iris: route pointer input per kind, so scroll falls through a hovered button

IRIS_TODO.md's "Input does not fall through by input type": run_sensors
treated "the cursor is over this widget" and "this widget consumed the
event" as the same check, so a widget registered only for click() still
blocked a Scroll meant for a list underneath it. Fixed by judging
consumption per input kind -- with nothing momentary happening this
frame the topmost hovered widget still wins (unchanged), but once a
scroll or a press/release is actually happening, only a widget whose
registered senses include a matching non-hover one (via the new
TypeEventManager::registered, which lists a widget's registrations
without running anything) can consume it.

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.

Co-Authored-By: Claude Sonnet <noreply@anthropic.com>

.gitignore

@@ -21,3 +21,6 @@ certs/
 config.ron
 config.json
 sessions/
+
+# iris, the in-house UI library, is vendored at iris/ and built by cargo.
+iris/target/

AGENTS.md

@@ -46,6 +46,10 @@ Module-by-module intent is in PLAN.md's "Backend layout".
   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

IRIS.md

@@ -0,0 +1,79 @@
+# 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

@@ -0,0 +1,66 @@
+# 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

@@ -0,0 +1,897 @@
+# 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.

TEXTURES.md

@@ -0,0 +1,496 @@
+# 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.

iris/.gitignore

@@ -0,0 +1,2 @@
+/target
+perf.data*

iris/Cargo.toml

@@ -0,0 +1,42 @@
+[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

@@ -0,0 +1,40 @@
+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

@@ -0,0 +1,12 @@
+[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

@@ -0,0 +1,23 @@
+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

@@ -0,0 +1,18 @@
+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

@@ -0,0 +1,169 @@
+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

@@ -0,0 +1,44 @@
+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

@@ -0,0 +1,34 @@
+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 {}

iris/core/src/lib.rs

@@ -0,0 +1,33 @@
+#![feature(macro_metavar_expr_concat)]
+#![feature(const_ops)]
+#![feature(const_trait_impl)]
+#![feature(const_convert)]
+#![feature(unboxed_closures)]
+#![feature(fn_traits)]
+#![feature(const_destruct)]
+#![feature(associated_type_defaults)]
+#![feature(unsize)]
+#![feature(coerce_unsized)]
+#![feature(option_into_flat_iter)]
+
+mod attr;
+mod event;
+mod num;
+mod orientation;
+mod primitive;
+mod render;
+mod ui;
+mod widget;
+
+pub mod util;
+
+pub use attr::*;
+pub use event::*;
+pub use num::*;
+pub use orientation::*;
+pub use primitive::*;
+pub use render::*;
+pub use ui::*;
+pub use widget::*;
+
+pub type UiColor = primitive::Color<u8>;

iris/core/src/num.rs

@@ -0,0 +1,49 @@
+use crate::util::Vec2;
+use std::marker::Destruct;
+
+pub const trait UiNum {
+    fn to_f32(self) -> f32;
+}
+
+const impl UiNum for f32 {
+    fn to_f32(self) -> f32 {
+        self
+    }
+}
+
+const impl UiNum for u32 {
+    fn to_f32(self) -> f32 {
+        self as f32
+    }
+}
+
+const impl UiNum for i32 {
+    fn to_f32(self) -> f32 {
+        self as f32
+    }
+}
+
+pub const fn vec2(x: impl const UiNum, y: impl const UiNum) -> Vec2 {
+    Vec2::new(x.to_f32(), y.to_f32())
+}
+
+const impl<T: const UiNum + Copy> From<T> for Vec2 {
+    fn from(v: T) -> Self {
+        Self {
+            x: v.to_f32(),
+            y: v.to_f32(),
+        }
+    }
+}
+
+const impl<T: const UiNum, U: const UiNum> From<(T, U)> for Vec2
+where
+    (T, U): const Destruct,
+{
+    fn from((x, y): (T, U)) -> Self {
+        Self {
+            x: x.to_f32(),
+            y: y.to_f32(),
+        }
+    }
+}

iris/core/src/orientation/align.rs

@@ -0,0 +1,200 @@
+use crate::vec2;
+
+use super::*;
+
+#[derive(Clone, Copy, PartialEq, Eq)]
+pub struct Align {
+    pub x: Option<AxisAlign>,
+    pub y: Option<AxisAlign>,
+}
+
+impl Align {
+    pub const TOP_LEFT: RegionAlign = RegionAlign::TOP_LEFT;
+    pub const TOP_CENTER: RegionAlign = RegionAlign::TOP_CENTER;
+    pub const TOP_RIGHT: RegionAlign = RegionAlign::TOP_RIGHT;
+    pub const CENTER_LEFT: RegionAlign = RegionAlign::CENTER_LEFT;
+    pub const CENTER: RegionAlign = RegionAlign::CENTER;
+    pub const CENTER_RIGHT: RegionAlign = RegionAlign::CENTER_RIGHT;
+    pub const BOT_LEFT: RegionAlign = RegionAlign::BOT_LEFT;
+    pub const BOT_CENTER: RegionAlign = RegionAlign::BOT_CENTER;
+    pub const BOT_RIGHT: RegionAlign = RegionAlign::BOT_RIGHT;
+    pub const LEFT: CardinalAlign = CardinalAlign::LEFT;
+    pub const H_CENTER: CardinalAlign = CardinalAlign::H_CENTER;
+    pub const RIGHT: CardinalAlign = CardinalAlign::RIGHT;
+    pub const TOP: CardinalAlign = CardinalAlign::TOP;
+    pub const V_CENTER: CardinalAlign = CardinalAlign::V_CENTER;
+    pub const BOT: CardinalAlign = CardinalAlign::BOT;
+
+    pub fn tuple(&self) -> (Option<AxisAlign>, Option<AxisAlign>) {
+        (self.x, self.y)
+    }
+}
+
+#[derive(Clone, Copy, PartialEq, Eq)]
+pub enum AxisAlign {
+    Neg,
+    Center,
+    Pos,
+}
+
+impl AxisAlign {
+    pub const fn rel(&self) -> f32 {
+        match self {
+            Self::Neg => 0.0,
+            Self::Center => 0.5,
+            Self::Pos => 1.0,
+        }
+    }
+}
+
+pub struct CardinalAlign {
+    axis: Axis,
+    align: AxisAlign,
+}
+
+impl CardinalAlign {
+    pub const LEFT: Self = Self::new(Axis::X, AxisAlign::Neg);
+    pub const H_CENTER: Self = Self::new(Axis::X, AxisAlign::Center);
+    pub const RIGHT: Self = Self::new(Axis::X, AxisAlign::Pos);
+    pub const TOP: Self = Self::new(Axis::Y, AxisAlign::Neg);
+    pub const V_CENTER: Self = Self::new(Axis::Y, AxisAlign::Center);
+    pub const BOT: Self = Self::new(Axis::Y, AxisAlign::Pos);
+
+    pub const fn new(axis: Axis, align: AxisAlign) -> Self {
+        Self { axis, align }
+    }
+}
+
+#[derive(Clone, Copy, PartialEq, Eq)]
+pub struct RegionAlign {
+    pub x: AxisAlign,
+    pub y: AxisAlign,
+}
+
+impl RegionAlign {
+    pub const TOP_LEFT: Self = Self::new(AxisAlign::Neg, AxisAlign::Neg);
+    pub const TOP_CENTER: Self = Self::new(AxisAlign::Center, AxisAlign::Neg);
+    pub const TOP_RIGHT: Self = Self::new(AxisAlign::Pos, AxisAlign::Neg);
+    pub const CENTER_LEFT: Self = Self::new(AxisAlign::Neg, AxisAlign::Center);
+    pub const CENTER: Self = Self::new(AxisAlign::Center, AxisAlign::Center);
+    pub const CENTER_RIGHT: Self = Self::new(AxisAlign::Pos, AxisAlign::Center);
+    pub const BOT_LEFT: Self = Self::new(AxisAlign::Neg, AxisAlign::Pos);
+    pub const BOT_CENTER: Self = Self::new(AxisAlign::Center, AxisAlign::Pos);
+    pub const BOT_RIGHT: Self = Self::new(AxisAlign::Pos, AxisAlign::Pos);
+
+    pub const fn new(x: AxisAlign, y: AxisAlign) -> Self {
+        Self { x, y }
+    }
+    pub const fn rel(&self) -> Vec2 {
+        vec2(self.x.rel(), self.y.rel())
+    }
+}
+
+impl UiVec2 {
+    pub fn partial_align(&self, align: Align) -> UiRegion {
+        UiRegion {
+            x: if let Some(align) = align.x {
+                self.x.align(align)
+            } else {
+                UiSpan::FULL
+            },
+            y: if let Some(align) = align.y {
+                self.y.align(align)
+            } else {
+                UiSpan::FULL
+            },
+        }
+    }
+
+    pub fn align(&self, align: RegionAlign) -> UiRegion {
+        UiRegion {
+            x: self.x.align(align.x),
+            y: self.y.align(align.y),
+        }
+    }
+}
+
+impl Vec2 {
+    pub fn partial_align(&self, align: Align) -> UiRegion {
+        let s = UiVec2::from(*self);
+        UiRegion {
+            x: if let Some(align) = align.x {
+                s.x.align(align)
+            } else {
+                UiSpan::FULL
+            },
+            y: if let Some(align) = align.y {
+                s.y.align(align)
+            } else {
+                UiSpan::FULL
+            },
+        }
+    }
+
+    pub fn align(&self, align: RegionAlign) -> UiRegion {
+        let s = UiVec2::from(*self);
+        UiRegion {
+            x: s.x.align(align.x),
+            y: s.y.align(align.y),
+        }
+    }
+}
+
+impl UiScalar {
+    pub const fn align(&self, align: AxisAlign) -> UiSpan {
+        let rel = align.rel();
+        let mut start = UiScalar::rel(rel);
+        start.abs -= self.abs * rel;
+        start.rel -= self.rel * rel;
+        let mut end = UiScalar::rel(rel);
+        end.abs += self.abs * (1.0 - rel);
+        end.rel += self.rel * (1.0 - rel);
+        UiSpan { start, end }
+    }
+}
+
+impl From<RegionAlign> for Align {
+    fn from(region: RegionAlign) -> Self {
+        Self {
+            x: Some(region.x),
+            y: Some(region.y),
+        }
+    }
+}
+
+impl From<Align> for RegionAlign {
+    fn from(align: Align) -> Self {
+        Self {
+            x: align.x.unwrap_or(AxisAlign::Center),
+            y: align.y.unwrap_or(AxisAlign::Center),
+        }
+    }
+}
+
+impl From<CardinalAlign> for RegionAlign {
+    fn from(align: CardinalAlign) -> Self {
+        Align::from(align).into()
+    }
+}
+
+impl From<CardinalAlign> for Align {
+    fn from(cardinal: CardinalAlign) -> Self {
+        let align = Some(cardinal.align);
+        match cardinal.axis {
+            Axis::X => Self { x: align, y: None },
+            Axis::Y => Self { x: None, y: align },
+        }
+    }
+}
+
+const impl From<RegionAlign> for UiVec2 {
+    fn from(align: RegionAlign) -> Self {
+        Self::rel(align.rel())
+    }
+}
+
+impl RegionAlign {
+    pub const fn pos(self) -> UiVec2 {
+        UiVec2::from(self)
+    }
+}

iris/core/src/orientation/axis.rs

@@ -0,0 +1,115 @@
+use super::*;
+
+#[derive(Copy, Clone, Eq, PartialEq)]
+pub enum Axis {
+    X,
+    Y,
+}
+
+impl std::ops::Not for Axis {
+    type Output = Self;
+
+    fn not(self) -> Self::Output {
+        match self {
+            Self::X => Self::Y,
+            Self::Y => Self::X,
+        }
+    }
+}
+
+#[derive(Clone, Copy, Eq, PartialEq)]
+pub struct Dir {
+    pub axis: Axis,
+    pub sign: Sign,
+}
+
+impl Dir {
+    pub const fn new(axis: Axis, dir: Sign) -> Self {
+        Self { axis, sign: dir }
+    }
+
+    pub const LEFT: Self = Self::new(Axis::X, Sign::Neg);
+    pub const RIGHT: Self = Self::new(Axis::X, Sign::Pos);
+    pub const UP: Self = Self::new(Axis::Y, Sign::Neg);
+    pub const DOWN: Self = Self::new(Axis::Y, Sign::Pos);
+}
+
+#[derive(Clone, Copy, Eq, PartialEq)]
+pub enum Sign {
+    Neg,
+    Pos,
+}
+
+impl Vec2 {
+    pub fn axis(&self, axis: Axis) -> f32 {
+        match axis {
+            Axis::X => self.x,
+            Axis::Y => self.y,
+        }
+    }
+
+    pub fn axis_mut(&mut self, axis: Axis) -> &mut f32 {
+        match axis {
+            Axis::X => &mut self.x,
+            Axis::Y => &mut self.y,
+        }
+    }
+
+    pub const fn from_axis(axis: Axis, aligned: f32, ortho: f32) -> Self {
+        Self {
+            x: match axis {
+                Axis::X => aligned,
+                Axis::Y => ortho,
+            },
+            y: match axis {
+                Axis::Y => aligned,
+                Axis::X => ortho,
+            },
+        }
+    }
+}
+
+pub const trait AxisT {
+    fn get() -> Axis;
+}
+
+pub struct XAxis;
+const impl AxisT for XAxis {
+    fn get() -> Axis {
+        Axis::X
+    }
+}
+
+pub struct YAxis;
+const impl AxisT for YAxis {
+    fn get() -> Axis {
+        Axis::Y
+    }
+}
+
+#[derive(Clone, Copy, Debug, Default)]
+pub struct BothAxis<T> {
+    pub x: T,
+    pub y: T,
+}
+
+impl<T> BothAxis<T> {
+    pub const fn axis<A: const AxisT>(&mut self) -> &mut T {
+        match A::get() {
+            Axis::X => &mut self.x,
+            Axis::Y => &mut self.y,
+        }
+    }
+    pub fn take_axis<A: const AxisT>(self) -> T {
+        match A::get() {
+            Axis::X => self.x,
+            Axis::Y => self.y,
+        }
+    }
+    pub fn axis_dyn(&mut self, axis: Axis) -> &mut T {
+        match axis {
+            Axis::X => &mut self.x,
+            Axis::Y => &mut self.y,
+        }
+    }
+}

iris/core/src/orientation/len.rs

@@ -0,0 +1,198 @@
+use super::*;
+use crate::{UiNum, util::impl_op};
+
+#[derive(Debug, Default, Clone, Copy, PartialEq)]
+pub struct Size {
+    pub x: Len,
+    pub y: Len,
+}
+
+#[derive(Debug, Clone, Copy, PartialEq)]
+pub struct Len {
+    pub abs: f32,
+    pub rel: f32,
+    pub rest: f32,
+}
+
+impl<N: UiNum> From<N> for Len {
+    fn from(value: N) -> Self {
+        Len::abs(value.to_f32())
+    }
+}
+
+impl<Nx: UiNum, Ny: UiNum> From<(Nx, Ny)> for Size {
+    fn from((x, y): (Nx, Ny)) -> Self {
+        Self {
+            x: x.into(),
+            y: y.into(),
+        }
+    }
+}
+
+impl From<Len> for Size {
+    fn from(value: Len) -> Self {
+        Self { x: value, y: value }
+    }
+}
+
+impl Size {
+    pub const ZERO: Self = Self {
+        x: Len::ZERO,
+        y: Len::ZERO,
+    };
+
+    pub const REST: Self = Self {
+        x: Len::REST,
+        y: Len::REST,
+    };
+
+    pub fn abs(v: Vec2) -> Self {
+        Self {
+            x: Len::abs(v.x),
+            y: Len::abs(v.y),
+        }
+    }
+
+    pub fn rel(v: Vec2) -> Self {
+        Self {
+            x: Len::rel(v.x),
+            y: Len::rel(v.y),
+        }
+    }
+
+    pub fn rest(v: Vec2) -> Self {
+        Self {
+            x: Len::rest(v.x),
+            y: Len::rest(v.y),
+        }
+    }
+
+    pub fn to_uivec2(self) -> UiVec2 {
+        UiVec2 {
+            x: self.x.apply_rest(),
+            y: self.y.apply_rest(),
+        }
+    }
+
+    pub fn from_axis(axis: Axis, aligned: Len, ortho: Len) -> Self {
+        match axis {
+            Axis::X => Self {
+                x: aligned,
+                y: ortho,
+            },
+            Axis::Y => Self {
+                x: ortho,
+                y: aligned,
+            },
+        }
+    }
+
+    pub fn axis(&self, axis: Axis) -> Len {
+        match axis {
+            Axis::X => self.x,
+            Axis::Y => self.y,
+        }
+    }
+}
+
+impl Len {
+    pub const ZERO: Self = Self {
+        abs: 0.0,
+        rel: 0.0,
+        rest: 0.0,
+    };
+
+    pub const REST: Self = Self {
+        abs: 0.0,
+        rel: 0.0,
+        rest: 1.0,
+    };
+
+    pub fn apply_rest(&self) -> UiScalar {
+        UiScalar {
+            rel: self.rel + if self.rest > 0.0 { 1.0 } else { 0.0 },
+            abs: self.abs,
+        }
+    }
+
+    pub fn abs(abs: impl UiNum) -> Self {
+        Self {
+            abs: abs.to_f32(),
+            rel: 0.0,
+            rest: 0.0,
+        }
+    }
+    pub fn rel(rel: impl UiNum) -> Self {
+        Self {
+            abs: 0.0,
+            rel: rel.to_f32(),
+            rest: 0.0,
+        }
+    }
+    pub fn rest(ratio: impl UiNum) -> Self {
+        Self {
+            abs: 0.0,
+            rel: 0.0,
+            rest: ratio.to_f32(),
+        }
+    }
+}
+
+pub mod len_fns {
+    use super::*;
+
+    pub fn abs(abs: impl UiNum) -> Len {
+        Len {
+            abs: abs.to_f32(),
+            rel: 0.0,
+            rest: 0.0,
+        }
+    }
+    pub fn rel(rel: impl UiNum) -> Len {
+        Len {
+            abs: 0.0,
+            rel: rel.to_f32(),
+            rest: 0.0,
+        }
+    }
+    pub fn rest(ratio: impl UiNum) -> Len {
+        Len {
+            abs: 0.0,
+            rel: 0.0,
+            rest: ratio.to_f32(),
+        }
+    }
+}
+
+impl_op!(Len Add add; abs rel rest);
+impl_op!(Len Sub sub; abs rel rest);
+
+impl_op!(Size Add add; x y);
+impl_op!(Size Sub sub; x y);
+
+impl Default for Len {
+    fn default() -> Self {
+        Self::rest(1.0)
+    }
+}
+
+impl std::fmt::Display for Size {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(f, "({}, {})", self.x, self.y)
+    }
+}
+
+impl std::fmt::Display for Len {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        if self.abs != 0.0 {
+            write!(f, "{} abs;", self.abs)?;
+        }
+        if self.rel != 0.0 {
+            write!(f, "{} rel;", self.rel)?;
+        }
+        if self.rest != 0.0 {
+            write!(f, "{} rest;", self.rest)?;
+        }
+        Ok(())
+    }
+}

iris/core/src/orientation/mod.rs

@@ -0,0 +1,11 @@
+mod align;
+mod axis;
+mod len;
+mod pos;
+
+use crate::util::Vec2;
+
+pub use align::*;
+pub use axis::*;
+pub use len::*;
+pub use pos::*;

iris/core/src/orientation/pos.rs

@@ -0,0 +1,464 @@
+use std::{fmt::Display, hash::Hash, marker::Destruct};
+
+use super::*;
+use crate::{
+    UiNum,
+    util::{LerpUtil, impl_op},
+};
+
+#[repr(C)]
+#[derive(Debug, Copy, Clone, PartialEq, Eq, Hash, bytemuck::Pod, bytemuck::Zeroable, Default)]
+pub struct UiVec2 {
+    pub x: UiScalar,
+    pub y: UiScalar,
+}
+
+impl UiVec2 {
+    pub const ZERO: Self = Self {
+        x: UiScalar::ZERO,
+        y: UiScalar::ZERO,
+    };
+
+    pub const fn new(x: UiScalar, y: UiScalar) -> Self {
+        Self { x, y }
+    }
+
+    pub const fn abs(abs: impl const Into<Vec2>) -> Self {
+        let abs = abs.into();
+        Self {
+            x: UiScalar::abs(abs.x),
+            y: UiScalar::abs(abs.y),
+        }
+    }
+
+    pub const fn rel(rel: impl const Into<Vec2>) -> Self {
+        let rel = rel.into();
+        Self {
+            x: UiScalar::rel(rel.x),
+            y: UiScalar::rel(rel.y),
+        }
+    }
+
+    pub const fn shift(&mut self, offset: impl const Into<UiVec2>) {
+        let offset = offset.into();
+        *self += offset;
+    }
+
+    pub const fn offset(mut self, offset: impl const Into<UiVec2>) -> Self {
+        self.shift(offset);
+        self
+    }
+
+    pub const fn within(&self, region: &UiRegion) -> UiVec2 {
+        UiVec2 {
+            x: self.x.within(&region.x),
+            y: self.y.within(&region.y),
+        }
+    }
+
+    pub const fn outside(&self, region: &UiRegion) -> UiVec2 {
+        UiVec2 {
+            x: self.x.outside(&region.x),
+            y: self.y.outside(&region.y),
+        }
+    }
+
+    pub fn axis_mut(&mut self, axis: Axis) -> &mut UiScalar {
+        match axis {
+            Axis::X => &mut self.x,
+            Axis::Y => &mut self.y,
+        }
+    }
+
+    pub fn axis(&self, axis: Axis) -> UiScalar {
+        match axis {
+            Axis::X => self.x,
+            Axis::Y => self.y,
+        }
+    }
+
+    pub fn to_abs(&self, rel: Vec2) -> Vec2 {
+        Vec2 {
+            x: self.x.to_abs(rel.x),
+            y: self.y.to_abs(rel.y),
+        }
+    }
+
+    pub const FULL_SIZE: Self = Self::rel(Vec2::ONE);
+
+    pub const fn from_axis(axis: Axis, aligned: UiScalar, ortho: UiScalar) -> Self {
+        match axis {
+            Axis::X => Self {
+                x: aligned,
+                y: ortho,
+            },
+            Axis::Y => Self {
+                x: ortho,
+                y: aligned,
+            },
+        }
+    }
+
+    pub fn get_abs(&self) -> Vec2 {
+        (self.x.abs, self.y.abs).into()
+    }
+
+    pub fn get_rel(&self) -> Vec2 {
+        (self.x.rel, self.y.rel).into()
+    }
+
+    pub fn abs_mut(&mut self) -> Vec2View<'_> {
+        Vec2View {
+            x: &mut self.x.abs,
+            y: &mut self.y.abs,
+        }
+    }
+}
+
+impl Display for UiVec2 {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(f, "rel{};abs{}", self.get_rel(), self.get_abs())
+    }
+}
+
+impl_op!(UiVec2 Add add; x y);
+impl_op!(UiVec2 Sub sub; x y);
+
+const impl From<Vec2> for UiVec2 {
+    fn from(abs: Vec2) -> Self {
+        Self::abs(abs)
+    }
+}
+
+const impl<T: const UiNum, U: const UiNum> From<(T, U)> for UiVec2
+where
+    (T, U): const Destruct,
+{
+    fn from(abs: (T, U)) -> Self {
+        Self::abs(abs)
+    }
+}
+
+#[repr(C)]
+#[derive(Debug, Copy, Clone, PartialEq, bytemuck::Pod, Default, bytemuck::Zeroable)]
+pub struct UiScalar {
+    pub rel: f32,
+    pub abs: f32,
+}
+
+impl Eq for UiScalar {}
+impl Hash for UiScalar {
+    fn hash<H: std::hash::Hasher>(&self, state: &mut H) {
+        state.write_u32(self.rel.to_bits());
+        state.write_u32(self.abs.to_bits());
+    }
+}
+
+impl_op!(UiScalar Add add; rel abs);
+impl_op!(UiScalar Sub sub; rel abs);
+
+impl UiScalar {
+    pub const ZERO: Self = Self { rel: 0.0, abs: 0.0 };
+    pub const FULL: Self = Self { rel: 1.0, abs: 0.0 };
+
+    pub const fn new(rel: f32, abs: f32) -> Self {
+        Self { rel, abs }
+    }
+
+    pub const fn rel(rel: f32) -> Self {
+        Self { rel, abs: 0.0 }
+    }
+
+    pub const fn abs(abs: f32) -> Self {
+        Self { rel: 0.0, abs }
+    }
+
+    pub const fn rel_min() -> Self {
+        Self::new(0.0, 0.0)
+    }
+
+    pub const fn rel_max() -> Self {
+        Self::new(1.0, 0.0)
+    }
+
+    pub const fn max(&self, other: Self) -> Self {
+        Self {
+            rel: self.rel.max(other.rel),
+            abs: self.abs.max(other.abs),
+        }
+    }
+
+    pub const fn min(&self, other: Self) -> Self {
+        Self {
+            rel: self.rel.min(other.rel),
+            abs: self.abs.min(other.abs),
+        }
+    }
+
+    pub const fn offset(mut self, amt: f32) -> Self {
+        self.abs += amt;
+        self
+    }
+
+    pub const fn within(&self, span: &UiSpan) -> Self {
+        let anchor = self.rel.lerp(span.start.rel, span.end.rel);
+        let offset = self.abs + self.rel.lerp(span.start.abs, span.end.abs);
+        Self {
+            rel: anchor,
+            abs: offset,
+        }
+    }
+
+    pub const fn outside(&self, span: &UiSpan) -> Self {
+        let rel = self.rel.lerp_inv(span.start.rel, span.end.rel);
+        let abs = self.abs - rel.lerp(span.start.abs, span.end.abs);
+        Self { rel, abs }
+    }
+
+    pub fn within_len(&self, len: UiScalar) -> Self {
+        self.within(&UiSpan {
+            start: UiScalar::ZERO,
+            end: len,
+        })
+    }
+
+    pub fn select_len(&self, len: UiScalar) -> Self {
+        len.within_len(*self)
+    }
+
+    pub const fn flip(&mut self) {
+        self.rel = 1.0 - self.rel;
+        self.abs = -self.abs;
+    }
+
+    pub const fn to(&self, end: Self) -> UiSpan {
+        UiSpan { start: *self, end }
+    }
+
+    pub const fn to_abs(&self, rel: f32) -> f32 {
+        self.rel * rel + self.abs
+    }
+}
+
+#[repr(C)]
+#[derive(Debug, Copy, Clone, PartialEq, bytemuck::Pod, bytemuck::Zeroable)]
+pub struct UiSpan {
+    pub start: UiScalar,
+    pub end: UiScalar,
+}
+
+impl UiSpan {
+    pub const FULL: Self = Self {
+        start: UiScalar::ZERO,
+        end: UiScalar::FULL,
+    };
+
+    pub const fn rel(rel: f32) -> Self {
+        Self {
+            start: UiScalar::rel(rel),
+            end: UiScalar::rel(rel),
+        }
+    }
+
+    pub const fn new(start: UiScalar, end: UiScalar) -> Self {
+        Self { start, end }
+    }
+
+    pub const fn flip(&mut self) {
+        self.start.flip();
+        self.end.flip();
+        std::mem::swap(&mut self.start.rel, &mut self.end.rel);
+        std::mem::swap(&mut self.start.abs, &mut self.end.abs);
+    }
+
+    pub const fn shift(&mut self, offset: UiScalar) {
+        self.start += offset;
+        self.end += offset;
+    }
+
+    pub const fn within(&self, parent: &Self) -> Self {
+        Self {
+            start: self.start.within(parent),
+            end: self.end.within(parent),
+        }
+    }
+
+    pub const fn outside(&self, parent: &Self) -> Self {
+        Self {
+            start: self.start.outside(parent),
+            end: self.end.outside(parent),
+        }
+    }
+
+    pub const fn len(&self) -> UiScalar {
+        self.end - self.start
+    }
+}
+
+#[repr(C)]
+#[derive(Debug, Copy, Clone, PartialEq, bytemuck::Pod, bytemuck::Zeroable)]
+pub struct UiRegion {
+    pub x: UiSpan,
+    pub y: UiSpan,
+}
+
+impl UiRegion {
+    pub const FULL: Self = Self {
+        x: UiSpan::FULL,
+        y: UiSpan::FULL,
+    };
+
+    pub const fn new(x: UiSpan, y: UiSpan) -> Self {
+        Self { x, y }
+    }
+
+    pub const fn rel(rel: Vec2) -> Self {
+        Self {
+            x: UiSpan::rel(rel.x),
+            y: UiSpan::rel(rel.y),
+        }
+    }
+    pub const fn within(&self, parent: &Self) -> Self {
+        Self {
+            x: self.x.within(&parent.x),
+            y: self.y.within(&parent.y),
+        }
+    }
+    pub const fn outside(&self, parent: &Self) -> Self {
+        Self {
+            x: self.x.outside(&parent.x),
+            y: self.y.outside(&parent.y),
+        }
+    }
+
+    pub const fn axis(&mut self, axis: Axis) -> &UiSpan {
+        match axis {
+            Axis::X => &self.x,
+            Axis::Y => &self.y,
+        }
+    }
+
+    pub const fn axis_mut(&mut self, axis: Axis) -> &mut UiSpan {
+        match axis {
+            Axis::X => &mut self.x,
+            Axis::Y => &mut self.y,
+        }
+    }
+
+    pub const fn flip(&mut self, axis: Axis) {
+        match axis {
+            Axis::X => self.x.flip(),
+            Axis::Y => self.y.flip(),
+        }
+    }
+
+    pub fn shift(&mut self, offset: impl Into<UiVec2>) {
+        let offset = offset.into();
+        self.x.shift(offset.x);
+        self.y.shift(offset.y);
+    }
+
+    pub fn offset(mut self, offset: impl Into<UiVec2>) -> Self {
+        self.shift(offset);
+        self
+    }
+
+    pub fn to_px(&self, size: Vec2) -> PixelRegion {
+        PixelRegion {
+            top_left: self.top_left().get_rel() * size + self.top_left().get_abs(),
+            bot_right: self.bot_right().get_rel() * size + self.bot_right().get_abs(),
+        }
+    }
+
+    pub const fn center(&self) -> UiVec2 {
+        Align::CENTER.pos().within(self)
+    }
+
+    pub const fn size(&self) -> UiVec2 {
+        UiVec2 {
+            x: self.x.len(),
+            y: self.y.len(),
+        }
+    }
+
+    pub const fn top_left(&self) -> UiVec2 {
+        UiVec2 {
+            x: self.x.start,
+            y: self.y.start,
+        }
+    }
+
+    pub const fn bot_right(&self) -> UiVec2 {
+        UiVec2 {
+            x: self.x.end,
+            y: self.y.end,
+        }
+    }
+
+    pub const fn from_axis(axis: Axis, aligned: UiSpan, ortho: UiSpan) -> Self {
+        Self {
+            x: match axis {
+                Axis::X => aligned,
+                Axis::Y => ortho,
+            },
+            y: match axis {
+                Axis::X => ortho,
+                Axis::Y => aligned,
+            },
+        }
+    }
+}
+
+impl Display for UiRegion {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(
+            f,
+            "{} -> {} (size: {})",
+            self.top_left(),
+            self.bot_right(),
+            self.size()
+        )
+    }
+}
+
+#[derive(Debug)]
+pub struct PixelRegion {
+    pub top_left: Vec2,
+    pub bot_right: Vec2,
+}
+
+impl PixelRegion {
+    pub fn contains(&self, pos: Vec2) -> bool {
+        pos.x >= self.top_left.x
+            && pos.x <= self.bot_right.x
+            && pos.y >= self.top_left.y
+            && pos.y <= self.bot_right.y
+    }
+
+    pub fn size(&self) -> Vec2 {
+        self.bot_right - self.top_left
+    }
+}
+
+impl Display for PixelRegion {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(f, "{} -> {}", self.top_left, self.bot_right)
+    }
+}
+
+pub struct Vec2View<'a> {
+    pub x: &'a mut f32,
+    pub y: &'a mut f32,
+}
+
+impl Vec2View<'_> {
+    pub fn set(&mut self, other: Vec2) {
+        *self.x = other.x;
+        *self.y = other.y;
+    }
+
+    pub fn add(&mut self, other: Vec2) {
+        *self.x += other.x;
+        *self.y += other.y;
+    }
+}

