map.get(&id).into_iter().flatten().map(|(e, _)| e)
}
pub fn run_fn<'a>(
&mut self,
id: impl IdLike,
) -> impl for<'b> FnOnce(EventCtx<'_, Rsc, E::Data<'b>>, &mut Rsc) + 'a {
let fs = self.map.get(&id.id()).cloned().unwrap_or_default();
move |ctx, rsc| {
for (e, f) in fs {
if let Some(data) = e.should_run(&ctx.data) {
f(
EventCtx {
state: ctx.state,
data,
},
rsc,
)
}
}
}
}
}
iris/core/src/event/mod.rs
-44
View File
@@ -1,44 +0,0 @@
mod ctx;
mod manager;
mod rsc;
pub use ctx::*;
pub use manager::*;
pub use rsc::*;
pub trait Event: Sized + 'static + Clone {
type Data<'a>: Clone = ();
type State: Default = ();
#[allow(unused_variables)]
fn should_run<'a>(&self, data: &Self::Data<'a>) -> Option<Self::Data<'a>> {
Some(data.clone())
}
}
pub trait EventLike {
type Event: Event;
fn into_event(self) -> Self::Event;
}
impl<E: Event> EventLike for E {
type Event = Self;
fn into_event(self) -> Self::Event {
self
}
}
pub trait EventFn<Rsc: HasEvents, Data>: Fn(EventCtx<Rsc, Data>, &mut Rsc) + 'static {}
impl<Rsc: HasEvents, F: Fn(EventCtx<Rsc, Data>, &mut Rsc) + 'static, Data> EventFn<Rsc, Data>
for F
{
}
pub trait WidgetEventFn<Rsc: HasEvents, Data, W: ?Sized>:
Fn(EventIdCtx<Rsc, Data, W>, &mut Rsc) + 'static
{
}
impl<Rsc: HasEvents, F: Fn(EventIdCtx<Rsc, Data, W>, &mut Rsc) + 'static, Data, W: ?Sized>
WidgetEventFn<Rsc, Data, W> for F
{
}
iris/core/src/event/rsc.rs
-34
View File
@@ -1,34 +0,0 @@
use crate::{
Event, EventCtx, EventLike, EventManager, IdLike, UiRsc, WeakWidget, Widget, WidgetEventFn,
};
pub trait HasState: 'static {
type State;
}
pub trait HasEvents: Sized + UiRsc + HasState {
fn events(&self) -> &EventManager<Self>;
fn events_mut(&mut self) -> &mut EventManager<Self>;
fn register_event<W: Widget + ?Sized, E: EventLike>(
&mut self,
id: WeakWidget<W>,
event: E,
f: impl for<'a> WidgetEventFn<Self, <E::Event as Event>::Data<'a>, W>,
) {
self.events_mut().register(id, event, f);
}
}
pub trait RunEvents: HasEvents {
fn run_event<E: EventLike>(
&mut self,
id: impl IdLike,
data: <E::Event as Event>::Data<'_>,
state: &mut Self::State,
) {
let f = self.events_mut().get_type::<E>().run_fn(id);
f(EventCtx { state, data }, self)
}
}
impl<T: HasEvents> RunEvents for T {}
Loaded 100 of 223 files, more files were not shown because too many files have changed in this diff. Show more