iris/core/src/primitive/color.rs

@@ -0,0 +1,172 @@
+use std::marker::Destruct;
+
+/// stored in linear for sane manipulation
+#[repr(C)]
+#[derive(Clone, Copy, Hash, PartialEq, Eq, bytemuck::Zeroable, Debug)]
+pub struct Color<T> {
+    pub r: T,
+    pub g: T,
+    pub b: T,
+    pub a: T,
+}
+
+/// Required by parley's `Brush`, which every text style is generic over. Opaque
+/// black rather than transparent: a brush that was never set should be visible
+/// and obviously unstyled, not invisible.
+impl<T: ColorNum> Default for Color<T> {
+    fn default() -> Self {
+        Self::BLACK
+    }
+}
+
+impl<T: ColorNum> Color<T> {
+    pub const BLACK: Self = Self::rgb(T::MIN, T::MIN, T::MIN);
+    pub const WHITE: Self = Self::rgb(T::MAX, T::MAX, T::MAX);
+    pub const GRAY: Self = Self::rgb(T::MID, T::MID, T::MID);
+
+    pub const RED: Self = Self::rgb(T::MAX, T::MIN, T::MIN);
+    pub const ORANGE: Self = Self::rgb(T::MAX, T::MID, T::MIN);
+    pub const YELLOW: Self = Self::rgb(T::MAX, T::MAX, T::MIN);
+    pub const LIME: Self = Self::rgb(T::MID, T::MAX, T::MIN);
+    pub const GREEN: Self = Self::rgb(T::MIN, T::MAX, T::MIN);
+    pub const TURQUOISE: Self = Self::rgb(T::MIN, T::MAX, T::MID);
+    pub const CYAN: Self = Self::rgb(T::MIN, T::MAX, T::MAX);
+    pub const SKY: Self = Self::rgb(T::MIN, T::MID, T::MAX);
+    pub const BLUE: Self = Self::rgb(T::MIN, T::MIN, T::MAX);
+    pub const PURPLE: Self = Self::rgb(T::MID, T::MIN, T::MAX);
+    pub const MAGENTA: Self = Self::rgb(T::MAX, T::MIN, T::MAX);
+
+    pub const NONE: Self = Self::new(T::MIN, T::MIN, T::MIN, T::MIN);
+
+    pub const fn new(r: T, g: T, b: T, a: T) -> Self {
+        Self { r, g, b, a }
+    }
+    pub const fn rgb(r: T, g: T, b: T) -> Self {
+        Self { r, g, b, a: T::MAX }
+    }
+    pub fn alpha(mut self, a: T) -> Self {
+        self.a = a;
+        self
+    }
+
+    pub fn as_arr(self) -> [T; 4] {
+        [self.r, self.g, self.b, self.a]
+    }
+}
+
+pub const trait F32Conversion {
+    fn to(self) -> f32;
+    fn from(x: f32) -> Self;
+}
+
+pub trait ColorNum {
+    const MIN: Self;
+    const MID: Self;
+    const MAX: Self;
+}
+
+macro_rules! map_rgb {
+    ($x:ident,$self:ident, $e:tt) => {
+        #[allow(unused_braces)]
+        Self {
+            r: {
+                let $x = $self.r;
+                $e
+            },
+            g: {
+                let $x = $self.g;
+                $e
+            },
+            b: {
+                let $x = $self.b;
+                $e
+            },
+            a: $self.a,
+        }
+    };
+}
+
+impl<T: ColorNum + const F32Conversion> Color<T>
+where
+    Self: const Destruct,
+{
+    pub const fn mul_rgb(self, amt: impl const F32Conversion) -> Self {
+        let amt = amt.to();
+        map_rgb!(x, self, { T::from(x.to() * amt) })
+    }
+
+    pub const fn add_rgb(self, amt: impl const F32Conversion) -> Self {
+        let amt = amt.to();
+        map_rgb!(x, self, { T::from(x.to() + amt) })
+    }
+
+    pub const fn darker(self, amt: f32) -> Self {
+        self.mul_rgb(1.0 - amt)
+    }
+
+    pub const fn brighter(self, amt: f32) -> Self {
+        map_rgb!(x, self, {
+            let x = x.to();
+            T::from(x + (1.0 - x) * amt)
+        })
+    }
+
+    pub fn map_rgb(self, f: impl Fn(T) -> T) -> Self {
+        Self {
+            r: f(self.r),
+            g: f(self.g),
+            b: f(self.b),
+            a: self.a,
+        }
+    }
+
+    pub fn srgb(r: T, g: T, b: T) -> Self {
+        Self {
+            r: s_to_l(r),
+            g: s_to_l(g),
+            b: s_to_l(b),
+            a: T::MAX,
+        }
+    }
+}
+
+fn s_to_l<T: F32Conversion>(x: T) -> T {
+    let x = x.to();
+    T::from(if x <= 0.0405 {
+        x / 12.92
+    } else {
+        ((x + 0.055) / 1.055).powf(2.4)
+    })
+}
+
+impl ColorNum for u8 {
+    const MIN: Self = u8::MIN;
+    const MID: Self = u8::MAX / 2;
+    const MAX: Self = u8::MAX;
+}
+
+impl ColorNum for f32 {
+    const MIN: Self = 0.0;
+    const MID: Self = 0.5;
+    const MAX: Self = 1.0;
+}
+
+unsafe impl bytemuck::Pod for Color<u8> {}
+
+const impl F32Conversion for f32 {
+    fn to(self) -> f32 {
+        self
+    }
+    fn from(x: f32) -> Self {
+        x
+    }
+}
+
+const impl F32Conversion for u8 {
+    fn to(self) -> f32 {
+        self as f32 / 255.0
+    }
+    fn from(x: f32) -> Self {
+        (x * 255.0).clamp(0.0, 255.0) as Self
+    }
+}

iris/core/src/primitive/layer.rs

@@ -0,0 +1,288 @@
+use std::ops::{Index, IndexMut};
+
+use crate::{
+    UiRegion, WidgetId,
+    render::{MaskIdx, MoveIdx, Primitive, PrimitiveHandle, PrimitiveInst, Primitives},
+    util::to_mut,
+};
+
+pub type LayerId = usize;
+
+struct LayerNode<T> {
+    next: Ptr,
+    prev: Ptr,
+    child: Option<Child>,
+    depth: usize,
+    data: T,
+}
+
+#[derive(Clone, Copy, Debug)]
+enum Ptr {
+    /// continue on same level
+    Next(usize),
+    /// go back to parent
+    Parent(usize),
+    /// end
+    None,
+}
+
+/// TODO: currently this does not ever free layers
+/// is that realistically desired?
+pub struct Layers<T> {
+    vec: Vec<LayerNode<T>>,
+    /// index of last layer at top level (start at first = 0)
+    last: usize,
+}
+
+#[derive(Clone, Copy)]
+struct Child {
+    head: usize,
+    tail: usize,
+}
+
+pub type PrimitiveLayers = Layers<Primitives>;
+
+impl<T: Default> Layers<T> {
+    pub fn new() -> Layers<T> {
+        Self {
+            vec: vec![LayerNode::head()],
+            last: 0,
+        }
+    }
+
+    pub fn clear(&mut self) {
+        self.vec.clear();
+        self.vec.push(LayerNode::head());
+    }
+
+    fn push(&mut self, node: LayerNode<T>) -> LayerId {
+        let i = self.vec.len();
+        self.vec.push(node);
+        i
+    }
+
+    pub fn next(&mut self, i: LayerId) -> LayerId {
+        if let Ptr::Next(i) = self.vec[i].next {
+            return i;
+        }
+        let i_new = self.push(LayerNode::new(
+            T::default(),
+            self.vec[i].next,
+            Ptr::Next(i),
+            self.vec[i].depth,
+        ));
+        self.vec[i].next = Ptr::Next(i_new);
+        self.vec[i_new].prev = Ptr::Next(i);
+        match self.vec[i_new].next {
+            Ptr::Next(i) => self.vec[i].prev = Ptr::Next(i_new),
+            Ptr::Parent(i) => self.vec[i].child.as_mut().unwrap().tail = i_new,
+            Ptr::None => self.last = i_new,
+        }
+        i_new
+    }
+
+    pub fn child(&mut self, i: LayerId) -> LayerId {
+        if let Some(c) = self.vec[i].child {
+            return c.head;
+        }
+        let i_child = self.push(LayerNode::new(
+            T::default(),
+            Ptr::Parent(i),
+            Ptr::Parent(i),
+            self.vec[i].depth + 1,
+        ));
+        self.vec[i].child = Some(Child {
+            head: i_child,
+            tail: i_child,
+        });
+        i_child
+    }
+
+    pub fn iter_mut(&mut self) -> LayerIteratorMut<'_, T> {
+        LayerIteratorMut::new(&mut self.vec, self.last)
+    }
+
+    pub fn iter_orderless_mut(&mut self) -> impl Iterator<Item = (usize, &mut T)> {
+        self.vec.iter_mut().map(|n| &mut n.data).enumerate()
+    }
+
+    pub fn iter(&self) -> impl Iterator<Item = (LayerId, &T)> {
+        self.indices().map(|i| (i, &self.vec[i].data))
+    }
+
+    pub fn iter_depth(&self) -> impl Iterator<Item = ((LayerId, usize), &T)> {
+        self.indices()
+            .map(|i| ((i, self.vec[i].depth), &self.vec[i].data))
+    }
+
+    pub fn indices(&self) -> LayerIndexIterator<'_, T> {
+        LayerIndexIterator::new(&self.vec, self.last)
+    }
+}
+
+impl PrimitiveLayers {
+    pub fn write<P: Primitive>(
+        &mut self,
+        layer: LayerId,
+        info: PrimitiveInst<P>,
+    ) -> PrimitiveHandle {
+        self[layer].write(layer, info)
+    }
+
+    pub fn free(&mut self, h: &PrimitiveHandle) -> MaskIdx {
+        self[h.layer].free(h)
+    }
+
+    pub fn write_image(
+        &mut self,
+        layer: LayerId,
+        id: WidgetId,
+        texture_idx: u32,
+        region: UiRegion,
+        mask_idx: MaskIdx,
+        move_idx: MoveIdx,
+    ) -> PrimitiveHandle {
+        self[layer].write_image(layer, id, texture_idx, region, mask_idx, move_idx)
+    }
+}
+
+impl<T: Default> Default for Layers<T> {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+impl<T> Index<LayerId> for Layers<T> {
+    type Output = T;
+
+    fn index(&self, index: LayerId) -> &Self::Output {
+        &self.vec[index].data
+    }
+}
+
+impl<T> IndexMut<LayerId> for Layers<T> {
+    fn index_mut(&mut self, index: LayerId) -> &mut Self::Output {
+        &mut self.vec[index].data
+    }
+}
+
+impl<T: Default> LayerNode<T> {
+    pub fn new(data: T, next: Ptr, prev: Ptr, depth: usize) -> Self {
+        Self {
+            next,
+            prev,
+            child: None,
+            data,
+            depth,
+        }
+    }
+
+    pub fn head() -> Self {
+        Self::new(T::default(), Ptr::None, Ptr::None, 0)
+    }
+}
+
+pub struct LayerIteratorMut<'a, T> {
+    inner: LayerIndexIterator<'a, T>,
+}
+
+impl<'a, T> Iterator for LayerIteratorMut<'a, T> {
+    type Item = (usize, &'a mut T);
+
+    fn next(&mut self) -> Option<Self::Item> {
+        let i = self.inner.next()?;
+        // SAFETY: requires index iterator to work properly
+        let layer = unsafe { to_mut(&self.inner.vec[i].data) };
+        Some((i, layer))
+    }
+}
+
+impl<'a, T> DoubleEndedIterator for LayerIteratorMut<'a, T> {
+    fn next_back(&mut self) -> Option<Self::Item> {
+        let i = self.inner.next_back()?;
+        // SAFETY: requires index iterator to work properly
+        let layer = unsafe { to_mut(&self.inner.vec[i].data) };
+        Some((i, layer))
+    }
+}
+
+impl<'a, T> LayerIteratorMut<'a, T> {
+    fn new(vec: &'a mut Vec<LayerNode<T>>, last: usize) -> Self {
+        Self {
+            inner: LayerIndexIterator::new(vec, last),
+        }
+    }
+}
+
+pub struct LayerIndexIterator<'a, T> {
+    next: Option<usize>,
+    next_back: Option<usize>,
+    vec: &'a Vec<LayerNode<T>>,
+}
+
+impl<'a, T> Iterator for LayerIndexIterator<'a, T> {
+    type Item = usize;
+
+    fn next(&mut self) -> Option<Self::Item> {
+        let ret_i = self.next?;
+        let node = &self.vec[ret_i];
+        self.next = if let Some(c) = node.child {
+            Some(c.head)
+        } else if let Ptr::Next(i) = node.next {
+            Some(i)
+        } else if let Ptr::Parent(i) = node.next {
+            let mut node = &self.vec[i];
+            while let Ptr::Parent(i) = node.next {
+                node = &self.vec[i];
+            }
+            if let Ptr::Next(i) = node.next {
+                Some(i)
+            } else {
+                None
+            }
+        } else {
+            None
+        };
+        if self.next_back.unwrap() == ret_i {
+            self.next = None;
+            self.next_back = None;
+        }
+        Some(ret_i)
+    }
+}
+
+impl<'a, T> DoubleEndedIterator for LayerIndexIterator<'a, T> {
+    fn next_back(&mut self) -> Option<Self::Item> {
+        let ret_i = self.next_back?;
+        let node = &self.vec[ret_i];
+        self.next_back = if let Ptr::Next(mut i) = node.prev {
+            while let Some(c) = self.vec[i].child {
+                i = c.tail
+            }
+            Some(i)
+        } else if let Ptr::Parent(i) = node.prev {
+            Some(i)
+        } else {
+            None
+        };
+        if self.next.unwrap() == ret_i {
+            self.next = None;
+            self.next_back = None;
+        }
+        Some(ret_i)
+    }
+}
+
+impl<'a, T> LayerIndexIterator<'a, T> {
+    fn new(vec: &'a Vec<LayerNode<T>>, last: usize) -> Self {
+        let mut last = last;
+        while let Some(c) = vec[last].child {
+            last = c.tail;
+        }
+        Self {
+            next: Some(0),
+            next_back: Some(last),
+            vec,
+        }
+    }
+}

iris/core/src/primitive/mod.rs

@@ -0,0 +1,9 @@
+mod color;
+mod layer;
+mod text;
+mod texture;
+
+pub use color::*;
+pub use layer::*;
+pub use text::*;
+pub use texture::*;

iris/core/src/primitive/text.rs

@@ -0,0 +1,277 @@
+use crate::{Align, GlyphAtlas, GlyphKey, PlacedGlyph, RegionAlign, Textures, UiColor, util::Vec2};
+use parley::{
+    Alignment, AlignmentOptions, FontContext, FontFamily, FontFamilyName, GenericFamily, Layout,
+    LayoutContext, LineHeight, PositionedLayoutItem, StyleProperty,
+};
+use swash::{
+    FontRef,
+    scale::{Render, ScaleContext, Source, StrikeWith},
+    zeno::{Format, Vector},
+};
+
+/// Everything text needs that outlives one string: the font collection, the
+/// layout scratch space, the glyph rasteriser and the atlas they fill.
+pub struct TextData {
+    pub font_cx: FontContext,
+    pub layout_cx: LayoutContext<UiColor>,
+    scale_cx: ScaleContext,
+    pub atlas: GlyphAtlas,
+}
+
+impl Default for TextData {
+    fn default() -> Self {
+        Self {
+            font_cx: FontContext::new(),
+            layout_cx: LayoutContext::new(),
+            scale_cx: ScaleContext::new(),
+            atlas: GlyphAtlas::default(),
+        }
+    }
+}
+
+/// Which family to ask for. Kept as an owned name rather than parley's
+/// borrowed `FontFamily<'_>` so that a widget can hold one without a lifetime.
+#[derive(Clone, PartialEq)]
+pub enum Family {
+    SansSerif,
+    Serif,
+    Monospace,
+    Named(String),
+}
+
+impl Family {
+    fn family(&self) -> FontFamily<'_> {
+        let name = match self {
+            Self::SansSerif => FontFamilyName::Generic(GenericFamily::SansSerif),
+            Self::Serif => FontFamilyName::Generic(GenericFamily::Serif),
+            Self::Monospace => FontFamilyName::Generic(GenericFamily::Monospace),
+            Self::Named(name) => FontFamilyName::Named(name.as_str().into()),
+        };
+        FontFamily::Single(name)
+    }
+}
+
+#[derive(Clone, PartialEq)]
+pub struct TextAttrs {
+    pub color: UiColor,
+    pub font_size: f32,
+    pub line_height: f32,
+    pub family: Family,
+    pub wrap: bool,
+    /// inner alignment of text region (within where it's drawn)
+    pub align: RegionAlign,
+}
+
+pub const LINE_HEIGHT_MULT: f32 = 1.1;
+
+impl Default for TextAttrs {
+    fn default() -> Self {
+        let size = 16.0;
+        Self {
+            color: UiColor::WHITE,
+            font_size: size,
+            line_height: size * LINE_HEIGHT_MULT,
+            family: Family::SansSerif,
+            wrap: false,
+            align: Align::CENTER_LEFT,
+        }
+    }
+}
+
+/// A string together with its laid-out form.
+///
+/// The text and the layout live in one place because parley's `Layout` borrows
+/// nothing but is only meaningful against the string it was built from: keeping
+/// them apart is how they get out of step.
+pub struct TextBuffer {
+    text: String,
+    layout: Layout<UiColor>,
+    /// What the current layout was built for, so `shape` can decline to redo
+    /// work that would come out the same.
+    shaped: Option<(TextAttrs, Option<f32>)>,
+}
+
+impl TextBuffer {
+    pub fn new(text: impl Into<String>) -> Self {
+        Self {
+            text: text.into(),
+            layout: Layout::new(),
+            shaped: None,
+        }
+    }
+
+    pub fn new_empty() -> Self {
+        Self::new("")
+    }
+
+    pub fn text(&self) -> &str {
+        &self.text
+    }
+
+    pub fn layout(&self) -> &Layout<UiColor> {
+        &self.layout
+    }
+
+    pub fn is_empty(&self) -> bool {
+        self.text.is_empty()
+    }
+
+    pub fn set_text(&mut self, text: impl Into<String>) {
+        let text = text.into();
+        if text != self.text {
+            self.text = text;
+            self.shaped = None;
+        }
+    }
+
+    /// Edit the string in place; invalidates the layout unconditionally, since
+    /// the caller is assumed to have changed something.
+    pub fn edit(&mut self) -> &mut String {
+        self.shaped = None;
+        &mut self.text
+    }
+
+    pub fn size(&self) -> Vec2 {
+        Vec2::new(self.layout.width(), self.layout.height())
+    }
+
+    /// Lay the text out, unless it is already laid out for these attributes and
+    /// this width.
+    pub fn shape(&mut self, data: &mut TextData, attrs: &TextAttrs, width: Option<f32>) {
+        if self.shaped.as_ref() == Some(&(attrs.clone(), width)) {
+            return;
+        }
+        let mut builder = data
+            .layout_cx
+            .ranged_builder(&mut data.font_cx, &self.text, 1.0, true);
+        builder.push_default(StyleProperty::FontFamily(attrs.family.family()));
+        builder.push_default(StyleProperty::FontSize(attrs.font_size));
+        builder.push_default(StyleProperty::LineHeight(LineHeight::Absolute(
+            attrs.line_height,
+        )));
+        builder.push_default(StyleProperty::Brush(attrs.color));
+        builder.build_into(&mut self.layout, &self.text);
+        self.layout.break_all_lines(width);
+        self.layout
+            .align(Alignment::Start, AlignmentOptions::default());
+        self.shaped = Some((attrs.clone(), width));
+    }
+}
+
+impl TextData {
+    /// Rasterise whatever of `buffer` is not in the atlas yet, and return where
+    /// each glyph goes relative to the text's top-left.
+    ///
+    /// Nothing is uploaded for a glyph already in the atlas, which is the point
+    /// of having one: a resize re-runs this and touches the GPU only if the new
+    /// width brought genuinely new glyphs into view.
+    pub fn place(&mut self, buffer: &TextBuffer, textures: &mut Textures) -> Vec<PlacedGlyph> {
+        let mut placed = Vec::new();
+        for line in buffer.layout.lines() {
+            for item in line.items() {
+                let PositionedLayoutItem::GlyphRun(run) = item else {
+                    continue;
+                };
+                let font = run.run().font();
+                let font_size = run.run().font_size();
+                let coords = run.run().normalized_coords();
+                let Some(font_ref) = FontRef::from_index(font.data.as_ref(), font.index as usize)
+                else {
+                    continue;
+                };
+                let coords_hash = hash_coords(coords);
+                // `font.data.id()` rather than the pointer, so the same font
+                // loaded twice is still one set of entries.
+                let font_id = font.data.id();
+
+                for glyph in run.positioned_glyphs() {
+                    let subpixel = ((glyph.x.fract() * 4.0).round() as i32).rem_euclid(4) as u8;
+                    let key = GlyphKey {
+                        font: font_id,
+                        glyph: glyph.id,
+                        size: (font_size * 16.0).round() as u32,
+                        subpixel,
+                        coords: coords_hash,
+                    };
+                    let entry = match self.atlas.get(&key) {
+                        Some(entry) => entry,
+                        None => {
+                            let mut scaler = self
+                                .scale_cx
+                                .builder(font_ref)
+                                .size(font_size)
+                                .hint(true)
+                                .normalized_coords(coords)
+                                .build();
+                            let image = Render::new(&[
+                                Source::ColorOutline(0),
+                                Source::ColorBitmap(StrikeWith::BestFit),
+                                Source::Outline,
+                            ])
+                            .format(Format::Alpha)
+                            .offset(Vector::new(subpixel as f32 / 4.0, 0.0))
+                            .render(&mut scaler, glyph.id as u16);
+                            match image {
+                                Some(image) => self.atlas.insert(key, &image, textures),
+                                None => {
+                                    self.atlas.insert_empty(key);
+                                    None
+                                }
+                            }
+                        }
+                    };
+                    let Some(entry) = entry else { continue };
+                    placed.push(PlacedGlyph {
+                        entry,
+                        offset: Vec2::new(
+                            glyph.x.floor() + entry.left as f32,
+                            glyph.y.floor() - entry.top as f32,
+                        ),
+                    });
+                }
+            }
+        }
+        placed
+    }
+}
+
+fn hash_coords(coords: &[i16]) -> u64 {
+    // FxHash over the coordinates; they are short and change rarely.
+    let mut h: u64 = 0xcbf2_9ce4_8422_2325;
+    for c in coords {
+        h ^= *c as u16 as u64;
+        h = h.wrapping_mul(0x1000_0000_01b3);
+    }
+    h
+}
+
+/// A laid-out string, ready to draw: where each glyph goes, how big the whole
+/// thing is, and what colour to tint the atlas with.
+///
+/// Cheap to clone and to keep, which is the point -- a widget holds one across
+/// frames and re-emits its quads without going near the rasteriser.
+#[derive(Clone)]
+pub struct RenderedText {
+    pub glyphs: std::sync::Arc<Vec<PlacedGlyph>>,
+    pub size: Vec2,
+    pub color: UiColor,
+}
+
+impl TextData {
+    /// Lay out and place in one step, which is what a widget wants.
+    pub fn render(
+        &mut self,
+        buffer: &mut TextBuffer,
+        attrs: &TextAttrs,
+        width: Option<f32>,
+        textures: &mut Textures,
+    ) -> RenderedText {
+        buffer.shape(self, attrs, width);
+        let glyphs = self.place(buffer, textures);
+        RenderedText {
+            glyphs: std::sync::Arc::new(glyphs),
+            size: buffer.size(),
+            color: attrs.color,
+        }
+    }
+}

iris/core/src/primitive/texture.rs

@@ -0,0 +1,226 @@
+use crate::util::{RefCounter, Vec2};
+use image::{DynamicImage, GenericImageView};
+use std::{
+    ops::Index,
+    sync::mpsc::{Receiver, Sender, channel},
+};
+
+/// Which of the two things a texture slot holds. See TEXTURES.md's
+/// "Recommended shape" for why these are drawn so differently: a page is a
+/// layer of one shared array texture and never gets its own bind group; a
+/// standalone image is the opposite, one texture and one bind group, never a
+/// layer.
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub enum TextureKind {
+    Image,
+    /// The array-texture layer this page was assigned. Chosen synchronously
+    /// by `Textures::add_page` rather than by the renderer, because glyph
+    /// insertion needs it in the same call, before any GPU sync happens.
+    Page {
+        layer: u32,
+    },
+}
+
+#[derive(Debug, Clone)]
+pub struct TextureHandle {
+    slot: u32,
+    kind: TextureKind,
+    size: Vec2,
+    counter: RefCounter,
+    send: Sender<(TextureKind, u32)>,
+}
+
+/// a texture manager for a ui
+/// note that this is heavily oriented towards wgpu's renderer so the primitives don't need mapped
+pub struct Textures {
+    free: Vec<u32>,
+    images: Vec<Option<DynamicImage>>,
+    /// Next layer to hand out to an atlas page. Pages are never freed (no
+    /// atlas eviction), so this only grows and `free` never holds one.
+    next_page_layer: u32,
+    updates: Vec<Update>,
+    send: Sender<(TextureKind, u32)>,
+    recv: Receiver<(TextureKind, u32)>,
+}
+
+pub enum TextureUpdate<'a> {
+    Push(TextureKind, &'a DynamicImage),
+    Set(TextureKind, u32, &'a DynamicImage),
+    /// Overwrite a rectangle of an existing texture, rather than replacing it.
+    /// The glyph atlas grows a glyph at a time, and re-uploading a whole atlas
+    /// per glyph is megabytes of copy for a few hundred bytes of change.
+    /// Only ever issued against a page -- a standalone image is never patched.
+    Patch(u32, PatchRect, &'a DynamicImage),
+    Free(u32),
+    PushFree(TextureKind),
+    SetFree,
+}
+
+#[derive(Debug, Clone, Copy)]
+pub struct PatchRect {
+    pub x: u32,
+    pub y: u32,
+    pub width: u32,
+    pub height: u32,
+}
+
+enum Update {
+    Push(TextureKind, u32),
+    Set(TextureKind, u32),
+    Patch(u32, PatchRect),
+    Free(u32),
+}
+
+impl Textures {
+    pub fn new() -> Self {
+        let (send, recv) = channel();
+        Self {
+            free: Vec::new(),
+            images: Vec::new(),
+            next_page_layer: 0,
+            updates: Vec::new(),
+            send,
+            recv,
+        }
+    }
+
+    pub fn add(&mut self, image: impl Into<DynamicImage>) -> TextureHandle {
+        let image = image.into();
+        let size = image.dimensions().into();
+        let kind = TextureKind::Image;
+        let slot = self.push(kind, image);
+        TextureHandle {
+            slot,
+            kind,
+            size,
+            counter: RefCounter::new(),
+            send: self.send.clone(),
+        }
+    }
+
+    /// Adds a page of the shared glyph atlas array. Only `atlas.rs` should
+    /// call this -- everything else wants `add`.
+    pub fn add_page(&mut self, image: impl Into<DynamicImage>) -> TextureHandle {
+        let image = image.into();
+        let size = image.dimensions().into();
+        let layer = self.next_page_layer;
+        self.next_page_layer += 1;
+        let kind = TextureKind::Page { layer };
+        let slot = self.push(kind, image);
+        TextureHandle {
+            slot,
+            kind,
+            size,
+            counter: RefCounter::new(),
+            send: self.send.clone(),
+        }
+    }
+
+    fn push(&mut self, kind: TextureKind, image: DynamicImage) -> u32 {
+        if let Some(i) = self.free.pop() {
+            self.images[i as usize] = Some(image);
+            self.updates.push(Update::Set(kind, i));
+            i
+        } else {
+            let i = self.images.len() as u32;
+            self.images.push(Some(image));
+            self.updates.push(Update::Push(kind, i));
+            i
+        }
+    }
+
+    /// The stored image for a handle, to be written into before `patch`.
+    pub fn image_mut(&mut self, handle: &TextureHandle) -> &mut DynamicImage {
+        self.images[handle.slot as usize]
+            .as_mut()
+            .expect("texture was freed while still held")
+    }
+
+    /// Queue an upload of just `rect`, after writing it with `image_mut`.
+    pub fn patch(&mut self, handle: &TextureHandle, rect: PatchRect) {
+        self.updates.push(Update::Patch(handle.slot, rect));
+    }
+
+    pub fn free(&mut self) {
+        for (kind, idx) in self.recv.try_iter() {
+            self.images[idx as usize] = None;
+            self.updates.push(Update::Free(idx));
+            // A page's slot is never reclaimed: `GlyphAtlas` never drops the
+            // handles it holds, and there is no eviction path for a hole in
+            // the middle of the array's layers. If that ever changes, this
+            // is where a freed page's layer would need to go on a free list
+            // of its own, separate from `free`, which only ever holds
+            // ordinary image slots today.
+            if kind == TextureKind::Image {
+                self.free.push(idx);
+            }
+        }
+    }
+
+    pub fn updates(&mut self) -> impl Iterator<Item = TextureUpdate<'_>> {
+        self.updates.drain(..).map(|u| match u {
+            Update::Push(kind, i) => self.images[i as usize]
+                .as_ref()
+                .map(|img| TextureUpdate::Push(kind, img))
+                .unwrap_or(TextureUpdate::PushFree(kind)),
+            Update::Set(kind, i) => self.images[i as usize]
+                .as_ref()
+                .map(|img| TextureUpdate::Set(kind, i, img))
+                .unwrap_or(TextureUpdate::SetFree),
+            Update::Patch(i, rect) => self.images[i as usize]
+                .as_ref()
+                .map(|img| TextureUpdate::Patch(i, rect, img))
+                .unwrap_or(TextureUpdate::SetFree),
+            Update::Free(i) => TextureUpdate::Free(i),
+        })
+    }
+}
+
+impl TextureHandle {
+    pub fn size(&self) -> Vec2 {
+        self.size
+    }
+
+    /// The bind-group index this handle draws with. Only valid for a
+    /// standalone image; an atlas page has no bind group of its own -- it
+    /// samples the shared array via `layer()` instead. Getting this wrong is
+    /// a caller bug (the wrong kind of handle reached the wrong draw path),
+    /// not a recoverable condition, so it panics rather than drawing garbage.
+    pub fn image_index(&self) -> u32 {
+        match self.kind {
+            TextureKind::Image => self.slot,
+            TextureKind::Page { .. } => panic!("image_index() called on an atlas page handle"),
+        }
+    }
+
+    /// The layer this page occupies in the shared atlas array texture.
+    /// Only valid for a page handle; see `image_index`'s note.
+    pub fn layer(&self) -> u32 {
+        match self.kind {
+            TextureKind::Page { layer } => layer,
+            TextureKind::Image => panic!("layer() called on a standalone image handle"),
+        }
+    }
+}
+
+impl Drop for TextureHandle {
+    fn drop(&mut self) {
+        if self.counter.drop() {
+            let _ = self.send.send((self.kind, self.slot));
+        }
+    }
+}
+
+impl Index<&TextureHandle> for Textures {
+    type Output = DynamicImage;
+
+    fn index(&self, index: &TextureHandle) -> &Self::Output {
+        self.images[index.slot as usize].as_ref().unwrap()
+    }
+}
+
+impl Default for Textures {
+    fn default() -> Self {
+        Self::new()
+    }
+}

iris/core/src/render/atlas.rs

@@ -0,0 +1,235 @@
+//! A glyph atlas: one texture holding many rasterised glyphs, so drawing text
+//! is a quad per glyph rather than a texture per string.
+//!
+//! What this replaces is why it exists. Text used to be rasterised into its own
+//! `RgbaImage` and uploaded as a whole texture, per text widget, every time
+//! anything about it changed -- so every window resize re-rasterised and
+//! re-uploaded every visible string, which is what the TODO meant by "resizing
+//! (per frame) is really slow". Here a glyph is rasterised once for a given
+//! font, size and subpixel offset and then reused by every string that contains
+//! it, and a resize re-emits quads without touching the GPU's copy at all.
+
+use crate::{
+    PatchRect, TextureHandle, Textures,
+    util::{HashMap, Vec2},
+};
+use image::RgbaImage;
+use swash::scale::image::{Content, Image};
+
+/// Side of one atlas page, in pixels. 1024 is 4 MB at RGBA8 -- enough for a
+/// few thousand glyphs at UI sizes, and small enough that a page nobody fills
+/// is not a big waste. Also the fixed width/height of every layer of the
+/// shared array texture in `render::texture` -- `pub(crate)` so that module
+/// can size it without a second constant to keep in sync.
+pub(crate) const PAGE: u32 = 1024;
+
+/// Transparent margin kept around every glyph, so that sampling one cannot
+/// pick up its neighbour along a shared edge.
+const PAD: u32 = 1;
+
+/// Identifies a rasterised glyph. Anything that changes the pixels has to be in
+/// here, or two different glyphs share one entry and the wrong one is drawn.
+#[derive(Clone, Copy, PartialEq, Eq, Hash)]
+pub struct GlyphKey {
+    pub font: u64,
+    pub glyph: u32,
+    /// Font size in 1/16 px, so sizes that round to the same pixels share a
+    /// raster instead of filling the atlas with near-duplicates.
+    pub size: u32,
+    /// Horizontal subpixel phase, in 1/4 px.
+    pub subpixel: u8,
+    /// Hash of the variation coordinates; a variable font at two weights is two
+    /// different sets of pixels from one glyph id.
+    pub coords: u64,
+}
+
+#[derive(Clone, Copy)]
+pub struct GlyphEntry {
+    pub uv_min: [f32; 2],
+    pub uv_max: [f32; 2],
+    /// Offset from the glyph's pen position to the top-left of its pixels.
+    pub left: i32,
+    pub top: i32,
+    pub width: u32,
+    pub height: u32,
+    pub is_color: bool,
+    /// The atlas array layer this glyph's page occupies.
+    pub layer: u32,
+}
+
+struct Page {
+    handle: TextureHandle,
+    /// Shelf packing: glyphs are placed left to right along a shelf whose
+    /// height is the tallest glyph on it, and a new shelf starts above when the
+    /// row runs out. Chosen over a real packer because glyphs at one size are
+    /// close to the same height, which is the case shelves are good at.
+    x: u32,
+    y: u32,
+    shelf_height: u32,
+}
+
+#[derive(Default)]
+pub struct GlyphAtlas {
+    pages: Vec<Page>,
+    /// `None` for a glyph that rasterised to nothing -- a space, say. Cached
+    /// too, so it is not re-rasterised on every layout.
+    entries: HashMap<GlyphKey, Option<GlyphEntry>>,
+}
+
+impl GlyphAtlas {
+    pub fn get(&self, key: &GlyphKey) -> Option<Option<GlyphEntry>> {
+        self.entries.get(key).copied()
+    }
+
+    /// Rasterised pixels in, a place in the atlas out. `None` means the glyph
+    /// has no pixels, which is a normal answer rather than a failure.
+    pub fn insert(
+        &mut self,
+        key: GlyphKey,
+        image: &Image,
+        textures: &mut Textures,
+    ) -> Option<GlyphEntry> {
+        let w = image.placement.width;
+        let h = image.placement.height;
+        if w == 0 || h == 0 {
+            self.entries.insert(key, None);
+            return None;
+        }
+        if w + PAD * 2 > PAGE || h + PAD * 2 > PAGE {
+            // A single glyph larger than a page. Refusing is better than
+            // silently drawing a cropped one; the caller draws nothing.
+            self.entries.insert(key, None);
+            return None;
+        }
+
+        let (page_idx, x, y) = self.allocate(w, h, textures);
+        let page = &self.pages[page_idx];
+
+        let img = textures.image_mut(&page.handle);
+        let rgba = img.as_mut_rgba8().expect("atlas page is rgba8");
+        write_glyph(rgba, image, x, y);
+
+        let handle = page.handle.clone();
+        let rect = PatchRect {
+            x,
+            y,
+            width: w,
+            height: h,
+        };
+        textures.patch(&handle, rect);
+
+        let page = &self.pages[page_idx];
+        let scale = 1.0 / PAGE as f32;
+        let entry = GlyphEntry {
+            uv_min: [x as f32 * scale, y as f32 * scale],
+            uv_max: [(x + w) as f32 * scale, (y + h) as f32 * scale],
+            left: image.placement.left,
+            top: image.placement.top,
+            width: w,
+            height: h,
+            is_color: matches!(image.content, Content::Color),
+            layer: page.handle.layer(),
+        };
+        self.entries.insert(key, Some(entry));
+        Some(entry)
+    }
+
+    /// A free `w`x`h` spot, opening a shelf or a page as needed.
+    fn allocate(&mut self, w: u32, h: u32, textures: &mut Textures) -> (usize, u32, u32) {
+        let need_w = w + PAD;
+        let need_h = h + PAD;
+        if let Some(i) = self.pages.iter().position(|p| fits(p, need_w, need_h)) {
+            let page = &mut self.pages[i];
+            if page.x + need_w > PAGE {
+                page.y += page.shelf_height;
+                page.x = PAD;
+                page.shelf_height = 0;
+            }
+            let (x, y) = (page.x, page.y);
+            page.x += need_w;
+            page.shelf_height = page.shelf_height.max(need_h);
+            return (i, x, y);
+        }
+
+        let handle = textures.add_page(RgbaImage::new(PAGE, PAGE));
+        self.pages.push(Page {
+            handle,
+            x: PAD + w + PAD,
+            y: PAD,
+            shelf_height: h + PAD,
+        });
+        (self.pages.len() - 1, PAD, PAD)
+    }
+
+    /// Record that a glyph has no pixels, so it is not re-rasterised.
+    pub fn insert_empty(&mut self, key: GlyphKey) {
+        self.entries.insert(key, None);
+    }
+
+    pub fn page_count(&self) -> usize {
+        self.pages.len()
+    }
+
+    pub fn glyph_count(&self) -> usize {
+        self.entries.len()
+    }
+}
+
+fn fits(page: &Page, need_w: u32, need_h: u32) -> bool {
+    // On the current shelf, or on a new one above it.
+    (page.x + need_w <= PAGE && page.y + need_h <= PAGE)
+        || (need_w + PAD <= PAGE && page.y + page.shelf_height + need_h <= PAGE)
+}
+
+/// Copy one rasterised glyph into the page image at `(x, y)`.
+///
+/// A mask glyph keeps its coverage in alpha with the colour left to the shader,
+/// so one raster serves text of any colour; a colour glyph carries its own.
+fn write_glyph(page: &mut RgbaImage, image: &Image, x: u32, y: u32) {
+    let w = image.placement.width;
+    let h = image.placement.height;
+    match image.content {
+        Content::Mask => {
+            for row in 0..h {
+                for col in 0..w {
+                    let a = image.data[(row * w + col) as usize];
+                    page.put_pixel(x + col, y + row, image::Rgba([255, 255, 255, a]));
+                }
+            }
+        }
+        Content::Color => {
+            for row in 0..h {
+                for col in 0..w {
+                    let i = ((row * w + col) * 4) as usize;
+                    let px = [
+                        image.data[i],
+                        image.data[i + 1],
+                        image.data[i + 2],
+                        image.data[i + 3],
+                    ];
+                    page.put_pixel(x + col, y + row, image::Rgba(px));
+                }
+            }
+        }
+        Content::SubpixelMask => {
+            // Not asked for: `Format::Alpha` is what the renderer requests, so
+            // reaching here means the request changed and this needs writing.
+            // Drawn as a plain mask from the green channel rather than dropped,
+            // so the text is readable rather than absent.
+            for row in 0..h {
+                for col in 0..w {
+                    let i = ((row * w + col) * 4) as usize;
+                    let a = image.data[i + 1];
+                    page.put_pixel(x + col, y + row, image::Rgba([255, 255, 255, a]));
+                }
+            }
+        }
+    }
+}
+
+/// Where a glyph goes on screen, in pixels relative to the text's origin.
+#[derive(Clone, Copy)]
+pub struct PlacedGlyph {
+    pub entry: GlyphEntry,
+    pub offset: Vec2,
+}

iris/core/src/render/data.rs

@@ -0,0 +1,92 @@
+use crate::{UiRegion, util::Id};
+use wgpu::*;
+
+#[repr(C)]
+#[derive(Copy, Clone, bytemuck::Pod, bytemuck::Zeroable, Default)]
+pub struct WindowUniform {
+    pub width: f32,
+    pub height: f32,
+}
+
+#[repr(C)]
+#[derive(Copy, Clone, bytemuck::Pod, bytemuck::Zeroable)]
+pub struct PrimitiveInstance {
+    pub region: UiRegion,
+    pub binding: u32,
+    pub idx: u32,
+    pub mask_idx: MaskIdx,
+    pub move_idx: MoveIdx,
+}
+
+impl PrimitiveInstance {
+    const ATTRIBS: [VertexAttribute; 8] = vertex_attr_array![
+        0 => Float32x2,
+        1 => Float32x2,
+        2 => Float32x2,
+        3 => Float32x2,
+        4 => Uint32,
+        5 => Uint32,
+        6 => Uint32,
+        7 => Uint32,
+    ];
+
+    pub fn desc() -> VertexBufferLayout<'static> {
+        VertexBufferLayout {
+            array_stride: std::mem::size_of::<Self>() as BufferAddress,
+            step_mode: VertexStepMode::Instance,
+            attributes: &Self::ATTRIBS,
+        }
+    }
+}
+
+pub type MaskIdx = Id<u32>;
+
+impl MaskIdx {
+    pub const NONE: Self = Self::preset(u32::MAX);
+}
+
+pub type MoveIdx = Id<u32>;
+
+#[repr(C)]
+#[derive(Debug, Copy, Clone, bytemuck::Pod, bytemuck::Zeroable)]
+pub struct Mask {
+    pub region: UiRegion,
+    /// The mask-owning widget's own move slot -- resolved in the fragment
+    /// shader against the same chain the vertex shader walks for a
+    /// primitive's own corners, so a mask and the content clipped by it
+    /// can move independently. See LAYOUT.md section 2b.
+    pub move_idx: MoveIdx,
+}
+
+/// One widget's cumulative on-screen translation, and the slot of the
+/// ancestor to add on top of it. `parent == u32::MAX` ends the chain. A
+/// pure abs-pixel delta, not a general `UiRegion` remap -- sufficient for
+/// every call site that moves a widget (`Scroll`, `Offset`) since both are
+/// translations of an already-drawn subtree. See LAYOUT.md section 2.
+///
+/// `_pad` matches WGSL's storage-buffer layout for `MoveOffset`: `delta` is
+/// a `vec2<f32>`, which gives the struct an 8-byte alignment and rounds its
+/// WGSL size up to 16 bytes even though `delta` + `parent` only total 12 --
+/// the same trap `GlyphPrimitive` documents below. `bytemuck` does not
+/// check this for us, and getting it wrong is a wgpu validation panic at
+/// draw time ("buffer bound ... with size 12 where the shader expects 16"),
+/// not a compile error.
+#[repr(C)]
+#[derive(Debug, Copy, Clone, bytemuck::Pod, bytemuck::Zeroable)]
+pub struct MoveOffset {
+    pub delta: [f32; 2],
+    pub parent: u32,
+    _pad: u32,
+}
+
+impl MoveOffset {
+    pub const NONE_PARENT: u32 = u32::MAX;
+
+    pub fn new(delta: [f32; 2], parent: u32) -> Self {
+        Self {
+            delta,
+            parent,
+            _pad: 0,
+        }
+    }
+}

iris/core/src/render/mod.rs

@@ -0,0 +1,444 @@
+use crate::{
+    UiData, UiRenderState,
+    render::{data::PrimitiveInstance, texture::GpuTextures, util::ArrBuf},
+    util::{HashMap, Vec2},
+};
+use data::WindowUniform;
+use wgpu::{
+    util::{BufferInitDescriptor, DeviceExt},
+    *,
+};
+
+mod atlas;
+mod data;
+mod primitive;
+mod texture;
+mod util;
+
+pub use atlas::*;
+pub use data::{Mask, MaskIdx, MoveIdx, MoveOffset};
+pub use primitive::*;
+
+const SHAPE_SHADER: &str = include_str!("./shader.wgsl");
+
+pub struct UiRenderNode {
+    uniform_group: BindGroup,
+    primitive_layout: BindGroupLayout,
+    rsc_layout: BindGroupLayout,
+    rsc_group: BindGroup,
+
+    pipeline: RenderPipeline,
+
+    layers: HashMap<usize, RenderLayer>,
+    active: Vec<usize>,
+    window_buffer: Buffer,
+    textures: GpuTextures,
+    masks: ArrBuf<Mask>,
+    move_offsets: ArrBuf<MoveOffset>,
+}
+
+struct RenderLayer {
+    instance: ArrBuf<PrimitiveInstance>,
+    primitives: PrimitiveBuffers,
+    primitive_group: BindGroup,
+    /// A standalone image's instances, kept apart from `instance` because
+    /// each one draws with its own bind group -- see `UiRenderNode::draw`.
+    image_instance: ArrBuf<PrimitiveInstance>,
+    /// The texture slot each entry of `image_instance` draws with, in the
+    /// same order, refreshed alongside it. Not stored in the vertex buffer
+    /// itself because it names a bind group, not shader data.
+    image_tex_indices: Vec<u32>,
+}
+
+impl UiRenderNode {
+    pub fn draw<'a>(&'a self, pass: &mut RenderPass<'a>) {
+        pass.set_pipeline(&self.pipeline);
+        pass.set_bind_group(0, &self.uniform_group, &[]);
+        for i in &self.active {
+            let layer = &self.layers[i];
+            if layer.instance.len() == 0 && layer.image_instance.len() == 0 {
+                continue;
+            }
+            pass.set_bind_group(1, &layer.primitive_group, &[]);
+            if layer.instance.len() > 0 {
+                pass.set_bind_group(2, &self.rsc_group, &[]);
+                pass.set_vertex_buffer(0, layer.instance.buffer.slice(..));
+                pass.draw(0..4, 0..layer.instance.len() as u32);
+            }
+            // Images draw after this layer's rects and glyphs, one draw call
+            // each with its own bind group. That draws every image "on top"
+            // within the layer, which loses nothing that currently exists:
+            // `Primitives::apply_free` frees with `swap_remove`, so a layer's
+            // draw order was already undefined before images had their own
+            // list -- nothing before this relied on interleaving a rect
+            // between two images at a particular position.
+            if layer.image_instance.len() > 0 {
+                pass.set_vertex_buffer(0, layer.image_instance.buffer.slice(..));
+                for (k, &tex_idx) in layer.image_tex_indices.iter().enumerate() {
+                    pass.set_bind_group(2, self.textures.image_bind_group(tex_idx), &[]);
+                    pass.draw(0..4, k as u32..k as u32 + 1);
+                }
+            }
+        }
+    }
+
+    pub fn update(
+        &mut self,
+        device: &Device,
+        queue: &Queue,
+        ui: &mut UiData,
+        ui_render: &mut UiRenderState,
+    ) {
+        self.active.clear();
+        for (i, primitives) in ui_render.layers.iter_mut() {
+            self.active.push(i);
+            for change in primitives.apply_free() {
+                if let Some(inst) = ui_render.active.get_mut(&change.id) {
+                    for h in &mut inst.primitives {
+                        // `is_image` disambiguates: `instances` and `images`
+                        // are separate lists with independent indices, so
+                        // without it a rect's renumbering could be applied to
+                        // an image handle that happened to share the same
+                        // (layer, inst_idx).
+                        if h.layer == i
+                            && h.inst_idx == change.old
+                            && (h.binding == IMAGE_BINDING) == change.is_image
+                        {
+                            h.inst_idx = change.new;
+                            break;
+                        }
+                    }
+                }
+            }
+            let rlayer = self.layers.entry(i).or_insert_with(|| {
+                let primitives = PrimitiveBuffers::new(device);
+                let primitive_group =
+                    Self::primitive_group(device, &self.primitive_layout, primitives.buffers());
+                RenderLayer {
+                    instance: ArrBuf::new(
+                        device,
+                        BufferUsages::VERTEX | BufferUsages::COPY_DST,
+                        "instance",
+                    ),
+                    primitives,
+                    primitive_group,
+                    image_instance: ArrBuf::new(
+                        device,
+                        BufferUsages::VERTEX | BufferUsages::COPY_DST,
+                        "image instance",
+                    ),
+                    image_tex_indices: Vec::new(),
+                }
+            });
+            if primitives.updated {
+                rlayer
+                    .instance
+                    .update(device, queue, primitives.instances());
+                rlayer.primitives.update(device, queue, primitives.data());
+                rlayer.primitive_group = Self::primitive_group(
+                    device,
+                    &self.primitive_layout,
+                    rlayer.primitives.buffers(),
+                );
+                rlayer
+                    .image_instance
+                    .update(device, queue, primitives.image_instances());
+                rlayer.image_tex_indices = primitives
+                    .image_instances()
+                    .iter()
+                    .map(|inst| inst.idx)
+                    .collect();
+                primitives.updated = false;
+            }
+        }
+        let masks_resized = if ui.masks.changed {
+            ui.masks.changed = false;
+            self.masks.update(device, queue, &ui.masks[..])
+        } else {
+            false
+        };
+        let moves_resized = if ui.move_offsets.changed {
+            ui.move_offsets.changed = false;
+            self.move_offsets
+                .update(device, queue, &ui.move_offsets[..])
+        } else {
+            false
+        };
+        let rebuild_main = self.textures.update(
+            &mut ui.textures,
+            &self.rsc_layout,
+            &self.masks,
+            &self.move_offsets,
+            masks_resized || moves_resized,
+        );
+        if rebuild_main {
+            self.rsc_group = Self::rsc_group(
+                device,
+                &self.rsc_layout,
+                &self.textures,
+                &self.masks,
+                &self.move_offsets,
+            );
+        }
+    }
+
+    /// Takes a size rather than a window type: this is the only thing the
+    /// core wanted from winit, and depending on a windowing backend for two
+    /// numbers is what put `android-activity` in the core's graph for an
+    /// Android build that is meant to go through android-view instead.
+    pub fn resize(&mut self, size: impl Into<Vec2>, queue: &Queue) {
+        let size = size.into();
+        let slice = &[WindowUniform {
+            width: size.x,
+            height: size.y,
+        }];
+        queue.write_buffer(&self.window_buffer, 0, bytemuck::cast_slice(slice));
+    }
+
+    pub fn new(device: &Device, queue: &Queue, config: &SurfaceConfiguration) -> Self {
+        let shader = device.create_shader_module(ShaderModuleDescriptor {
+            label: Some("UI Shape Shader"),
+            source: ShaderSource::Wgsl(SHAPE_SHADER.into()),
+        });
+
+        let window_uniform = WindowUniform::default();
+        let window_buffer = device.create_buffer_init(&BufferInitDescriptor {
+            label: Some("window"),
+            contents: bytemuck::cast_slice(&[window_uniform]),
+            usage: BufferUsages::UNIFORM | BufferUsages::COPY_DST,
+        });
+
+        let uniform_layout = device.create_bind_group_layout(&BindGroupLayoutDescriptor {
+            entries: &[BindGroupLayoutEntry {
+                binding: 0,
+                visibility: ShaderStages::VERTEX | ShaderStages::FRAGMENT,
+                ty: BindingType::Buffer {
+                    ty: BufferBindingType::Uniform,
+                    has_dynamic_offset: false,
+                    min_binding_size: None,
+                },
+                count: None,
+            }],
+            label: Some("window"),
+        });
+
+        let uniform_group = Self::bind_group_0(device, &uniform_layout, &window_buffer);
+
+        let primitive_layout = device.create_bind_group_layout(&BindGroupLayoutDescriptor {
+            entries: &PrimitiveBuffers::BINDINGS.map(|binding| BindGroupLayoutEntry {
+                binding,
+                visibility: ShaderStages::FRAGMENT,
+                ty: BindingType::Buffer {
+                    ty: BufferBindingType::Storage { read_only: true },
+                    has_dynamic_offset: false,
+                    min_binding_size: None,
+                },
+                count: None,
+            }),
+            label: Some("primitive"),
+        });
+
+        let tex_manager = GpuTextures::new(device, queue);
+        let masks = ArrBuf::new(
+            device,
+            BufferUsages::STORAGE | BufferUsages::COPY_DST,
+            "ui masks",
+        );
+        let move_offsets = ArrBuf::new(
+            device,
+            BufferUsages::STORAGE | BufferUsages::COPY_DST,
+            "ui move offsets",
+        );
+
+        let rsc_layout = Self::rsc_layout(device);
+        let rsc_group = Self::rsc_group(device, &rsc_layout, &tex_manager, &masks, &move_offsets);
+
+        let pipeline_layout = device.create_pipeline_layout(&PipelineLayoutDescriptor {
+            label: Some("UI Shape Pipeline Layout"),
+            bind_group_layouts: &[&uniform_layout, &primitive_layout, &rsc_layout],
+            immediate_size: 0,
+        });
+        let pipeline = device.create_render_pipeline(&RenderPipelineDescriptor {
+            label: Some("UI Shape Pipeline"),
+            layout: Some(&pipeline_layout),
+            vertex: VertexState {
+                module: &shader,
+                entry_point: Some("vs_main"),
+                buffers: &[PrimitiveInstance::desc()],
+                compilation_options: Default::default(),
+            },
+            fragment: Some(FragmentState {
+                module: &shader,
+                entry_point: Some("fs_main"),
+                targets: &[Some(ColorTargetState {
+                    format: config.format,
+                    blend: Some(BlendState::ALPHA_BLENDING),
+                    write_mask: ColorWrites::ALL,
+                })],
+                compilation_options: Default::default(),
+            }),
+            primitive: PrimitiveState {
+                topology: PrimitiveTopology::TriangleStrip,
+                strip_index_format: None,
+                front_face: FrontFace::Cw,
+                cull_mode: Some(Face::Back),
+                polygon_mode: PolygonMode::Fill,
+                unclipped_depth: false,
+                conservative: false,
+            },
+            depth_stencil: None,
+            multisample: MultisampleState {
+                count: 1,
+                mask: !0,
+                alpha_to_coverage_enabled: false,
+            },
+            multiview_mask: None,
+            cache: None,
+        });
+
+        Self {
+            uniform_group,
+            primitive_layout,
+            rsc_layout,
+            rsc_group,
+            pipeline,
+            window_buffer,
+            layers: HashMap::default(),
+            active: Vec::new(),
+            textures: tex_manager,
+            masks,
+            move_offsets,
+        }
+    }
+
+    fn bind_group_0(
+        device: &Device,
+        layout: &BindGroupLayout,
+        window_buffer: &Buffer,
+    ) -> BindGroup {
+        device.create_bind_group(&BindGroupDescriptor {
+            layout,
+            entries: &[BindGroupEntry {
+                binding: 0,
+                resource: window_buffer.as_entire_binding(),
+            }],
+            label: Some("ui window"),
+        })
+    }
+
+    fn primitive_group(
+        device: &Device,
+        layout: &BindGroupLayout,
+        buffers: [(u32, &Buffer); PrimitiveBuffers::LEN],
+    ) -> BindGroup {
+        device.create_bind_group(&BindGroupDescriptor {
+            layout,
+            entries: &buffers.map(|(binding, buf)| BindGroupEntry {
+                binding,
+                resource: buf.as_entire_binding(),
+            }),
+            label: Some("ui primitives"),
+        })
+    }
+
+    /// Group 2: the shared atlas array, one standalone-image slot (a null
+    /// view for the main draw, a real one for each image's own bind group --
+    /// see `GpuTextures`), one sampler and the masks buffer. No `count` on
+    /// any entry: this needs nothing beyond plain Vulkan 1.0 / GLES
+    /// sampling, unlike the `binding_array` layout it replaced (see
+    /// TEXTURES.md's "Recommended shape").
+    fn rsc_layout(device: &Device) -> BindGroupLayout {
+        device.create_bind_group_layout(&BindGroupLayoutDescriptor {
+            entries: &[
+                BindGroupLayoutEntry {
+                    binding: 0,
+                    visibility: ShaderStages::FRAGMENT,
+                    ty: BindingType::Texture {
+                        sample_type: TextureSampleType::Float { filterable: false },
+                        view_dimension: TextureViewDimension::D2Array,
+                        multisampled: false,
+                    },
+                    count: None,
+                },
+                BindGroupLayoutEntry {
+                    binding: 1,
+                    visibility: ShaderStages::FRAGMENT,
+                    ty: BindingType::Texture {
+                        sample_type: TextureSampleType::Float { filterable: false },
+                        view_dimension: TextureViewDimension::D2,
+                        multisampled: false,
+                    },
+                    count: None,
+                },
+                BindGroupLayoutEntry {
+                    binding: 2,
+                    visibility: ShaderStages::FRAGMENT,
+                    ty: BindingType::Sampler(SamplerBindingType::NonFiltering),
+                    count: None,
+                },
+                BindGroupLayoutEntry {
+                    binding: 3,
+                    visibility: ShaderStages::FRAGMENT,
+                    ty: BindingType::Buffer {
+                        ty: BufferBindingType::Storage { read_only: true },
+                        has_dynamic_offset: false,
+                        min_binding_size: None,
+                    },
+                    count: None,
+                },
+                BindGroupLayoutEntry {
+                    binding: 4,
+                    visibility: ShaderStages::VERTEX | ShaderStages::FRAGMENT,
+                    ty: BindingType::Buffer {
+                        ty: BufferBindingType::Storage { read_only: true },
+                        has_dynamic_offset: false,
+                        min_binding_size: None,
+                    },
+                    count: None,
+                },
+            ],
+            label: Some("ui rsc"),
+        })
+    }
+
+    /// The main group: rects and glyphs never sample the image slot, so it
+    /// gets a 1x1 null view rather than any live standalone image's.
+    fn rsc_group(
+        device: &Device,
+        layout: &BindGroupLayout,
+        tex_manager: &GpuTextures,
+        masks: &ArrBuf<Mask>,
+        move_offsets: &ArrBuf<MoveOffset>,
+    ) -> BindGroup {
+        device.create_bind_group(&BindGroupDescriptor {
+            layout,
+            entries: &[
+                BindGroupEntry {
+                    binding: 0,
+                    resource: BindingResource::TextureView(tex_manager.array_view()),
+                },
+                BindGroupEntry {
+                    binding: 1,
+                    resource: BindingResource::TextureView(tex_manager.null_view()),
+                },
+                BindGroupEntry {
+                    binding: 2,
+                    resource: BindingResource::Sampler(tex_manager.sampler()),
+                },
+                BindGroupEntry {
+                    binding: 3,
+                    resource: masks.buffer.as_entire_binding(),
+                },
+                BindGroupEntry {
+                    binding: 4,
+                    resource: move_offsets.buffer.as_entire_binding(),
+                },
+            ],
+            label: Some("ui rsc"),
+        })
+    }
+
+    pub fn view_count(&self) -> usize {
+        self.textures.view_count()
+    }
+}

iris/core/src/render/primitive.rs

@@ -0,0 +1,446 @@
+use std::ops::{Deref, DerefMut};
+
+use crate::{
+    Color, UiRegion, WidgetId,
+    render::{
+        ArrBuf,
+        data::{MaskIdx, MoveIdx, PrimitiveInstance},
+    },
+};
+use bytemuck::Pod;
+use wgpu::*;
+
+pub struct Primitives {
+    instances: Vec<PrimitiveInstance>,
+    assoc: Vec<WidgetId>,
+    data: PrimitiveData,
+    free: Vec<usize>,
+
+    /// Standalone images, kept apart from `instances` because each one draws
+    /// with its own bind group rather than sharing the layer's one instanced
+    /// draw -- see TEXTURES.md's "Recommended shape". `idx` on each
+    /// `PrimitiveInstance` here is the texture's slot in `Textures`/
+    /// `GpuTextures`, not an index into `data`; there is no per-image entry
+    /// in `data` because a bind group already picks the texture; nothing
+    /// left to look up per-instance.
+    images: Vec<PrimitiveInstance>,
+    image_assoc: Vec<WidgetId>,
+    image_free: Vec<usize>,
+
+    pub updated: bool,
+}
+
+impl Default for Primitives {
+    fn default() -> Self {
+        Self {
+            instances: Default::default(),
+            assoc: Default::default(),
+            data: Default::default(),
+            free: Vec::new(),
+            images: Default::default(),
+            image_assoc: Default::default(),
+            image_free: Vec::new(),
+            updated: true,
+        }
+    }
+}
+
+/// The `binding` tag `Painter` writes on an image instance. Distinct from any
+/// `Primitive::BINDING` because images have no `PrimitiveData` entry to key
+/// one from -- a bind group already selects the texture -- so this only ever
+/// has to match the shader's `TEXTURE` constant and flag "this instance lives
+/// in `Primitives::images`, not `Primitives::instances`" to the code below.
+pub const IMAGE_BINDING: u32 = 1;
+
+pub trait Primitive: Pod {
+    const BINDING: u32;
+    fn vec(data: &mut PrimitiveData) -> &mut PrimitiveVec<Self>;
+}
+
+macro_rules! primitives {
+    ($($name:ident: $ty:ty => $binding:expr,)*) => {
+        #[derive(Default)]
+        pub struct PrimitiveData {
+            $(pub(crate) $name: PrimitiveVec<$ty>,)*
+        }
+
+        pub struct PrimitiveBuffers {
+            $($name: ArrBuf<$ty>,)*
+        }
+
+        impl PrimitiveBuffers {
+            pub fn update(&mut self, device: &Device, queue: &Queue, data: &PrimitiveData) {
+                $(self.$name.update(device, queue, &data.$name);)*
+            }
+        }
+
+        impl PrimitiveBuffers {
+            pub const LEN: usize = primitives!(@count $($name)*);
+            /// The group-1 binding number each primitive's storage buffer
+            /// sits at, in declaration order. Not `0..LEN`: a primitive's
+            /// `BINDING` also tags its instances for the shader's dispatch
+            /// switch, and a removed primitive (as `TEXTURE` was, once
+            /// images stopped needing a per-instance storage entry) can
+            /// leave a gap, so the pipeline layout has to ask for these
+            /// exact numbers rather than assuming they are contiguous.
+            pub const BINDINGS: [u32; Self::LEN] = [$(<$ty>::BINDING,)*];
+            pub fn buffers(&self) -> [(u32, &Buffer); Self::LEN] {
+                [
+                    $((<$ty>::BINDING, &self.$name.buffer),)*
+                ]
+            }
+            pub fn new(device: &Device) -> Self {
+                Self {
+                    $($name: ArrBuf::new(
+                        device,
+                        BufferUsages::STORAGE | BufferUsages::COPY_DST,
+                        stringify!($name),
+                    ),)*
+                }
+            }
+        }
+
+        impl PrimitiveData {
+            pub fn clear(&mut self) {
+                $(self.$name.clear();)*
+            }
+            pub fn free(&mut self, binding: u32, idx: usize) {
+                match binding {
+                    $(<$ty>::BINDING => self.$name.free(idx),)*
+                    _ => unreachable!()
+                }
+            }
+        }
+
+        $(
+            unsafe impl bytemuck::Pod for $ty {}
+            unsafe impl bytemuck::Zeroable for $ty {}
+            impl Primitive for $ty {
+                const BINDING: u32 = $binding;
+                fn vec(data: &mut PrimitiveData) -> &mut PrimitiveVec<Self> {
+                    &mut data.$name
+                }
+            }
+        )*
+    };
+    // The recursion has to hand back the same shape it matches -- space
+    // separated, not comma separated. Written with `$($t),+` it re-entered
+    // with a comma as the first token and never terminated, which happened to
+    // work only because there were exactly two primitives: the first step left
+    // a single token, and a single token matches the base case whichever
+    // separator it was written with.
+    (@count $t1:tt $($t:tt)+) => { 1 + primitives!(@count $($t)+) };
+    (@count $t:tt) => { 1 };
+}
+
+pub struct PrimitiveInst<P> {
+    pub id: WidgetId,
+    pub primitive: P,
+    pub region: UiRegion,
+    pub mask_idx: MaskIdx,
+    pub move_idx: MoveIdx,
+}
+
+impl Primitives {
+    pub fn write<P: Primitive>(
+        &mut self,
+        layer: usize,
+        PrimitiveInst {
+            id,
+            primitive,
+            region,
+            mask_idx,
+            move_idx,
+        }: PrimitiveInst<P>,
+    ) -> PrimitiveHandle {
+        self.updated = true;
+        let vec = P::vec(&mut self.data);
+        let i = vec.add(primitive);
+        let inst = PrimitiveInstance {
+            region,
+            idx: i as u32,
+            mask_idx,
+            move_idx,
+            binding: P::BINDING,
+        };
+        let inst_i = if let Some(i) = self.free.pop() {
+            self.instances[i] = inst;
+            self.assoc[i] = id;
+            i
+        } else {
+            let i = self.instances.len();
+            self.instances.push(inst);
+            self.assoc.push(id);
+            i
+        };
+        PrimitiveHandle::new::<P>(layer, inst_i, i)
+    }
+
+    /// Writes an image instance directly -- there is no `Primitive` impl for
+    /// it to go through `write`, since it has nowhere in `PrimitiveData` to
+    /// put a per-instance entry. `texture_idx` is the slot the bind group at
+    /// draw time is chosen from, carried in the otherwise-unused `idx` field.
+    pub fn write_image(
+        &mut self,
+        layer: usize,
+        id: WidgetId,
+        texture_idx: u32,
+        region: UiRegion,
+        mask_idx: MaskIdx,
+        move_idx: MoveIdx,
+    ) -> PrimitiveHandle {
+        self.updated = true;
+        let inst = PrimitiveInstance {
+            region,
+            idx: texture_idx,
+            mask_idx,
+            move_idx,
+            binding: IMAGE_BINDING,
+        };
+        let inst_i = if let Some(i) = self.image_free.pop() {
+            self.images[i] = inst;
+            self.image_assoc[i] = id;
+            i
+        } else {
+            let i = self.images.len();
+            self.images.push(inst);
+            self.image_assoc.push(id);
+            i
+        };
+        PrimitiveHandle {
+            layer,
+            inst_idx: inst_i,
+            data_idx: 0,
+            binding: IMAGE_BINDING,
+        }
+    }
+
+    pub fn image_instances(&self) -> &Vec<PrimitiveInstance> {
+        &self.images
+    }
+
+    /// returns (old index, new index) for both lists this layer keeps --
+    /// `PrimitiveChange::is_image` says which, since the two have separate
+    /// index spaces and `old`/`new` alone would collide between them.
+    ///
+    /// Both lists free with `swap_remove`, so a layer's draw order was
+    /// already undefined before images existed: nothing here may assume one
+    /// primitive stays adjacent to another once anything in the layer has
+    /// been freed.
+    pub fn apply_free(&mut self) -> Vec<PrimitiveChange> {
+        let mut changes =
+            Self::apply_free_list(&mut self.free, &mut self.instances, &mut self.assoc, false);
+        changes.extend(Self::apply_free_list(
+            &mut self.image_free,
+            &mut self.images,
+            &mut self.image_assoc,
+            true,
+        ));
+        changes
+    }
+
+    fn apply_free_list(
+        free: &mut Vec<usize>,
+        instances: &mut Vec<PrimitiveInstance>,
+        assoc: &mut Vec<WidgetId>,
+        is_image: bool,
+    ) -> Vec<PrimitiveChange> {
+        free.sort_by(|a, b| b.cmp(a));
+        free.drain(..)
+            .filter_map(|i| {
+                instances.swap_remove(i);
+                assoc.swap_remove(i);
+                if i == instances.len() {
+                    return None;
+                }
+                let id = assoc[i];
+                let old = instances.len();
+                Some(PrimitiveChange {
+                    id,
+                    is_image,
+                    old,
+                    new: i,
+                })
+            })
+            .collect()
+    }
+
+    pub fn free(&mut self, h: &PrimitiveHandle) -> MaskIdx {
+        self.updated = true;
+        if h.binding == IMAGE_BINDING {
+            self.image_free.push(h.inst_idx);
+            self.images[h.inst_idx].mask_idx
+        } else {
+            self.data.free(h.binding, h.data_idx);
+            self.free.push(h.inst_idx);
+            self.instances[h.inst_idx].mask_idx
+        }
+    }
+
+    pub fn data(&self) -> &PrimitiveData {
+        &self.data
+    }
+
+    pub fn instances(&self) -> &Vec<PrimitiveInstance> {
+        &self.instances
+    }
+
+    pub fn region_mut(&mut self, h: &PrimitiveHandle) -> &mut UiRegion {
+        self.updated = true;
+        if h.binding == IMAGE_BINDING {
+            &mut self.images[h.inst_idx].region
+        } else {
+            &mut self.instances[h.inst_idx].region
+        }
+    }
+}
+
+pub struct PrimitiveChange {
+    pub id: WidgetId,
+    /// Which of `Primitives::instances`/`Primitives::images` this change
+    /// belongs to -- their `old`/`new` indices are independent, so a
+    /// consumer matching only on `(layer, inst_idx)` could apply an image's
+    /// renumbering to a rect's handle that happens to share the same index.
+    pub is_image: bool,
+    pub old: usize,
+    pub new: usize,
+}
+
+#[derive(Debug)]
+pub struct PrimitiveHandle {
+    pub layer: usize,
+    pub inst_idx: usize,
+    pub data_idx: usize,
+    pub binding: u32,
+}
+
+impl PrimitiveHandle {
+    fn new<P: Primitive>(layer: usize, inst_idx: usize, data_idx: usize) -> Self {
+        Self {
+            layer,
+            inst_idx,
+            data_idx,
+            binding: P::BINDING,
+        }
+    }
+}
+
+primitives!(
+    rects: RectPrimitive => 0,
+    glyphs: GlyphPrimitive => 2,
+);
+
+#[repr(C)]
+#[derive(Copy, Clone)]
+pub struct RectPrimitive {
+    pub color: Color<u8>,
+    pub radius: f32,
+    pub thickness: f32,
+    pub inner_radius: f32,
+}
+
+impl RectPrimitive {
+    pub fn color(color: Color<u8>) -> Self {
+        Self {
+            color,
+            radius: 0.0,
+            thickness: 0.0,
+            inner_radius: 0.0,
+        }
+    }
+}
+
+/// One glyph, drawn as a sub-rectangle of the glyph atlas array.
+///
+/// `color` is the text colour and is multiplied by the atlas's alpha for an
+/// ordinary mask glyph; a colour glyph (emoji) carries its own colour and
+/// takes the atlas texel unchanged, which is what `IS_COLOR` selects.
+#[repr(C)]
+#[derive(Debug, Copy, Clone)]
+pub struct GlyphPrimitive {
+    pub uv_min: [f32; 2],
+    pub uv_max: [f32; 2],
+    /// Layer of the shared atlas array texture this glyph's page occupies --
+    /// not a bind-group or view index, since a page never gets one of its
+    /// own. See TEXTURES.md's "Recommended shape".
+    pub layer: u32,
+    pub color: Color<u8>,
+    pub flags: u32,
+    /// Pads this struct's Rust size to match WGSL's storage-buffer layout for
+    /// `GlyphInfo`: two `vec2<f32>` members give the struct an 8-byte
+    /// alignment, which rounds the WGSL size up to 32 bytes even though the
+    /// fields above only total 28. `bytemuck` does not check this for us.
+    _pad: u32,
+}
+
+impl GlyphPrimitive {
+    pub const IS_COLOR: u32 = 1;
+
+    pub fn new(
+        uv_min: [f32; 2],
+        uv_max: [f32; 2],
+        layer: u32,
+        color: Color<u8>,
+        flags: u32,
+    ) -> Self {
+        Self {
+            uv_min,
+            uv_max,
+            layer,
+            color,
+            flags,
+            _pad: 0,
+        }
+    }
+}
+
+pub struct PrimitiveVec<T> {
+    vec: Vec<T>,
+    free: Vec<usize>,
+}
+
+impl<T> PrimitiveVec<T> {
+    pub fn new() -> Self {
+        Self {
+            vec: Vec::new(),
+            free: Vec::new(),
+        }
+    }
+    pub fn add(&mut self, t: T) -> usize {
+        if let Some(i) = self.free.pop() {
+            self.vec[i] = t;
+            i
+        } else {
+            let i = self.vec.len();
+            self.vec.push(t);
+            i
+        }
+    }
+    pub fn free(&mut self, i: usize) {
+        self.free.push(i);
+    }
+    pub fn clear(&mut self) {
+        self.free.clear();
+        self.vec.clear();
+    }
+}
+
+impl<T> Default for PrimitiveVec<T> {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+impl<T> Deref for PrimitiveVec<T> {
+    type Target = Vec<T>;
+
+    fn deref(&self) -> &Self::Target {
+        &self.vec
+    }
+}
+
+impl<T> DerefMut for PrimitiveVec<T> {
+    fn deref_mut(&mut self) -> &mut Self::Target {
+        &mut self.vec
+    }
+}

iris/core/src/render/shader.wgsl

@@ -0,0 +1,247 @@
+const RECT: u32 = 0u;
+// TEXTURE has no entry in group 1: a standalone image draws with its own
+// bind group (see UiRenderNode::draw), so there is nothing per-instance left
+// to look up here -- the bind group already picked the texture.
+const TEXTURE: u32 = 1u;
+const GLYPH: u32 = 2u;
+
+@group(0) @binding(0)
+var<uniform> window: WindowUniform;
+@group(1) @binding(RECT)
+var<storage> rects: array<Rect>;
+@group(1) @binding(GLYPH)
+var<storage> glyphs: array<GlyphInfo>;
+
+struct Rect {
+    color: u32,
+    radius: f32,
+    thickness: f32,
+    inner_radius: f32,
+}
+
+struct GlyphInfo {
+    uv_min: vec2<f32>,
+    uv_max: vec2<f32>,
+    // Layer of the shared atlas array texture, not a view or bind-group
+    // index -- a page never gets its own bind group. See TEXTURES.md's
+    // "Recommended shape".
+    layer: u32,
+    color: u32,
+    flags: u32,
+}
+
+struct Mask {
+    x: UiSpan,
+    y: UiSpan,
+    move_idx: u32,
+}
+
+/// One widget's cumulative on-screen translation and the slot of the
+/// ancestor to add on top of it. Mirrors `MoveOffset` in data.rs.
+struct MoveOffset {
+    delta: vec2<f32>,
+    parent: u32,
+}
+
+struct UiSpan {
+    start: UiScalar,
+    end: UiScalar,
+}
+
+struct UiScalar {
+    rel: f32,
+    abs: f32,
+}
+
+struct UiVec2 {
+    rel: vec2<f32>,
+    abs: vec2<f32>,
+}
+
+// The shared glyph atlas: every page is one layer. Growing it recreates this
+// texture with headroom and copies the old layers across -- see
+// GpuTextures::grow_array -- rather than the binding_array<texture_2d<f32>>
+// this replaced, which needed VK_EXT_descriptor_indexing and does not survive
+// a real share of Android GPUs (see TEXTURES.md).
+@group(2) @binding(0)
+var atlas: texture_2d_array<f32>;
+// One standalone image's texture. The main draw (rects and glyphs) binds a
+// 1x1 null texture here, since neither samples it; each image draw call
+// binds its own -- see UiRenderNode::draw.
+@group(2) @binding(1)
+var image_texture: texture_2d<f32>;
+@group(2) @binding(2)
+var samp: sampler;
+@group(2) @binding(3)
+var<storage> masks: array<Mask>;
+@group(2) @binding(4)
+var<storage> move_offsets: array<MoveOffset>;
+
+// A move chain more than this deep means something else is wrong (an
+// accidental cycle) -- kept in step with `MOVE_CHAIN_LIMIT` in
+// render_state.rs, which walks the identical bound on the CPU side for
+// hit-testing. Bounded so a malformed chain cannot hang the GPU.
+const MOVE_CHAIN_LIMIT: u32 = 16u;
+
+/// Sums the pixel delta along the parent chain starting at `idx`, shared by
+/// the vertex stage (a primitive's own corners) and the fragment stage (its
+/// mask's corners) so the walk is written once. See LAYOUT.md section 2b.
+fn resolve_move(idx: u32) -> vec2<f32> {
+    var total = vec2<f32>(0.0, 0.0);
+    var i = idx;
+    for (var step = 0u; step < MOVE_CHAIN_LIMIT; step++) {
+        let entry = move_offsets[i];
+        total += entry.delta;
+        if entry.parent == 4294967295u {
+            break;
+        }
+        i = entry.parent;
+    }
+    return total;
+}
+
+struct WindowUniform {
+    dim: vec2<f32>,
+};
+
+struct InstanceInput {
+    @location(0) x_start: vec2<f32>,
+    @location(1) x_end: vec2<f32>,
+    @location(2) y_start: vec2<f32>,
+    @location(3) y_end: vec2<f32>,
+    @location(4) binding: u32,
+    @location(5) idx: u32,
+    @location(6) mask_idx: u32,
+    @location(7) move_idx: u32,
+}
+
+struct VertexOutput {
+    @location(0) top_left: vec2<f32>,
+    @location(1) bot_right: vec2<f32>,
+    @location(2) uv: vec2<f32>,
+    @location(3) binding: u32,
+    @location(4) idx: u32,
+    @location(5) mask_idx: u32,
+    @builtin(position) clip_position: vec4<f32>,
+};
+
+struct Region {
+    pos: vec2<f32>,
+    uv: vec2<f32>,
+    top_left: vec2<f32>,
+    bot_right: vec2<f32>,
+}
+
+@vertex
+fn vs_main(
+    @builtin(vertex_index) vi: u32,
+    in: InstanceInput,
+) -> VertexOutput {
+    var out: VertexOutput;
+
+    let top_left_rel = vec2(in.x_start.x, in.y_start.x);
+    let top_left_abs = vec2(in.x_start.y, in.y_start.y);
+    let bot_right_rel = vec2(in.x_end.x, in.y_end.x);
+    let bot_right_abs = vec2(in.x_end.y, in.y_end.y);
+
+    let move_delta = resolve_move(in.move_idx);
+    let top_left = floor(top_left_rel * window.dim) + floor(top_left_abs) + move_delta;
+    let bot_right = floor(bot_right_rel * window.dim) + floor(bot_right_abs) + move_delta;
+    let size = bot_right - top_left;
+
+    let uv = vec2<f32>(
+        f32(vi % 2u),
+        f32(vi / 2u)
+    );
+    let pos = (top_left + uv * size) / window.dim * 2.0 - 1.0;
+    out.clip_position = vec4<f32>(pos.x, -pos.y, 0.0, 1.0);
+    out.uv = uv;
+    out.binding = in.binding;
+    out.idx = in.idx;
+    out.top_left = top_left;
+    out.bot_right = bot_right;
+    out.mask_idx = in.mask_idx;
+
+    return out;
+}
+
+@fragment
+fn fs_main(
+    in: VertexOutput
+) -> @location(0) vec4<f32> {
+    let pos = in.clip_position.xy;
+    let region = Region(pos, in.uv, in.top_left, in.bot_right);
+    let i = in.idx;
+    var color: vec4<f32>;
+    switch in.binding {
+        case RECT: {
+            color = draw_rounded_rect(region, rects[i]);
+        }
+        case TEXTURE: {
+            color = draw_texture(region);
+        }
+        case GLYPH: {
+            color = draw_glyph(region, glyphs[i]);
+        }
+        default: {
+            color = vec4(1.0, 0.0, 1.0, 1.0);
+        }
+    }
+    if in.mask_idx != 4294967295u {
+        let mask = masks[in.mask_idx];
+        let mask_delta = resolve_move(mask.move_idx);
+        let tl = UiVec2(vec2(mask.x.start.rel, mask.y.start.rel), vec2(mask.x.start.abs, mask.y.start.abs));
+        let br = UiVec2(vec2(mask.x.end.rel, mask.y.end.rel), vec2(mask.x.end.abs, mask.y.end.abs));
+
+        let top_left = floor(tl.rel * window.dim) + floor(tl.abs) + mask_delta;
+        let bot_right = floor(br.rel * window.dim) + floor(br.abs) + mask_delta;
+        if pos.x < top_left.x || pos.x > bot_right.x || pos.y < top_left.y || pos.y > bot_right.y {
+            color *= 0.0;
+        }
+    }
+    return color;
+}
+
+fn draw_texture(region: Region) -> vec4<f32> {
+    return textureSample(image_texture, samp, region.uv);
+}
+
+fn draw_glyph(region: Region, g: GlyphInfo) -> vec4<f32> {
+    let uv = mix(g.uv_min, g.uv_max, region.uv);
+    let texel = textureSample(atlas, samp, uv, i32(g.layer));
+    if (g.flags & 1u) != 0u {
+        return texel;
+    }
+    var color = unpack4x8unorm(g.color);
+    color.a *= texel.a;
+    return color;
+}
+
+fn draw_rounded_rect(region: Region, rect: Rect) -> vec4<f32> {
+    var color = unpack4x8unorm(rect.color);
+
+    let edge = 0.5;
+
+    let size = region.bot_right - region.top_left;
+    let corner = size / 2.0;
+    let center = region.top_left + corner;
+
+    let dist = distance_from_rect(region.pos, center, corner, rect.radius);
+    color.a *= 1.0 - smoothstep(-min(edge, rect.radius), edge, dist);
+
+    if rect.thickness > 0.0 {
+        let dist2 = distance_from_rect(region.pos, center, corner - rect.thickness, rect.inner_radius);
+        color.a *= smoothstep(-min(edge, rect.inner_radius), edge, dist2);
+    }
+
+    return color;
+}
+
+fn distance_from_rect(pixel_pos: vec2<f32>, rect_center: vec2<f32>, rect_corner: vec2<f32>, radius: f32) -> f32 {
+    // vec from center to pixel
+    let p = pixel_pos - rect_center;
+    // vec from inner rect corner to pixel
+    let q = abs(p) - (rect_corner - radius);
+    return length(max(q, vec2(0.0))) - radius;
+}
+

iris/core/src/render/texture.rs

@@ -0,0 +1,492 @@
+use image::{DynamicImage, EncodableLayout, GenericImageView};
+use wgpu::{util::DeviceExt, *};
+
+use crate::{
+    Mask, MoveOffset, PatchRect, TextureKind, TextureUpdate, Textures, render::util::ArrBuf,
+};
+
+use super::atlas::PAGE;
+
+/// What one texture slot is, GPU-side. Parallel to `Textures`' own slot
+/// numbering (`TextureKind`'s `Image`/`Page`), so a slot's index means the
+/// same thing on both sides without a second map to keep in sync.
+enum Slot {
+    /// A slot that was freed, or pushed and freed within the same batch
+    /// before ever reaching here.
+    Empty,
+    Image(ImageGpu),
+    /// The array layer a page occupies. Pages are never freed (see
+    /// `Textures::free`), so this is the only variant that outlives a `Free`.
+    Page(u32),
+}
+
+struct ImageGpu {
+    /// Kept alive alongside `view`/`bind_group`, which borrow from it only in
+    /// the sense that dropping this drops the GPU resource they point to.
+    #[allow(dead_code)]
+    texture: Texture,
+    view: TextureView,
+    bind_group: BindGroup,
+}
+
+/// Owns the two kinds of texture iris draws:
+///
+/// - **The glyph atlas**, one `texture_2d_array` whose layers are pages
+///   (`Slot::Page`), grown by recreating the array with headroom and
+///   `copy_texture_to_texture`-ing the old layers across. No feature beyond
+///   Vulkan 1.0/GLES sampling is needed for this -- a layer index is an
+///   ordinary sampling operand.
+/// - **Standalone images** (`Slot::Image`), each its own `Texture` and
+///   `BindGroup`, drawn one `draw()` call at a time with that bind group
+///   bound -- see `UiRenderNode::draw`.
+///
+/// See TEXTURES.md's "Recommended shape" for why, and RUST.md's
+/// "iris's binding array does not survive real Android hardware" for what
+/// this replaced (one giant `binding_array<texture_2d<f32>>` needing
+/// `VK_EXT_descriptor_indexing`, which a real share of Android GPUs lack).
+pub struct GpuTextures {
+    device: Device,
+    queue: Queue,
+
+    slots: Vec<Slot>,
+
+    array_texture: Texture,
+    array_view: TextureView,
+    array_capacity: u32,
+    /// Layers actually written. Only grows -- see `Slot::Page`.
+    page_count: u32,
+
+    sampler: Sampler,
+    /// Bound in the image slot of the main draw's bind group, which has
+    /// nothing of its own to put there: rects and glyphs never sample it,
+    /// but the layout requires something bound regardless.
+    null_view: TextureView,
+}
+
+impl GpuTextures {
+    /// Applies queued `Textures` updates, then reports whether the *main*
+    /// bind group (the one rects and glyphs draw with) needs rebuilding --
+    /// true when the atlas array was recreated (its view identity changed)
+    /// or the masks buffer was, since both are bound there. Pushing or
+    /// freeing a standalone image never touches that group: it built or drops
+    /// its own.
+    pub fn update(
+        &mut self,
+        textures: &mut Textures,
+        rsc_layout: &BindGroupLayout,
+        masks: &ArrBuf<Mask>,
+        move_offsets: &ArrBuf<MoveOffset>,
+        masks_resized: bool,
+    ) -> bool {
+        let mut rebuild_main = masks_resized;
+        if masks_resized {
+            // The masks or move-offsets buffer just moved, so every bind
+            // group holding a reference to either -- one per live
+            // standalone image -- is stale.
+            self.rebuild_image_bind_groups(rsc_layout, masks, move_offsets);
+        }
+        for update in textures.updates() {
+            match update {
+                TextureUpdate::Push(kind, image) => {
+                    rebuild_main |= self.push(kind, image, rsc_layout, masks, move_offsets);
+                }
+                TextureUpdate::Set(kind, i, image) => {
+                    rebuild_main |= self.set(kind, i, image, rsc_layout, masks, move_offsets);
+                }
+                // A patch changes texture contents, not which layer or bind
+                // group exists, so it never asks for a rebuild -- rebuilding
+                // per glyph is exactly the cost this exists to avoid.
+                TextureUpdate::Patch(i, rect, image) => self.patch(i, rect, image),
+                TextureUpdate::SetFree => {}
+                TextureUpdate::Free(i) => self.free(i),
+                TextureUpdate::PushFree(_kind) => self.slots.push(Slot::Empty),
+            }
+        }
+        rebuild_main
+    }
+
+    fn push(
+        &mut self,
+        kind: TextureKind,
+        image: &DynamicImage,
+        rsc_layout: &BindGroupLayout,
+        masks: &ArrBuf<Mask>,
+        move_offsets: &ArrBuf<MoveOffset>,
+    ) -> bool {
+        let (slot, rebuilt) = self.make_slot(kind, image, rsc_layout, masks, move_offsets);
+        self.slots.push(slot);
+        rebuilt
+    }
+
+    fn set(
+        &mut self,
+        kind: TextureKind,
+        i: u32,
+        image: &DynamicImage,
+        rsc_layout: &BindGroupLayout,
+        masks: &ArrBuf<Mask>,
+        move_offsets: &ArrBuf<MoveOffset>,
+    ) -> bool {
+        let (slot, rebuilt) = self.make_slot(kind, image, rsc_layout, masks, move_offsets);
+        self.slots[i as usize] = slot;
+        rebuilt
+    }
+
+    fn make_slot(
+        &mut self,
+        kind: TextureKind,
+        image: &DynamicImage,
+        rsc_layout: &BindGroupLayout,
+        masks: &ArrBuf<Mask>,
+        move_offsets: &ArrBuf<MoveOffset>,
+    ) -> (Slot, bool) {
+        match kind {
+            TextureKind::Image => {
+                let gpu = self.create_image(image, rsc_layout, masks, move_offsets);
+                (Slot::Image(gpu), false)
+            }
+            TextureKind::Page { layer } => {
+                let mut rebuilt = false;
+                if layer >= self.array_capacity {
+                    self.grow_array(rsc_layout, masks, move_offsets);
+                    rebuilt = true;
+                }
+                self.write_full_layer(layer, image);
+                self.page_count = self.page_count.max(layer + 1);
+                (Slot::Page(layer), rebuilt)
+            }
+        }
+    }
+
+    fn free(&mut self, i: u32) {
+        if let Some(slot) = self.slots.get_mut(i as usize) {
+            *slot = Slot::Empty;
+        }
+        // A page's layer is not reclaimed here either -- see `Slot::Page`.
+    }
+
+    fn patch(&mut self, i: u32, rect: PatchRect, image: &DynamicImage) {
+        let Some(&Slot::Page(layer)) = self.slots.get(i as usize) else {
+            return;
+        };
+        if rect.width == 0 || rect.height == 0 {
+            return;
+        }
+        // Cropped rather than written straight from the atlas, because
+        // write_texture wants tightly packed rows and the atlas rows are as
+        // wide as the atlas. A glyph is small, so the copy is too.
+        let sub = image
+            .view(rect.x, rect.y, rect.width, rect.height)
+            .to_image();
+        self.queue.write_texture(
+            TexelCopyTextureInfo {
+                texture: &self.array_texture,
+                mip_level: 0,
+                origin: Origin3d {
+                    x: rect.x,
+                    y: rect.y,
+                    z: layer,
+                },
+                aspect: TextureAspect::All,
+            },
+            sub.as_bytes(),
+            TexelCopyBufferLayout {
+                offset: 0,
+                bytes_per_row: Some(rect.width * 4),
+                rows_per_image: Some(rect.height),
+            },
+            Extent3d {
+                width: rect.width,
+                height: rect.height,
+                depth_or_array_layers: 1,
+            },
+        );
+    }
+
+    fn write_full_layer(&self, layer: u32, image: &DynamicImage) {
+        // Every page is created as exactly PAGE x PAGE (`GlyphAtlas::allocate`),
+        // so this is always a whole-layer write, never a crop.
+        let rgba = image.to_rgba8();
+        self.queue.write_texture(
+            TexelCopyTextureInfo {
+                texture: &self.array_texture,
+                mip_level: 0,
+                origin: Origin3d {
+                    x: 0,
+                    y: 0,
+                    z: layer,
+                },
+                aspect: TextureAspect::All,
+            },
+            rgba.as_bytes(),
+            TexelCopyBufferLayout {
+                offset: 0,
+                bytes_per_row: Some(PAGE * 4),
+                rows_per_image: Some(PAGE),
+            },
+            Extent3d {
+                width: PAGE,
+                height: PAGE,
+                depth_or_array_layers: 1,
+            },
+        );
+    }
+
+    /// Doubles the array's layer capacity (headroom, so this is rare) and
+    /// copies the old layers across GPU-side -- no readback. Recreates the
+    /// array's view, which invalidates every bind group that referenced it,
+    /// so this also rebuilds all of them before returning.
+    fn grow_array(
+        &mut self,
+        rsc_layout: &BindGroupLayout,
+        masks: &ArrBuf<Mask>,
+        move_offsets: &ArrBuf<MoveOffset>,
+    ) {
+        let new_capacity = self.array_capacity * 2;
+        let new_texture = Self::create_array_texture(&self.device, new_capacity);
+        if self.page_count > 0 {
+            let mut encoder = self
+                .device
+                .create_command_encoder(&CommandEncoderDescriptor {
+                    label: Some("atlas array grow"),
+                });
+            encoder.copy_texture_to_texture(
+                TexelCopyTextureInfo {
+                    texture: &self.array_texture,
+                    mip_level: 0,
+                    origin: Origin3d::ZERO,
+                    aspect: TextureAspect::All,
+                },
+                TexelCopyTextureInfo {
+                    texture: &new_texture,
+                    mip_level: 0,
+                    origin: Origin3d::ZERO,
+                    aspect: TextureAspect::All,
+                },
+                Extent3d {
+                    width: PAGE,
+                    height: PAGE,
+                    depth_or_array_layers: self.page_count,
+                },
+            );
+            self.queue.submit(std::iter::once(encoder.finish()));
+        }
+        self.array_texture = new_texture;
+        self.array_view = self.array_texture.create_view(&TextureViewDescriptor {
+            dimension: Some(TextureViewDimension::D2Array),
+            ..Default::default()
+        });
+        self.array_capacity = new_capacity;
+        self.rebuild_image_bind_groups(rsc_layout, masks, move_offsets);
+    }
+
+    fn rebuild_image_bind_groups(
+        &mut self,
+        rsc_layout: &BindGroupLayout,
+        masks: &ArrBuf<Mask>,
+        move_offsets: &ArrBuf<MoveOffset>,
+    ) {
+        for slot in &mut self.slots {
+            if let Slot::Image(gpu) = slot {
+                gpu.bind_group = Self::make_image_bind_group(
+                    &self.device,
+                    rsc_layout,
+                    &self.array_view,
+                    &gpu.view,
+                    &self.sampler,
+                    masks,
+                    move_offsets,
+                );
+            }
+        }
+    }
+
+    fn create_image(
+        &self,
+        image: &DynamicImage,
+        rsc_layout: &BindGroupLayout,
+        masks: &ArrBuf<Mask>,
+        move_offsets: &ArrBuf<MoveOffset>,
+    ) -> ImageGpu {
+        let rgba = image.to_rgba8();
+        let (width, height) = rgba.dimensions();
+        let texture = self.device.create_texture_with_data(
+            &self.queue,
+            &TextureDescriptor {
+                label: Some("image"),
+                size: Extent3d {
+                    width,
+                    height,
+                    depth_or_array_layers: 1,
+                },
+                mip_level_count: 1,
+                sample_count: 1,
+                dimension: TextureDimension::D2,
+                format: TextureFormat::Rgba8Unorm,
+                usage: TextureUsages::TEXTURE_BINDING | TextureUsages::COPY_DST,
+                view_formats: &[],
+            },
+            wgt::TextureDataOrder::MipMajor,
+            rgba.as_bytes(),
+        );
+        let view = texture.create_view(&TextureViewDescriptor::default());
+        let bind_group = Self::make_image_bind_group(
+            &self.device,
+            rsc_layout,
+            &self.array_view,
+            &view,
+            &self.sampler,
+            masks,
+            move_offsets,
+        );
+        ImageGpu {
+            texture,
+            view,
+            bind_group,
+        }
+    }
+
+    /// Builds group 2 for one standalone image: the shared atlas array, this
+    /// image's own view, the shared sampler, and the shared masks buffer --
+    /// the same layout the main draw uses with a null view in the image slot.
+    fn make_image_bind_group(
+        device: &Device,
+        rsc_layout: &BindGroupLayout,
+        array_view: &TextureView,
+        image_view: &TextureView,
+        sampler: &Sampler,
+        masks: &ArrBuf<Mask>,
+        move_offsets: &ArrBuf<MoveOffset>,
+    ) -> BindGroup {
+        device.create_bind_group(&BindGroupDescriptor {
+            layout: rsc_layout,
+            entries: &[
+                BindGroupEntry {
+                    binding: 0,
+                    resource: BindingResource::TextureView(array_view),
+                },
+                BindGroupEntry {
+                    binding: 1,
+                    resource: BindingResource::TextureView(image_view),
+                },
+                BindGroupEntry {
+                    binding: 2,
+                    resource: BindingResource::Sampler(sampler),
+                },
+                BindGroupEntry {
+                    binding: 3,
+                    resource: masks.buffer.as_entire_binding(),
+                },
+                BindGroupEntry {
+                    binding: 4,
+                    resource: move_offsets.buffer.as_entire_binding(),
+                },
+            ],
+            label: Some("ui rsc image"),
+        })
+    }
+
+    fn create_array_texture(device: &Device, capacity: u32) -> Texture {
+        device.create_texture(&TextureDescriptor {
+            label: Some("glyph atlas array"),
+            size: Extent3d {
+                width: PAGE,
+                height: PAGE,
+                depth_or_array_layers: capacity,
+            },
+            mip_level_count: 1,
+            sample_count: 1,
+            dimension: TextureDimension::D2,
+            format: TextureFormat::Rgba8Unorm,
+            usage: TextureUsages::TEXTURE_BINDING
+                | TextureUsages::COPY_DST
+                | TextureUsages::COPY_SRC,
+            view_formats: &[],
+        })
+    }
+
+    pub fn new(device: &Device, queue: &Queue) -> Self {
+        let sampler = default_sampler(device);
+        let null_view = null_texture_view(device);
+        let array_capacity = 1;
+        let array_texture = Self::create_array_texture(device, array_capacity);
+        let array_view = array_texture.create_view(&TextureViewDescriptor {
+            dimension: Some(TextureViewDimension::D2Array),
+            ..Default::default()
+        });
+        Self {
+            device: device.clone(),
+            queue: queue.clone(),
+            slots: Vec::new(),
+            array_texture,
+            array_view,
+            array_capacity,
+            page_count: 0,
+            sampler,
+            null_view,
+        }
+    }
+
+    pub fn array_view(&self) -> &TextureView {
+        &self.array_view
+    }
+
+    pub fn null_view(&self) -> &TextureView {
+        &self.null_view
+    }
+
+    pub fn sampler(&self) -> &Sampler {
+        &self.sampler
+    }
+
+    /// The bind group a standalone image draws with. Panics if `idx` names an
+    /// atlas page or a freed slot instead -- either is a caller bug (the
+    /// wrong kind of instance reached this draw path), not a condition to
+    /// recover from.
+    pub fn image_bind_group(&self, idx: u32) -> &BindGroup {
+        match self.slots.get(idx as usize) {
+            Some(Slot::Image(gpu)) => &gpu.bind_group,
+            other => panic!("texture slot {idx} is not a live standalone image: {other:?}"),
+        }
+    }
+
+    pub fn view_count(&self) -> usize {
+        self.slots
+            .iter()
+            .filter(|s| !matches!(s, Slot::Empty))
+            .count()
+    }
+}
+
+impl std::fmt::Debug for Slot {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            Slot::Empty => write!(f, "Empty"),
+            Slot::Image(_) => write!(f, "Image"),
+            Slot::Page(layer) => write!(f, "Page(layer={layer})"),
+        }
+    }
+}
+
+pub fn null_texture_view(device: &Device) -> TextureView {
+    device
+        .create_texture(&TextureDescriptor {
+            label: Some("null"),
+            size: Extent3d {
+                width: 1,
+                height: 1,
+                depth_or_array_layers: 1,
+            },
+            mip_level_count: 1,
+            sample_count: 1,
+            dimension: TextureDimension::D2,
+            format: TextureFormat::Rgba8Unorm,
+            usage: TextureUsages::TEXTURE_BINDING,
+            view_formats: &[],
+        })
+        .create_view(&TextureViewDescriptor::default())
+}
+
+pub fn default_sampler(device: &Device) -> Sampler {
+    device.create_sampler(&SamplerDescriptor::default())
+}

iris/core/src/render/util/mod.rs

@@ -0,0 +1,53 @@
+use std::marker::PhantomData;
+
+use bytemuck::Pod;
+use wgpu::*;
+
+pub struct ArrBuf<T: Pod> {
+    label: &'static str,
+    usage: BufferUsages,
+    pub buffer: Buffer,
+    len: usize,
+    _pd: PhantomData<T>,
+}
+
+impl<T: Pod> ArrBuf<T> {
+    pub fn new(device: &Device, usage: BufferUsages, label: &'static str) -> Self {
+        Self {
+            label,
+            usage,
+            buffer: Self::init_buf(device, 0, usage, label),
+            len: 0,
+            _pd: PhantomData,
+        }
+    }
+    /// Returns whether the underlying `Buffer` was recreated -- a caller that
+    /// cached a `BindGroup` referencing it (as `GpuTextures` does for the
+    /// masks buffer) needs to know to rebuild that too.
+    pub fn update(&mut self, device: &Device, queue: &Queue, data: &[T]) -> bool {
+        let resized = self.len != data.len();
+        if resized {
+            self.len = data.len();
+            self.buffer =
+                Self::init_buf(device, std::mem::size_of_val(data), self.usage, self.label);
+        }
+        queue.write_buffer(&self.buffer, 0, bytemuck::cast_slice(data));
+        resized
+    }
+    fn init_buf(device: &Device, size: usize, usage: BufferUsages, label: &'static str) -> Buffer {
+        let mut size = size as u64;
+        if usage.contains(BufferUsages::STORAGE) {
+            size = size.max(std::mem::size_of::<T>() as u64);
+        }
+        device.create_buffer(&BufferDescriptor {
+            label: Some(label),
+            size,
+            mapped_at_creation: false,
+            usage,
+        })
+    }
+    #[allow(clippy::len_without_is_empty)]
+    pub fn len(&self) -> usize {
+        self.len
+    }
+}

iris/core/src/ui/active.rs

@@ -0,0 +1,24 @@
+use crate::{LayerId, MaskIdx, MoveIdx, PrimitiveHandle, Size, TextureHandle, UiRegion, WidgetId};
+
+/// important non rendering data for retained drawing
+#[derive(Debug)]
+pub struct ActiveData {
+    pub id: WidgetId,
+    pub region: UiRegion,
+    pub parent: Option<WidgetId>,
+    pub textures: Vec<TextureHandle>,
+    pub primitives: Vec<PrimitiveHandle>,
+    pub children: Vec<WidgetId>,
+    pub mask: MaskIdx,
+    pub layer: LayerId,
+    /// What `Widget::draw` returned the last time this widget was actually
+    /// drawn -- read by a parent placing this widget again without
+    /// redrawing it, replacing `Cache.size`'s old role. See LAYOUT.md
+    /// section 5.
+    pub size: Size,
+    /// This widget's slot in `UiData::move_offsets`, assigned on its first
+    /// draw and kept for the rest of its life (redraws reuse it in place
+    /// so a retained child's `parent` link never goes stale). See
+    /// LAYOUT.md section 2.
+    pub move_slot: MoveIdx,
+}

iris/core/src/ui/mod.rs

@@ -0,0 +1,52 @@
+use crate::{
+    Mask, MoveOffset, TextData, Textures, WeakWidget, WidgetId, Widgets, util::TrackedArena,
+};
+
+mod active;
+mod painter;
+mod render_state;
+
+pub use active::*;
+pub use painter::Painter;
+pub use render_state::*;
+
+#[derive(Default)]
+pub struct UiData {
+    pub widgets: Widgets,
+    pub textures: Textures,
+    pub text: TextData,
+    pub masks: TrackedArena<Mask, u32>,
+    /// One entry per widget ever drawn, forming the parent-linked chain
+    /// `resolve_move` walks in both shader stages. Allocated once on a
+    /// widget's first draw and reused for every later redraw of the same
+    /// id (never reallocated), so a retained descendant's `parent` index
+    /// never goes stale -- see LAYOUT.md section 2.
+    pub move_offsets: TrackedArena<MoveOffset, u32>,
+}
+
+pub trait UiRsc {
+    fn ui(&self) -> &UiData;
+    fn ui_mut(&mut self) -> &mut UiData;
+
+    #[allow(unused_variables)]
+    fn on_add(&mut self, id: WeakWidget) {}
+    #[allow(unused_variables)]
+    fn on_remove(&mut self, id: WidgetId) {}
+    #[allow(unused_variables)]
+    fn on_draw(&mut self, active: &ActiveData) {}
+    #[allow(unused_variables)]
+    fn on_undraw(&mut self, active: &ActiveData) {}
+
+    fn widgets(&self) -> &Widgets {
+        &self.ui().widgets
+    }
+    fn widgets_mut(&mut self) -> &mut Widgets {
+        &mut self.ui_mut().widgets
+    }
+    fn free(&mut self) {
+        while let Some(id) = self.widgets_mut().free_next() {
+            self.on_remove(id);
+        }
+        self.ui_mut().textures.free();
+    }
+}

iris/core/src/ui/painter.rs

@@ -0,0 +1,236 @@
+use crate::{
+    RenderedText, Size, StrongWidget, TextAttrs, TextBuffer, TextData, TextureHandle, UiRegion,
+    UiRenderState, UiRsc, UiScalar, UiVec2, WidgetId,
+    render::{GlyphPrimitive, Mask, MaskIdx, MoveIdx, Primitive, PrimitiveHandle, PrimitiveInst},
+    util::Vec2,
+};
+
+/// makes your surfaces look pretty
+pub struct Painter<'a> {
+    pub(super) state: &'a mut UiRenderState,
+    pub(super) rsc: &'a mut dyn UiRsc,
+
+    pub(super) region: UiRegion,
+    pub(super) mask: MaskIdx,
+    pub(super) move_slot: MoveIdx,
+    pub(super) textures: Vec<TextureHandle>,
+    pub(super) primitives: Vec<PrimitiveHandle>,
+    pub(super) children: Vec<WidgetId>,
+    pub layer: usize,
+    pub(super) id: WidgetId,
+}
+
+impl<'a> Painter<'a> {
+    fn primitive_at<P: Primitive>(&mut self, primitive: P, region: UiRegion) {
+        let h = self.state.layers.write(
+            self.layer,
+            PrimitiveInst {
+                id: self.id,
+                primitive,
+                region,
+                mask_idx: self.mask,
+                move_idx: self.move_slot,
+            },
+        );
+        if self.mask != MaskIdx::NONE {
+            // TODO: I have no clue if this works at all :joy:
+            self.rsc.ui_mut().masks.push_ref(self.mask);
+        }
+        self.primitives.push(h);
+    }
+
+    /// Writes a primitive to be rendered
+    pub fn primitive<P: Primitive>(&mut self, primitive: P) {
+        self.primitive_at(primitive, self.region)
+    }
+
+    pub fn primitive_within<P: Primitive>(&mut self, primitive: P, region: UiRegion) {
+        self.primitive_at(primitive, region.within(&self.region));
+    }
+
+    pub fn set_mask(&mut self, region: UiRegion) {
+        assert!(self.mask == MaskIdx::NONE);
+        self.mask = self.rsc.ui_mut().masks.push(Mask {
+            region,
+            move_idx: self.move_slot,
+        });
+    }
+
+    /// Draws a widget within this widget's region, returning the size it
+    /// reported using.
+    pub fn widget<W: ?Sized>(&mut self, id: &StrongWidget<W>) -> Size {
+        self.widget_at(id, self.region)
+    }
+
+    /// Draws a widget somewhere within this one.
+    /// Useful for drawing child widgets in select areas.
+    pub fn widget_within<W: ?Sized>(&mut self, id: &StrongWidget<W>, region: UiRegion) -> Size {
+        self.widget_at(id, region.within(&self.region))
+    }
+
+    fn widget_at<W: ?Sized>(&mut self, id: &StrongWidget<W>, region: UiRegion) -> Size {
+        self.children.push(id.id());
+        // Passed directly rather than looked up from `self.active`: this
+        // widget's own `ActiveData` (which would carry its `move_slot`) is
+        // not inserted there until *after* its own `Widget::draw` returns,
+        // so a lookup here -- for a child drawn partway through that same
+        // call -- would always find nothing. `self.move_slot` is this
+        // widget's own slot, already known, and always correct regardless
+        // of insertion order. See `UiRenderState::move_parent_of`.
+        self.state.draw_inner(
+            self.layer,
+            id.id(),
+            region,
+            Some(self.id),
+            self.move_slot.idx() as u32,
+            self.mask,
+            None,
+            None,
+            self.rsc,
+        );
+        self.state
+            .active
+            .get(&id.id())
+            .map(|a| a.size)
+            .unwrap_or_default()
+    }
+
+    /// Move an already-drawn child from wherever it currently sits to
+    /// `region` (resolved against this widget's own region, matching
+    /// `widget_within`) without a second draw -- an O(1) offset write via
+    /// `UiRenderState::mov`. For a container that draws a child
+    /// provisionally to learn its size (e.g. `Aligned`) and then places it
+    /// for real. Only valid when the target keeps the child's drawn size;
+    /// if the shape actually changes, the normal `widget_within` dispatch
+    /// (which detects that from the stored region) does the right thing
+    /// instead.
+    pub fn reposition<W: ?Sized>(&mut self, id: &StrongWidget<W>, region: UiRegion) {
+        let region = region.within(&self.region);
+        self.state.reposition(id.id(), region, self.rsc);
+    }
+
+    /// 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<W: ?Sized>(
+        &mut self,
+        id: &StrongWidget<W>,
+        first: UiRegion,
+        second: impl FnOnce(Size) -> UiRegion,
+    ) -> Size {
+        let used = self.widget_within(id, first);
+        let region = second(used);
+        self.widget_within(id, region)
+    }
+
+    pub fn texture_within(&mut self, handle: &TextureHandle, region: UiRegion) {
+        self.textures.push(handle.clone());
+        self.write_image(handle.image_index(), region.within(&self.region));
+    }
+
+    pub fn texture(&mut self, handle: &TextureHandle) {
+        self.textures.push(handle.clone());
+        self.write_image(handle.image_index(), self.region);
+    }
+
+    pub fn texture_at(&mut self, handle: &TextureHandle, region: UiRegion) {
+        self.textures.push(handle.clone());
+        self.write_image(handle.image_index(), region);
+    }
+
+    /// A standalone image draws with its own bind group rather than sharing
+    /// the layer's one instanced draw, so it goes through
+    /// `Primitives::write_image` instead of `primitive_at`/`Primitive::vec`.
+    fn write_image(&mut self, texture_idx: u32, region: UiRegion) {
+        let h = self.state.layers.write_image(
+            self.layer,
+            self.id,
+            texture_idx,
+            region,
+            self.mask,
+            self.move_slot,
+        );
+        if self.mask != MaskIdx::NONE {
+            self.rsc.ui_mut().masks.push_ref(self.mask);
+        }
+        self.primitives.push(h);
+    }
+
+    pub fn render_text(
+        &mut self,
+        buffer: &mut TextBuffer,
+        attrs: &TextAttrs,
+        width: Option<f32>,
+    ) -> RenderedText {
+        let ui = self.rsc.ui_mut();
+        ui.text.render(buffer, attrs, width, &mut ui.textures)
+    }
+
+    /// Draw a laid-out string: one quad per glyph, all sampling the atlas.
+    ///
+    /// `origin` is where the text's top-left goes; every glyph is placed at an
+    /// absolute pixel offset from it, so re-drawing after a resize is this loop
+    /// and nothing else.
+    pub fn glyphs(&mut self, text: &RenderedText, origin: UiRegion) {
+        let flags_for = |is_color| {
+            if is_color {
+                GlyphPrimitive::IS_COLOR
+            } else {
+                0
+            }
+        };
+        for glyph in text.glyphs.iter() {
+            let mut region = origin;
+            region.x.end = region.x.start;
+            region.y.end = region.y.start;
+            let mut region = region.offset(UiVec2::abs(glyph.offset));
+            region.x.end = region.x.start + UiScalar::abs(glyph.entry.width as f32);
+            region.y.end = region.y.start + UiScalar::abs(glyph.entry.height as f32);
+            self.primitive_at(
+                GlyphPrimitive::new(
+                    glyph.entry.uv_min,
+                    glyph.entry.uv_max,
+                    glyph.entry.layer,
+                    text.color,
+                    flags_for(glyph.entry.is_color),
+                ),
+                region,
+            );
+        }
+    }
+
+    pub fn region(&self) -> UiRegion {
+        self.region
+    }
+
+    pub fn output_size(&self) -> Vec2 {
+        self.state.output_size
+    }
+
+    pub fn px_size(&mut self) -> Vec2 {
+        self.region.size().to_abs(self.state.output_size)
+    }
+
+    pub fn text_data(&mut self) -> &mut TextData {
+        &mut self.rsc.ui_mut().text
+    }
+
+    pub fn child_layer(&mut self) {
+        self.layer = self.state.layers.child(self.layer);
+    }
+
+    pub fn next_layer(&mut self) {
+        self.layer = self.state.layers.next(self.layer);
+    }
+
+    pub fn label(&self) -> &str {
+        &self.rsc.widgets().data(self.id).unwrap().label
+    }
+
+    pub fn id(&self) -> &WidgetId {
+        &self.id
+    }
+}

iris/core/src/ui/render_state.rs

@@ -0,0 +1,518 @@
+use crate::{
+    ActiveData, IdLike, MaskIdx, MoveIdx, Painter, PixelRegion, PrimitiveLayers, RegionAlign,
+    StrongWidget, UiRegion, UiRsc, UiVec2, WidgetId, Widgets,
+    render::MoveOffset,
+    util::{HashMap, HashSet, Id, Vec2},
+};
+
+pub struct UiRenderState {
+    pub active: HashMap<WidgetId, ActiveData>,
+    pub layers: PrimitiveLayers,
+    pub(super) output_size: Vec2,
+
+    old_root: Option<WidgetId>,
+    resized: bool,
+    draw_started: HashSet<WidgetId>,
+
+    /// `Widget::draw` calls and `Primitives::region_mut` rewrites since the
+    /// last `take_counters`. LAYOUT.md section 8's pass conditions are
+    /// stated in terms of these two: an unchanged frame must cost 0 of
+    /// each, and moving one widget must cost 0 draws and 0 rewrites
+    /// regardless of how many primitives are in its subtree.
+    draw_count: u64,
+    region_mut_count: u64,
+    mov_count: u64,
+}
+
+/// A move chain more than this deep would mean something else is wrong
+/// (an accidental cycle) -- see `resolve_move` in shader.wgsl, which walks
+/// the identical bound and must be kept in step with this constant.
+pub const MOVE_CHAIN_LIMIT: usize = 16;
+
+impl UiRenderState {
+    pub fn new() -> Self {
+        Self {
+            active: Default::default(),
+            layers: Default::default(),
+            output_size: Vec2::ZERO,
+            old_root: None,
+            resized: false,
+            draw_started: Default::default(),
+            draw_count: 0,
+            region_mut_count: 0,
+            mov_count: 0,
+        }
+    }
+
+    /// Reads and zeroes the (draws, region_mut rewrites, move_offsets
+    /// writes) counters -- call once per frame before `update()` to
+    /// measure exactly that frame, per LAYOUT.md section 8.
+    pub fn take_counters(&mut self) -> (u64, u64, u64) {
+        (
+            std::mem::take(&mut self.draw_count),
+            std::mem::take(&mut self.region_mut_count),
+            std::mem::take(&mut self.mov_count),
+        )
+    }
+
+    pub fn resize(&mut self, size: impl Into<Vec2>) {
+        self.output_size = size.into();
+        self.resized = true;
+    }
+
+    pub fn update<'a>(&mut self, root: impl Into<Option<&'a StrongWidget>>, rsc: &mut dyn UiRsc) {
+        // safety mechanism for memory leaks; might wanna return a result instead so user can
+        // decide whether to panic or not
+        if !rsc.widgets().waiting.is_empty() {
+            let widgets = rsc.widgets();
+            let len = widgets.waiting.len();
+            let all: Vec<_> = widgets
+                .waiting
+                .iter()
+                .map(|&w| format!("'{}' ({w:?})", widgets.label(w)))
+                .collect();
+            panic!(
+                "{len} widget(s) were never upgraded\n\
+                this is likely a memory leak; consider upgrading to strong if you plan on using it later\n\
+                weak widgets: {all:#?}"
+            );
+        }
+        let root = root.into();
+        if self.needs_redraw_all(root) {
+            self.redraw_all(root, rsc);
+            self.old_root = root.map(|r| r.id());
+            self.resized = false;
+        } else if rsc.widgets().has_updates() {
+            self.redraw_updates(rsc);
+        }
+    }
+
+    fn redraw_all(&mut self, root: Option<&StrongWidget>, rsc: &mut dyn UiRsc) {
+        self.clear(rsc);
+        // free all resources & cache
+        if let Some(id) = root {
+            self.draw_inner(
+                0,
+                id.id(),
+                UiRegion::FULL,
+                None,
+                MoveOffset::NONE_PARENT,
+                MaskIdx::NONE,
+                None,
+                None,
+                rsc,
+            );
+        }
+    }
+
+    /// The slot an *already-active* widget's `move_offsets` entry chains
+    /// to, read back from `self.active`. Only valid where the parent is
+    /// guaranteed to already be in `self.active` -- true for `redraw()`,
+    /// which targets a widget that was fully drawn on some earlier update,
+    /// but **not** for a widget being drawn as part of its own parent's
+    /// `Widget::draw` call: that parent's `ActiveData` is not inserted
+    /// until its `draw` returns (below), so a child drawn partway through
+    /// it would always read back "no parent" here. `Painter::widget_at`
+    /// avoids that trap by passing its own already-known `move_slot`
+    /// straight through instead of asking `self.active` to look it up.
+    fn move_parent_of(&self, parent: Option<WidgetId>) -> u32 {
+        parent
+            .and_then(|p| self.active.get(&p))
+            .map(|p| p.move_slot.idx() as u32)
+            .unwrap_or(MoveOffset::NONE_PARENT)
+    }
+
+    // TODO: should prolly make a DrawInfo struct or smth for everything other than rsc
+    #[allow(clippy::too_many_arguments)]
+    pub(super) fn draw_inner(
+        &mut self,
+        layer: usize,
+        id: WidgetId,
+        region: UiRegion,
+        parent: Option<WidgetId>,
+        parent_move_slot: u32,
+        mask: MaskIdx,
+        old_children: Option<Vec<WidgetId>>,
+        old_move_slot: Option<MoveIdx>,
+        rsc: &mut dyn UiRsc,
+    ) {
+        let mut old_children = old_children.unwrap_or_default();
+        let mut old_move_slot = old_move_slot;
+        if let Some(active) = self.active.get_mut(&id)
+            && !rsc.widgets().needs_redraw.contains(&id)
+        {
+            // check to see if we can skip drawing first
+            if active.region == region {
+                return;
+            } else if active.region.size() == region.size() {
+                // TODO: epsilon?
+                let from = active.region;
+                self.mov(id, from, region, rsc);
+                return;
+            } else if rsc
+                .widgets()
+                .get_dyn(id)
+                .map(|w| w.is_size_independent())
+                .unwrap_or(false)
+            {
+                // The offered region changed shape, but this widget's own
+                // drawn output does not depend on it (a fixed-size leaf) --
+                // rewrite its own primitives' regions in place (O(primitives
+                // owned directly by this widget, which for a leaf is O(1))
+                // instead of redrawing. See LAYOUT.md section 3.
+                let from = active.region;
+                for h in &active.primitives {
+                    let r = self.layers[h.layer].region_mut(h);
+                    *r = r.outside(&from).within(&region);
+                    self.region_mut_count += 1;
+                }
+                active.region = region;
+                return;
+            }
+            // if not, then maintain resize and track old children to remove unneeded
+            let active = self.remove(id, false, rsc).unwrap();
+            old_children = active.children;
+            old_move_slot = Some(active.move_slot);
+        }
+
+        // draw widget
+        self.draw_started.insert(id);
+
+        let move_slot = match old_move_slot {
+            // Reused across a real redraw of the same id: the fresh
+            // geometry this draw is about to write is placed at its
+            // correct absolute position by `region` itself, so any delta
+            // accumulated before this redraw is now stale and would
+            // double-offset it if left in place. The chain link (`parent`)
+            // is untouched -- the logical parent has not changed.
+            Some(slot) => {
+                let entry = rsc.ui_mut().move_offsets.get_mut(slot);
+                entry.delta = [0.0, 0.0];
+                slot
+            }
+            None => {
+                let slot = rsc
+                    .ui_mut()
+                    .move_offsets
+                    .push(MoveOffset::new([0.0, 0.0], parent_move_slot));
+                rsc.ui_mut().move_offsets.push_ref(slot);
+                if parent_move_slot != MoveOffset::NONE_PARENT {
+                    rsc.ui_mut()
+                        .move_offsets
+                        .push_ref(Id::preset(parent_move_slot));
+                }
+                slot
+            }
+        };
+
+        let mut painter = Painter {
+            state: self,
+            region,
+            mask,
+            move_slot,
+            layer,
+            id,
+            textures: Vec::new(),
+            primitives: Vec::new(),
+            children: Vec::new(),
+            rsc,
+        };
+
+        let mut widget = painter.rsc.widgets().get_dyn_dynamic(id);
+        painter.state.draw_count += 1;
+        let size = widget.draw(&mut painter);
+        drop(widget);
+
+        let Painter {
+            state: _,
+            rsc: _,
+            region,
+            mask,
+            move_slot,
+            textures,
+            primitives,
+            children,
+            layer,
+            id,
+        } = painter;
+
+        // add to active
+        let active = ActiveData {
+            id,
+            region,
+            parent,
+            textures,
+            primitives,
+            children,
+            mask,
+            layer,
+            size,
+            move_slot,
+        };
+
+        // remove old children that weren't kept
+        for c in &old_children {
+            if !active.children.contains(c) {
+                self.remove_rec(*c, rsc);
+            }
+        }
+
+        rsc.on_draw(&active);
+        self.active.insert(id, active);
+    }
+
+    /// O(1): write the delta for this widget's own slot in
+    /// `move_offsets`. No primitive is touched and there is no recursion --
+    /// every descendant's primitive references this slot transitively
+    /// through the parent chain the shader walks (`resolve_move`), so it
+    /// picks the new delta up for free. See LAYOUT.md section 2.
+    fn mov(&mut self, id: WidgetId, from: UiRegion, to: UiRegion, rsc: &mut dyn UiRsc) {
+        let Some(active) = self.active.get_mut(&id) else {
+            return;
+        };
+        let slot = active.move_slot;
+        active.region = to;
+        let from_px = from.top_left().to_abs(self.output_size);
+        let to_px = to.top_left().to_abs(self.output_size);
+        let delta = to_px - from_px;
+        let entry = rsc.ui_mut().move_offsets.get_mut(slot);
+        entry.delta[0] += delta.x;
+        entry.delta[1] += delta.y;
+        self.mov_count += 1;
+    }
+
+    /// Move an already-active widget to `to`. Used by `Painter::reposition`,
+    /// for a parent that drew a child provisionally (at the whole region it
+    /// was offered) and now knows where the child actually belongs.
+    ///
+    /// Unlike `mov` (called by `draw_inner`'s own dispatch, where the
+    /// *offered* region really did move and `active.region` already tracks
+    /// it), the child here was not offered a smaller region -- it was
+    /// offered everything and chose, on its own, to occupy only
+    /// `active.size` of it. By convention every widget in this crate that
+    /// does that anchors its own content at the top-left of whatever it
+    /// was given (`Rect`/`Image`/`Sized`/`MaxSize` -- see their `draw`
+    /// bodies), so that is where this assumes the child was actually
+    /// painted, not `active.region` itself (which is the *offered* box,
+    /// usually bigger). A nested `Aligned` whose own child is not top-left
+    /// anchored -- i.e. `Aligned` wrapping `Aligned` -- is the one shape
+    /// this does not cover; none of iris's widgets or examples build that
+    /// today. See LAYOUT.md's "Rejected, and why" / deviations for the
+    /// full reasoning.
+    ///
+    /// The delta is overwritten, not accumulated like `mov`'s: `from` is
+    /// recomputed fresh from `active.size`/`active.region` every call, so
+    /// repeating the same `reposition` (e.g. an unrelated redraw elsewhere
+    /// re-running this widget's parent without its own layout changing)
+    /// must land on the same answer, not drift further each time.
+    pub(super) fn reposition(&mut self, id: WidgetId, to: UiRegion, rsc: &mut dyn UiRsc) {
+        let Some(active) = self.active.get(&id) else {
+            return;
+        };
+        let from = active
+            .size
+            .to_uivec2()
+            .align(RegionAlign::TOP_LEFT)
+            .within(&active.region);
+        let slot = active.move_slot;
+        let from_px = from.top_left().to_abs(self.output_size);
+        let to_px = to.top_left().to_abs(self.output_size);
+        let delta = to_px - from_px;
+        let entry = rsc.ui_mut().move_offsets.get_mut(slot);
+        entry.delta = [delta.x, delta.y];
+        self.mov_count += 1;
+    }
+
+    /// NOTE: instance textures are cleared and self.textures freed
+    fn remove(&mut self, id: WidgetId, undraw: bool, rsc: &mut dyn UiRsc) -> Option<ActiveData> {
+        let mut active = self.active.remove(&id);
+        if let Some(active) = &mut active {
+            for h in &active.primitives {
+                let mask = self.layers.free(h);
+                if mask != MaskIdx::NONE {
+                    rsc.ui_mut().masks.remove(mask);
+                }
+            }
+            active.textures.clear();
+            rsc.ui_mut().textures.free();
+            if undraw {
+                // Permanent removal: retire this widget's own move slot
+                // (the self-ownership ref taken when it was allocated) and
+                // the up-link ref it held on its parent's slot -- read from
+                // the arena entry itself, not from `active.parent`, since
+                // the parent's own `ActiveData` may already be gone by the
+                // time a deep descendant is retired (see LAYOUT.md
+                // section 2's lifecycle note).
+                let parent_slot = rsc.ui_mut().move_offsets[active.move_slot.idx()].parent;
+                rsc.ui_mut().move_offsets.remove(active.move_slot);
+                if parent_slot != MoveOffset::NONE_PARENT {
+                    rsc.ui_mut().move_offsets.remove(Id::preset(parent_slot));
+                }
+                rsc.on_undraw(active);
+            }
+        }
+        active
+    }
+
+    fn remove_rec(&mut self, id: WidgetId, rsc: &mut dyn UiRsc) -> Option<ActiveData> {
+        let inst = self.remove(id, true, rsc);
+        if let Some(inst) = &inst {
+            for c in &inst.children {
+                self.remove_rec(*c, rsc);
+            }
+        }
+        inst
+    }
+
+    fn clear(&mut self, rsc: &mut dyn UiRsc) {
+        for (_, active) in self.active.drain() {
+            rsc.on_undraw(&active);
+        }
+        self.layers.clear();
+        rsc.widgets_mut().needs_redraw.clear();
+        rsc.free();
+    }
+
+    pub fn redraw_updates(&mut self, rsc: &mut dyn UiRsc) {
+        while let Some(&id) = rsc.widgets().needs_redraw.iter().next() {
+            self.redraw(id, rsc);
+        }
+        rsc.free();
+    }
+
+    pub fn root_changed<'a>(&self, root: impl Into<Option<&'a StrongWidget>>) -> bool {
+        root.into().map(|r| r.id()) != self.old_root
+    }
+
+    /// What `update` will redraw everything for. Named and shared with
+    /// `needs_redraw` rather than written out twice, because the two must
+    /// agree: `needs_redraw` is what asks for the frame that `update` would
+    /// draw, so a condition in one and not the other is a frame nobody
+    /// requests and a stale window. `resized` was missing from `needs_redraw`,
+    /// which is latent on Wayland only because winit asks for a redraw after a
+    /// resize by itself -- a resize changes neither the root nor any widget,
+    /// so nothing else here would have asked.
+    fn needs_redraw_all<'a>(&self, root: impl Into<Option<&'a StrongWidget>>) -> bool {
+        self.root_changed(root) || self.resized
+    }
+
+    pub fn needs_redraw<'a>(
+        &self,
+        root: impl Into<Option<&'a StrongWidget>>,
+        widgets: &Widgets,
+    ) -> bool {
+        self.needs_redraw_all(root) || widgets.has_updates()
+    }
+
+    pub fn active_widgets(&self) -> usize {
+        self.active.len()
+    }
+
+    pub fn debug(&self, widgets: &Widgets, label: &str) -> impl Iterator<Item = &ActiveData> {
+        self.active.iter().filter_map(move |(&id, inst)| {
+            let l = widgets.label(id);
+            if l == label { Some(inst) } else { None }
+        })
+    }
+
+    pub fn debug_layers(&self) {
+        for ((idx, depth), primitives) in self.layers.iter_depth() {
+            let indent = " ".repeat(depth * 2);
+            let len = primitives.instances().len();
+            print!("{indent}{idx}: {len} primitives");
+            if len >= 1 {
+                print!(" ({})", primitives.instances()[0].binding);
+            }
+            println!();
+        }
+    }
+
+    /// `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). See LAYOUT.md
+    /// section 2b.
+    pub fn resolved_region(&self, id: &impl IdLike, rsc: &dyn UiRsc) -> Option<UiRegion> {
+        let active = self.active.get(&id.id())?;
+        let delta = self.resolve_move_chain(active.move_slot, rsc);
+        Some(active.region.offset(UiVec2::abs(delta)))
+    }
+
+    /// The plain-Rust twin of `resolve_move` in shader.wgsl: sums the
+    /// pixel delta along the parent chain starting at `slot`. Both walks
+    /// share `MOVE_CHAIN_LIMIT` as their bound so the two cannot disagree
+    /// about where the chain ends.
+    fn resolve_move_chain(&self, mut slot: MoveIdx, rsc: &dyn UiRsc) -> Vec2 {
+        let offsets = &rsc.ui().move_offsets;
+        let mut delta = Vec2::ZERO;
+        for i in 0..MOVE_CHAIN_LIMIT {
+            let entry = &offsets[slot.idx()];
+            delta.x += entry.delta[0];
+            delta.y += entry.delta[1];
+            if entry.parent == MoveOffset::NONE_PARENT {
+                return delta;
+            }
+            slot = Id::preset(entry.parent);
+            debug_assert!(
+                i + 1 < MOVE_CHAIN_LIMIT,
+                "move offset chain exceeded MOVE_CHAIN_LIMIT; a widget's `parent` link is \
+                 probably cyclic"
+            );
+        }
+        delta
+    }
+
+    pub fn window_region(&self, id: &impl IdLike, rsc: &dyn UiRsc) -> Option<PixelRegion> {
+        let region = self.resolved_region(id, rsc)?;
+        Some(region.to_px(self.output_size))
+    }
+
+    /// redraws a widget that's currently active (drawn)
+    pub fn redraw(&mut self, id: WidgetId, rsc: &mut dyn UiRsc) {
+        rsc.widgets_mut().needs_redraw.remove(&id);
+        self.draw_started.remove(&id);
+        if self.draw_started.contains(&id) {
+            return;
+        }
+
+        let Some(active) = self.remove(id, false, rsc) else {
+            return;
+        };
+        let old_size = active.size;
+        let parent = active.parent;
+        // `old_move_slot` being `Some` below means the slot is reused in
+        // place rather than freshly parented, so this is only reached for
+        // logging/clarity's sake, never actually used to link a new slot.
+        let parent_move_slot = self.move_parent_of(parent);
+
+        self.draw_inner(
+            active.layer,
+            id,
+            active.region,
+            parent,
+            parent_move_slot,
+            active.mask,
+            Some(active.children),
+            Some(active.move_slot),
+            rsc,
+        );
+
+        // If this widget's own reported size changed, its parent's layout
+        // (which placed it using the old size) is now stale and needs to
+        // relay out too. Checked after the real draw, not before it --
+        // there is no query left that answers "what size would this be"
+        // without actually drawing (LAYOUT.md section 5).
+        if let Some(pid) = parent {
+            let new_size = self.active.get(&id).map(|a| a.size);
+            if new_size != Some(old_size) {
+                self.redraw(pid, rsc);
+            }
+        }
+    }
+}
+
+impl Default for UiRenderState {
+    fn default() -> Self {
+        Self::new()
+    }
+}

iris/core/src/util/arena.rs

@@ -0,0 +1,118 @@
+use std::ops::Deref;
+
+use crate::util::{Id, IdNum, IdTracker};
+
+pub struct Arena<T, I> {
+    data: Vec<T>,
+    tracker: IdTracker<I>,
+}
+
+impl<T, I: IdNum> Arena<T, I> {
+    pub fn new() -> Self {
+        Self {
+            data: Vec::new(),
+            tracker: IdTracker::default(),
+        }
+    }
+
+    pub fn push(&mut self, value: T) -> Id<I> {
+        let id = self.tracker.next();
+        let i = id.idx();
+        if i == self.data.len() {
+            self.data.push(value);
+        } else {
+            self.data[i] = value;
+        }
+        id
+    }
+
+    pub fn remove(&mut self, id: Id<I>) -> T
+    where
+        T: Copy,
+    {
+        let i = id.idx();
+        self.tracker.free(id);
+        self.data[i]
+    }
+}
+
+impl<T, I: IdNum> Default for Arena<T, I> {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+pub struct TrackedArena<T, I> {
+    inner: Arena<T, I>,
+    refs: Vec<u32>,
+    pub changed: bool,
+}
+
+impl<T, I: IdNum> TrackedArena<T, I> {
+    pub fn new() -> Self {
+        Self {
+            inner: Arena::default(),
+            refs: Vec::new(),
+            changed: true,
+        }
+    }
+
+    pub fn push(&mut self, value: T) -> Id<I> {
+        self.changed = true;
+        let id = self.inner.push(value);
+        let i = id.idx();
+        if i == self.refs.len() {
+            self.refs.push(0);
+        }
+        id
+    }
+
+    pub fn push_ref(&mut self, i: Id<I>) {
+        self.refs[i.idx()] += 1;
+    }
+
+    /// Mutable access to an existing entry, for the rare case (the move
+    /// offset chain) where an already-allocated slot is updated in place
+    /// rather than replaced. Marks the arena changed so the GPU copy is
+    /// re-uploaded.
+    pub fn get_mut(&mut self, id: Id<I>) -> &mut T {
+        self.changed = true;
+        &mut self.inner.data[id.idx()]
+    }
+
+    pub fn remove(&mut self, id: Id<I>) -> T
+    where
+        T: Copy,
+    {
+        let i = id.idx();
+        self.refs[i] -= 1;
+        if self.refs[i] == 0 {
+            self.changed = true;
+            self.inner.remove(id)
+        } else {
+            self[i]
+        }
+    }
+}
+
+impl<T, I: IdNum> Default for TrackedArena<T, I> {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+impl<T, I> Deref for TrackedArena<T, I> {
+    type Target = Vec<T>;
+
+    fn deref(&self) -> &Self::Target {
+        &self.inner.data
+    }
+}
+
+impl<T, I> Deref for Arena<T, I> {
+    type Target = Vec<T>;
+
+    fn deref(&self) -> &Self::Target {
+        &self.data
+    }
+}

iris/core/src/util/borrow.rs

@@ -0,0 +1,35 @@
+use std::ops::{Deref, DerefMut};
+
+pub struct DynBorrower<'a, T: ?Sized> {
+    data: &'a mut T,
+    borrowed: &'a mut bool,
+}
+
+impl<'a, T: ?Sized> DynBorrower<'a, T> {
+    pub fn new(data: &'a mut T, borrowed: &'a mut bool) -> Self {
+        if *borrowed {
+            panic!("tried to mutably borrow the same thing twice");
+        }
+        Self { data, borrowed }
+    }
+}
+
+impl<T: ?Sized> Drop for DynBorrower<'_, T> {
+    fn drop(&mut self) {
+        *self.borrowed = false;
+    }
+}
+
+impl<T: ?Sized> Deref for DynBorrower<'_, T> {
+    type Target = T;
+
+    fn deref(&self) -> &Self::Target {
+        self.data
+    }
+}
+
+impl<T: ?Sized> DerefMut for DynBorrower<'_, T> {
+    fn deref_mut(&mut self) -> &mut Self::Target {
+        self.data
+    }
+}

iris/core/src/util/change.rs

@@ -0,0 +1,30 @@
+use std::ops::{Deref, DerefMut};
+
+pub struct MutDetect<T> {
+    inner: T,
+    pub changed: bool,
+}
+
+impl<T> Deref for MutDetect<T> {
+    type Target = T;
+
+    fn deref(&self) -> &Self::Target {
+        &self.inner
+    }
+}
+
+impl<T> DerefMut for MutDetect<T> {
+    fn deref_mut(&mut self) -> &mut Self::Target {
+        self.changed = true;
+        &mut self.inner
+    }
+}
+
+impl<T> From<T> for MutDetect<T> {
+    fn from(inner: T) -> Self {
+        MutDetect {
+            inner,
+            changed: true,
+        }
+    }
+}

iris/core/src/util/id.rs

@@ -0,0 +1,87 @@
+#[repr(C)]
+#[derive(Eq, Hash, PartialEq, Debug, Clone, Copy, bytemuck::Zeroable)]
+pub struct Id<I = u64>(I);
+
+unsafe impl<I: Copy + bytemuck::Zeroable + 'static> bytemuck::Pod for Id<I> {}
+
+pub struct IdTracker<I = u64> {
+    free: Vec<Id<I>>,
+    cur: Id<I>,
+}
+
+impl<I: IdNum> IdTracker<I> {
+    #[allow(clippy::should_implement_trait)]
+    pub fn next(&mut self) -> Id<I> {
+        if let Some(id) = self.free.pop() {
+            return id;
+        }
+        let next = self.cur.next();
+        std::mem::replace(&mut self.cur, next)
+    }
+
+    #[allow(dead_code)]
+    pub fn free(&mut self, id: Id<I>) {
+        self.free.push(id);
+    }
+}
+
+impl<I: IdNum> Id<I> {
+    #[allow(dead_code)]
+    /// for debug purposes; should this be exposed?
+    /// generally you want to use labels with widgets
+    pub(crate) fn raw(id: I) -> Self {
+        Self(id)
+    }
+    pub fn idx(&self) -> usize {
+        self.0.idx()
+    }
+    pub fn next(&self) -> Id<I> {
+        Self(self.0.next())
+    }
+    pub const fn preset(value: I) -> Self {
+        Self(value)
+    }
+}
+
+impl<I: IdNum> Default for IdTracker<I> {
+    fn default() -> Self {
+        Self {
+            free: Vec::new(),
+            cur: Id(I::first()),
+        }
+    }
+}
+
+pub trait IdNum {
+    fn first() -> Self;
+    fn next(&self) -> Self;
+    fn idx(&self) -> usize;
+}
+
+impl IdNum for u64 {
+    fn first() -> Self {
+        0
+    }
+
+    fn next(&self) -> Self {
+        self + 1
+    }
+
+    fn idx(&self) -> usize {
+        *self as usize
+    }
+}
+
+impl IdNum for u32 {
+    fn first() -> Self {
+        0
+    }
+
+    fn next(&self) -> Self {
+        self + 1
+    }
+
+    fn idx(&self) -> usize {
+        *self as usize
+    }
+}

iris/core/src/util/math.rs

@@ -0,0 +1,88 @@
+use std::ops::*;
+
+pub const trait LerpUtil {
+    fn lerp(self, from: Self, to: Self) -> Self;
+    fn lerp_inv(self, from: Self, to: Self) -> Self;
+}
+
+pub const trait DivOr {
+    fn div_or(self, rhs: Self, other: Self) -> Self;
+}
+
+const impl DivOr for f32 {
+    fn div_or(self, rhs: Self, other: Self) -> Self {
+        let res = self / rhs;
+        if res.is_nan() { other } else { res }
+    }
+}
+
+const impl<
+    T: const Add<Output = T> + const Sub<Output = T> + const Mul<Output = T> + const DivOr + Copy,
+> LerpUtil for T
+{
+    /// linear interpolation
+    /// from * (1.0 - self) + to * self
+    fn lerp(self, from: Self, to: Self) -> Self {
+        from + (to - from) * self
+    }
+    /// inverse of lerp
+    fn lerp_inv(self, from: Self, to: Self) -> Self {
+        (self - from).div_or(to - from, from)
+    }
+}
+
+macro_rules! impl_op {
+    ($T:ident $op:ident $fn:ident $opa:ident $fna:ident; $($field:ident)*) => {
+        #[allow(non_snake_case)]
+        mod ${concat($T, _op_, $fn, _impl)} {
+            use super::*;
+            #[allow(unused_imports)]
+            use std::ops::*;
+            const impl $op for $T {
+                type Output = Self;
+
+                fn $fn(self, rhs: Self) -> Self::Output {
+                    Self {
+                        $($field: self.$field.$fn(rhs.$field),)*
+                    }
+                }
+            }
+            const impl $opa for $T {
+                fn $fna(&mut self, rhs: Self) {
+                    *self = self.$fn(rhs);
+                }
+            }
+            const impl $op<f32> for $T {
+                type Output = Self;
+
+                fn $fn(self, rhs: f32) -> Self::Output {
+                    Self {
+                        $($field: self.$field.$fn(rhs),)*
+                    }
+                }
+            }
+            const impl $op<$T> for f32 {
+                type Output = $T;
+
+                fn $fn(self, rhs: $T) -> Self::Output {
+                    $T {
+                        $($field: self.$fn(rhs.$field),)*
+                    }
+                }
+            }
+            const impl $opa<f32> for $T {
+                fn $fna(&mut self, rhs: f32) {
+                    *self = self.$fn(rhs);
+                }
+            }
+        }
+    };
+    ($T:ident $op:ident $fn:ident; $($field:ident)*) => {
+        impl_op!($T $op $fn ${concat($op,Assign)} ${concat($fn,_assign)}; $($field)*);
+    };
+    (impl $op:ident for $T:ident: $fn:ident $($field:ident)*) => {
+        impl_op!($T $op $fn ${concat($op,Assign)} ${concat($fn,_assign)}; $($field)*);
+    };
+}
+
+pub(crate) use impl_op;

iris/core/src/util/mod.rs

@@ -0,0 +1,24 @@
+mod arena;
+mod borrow;
+mod change;
+mod id;
+mod math;
+mod refcount;
+mod slot;
+mod trust;
+mod typemap;
+mod vec2;
+
+pub use arena::*;
+pub use borrow::*;
+pub use change::*;
+pub use id::*;
+pub use math::*;
+pub use refcount::*;
+pub use slot::*;
+pub use trust::*;
+pub use typemap::*;
+pub use vec2::*;
+
+pub type HashMap<K, V> = fxhash::FxHashMap<K, V>;
+pub type HashSet<K> = fxhash::FxHashSet<K>;

iris/core/src/util/refcount.rs

@@ -0,0 +1,36 @@
+use std::sync::{
+    Arc,
+    atomic::{AtomicU32, Ordering},
+};
+
+#[derive(Debug)]
+pub struct RefCounter(Arc<AtomicU32>);
+
+impl RefCounter {
+    pub fn new() -> Self {
+        Self(Arc::new(0.into()))
+    }
+    pub fn refs(&self) -> u32 {
+        self.0.load(Ordering::Acquire)
+    }
+    pub fn drop(&mut self) -> bool {
+        let refs = self.0.fetch_sub(1, Ordering::Release);
+        refs == 0
+    }
+    pub fn quiet_clone(&self) -> Self {
+        Self(self.0.clone())
+    }
+}
+
+impl Default for RefCounter {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+impl Clone for RefCounter {
+    fn clone(&self) -> Self {
+        self.0.fetch_add(1, Ordering::Release);
+        Self(self.0.clone())
+    }
+}

iris/core/src/util/slot.rs

@@ -0,0 +1,69 @@
+#[derive(Clone, Copy, Debug, PartialEq, Eq, Hash)]
+pub struct SlotId {
+    idx: u32,
+    genr: u32,
+}
+
+pub struct SlotVec<T> {
+    data: Vec<(u32, Option<T>)>,
+    free: Vec<u32>,
+}
+
+impl<T> SlotVec<T> {
+    pub fn new() -> Self {
+        Self {
+            data: Default::default(),
+            free: Default::default(),
+        }
+    }
+
+    pub fn add(&mut self, x: T) -> SlotId {
+        if let Some(idx) = self.free.pop() {
+            let (genr, data) = &mut self.data[idx as usize];
+            *data = Some(x);
+            SlotId { idx, genr: *genr }
+        } else {
+            let idx = self.data.len() as u32;
+            let genr = 0;
+            self.data.push((genr, Some(x)));
+            SlotId { idx, genr }
+        }
+    }
+
+    pub fn free(&mut self, id: SlotId) {
+        let (genr, data) = &mut self.data[id.idx as usize];
+        *genr += 1;
+        *data = None;
+        self.free.push(id.idx);
+    }
+
+    pub fn get(&self, id: SlotId) -> Option<&T> {
+        let slot = &self.data[id.idx as usize];
+        if slot.0 != id.genr {
+            return None;
+        }
+        slot.1.as_ref()
+    }
+
+    pub fn get_mut(&mut self, id: SlotId) -> Option<&mut T> {
+        let slot = &mut self.data[id.idx as usize];
+        if slot.0 != id.genr {
+            return None;
+        }
+        slot.1.as_mut()
+    }
+
+    pub fn len(&self) -> usize {
+        self.data.len() - self.free.len()
+    }
+
+    pub fn is_empty(&self) -> bool {
+        self.len() == 0
+    }
+}
+
+impl<T> Default for SlotVec<T> {
+    fn default() -> Self {
+        Self::new()
+    }
+}

iris/core/src/util/trust.rs

@@ -0,0 +1,17 @@
+#[allow(clippy::missing_safety_doc)]
+pub unsafe fn forget_ref<'a, T>(x: &T) -> &'a T {
+    unsafe { std::mem::transmute::<&T, &T>(x) }
+}
+
+#[allow(clippy::missing_safety_doc)]
+pub unsafe fn forget_mut<'a, T>(x: &mut T) -> &'a mut T {
+    unsafe { std::mem::transmute::<&mut T, &mut T>(x) }
+}
+
+#[allow(clippy::mut_from_ref, clippy::missing_safety_doc)]
+pub unsafe fn to_mut<T>(x: &T) -> &mut T {
+    #[allow(mutable_transmutes)]
+    unsafe {
+        std::mem::transmute::<&T, &mut T>(x)
+    }
+}

iris/core/src/util/typemap.rs

@@ -0,0 +1,56 @@
+use crate::util::HashMap;
+use std::{
+    any::TypeId,
+    marker::Unsize,
+    ops::{Deref, DerefMut},
+};
+
+pub struct TypeMap<Trait: ?Sized> {
+    map: HashMap<TypeId, Box<Trait>>,
+}
+
+impl<Trait: ?Sized> TypeMap<Trait> {
+    pub fn set_type<T: Unsize<Trait> + 'static>(&mut self, val: T) {
+        self.map
+            .insert(TypeId::of::<T>(), Box::new(val) as Box<Trait>);
+    }
+
+    pub fn type_mut<T: Unsize<Trait> + 'static>(&mut self) -> Option<&mut T> {
+        Some(Self::convert_mut(self.map.get_mut(&TypeId::of::<T>())?))
+    }
+
+    pub fn type_or_default<T: Default + Unsize<Trait> + 'static>(&mut self) -> &mut T {
+        Self::convert_mut(
+            self.map
+                .entry(TypeId::of::<T>())
+                .or_insert(Box::new(T::default()) as Box<Trait>),
+        )
+    }
+
+    fn convert_mut<T: Unsize<Trait>>(entry: &mut Box<Trait>) -> &mut T {
+        // allegedly this is just what Any does...
+        unsafe { &mut *(entry.as_mut() as *mut Trait as *mut T) }
+    }
+}
+
+impl<T: ?Sized> Deref for TypeMap<T> {
+    type Target = HashMap<TypeId, Box<T>>;
+
+    fn deref(&self) -> &Self::Target {
+        &self.map
+    }
+}
+
+impl<T: ?Sized> DerefMut for TypeMap<T> {
+    fn deref_mut(&mut self) -> &mut Self::Target {
+        &mut self.map
+    }
+}
+
+impl<T: ?Sized> Default for TypeMap<T> {
+    fn default() -> Self {
+        Self {
+            map: Default::default(),
+        }
+    }
+}

iris/core/src/util/vec2.rs

@@ -0,0 +1,99 @@
+use crate::util::{DivOr, impl_op};
+use std::{hash::Hash, ops::*};
+
+#[repr(C)]
+#[derive(Clone, Copy, PartialEq, Default, bytemuck::Pod, bytemuck::Zeroable)]
+pub struct Vec2 {
+    pub x: f32,
+    pub y: f32,
+}
+
+impl Eq for Vec2 {}
+
+impl Hash for Vec2 {
+    fn hash<H: std::hash::Hasher>(&self, state: &mut H) {
+        state.write_u32(self.x.to_bits());
+        state.write_u32(self.y.to_bits());
+    }
+}
+
+impl Vec2 {
+    pub const ZERO: Self = Self::new(0.0, 0.0);
+    pub const ONE: Self = Self::new(1.0, 1.0);
+
+    pub const fn new(x: f32, y: f32) -> Self {
+        Self { x, y }
+    }
+
+    pub const fn round(self) -> Self {
+        Self {
+            x: self.x.round(),
+            y: self.y.round(),
+        }
+    }
+
+    pub const fn floor(self) -> Self {
+        Self {
+            x: self.x.floor(),
+            y: self.y.floor(),
+        }
+    }
+
+    pub const fn ceil(self) -> Self {
+        Self {
+            x: self.x.ceil(),
+            y: self.y.ceil(),
+        }
+    }
+
+    pub const fn tuple(&self) -> (f32, f32) {
+        (self.x, self.y)
+    }
+
+    pub const fn with_x(mut self, x: f32) -> Self {
+        self.x = x;
+        self
+    }
+
+    pub const fn with_y(mut self, y: f32) -> Self {
+        self.y = y;
+        self
+    }
+}
+
+// this version looks kinda cool... is it more readable? more annoying to copy and change though
+impl_op!(impl Add for Vec2: add x y);
+impl_op!(Vec2 Sub sub; x y);
+impl_op!(Vec2 Mul mul; x y);
+impl_op!(Vec2 Div div; x y);
+
+const impl DivOr for Vec2 {
+    fn div_or(self, rhs: Self, other: Self) -> Self {
+        Self {
+            x: self.x.div_or(rhs.x, other.x),
+            y: self.y.div_or(rhs.y, other.y),
+        }
+    }
+}
+
+impl Neg for Vec2 {
+    type Output = Self;
+
+    fn neg(mut self) -> Self::Output {
+        self.x = -self.x;
+        self.y = -self.y;
+        self
+    }
+}
+
+impl std::fmt::Debug for Vec2 {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(f, "({}, {})", self.x, self.y)
+    }
+}
+
+impl std::fmt::Display for Vec2 {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(f, "({}, {})", self.x, self.y)
+    }
+}

iris/core/src/widget/data.rs

@@ -0,0 +1,22 @@
+use crate::Widget;
+
+pub struct WidgetData {
+    pub widget: Box<dyn Widget>,
+    pub label: String,
+    /// dynamic borrow checking
+    pub borrowed: bool,
+}
+
+impl WidgetData {
+    pub fn new<W: Widget>(widget: W) -> Self {
+        let mut label = std::any::type_name::<W>().to_string();
+        if let (Some(first), Some(last)) = (label.find(":"), label.rfind(":")) {
+            label = label.split_at(first).0.to_string() + "::" + label.split_at(last + 1).1;
+        }
+        Self {
+            widget: Box::new(widget),
+            label,
+            borrowed: false,
+        }
+    }
+}

iris/core/src/widget/handle.rs

@@ -0,0 +1,164 @@
+use std::{marker::Unsize, ops::CoerceUnsized, sync::mpsc::Sender};
+
+use crate::{
+    UiRsc, Widget,
+    util::{RefCounter, SlotId},
+};
+
+pub type WidgetId = SlotId;
+
+/// An identifier for a widget that can index a UI or event ctx to get it.
+/// This is a strong handle that does not impl Clone, and when it is dropped,
+/// a signal is sent to the owning UI to clean up the resources.
+///
+/// TODO: ergonomic clones when they get put in rust-analyzer & don't cause ICEs?
+pub struct StrongWidget<W: ?Sized = dyn Widget> {
+    pub(super) id: WidgetId,
+    counter: RefCounter,
+    send: Sender<WidgetId>,
+    ty: *const W,
+}
+
+/// A weak handle to a widget.
+/// Will not keep it alive, but can still be used for indexing like WidgetHandle.
+pub struct WeakWidget<W: ?Sized = dyn Widget> {
+    pub(super) id: WidgetId,
+    #[allow(unused)]
+    ty: *const W,
+}
+
+impl<W: Widget + ?Sized + Unsize<dyn Widget>> StrongWidget<W> {
+    pub fn any(self) -> StrongWidget<dyn Widget> {
+        self
+    }
+}
+
+impl<W: ?Sized> StrongWidget<W> {
+    pub(crate) fn new(id: WidgetId, send: Sender<WidgetId>) -> Self {
+        Self {
+            id,
+            counter: RefCounter::new(),
+            send,
+            ty: null_ptr(),
+        }
+    }
+
+    pub fn id(&self) -> WidgetId {
+        self.id
+    }
+
+    pub fn refs(&self) -> u32 {
+        self.counter.refs()
+    }
+
+    pub fn weak(&self) -> WeakWidget<W> {
+        let Self { ty, id, .. } = *self;
+        WeakWidget { ty, id }
+    }
+}
+
+impl<W: ?Sized> WeakWidget<W> {
+    pub(crate) fn new(id: WidgetId) -> Self {
+        Self { id, ty: null_ptr() }
+    }
+
+    pub fn id(&self) -> WidgetId {
+        self.id
+    }
+
+    #[track_caller]
+    pub fn upgrade(self, ui: &mut impl UiRsc) -> StrongWidget<W> {
+        ui.widgets_mut().upgrade(self)
+    }
+}
+
+impl<W: ?Sized> Drop for StrongWidget<W> {
+    fn drop(&mut self) {
+        if self.counter.drop() {
+            let _ = self.send.send(self.id);
+        }
+    }
+}
+
+pub trait WidgetIdFn<Rsc, W: ?Sized = dyn Widget>: FnOnce(&mut Rsc) -> WeakWidget<W> {}
+impl<Rsc, W: ?Sized, F: FnOnce(&mut Rsc) -> WeakWidget<W>> WidgetIdFn<Rsc, W> for F {}
+
+pub trait IdLike {
+    type Widget: ?Sized;
+    fn id(&self) -> WidgetId;
+}
+
+impl<W: ?Sized> IdLike for &StrongWidget<W> {
+    type Widget = W;
+    fn id(&self) -> WidgetId {
+        self.id
+    }
+}
+
+impl<W: ?Sized> IdLike for StrongWidget<W> {
+    type Widget = W;
+    fn id(&self) -> WidgetId {
+        self.id
+    }
+}
+
+impl<W: ?Sized> IdLike for WeakWidget<W> {
+    type Widget = W;
+    fn id(&self) -> WidgetId {
+        self.id
+    }
+}
+
+impl IdLike for WidgetId {
+    type Widget = dyn Widget;
+    fn id(&self) -> WidgetId {
+        *self
+    }
+}
+
+impl<T: ?Sized + Unsize<U>, U: ?Sized> CoerceUnsized<StrongWidget<U>> for StrongWidget<T> {}
+impl<T: ?Sized + Unsize<U>, U: ?Sized> CoerceUnsized<WeakWidget<U>> for WeakWidget<T> {}
+
+impl<W: ?Sized> Clone for WeakWidget<W> {
+    fn clone(&self) -> Self {
+        *self
+    }
+}
+impl<W: ?Sized> Copy for WeakWidget<W> {}
+impl<W: ?Sized> PartialEq for WeakWidget<W> {
+    fn eq(&self, other: &Self) -> bool {
+        self.id == other.id
+    }
+}
+
+impl<W> PartialEq for StrongWidget<W> {
+    fn eq(&self, other: &Self) -> bool {
+        self.id == other.id
+    }
+}
+
+impl<W> std::fmt::Debug for StrongWidget<W> {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        self.id.fmt(f)
+    }
+}
+
+impl<'a, W: Widget + 'a, State: UiRsc> FnOnce<(&'a mut State,)> for WeakWidget<W> {
+    type Output = &'a mut W;
+
+    extern "rust-call" fn call_once(self, args: (&'a mut State,)) -> Self::Output {
+        &mut args.0.widgets_mut()[self]
+    }
+}
+
+fn null_ptr<W: ?Sized>() -> *const W {
+    if size_of::<&W>() == size_of::<*const dyn Widget>() {
+        let w: *const dyn Widget = &();
+        unsafe { std::mem::transmute_copy(&w) }
+    } else {
+        unsafe { std::mem::transmute_copy(&[0usize; 1]) }
+    }
+}
+
+unsafe impl<W: ?Sized> Send for WeakWidget<W> {}
+unsafe impl<W: ?Sized> Sync for WeakWidget<W> {}

iris/core/src/widget/like.rs

@@ -0,0 +1,75 @@
+use crate::UiRsc;
+
+use super::*;
+use std::marker::Unsize;
+
+pub trait WidgetLike<Rsc: UiRsc, Tag>: Sized {
+    type Widget: Widget + ?Sized + Unsize<dyn Widget>;
+
+    fn add(self, rsc: &mut Rsc) -> WeakWidget<Self::Widget>;
+
+    fn add_strong(self, rsc: &mut Rsc) -> StrongWidget<Self::Widget> {
+        self.add(rsc).upgrade(rsc)
+    }
+
+    fn with_id<W2>(
+        self,
+        f: impl FnOnce(&mut Rsc, WeakWidget<Self::Widget>) -> WeakWidget<W2>,
+    ) -> impl WidgetIdFn<Rsc, W2> {
+        move |state| {
+            let id = self.add(state);
+            f(state, id)
+        }
+    }
+
+    fn set_root(self, rsc: &mut Rsc, root: &mut impl HasRoot) {
+        let id = self.add_strong(rsc);
+        root.set_root(id);
+    }
+}
+
+pub trait HasRoot {
+    fn set_root(&mut self, root: StrongWidget);
+}
+
+pub trait WidgetArrLike<Rsc, const LEN: usize, Tag> {
+    #[track_caller]
+    fn add(self, state: &mut Rsc) -> WidgetArr<LEN>;
+}
+
+impl<Rsc, const LEN: usize> WidgetArrLike<Rsc, LEN, ArrTag> for WidgetArr<LEN> {
+    fn add(self, _: &mut Rsc) -> WidgetArr<LEN> {
+        self
+    }
+}
+
+// variadic generics please save us
+macro_rules! impl_widget_arr {
+    ($n:expr;$($W:ident)*) => {
+        impl_widget_arr!($n;$($W)*;$(${concat($W,Tag)})*);
+    };
+    ($n:expr;$($W:ident)*;$($Tag:ident)*) => {
+        impl<Rsc: UiRsc, $($W: WidgetLike<Rsc, $Tag>,$Tag,)*> WidgetArrLike<Rsc, $n, ($($Tag,)*)> for ($($W,)*) {
+            fn add(self, rsc: &mut Rsc) -> WidgetArr<$n> {
+                #[allow(non_snake_case)]
+                let ($($W,)*) = self;
+                WidgetArr::new(
+                    [$($W.add(rsc).upgrade(rsc),)*],
+                )
+            }
+        }
+    };
+}
+
+impl_widget_arr!(1;A);
+impl_widget_arr!(2;A B);
+impl_widget_arr!(3;A B C);
+impl_widget_arr!(4;A B C D);
+impl_widget_arr!(5;A B C D E);
+impl_widget_arr!(6;A B C D E F);
+impl_widget_arr!(7;A B C D E F G);
+impl_widget_arr!(8;A B C D E F G H);
+impl_widget_arr!(9;A B C D E F G H I);
+impl_widget_arr!(10;A B C D E F G H I J);
+impl_widget_arr!(11;A B C D E F G H I J K);
+impl_widget_arr!(12;A B C D E F G H I J K L);

iris/core/src/widget/mod.rs

@@ -0,0 +1,84 @@
+use crate::{Painter, Size};
+use std::any::Any;
+
+mod data;
+mod handle;
+mod like;
+mod tag;
+mod view;
+mod widgets;
+
+pub use data::*;
+pub use handle::*;
+pub use like::*;
+pub use tag::*;
+pub use view::*;
+pub use widgets::*;
+
+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. See LAYOUT.md.
+    fn is_size_independent(&self) -> bool {
+        false
+    }
+}
+
+impl Widget for () {
+    fn draw(&mut self, _: &mut Painter) -> Size {
+        Size::ZERO
+    }
+
+    fn is_size_independent(&self) -> bool {
+        true
+    }
+}
+
+impl dyn Widget {
+    pub fn as_any(&self) -> &dyn Any {
+        self
+    }
+
+    pub fn as_any_mut(&mut self) -> &mut dyn Any {
+        self
+    }
+}
+
+/// A function that returns a widget given a UI.
+/// Useful for defining trait functions on widgets that create a parent widget so that the children
+/// don't need to be IDs yet
+pub trait WidgetFn<State, W: Widget + ?Sized>: FnOnce(&mut State) -> W {}
+impl<State, W: Widget + ?Sized, F: FnOnce(&mut State) -> W> WidgetFn<State, W> for F {}
+
+pub struct WidgetArr<const LEN: usize> {
+    pub arr: [StrongWidget; LEN],
+}
+
+impl<const LEN: usize> WidgetArr<LEN> {
+    pub fn new(arr: [StrongWidget; LEN]) -> Self {
+        Self { arr }
+    }
+}
+
+pub trait WidgetOption<State> {
+    fn get(self, state: &mut State) -> Option<StrongWidget>;
+}
+
+impl<State> WidgetOption<State> for () {
+    fn get(self, _: &mut State) -> Option<StrongWidget> {
+        None
+    }
+}
+
+impl<State, F: FnOnce(&mut State) -> Option<StrongWidget>> WidgetOption<State> for F {
+    fn get(self, state: &mut State) -> Option<StrongWidget> {
+        self(state)
+    }
+}

iris/core/src/widget/tag.rs

@@ -0,0 +1,64 @@
+use super::*;
+use crate::UiRsc;
+use std::marker::Unsize;
+
+pub struct WidgetTag;
+impl<Rsc: UiRsc, W: Widget> WidgetLike<Rsc, WidgetTag> for W {
+    type Widget = W;
+    fn add(self, rsc: &mut Rsc) -> WeakWidget<W> {
+        let w = rsc.ui_mut().widgets.add_weak(self);
+        rsc.on_add(w);
+        w
+    }
+}
+
+pub struct FnTag;
+impl<Rsc: UiRsc, W: Widget, F: FnOnce(&mut Rsc) -> W> WidgetLike<Rsc, FnTag> for F {
+    type Widget = W;
+    fn add(self, rsc: &mut Rsc) -> WeakWidget<W> {
+        self(rsc).add(rsc)
+    }
+}
+
+pub trait WidgetFnTrait<Rsc> {
+    type Widget: Widget;
+    fn run(self, rsc: &mut Rsc) -> Self::Widget;
+}
+pub struct FnTraitTag;
+impl<Rsc: UiRsc, T: WidgetFnTrait<Rsc>> WidgetLike<Rsc, FnTraitTag> for T {
+    type Widget = T::Widget;
+    #[track_caller]
+    fn add(self, rsc: &mut Rsc) -> WeakWidget<T::Widget> {
+        self.run(rsc).add(rsc)
+    }
+}
+
+pub struct RefTag;
+impl<Rsc: UiRsc, W: ?Sized + Widget + Unsize<dyn Widget>> WidgetLike<Rsc, RefTag>
+    for WeakWidget<W>
+{
+    type Widget = W;
+    fn add(self, _: &mut Rsc) -> WeakWidget<W> {
+        self
+    }
+}
+
+pub struct RefFnTag;
+impl<Rsc: UiRsc, W: ?Sized + Widget + Unsize<dyn Widget>, F: FnOnce(&mut Rsc) -> WeakWidget<W>>
+    WidgetLike<Rsc, RefFnTag> for F
+{
+    type Widget = W;
+    fn add(self, rsc: &mut Rsc) -> WeakWidget<W> {
+        self(rsc)
+    }
+}
+
+pub struct ViewTag;
+impl<Rsc: UiRsc, V: WidgetView> WidgetLike<Rsc, ViewTag> for V {
+    type Widget = V::Widget;
+    fn add(self, _: &mut Rsc) -> WeakWidget<Self::Widget> {
+        self.root()
+    }
+}
+
+pub struct ArrTag;

iris/core/src/widget/view.rs

@@ -0,0 +1,24 @@
+use std::marker::Unsize;
+
+use crate::{IdLike, WeakWidget, Widget};
+
+pub trait WidgetView {
+    type Widget: Widget + ?Sized + Unsize<dyn Widget>;
+    fn root(&self) -> WeakWidget<Self::Widget>;
+}
+
+pub trait HasWidget {
+    type Widget: Widget + ?Sized + Unsize<dyn Widget>;
+}
+
+impl<W: Widget + Unsize<dyn Widget> + ?Sized> HasWidget for WeakWidget<W> {
+    type Widget = W;
+}
+
+impl<WV: WidgetView> IdLike for WV {
+    type Widget = WV::Widget;
+
+    fn id(&self) -> super::WidgetId {
+        self.root().id
+    }
+}

iris/core/src/widget/widgets.rs

@@ -0,0 +1,145 @@
+use std::sync::mpsc::{Receiver, Sender, channel};
+
+use crate::{
+    IdLike, StrongWidget, WeakWidget, Widget, WidgetData, WidgetId,
+    util::{DynBorrower, HashSet, SlotVec, forget_mut, to_mut},
+};
+
+pub struct Widgets {
+    pub needs_redraw: HashSet<WidgetId>,
+    vec: SlotVec<WidgetData>,
+    send: Sender<WidgetId>,
+    recv: Receiver<WidgetId>,
+    pub(crate) waiting: HashSet<WidgetId>,
+}
+
+impl Widgets {
+    pub fn new() -> Self {
+        let (send, recv) = channel();
+        Self {
+            needs_redraw: Default::default(),
+            vec: Default::default(),
+            waiting: Default::default(),
+            send,
+            recv,
+        }
+    }
+
+    pub fn has_updates(&self) -> bool {
+        !self.needs_redraw.is_empty()
+    }
+
+    pub fn get_dyn(&self, id: WidgetId) -> Option<&dyn Widget> {
+        Some(self.vec.get(id)?.widget.as_ref())
+    }
+
+    pub fn get_dyn_mut(&mut self, id: WidgetId) -> Option<&mut dyn Widget> {
+        self.needs_redraw.insert(id);
+        Some(self.vec.get_mut(id)?.widget.as_mut())
+    }
+
+    /// get_dyn but dynamic borrow checking of widgets
+    /// lets you do recursive (tree) operations, like the painter does
+    pub(crate) fn get_dyn_dynamic<'a>(&self, id: WidgetId) -> WidgetWrapper<'a> {
+        // SAFETY: must guarantee no other mutable references to this widget exist
+        // done through the borrow variable
+        let data = unsafe { forget_mut(to_mut(self.vec.get(id).unwrap())) };
+        if data.borrowed {
+            panic!("tried to mutably borrow the same widget twice");
+        }
+        WidgetWrapper::new(data.widget.as_mut(), &mut data.borrowed)
+    }
+
+    pub fn get<I: IdLike>(&self, id: &I) -> Option<&I::Widget>
+    where
+        I::Widget: Sized + Widget,
+    {
+        self.get_dyn(id.id())?.as_any().downcast_ref()
+    }
+
+    pub fn get_mut<I: IdLike>(&mut self, id: &I) -> Option<&mut I::Widget>
+    where
+        I::Widget: Sized + Widget,
+    {
+        self.get_dyn_mut(id.id())?.as_any_mut().downcast_mut()
+    }
+
+    pub fn add_strong<W: Widget>(&mut self, widget: W) -> StrongWidget<W> {
+        let id = self.vec.add(WidgetData::new(widget));
+        StrongWidget::new(id, self.send.clone())
+    }
+
+    pub fn add_weak<W: Widget>(&mut self, widget: W) -> WeakWidget<W> {
+        let id = self.vec.add(WidgetData::new(widget));
+        self.waiting.insert(id);
+        WeakWidget::new(id)
+    }
+
+    #[track_caller]
+    pub fn upgrade<W: ?Sized>(&mut self, rf: WeakWidget<W>) -> StrongWidget<W> {
+        if !self.waiting.remove(&rf.id()) {
+            let label = self.label(rf);
+            let id = rf.id();
+            panic!(
+                "widget '{label}' ({id:?}) was already added\ncannot add a widget twice; consider creating two"
+            )
+        }
+        StrongWidget::new(rf.id(), self.send.clone())
+    }
+
+    pub fn data(&self, id: impl IdLike) -> Option<&WidgetData> {
+        self.vec.get(id.id())
+    }
+
+    pub fn label(&self, id: impl IdLike) -> &String {
+        &self.data(id.id()).unwrap().label
+    }
+
+    /// useful for debugging
+    pub fn set_label(&mut self, id: impl IdLike, label: String) {
+        self.data_mut(id.id()).unwrap().label = label;
+    }
+
+    pub fn data_mut(&mut self, id: impl IdLike) -> Option<&mut WidgetData> {
+        self.vec.get_mut(id.id())
+    }
+
+    pub fn free_next(&mut self) -> Option<WidgetId> {
+        let next = self.recv.try_recv().ok()?;
+        self.vec.free(next);
+        Some(next)
+    }
+
+    #[allow(clippy::len_without_is_empty)]
+    pub fn len(&self) -> usize {
+        self.vec.len()
+    }
+}
+
+impl Default for Widgets {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+pub type WidgetWrapper<'a> = DynBorrower<'a, dyn Widget>;
+
+impl<I: IdLike> std::ops::Index<I> for Widgets
+where
+    I::Widget: Sized + Widget,
+{
+    type Output = I::Widget;
+
+    fn index(&self, id: I) -> &Self::Output {
+        self.get(&id).unwrap()
+    }
+}
+
+impl<I: IdLike> std::ops::IndexMut<I> for Widgets
+where
+    I::Widget: Sized + Widget,
+{
+    fn index_mut(&mut self, id: I) -> &mut Self::Output {
+        self.get_mut(&id).unwrap()
+    }
+}

iris/examples/minimal.rs

@@ -0,0 +1,21 @@
+use iris::prelude::*;
+
+fn main() {
+    DefaultApp::<State>::run();
+}
+
+#[derive(DefaultUiState)]
+struct State {
+    ui_state: DefaultUiState,
+}
+
+impl DefaultAppState for State {
+    fn new(
+        mut ui_state: DefaultUiState,
+        rsc: &mut DefaultRsc<Self>,
+        _: Proxy<Self::Event>,
+    ) -> Self {
+        rect(Color::RED).set_root(rsc, &mut ui_state);
+        Self { ui_state }
+    }
+}

iris/examples/tabs/main.rs

@@ -0,0 +1,224 @@
+use std::{cell::RefCell, rc::Rc};
+use winit::event::WindowEvent;
+
+use iris::prelude::*;
+type ClientRsc = DefaultRsc<Client>;
+
+fn main() {
+    DefaultApp::<Client>::run();
+}
+
+#[derive(DefaultUiState)]
+pub struct Client {
+    ui_state: DefaultUiState,
+    info: WeakWidget<Text>,
+}
+
+impl DefaultAppState for Client {
+    fn new(
+        mut ui_state: DefaultUiState,
+        rsc: &mut DefaultRsc<Self>,
+        _: Proxy<Self::Event>,
+    ) -> Self {
+        let rrect = rect(Color::WHITE).radius(20);
+        let pad_test = (
+            rrect.color(Color::BLUE),
+            (
+                rrect
+                    .color(Color::RED)
+                    .sized((100, 100))
+                    .center()
+                    .width(rest(2)),
+                (
+                    rrect.color(Color::ORANGE),
+                    rrect.color(Color::LIME).pad(10.0),
+                )
+                    .span(Dir::RIGHT)
+                    .width(rest(2)),
+                rrect.color(Color::YELLOW),
+            )
+                .span(Dir::RIGHT)
+                .pad(10)
+                .width(rest(3)),
+        )
+            .span(Dir::RIGHT)
+            .add(rsc);
+
+        let span_test = (
+            rrect.color(Color::GREEN).width(100),
+            rrect.color(Color::ORANGE),
+            rrect.color(Color::CYAN),
+            rrect.color(Color::BLUE).width(rel(0.5)),
+            rrect.color(Color::MAGENTA).width(100),
+            rrect.color(Color::RED).width(100),
+        )
+            .span(Dir::LEFT)
+            .add(rsc);
+
+        let span_add = Span::empty(Dir::RIGHT).add(rsc);
+
+        let add_button = rect(Color::LIME)
+            .radius(30)
+            .on(CursorSense::click(), move |_, rsc| {
+                let child = image(include_bytes!("assets/sungals.png"))
+                    .center()
+                    .add_strong(rsc);
+                span_add(rsc).push(child);
+            })
+            .sized((150, 150))
+            .align(Align::BOT_RIGHT);
+
+        let del_button = rect(Color::RED)
+            .radius(30)
+            .on(CursorSense::click(), move |_, rsc| {
+                span_add(rsc).pop();
+            })
+            .sized((150, 150))
+            .align(Align::BOT_LEFT);
+
+        let span_add_test = (span_add, add_button, del_button).stack().add(rsc);
+
+        let btext = |content| wtext(content).size(30);
+
+        let text_test = (
+            btext("this is a").align(Align::LEFT),
+            btext("teeeeeeeest").align(Align::RIGHT),
+            btext("okkk\nokkkkkk!").align(Align::LEFT),
+            btext("hmm"),
+            btext("a"),
+            (
+                btext("'").family(Family::Monospace).align(Align::TOP),
+                btext("'").family(Family::Monospace),
+                btext(":gamer mode").family(Family::Monospace),
+                rect(Color::CYAN).sized((10, 10)).center(),
+                rect(Color::RED).sized((100, 100)).center(),
+                rect(Color::PURPLE).sized((50, 50)).align(Align::TOP),
+            )
+                .span(Dir::RIGHT)
+                .center(),
+            wtext("pretty cool right?").size(50),
+        )
+            .span(Dir::DOWN)
+            .add(rsc);
+
+        let texts = Span::empty(Dir::DOWN).gap(10).add(rsc);
+        let msg_area = texts.scrollable().masked().background(rect(Color::SKY));
+        let add_text = wtext("add")
+            .editable(EditMode::MultiLine)
+            .text_align(Align::LEFT)
+            .size(30)
+            .attr::<Selectable>(())
+            .on(Submit, move |ctx, rsc| {
+                let w = ctx.widget;
+                let content = w.edit(rsc).take();
+                let text = wtext(content)
+                    .editable(EditMode::MultiLine)
+                    .size(30)
+                    .text_align(Align::LEFT)
+                    .wrap(true)
+                    .attr::<Selectable>(());
+                let msg_box = text
+                    .background(rect(Color::WHITE.darker(0.5)))
+                    .add_strong(rsc);
+                texts(rsc).push(msg_box);
+            })
+            .add(rsc);
+
+        let text_edit_scroll = (
+            msg_area.height(rest(1)),
+            (
+                Rect::new(Color::WHITE.darker(0.9)),
+                (
+                    add_text.width(rest(1)),
+                    Rect::new(Color::GREEN)
+                        .on(CursorSense::click(), move |ctx, rsc: &mut ClientRsc| {
+                            rsc.run_event::<Submit>(add_text, (), ctx.state);
+                        })
+                        .sized((40, 40)),
+                )
+                    .span(Dir::RIGHT)
+                    .pad(10),
+            )
+                .stack()
+                .size(StackSize::Child(1))
+                .layer_offset(1)
+                .align(Align::BOT),
+        )
+            .span(Dir::DOWN)
+            .add(rsc);
+
+        let main = WidgetPtr::new().add(rsc);
+
+        let vals = Rc::new(RefCell::new((0, Vec::new())));
+        let mut switch_button = |color, to: WeakWidget, label| {
+            let to = to.upgrade(rsc);
+            let vec = &mut vals.borrow_mut().1;
+            let i = vec.len();
+            if vec.is_empty() {
+                vec.push(None);
+                main(rsc).set(to);
+            } else {
+                vec.push(Some(to));
+            }
+            let vals = vals.clone();
+            let rect = rect(color)
+                .on(CursorSense::click(), move |ctx, rsc| {
+                    let (prev, vec) = &mut *vals.borrow_mut();
+                    if let Some(h) = vec[i].take() {
+                        vec[*prev] = main(rsc).replace(h);
+                        *prev = i;
+                    }
+                    ctx.widget(rsc).color = color.darker(0.3);
+                })
+                .on(
+                    CursorSense::HoverStart | CursorSense::unclick(),
+                    move |ctx, rsc| {
+                        ctx.widget(rsc).color = color.brighter(0.2);
+                    },
+                )
+                .on(CursorSense::HoverEnd, move |ctx, rsc| {
+                    ctx.widget(rsc).color = color;
+                });
+            (rect, wtext(label).size(30).text_align(Align::CENTER)).stack()
+        };
+
+        let tabs = (
+            switch_button(Color::RED, pad_test, "pad"),
+            switch_button(Color::GREEN, span_test, "span"),
+            switch_button(Color::BLUE, span_add_test, "image span"),
+            switch_button(Color::MAGENTA, text_test, "text layout"),
+            switch_button(
+                Color::YELLOW.mul_rgb(0.5),
+                text_edit_scroll,
+                "text edit scroll",
+            ),
+        )
+            .span(Dir::RIGHT);
+
+        let info = wtext("").add(rsc);
+        let info_sect = info.pad(10).align(Align::RIGHT);
+
+        ((tabs.height(40), main.pad(10)).span(Dir::DOWN), info_sect)
+            .stack()
+            .set_root(rsc, &mut ui_state);
+
+        Self { ui_state, info }
+    }
+
+    fn window_event(
+        &mut self,
+        _: WindowEvent,
+        rsc: &mut DefaultRsc<Self>,
+        render: &mut UiRenderState,
+    ) {
+        let new = format!(
+            "widgets: {}\nactive: {}\nviews: {}",
+            rsc.widgets().len(),
+            render.active_widgets(),
+            self.ui_state.renderer.ui.view_count(),
+        );
+        if new != *rsc.widgets()[self.info].content {
+            *rsc.widgets_mut()[self.info].content = new;
+        }
+    }
+}

iris/examples/task.rs

@@ -0,0 +1,34 @@
+use iris::prelude::*;
+use std::time::Duration;
+
+fn main() {
+    DefaultApp::<State>::run();
+}
+
+#[derive(DefaultUiState)]
+struct State {
+    ui_state: DefaultUiState,
+}
+
+impl DefaultAppState for State {
+    fn new(
+        mut ui_state: DefaultUiState,
+        rsc: &mut DefaultRsc<Self>,
+        _: Proxy<Self::Event>,
+    ) -> Self {
+        let rect = rect(Color::RED).add(rsc);
+        rect.task_on(CursorSense::click(), async move |mut ctx| {
+            tokio::time::sleep(Duration::from_secs(1)).await;
+            ctx.update(move |_, rsc| {
+                let rect = rect(rsc);
+                if rect.color == Color::RED {
+                    rect.color = Color::BLUE;
+                } else {
+                    rect.color = Color::RED;
+                }
+            });
+        })
+        .set_root(rsc, &mut ui_state);
+        Self { ui_state }
+    }
+}

iris/examples/view.rs

@@ -0,0 +1,53 @@
+use iris::prelude::*;
+
+fn main() {
+    DefaultApp::<State>::run();
+}
+
+#[derive(DefaultUiState)]
+struct State {
+    ui_state: DefaultUiState,
+}
+
+type Rsc = DefaultRsc<State>;
+
+#[derive(Clone, Copy, WidgetView)]
+struct Test {
+    #[root]
+    root: WeakWidget<Rect>,
+    cur: WeakState<bool>,
+}
+
+impl Test {
+    pub fn new(rsc: &mut Rsc) -> Self {
+        let root = rect(Color::RED).add(rsc);
+        let cur = rsc.create_state(root, false);
+        Self { root, cur }
+    }
+    pub fn toggle(&self, rsc: &mut Rsc) {
+        let cur = &mut rsc[self.cur];
+        *cur = !*cur;
+        if *cur {
+            rsc[self.root].color = Color::BLUE;
+        } else {
+            rsc[self.root].color = Color::RED;
+        }
+    }
+}
+
+impl DefaultAppState for State {
+    fn new(
+        mut ui_state: DefaultUiState,
+        rsc: &mut DefaultRsc<Self>,
+        _: Proxy<Self::Event>,
+    ) -> Self {
+        let test = Test::new(rsc);
+
+        test.on(CursorSense::click(), move |_, rsc| {
+            test.toggle(rsc);
+        })
+        .set_root(rsc, &mut ui_state);
+
+        Self { ui_state }
+    }
+}

iris/headless.conf

@@ -0,0 +1,14 @@
+# The compositor `run-headless.sh` starts, because this machine has no
+# display. Nothing here is meant to be looked at directly; `grim` is.
+#
+# No Xwayland: winit talks Wayland natively, and starting an X server is a
+# second thing to go wrong for no gain. (`emu`'s config forces it because the
+# Android emulator's renderer speaks GLX.)
+xwayland disable
+
+# A desktop-shaped output, since this is the desktop half of the port. Larger
+# than the window an example opens, so nothing is scaled or clipped.
+output HEADLESS-1 mode 1920x1200@60Hz
+
+default_border none
+focus_follows_mouse no

iris/macro/Cargo.toml

@@ -0,0 +1,12 @@
+[package]
+name = "iris-macro"
+version.workspace = true
+edition.workspace = true
+
+[dependencies]
+proc-macro2 = "1.0.103"
+quote = "1.0.42"
+syn = { version = "2.0.111", features = ["full"] }
+
+[lib]
+proc-macro = true

iris/macro/src/lib.rs

@@ -0,0 +1,196 @@
+extern crate proc_macro;
+use proc_macro::TokenStream;
+use quote::quote;
+use syn::{
+    Attribute, Block, Error, GenericParam, Generics, Ident, ItemStruct, ItemTrait, Signature,
+    Token, Type, Visibility,
+    parse::{Parse, ParseStream, Result},
+    parse_macro_input, parse_quote,
+    spanned::Spanned,
+};
+
+struct Input {
+    attrs: Vec<Attribute>,
+    vis: Visibility,
+    name: Ident,
+    generics: Generics,
+    fns: Vec<InputFn>,
+}
+
+struct InputFn {
+    sig: Signature,
+    body: Block,
+}
+
+impl Parse for Input {
+    fn parse(input: ParseStream) -> Result<Self> {
+        let attrs = input.call(Attribute::parse_outer)?;
+        let vis = input.parse()?;
+        input.parse::<Token![trait]>()?;
+        let name = input.parse()?;
+        let generics = input.parse::<Generics>()?;
+        input.parse::<Token![;]>()?;
+        let mut fns = Vec::new();
+        while !input.is_empty() {
+            let sig = input.parse()?;
+            let body = input.parse()?;
+            fns.push(InputFn { sig, body })
+        }
+        if !input.is_empty() {
+            input.error("function expected");
+        }
+        Ok(Input {
+            attrs,
+            vis,
+            name,
+            generics,
+            fns,
+        })
+    }
+}
+
+#[proc_macro]
+pub fn widget_trait(input: TokenStream) -> TokenStream {
+    let Input {
+        attrs,
+        vis,
+        name,
+        mut generics,
+        fns,
+    } = parse_macro_input!(input as Input);
+
+    let sigs: Vec<_> = fns.iter().map(|f| f.sig.clone()).collect();
+    let impls: Vec<_> = fns
+        .iter()
+        .map(|InputFn { sig, body }| quote! { #sig #body })
+        .collect();
+
+    let Some(GenericParam::Type(state)) = generics.params.first() else {
+        return Error::new(name.span(), "expected state generic parameter")
+            .into_compile_error()
+            .into();
+    };
+
+    let state = &state.ident;
+
+    generics
+        .params
+        .push(parse_quote!(WL: WidgetLike<#state, Tag>));
+    generics.params.push(parse_quote!(Tag));
+
+    let mut trai: ItemTrait = parse_quote!(
+        #vis trait #name #generics {
+            #(#sigs;)*
+        }
+    );
+
+    trai.attrs = attrs;
+
+    quote! {
+        #trai
+
+        impl #generics #name<Rsc, WL, Tag> for WL {
+            #(#impls)*
+        }
+    }
+    .into()
+}
+
+#[proc_macro_derive(DefaultUiState, attributes(default_ui_state))]
+pub fn derive_default_ui_state(input: TokenStream) -> TokenStream {
+    let mut output = proc_macro2::TokenStream::new();
+
+    let state: ItemStruct = parse_macro_input!(input);
+
+    let mut found_attr = false;
+    let mut state_field = None;
+    for field in &state.fields {
+        if !found_attr
+            && let Type::Path(path) = &field.ty
+            && path.path.is_ident("DefaultUiState")
+        {
+            state_field = Some(field);
+        }
+        let Some(attr) = field
+            .attrs
+            .iter()
+            .find(|a| a.path().is_ident("default_ui_state"))
+        else {
+            continue;
+        };
+        if found_attr {
+            output.extend(
+                Error::new(
+                    attr.span(),
+                    "cannot have more than one default_ui_state attribute",
+                )
+                .into_compile_error(),
+            );
+            continue;
+        }
+        found_attr = true;
+        state_field = Some(field);
+    }
+    let Some(field) = state_field else {
+        output.extend(
+            Error::new(state.ident.span(), "no DefaultUiState field found").into_compile_error(),
+        );
+        return output.into();
+    };
+    let sname = &state.ident;
+    let fname = field.ident.as_ref().unwrap();
+    output.extend(quote! {
+        impl iris::default::HasDefaultUiState for #sname {
+            fn default_state(&self) -> &iris::default::DefaultUiState {
+                &self.#fname
+            }
+            fn default_state_mut(&mut self) -> &mut iris::default::DefaultUiState {
+                &mut self.#fname
+            }
+        }
+    });
+    output.into()
+}
+
+#[proc_macro_derive(WidgetView, attributes(root))]
+pub fn derive_widget_view(input: TokenStream) -> TokenStream {
+    let mut output = proc_macro2::TokenStream::new();
+
+    let state: ItemStruct = parse_macro_input!(input);
+
+    let mut found_attr = false;
+    let mut state_field = None;
+    for field in &state.fields {
+        let Some(attr) = field.attrs.iter().find(|a| a.path().is_ident("root")) else {
+            continue;
+        };
+        if found_attr {
+            output.extend(
+                Error::new(attr.span(), "cannot have more than one root widget")
+                    .into_compile_error(),
+            );
+            continue;
+        }
+        found_attr = true;
+        state_field = Some(field);
+    }
+    let Some(field) = state_field else {
+        output.extend(
+            Error::new(state.ident.span(), "no root widget field found (#[root])")
+                .into_compile_error(),
+        );
+        return output.into();
+    };
+    let sname = &state.ident;
+    let fname = field.ident.as_ref().unwrap();
+    let fty = &field.ty;
+    output.extend(quote! {
+        impl iris::core::WidgetView for #sname {
+            type Widget = <#fty as iris::core::HasWidget>::Widget;
+            fn root(&self) -> #fty {
+                self.#fname
+            }
+        }
+    });
+    output.into()
+}

iris/readme.md

@@ -0,0 +1,31 @@
+# iris
+
+My experimental attempt at a rust ui library (also my first ui library).
+
+It's currently designed around using retained data structures (widgets), rather than diffing generated trees from data like xilem or iced. This is an experiment and I'm not sure if it's a good idea or not.
+
+Examples are in `examples`, eg. `cargo run --example tabs`.
+
+Goals, in general order:
+1. does what I want it to (text, images, video, animations)
+2. very easy to use ignoring ergonomic ref counting
+3. reasonably fast / efficient (a lot faster than electron, save battery life, try to beat iced and xilem)
+
+## dev details
+
+not targeting web rn cause wanna use actual nice gpu features & entire point of this is to make desktop apps / not need a web browser
+
+general ideas trynna use rn / experiment with:
+- retained mode
+- specifically designed around wgpu so there's no translation
+- postfix functions for most things to prevent unreadable indentation (going very well)
+- events can be done directly where you draw the widgets
+- almost no macros in user code & actual LSP typechecking (variadic generics if you can hear me please save us)
+- relative anchor + absolute offset coord system (+ "rest" / leftover during widget layout)
+- single threaded ui & pass context around to make non async usage straightforward (pretty unsure about this)
+- widgets store outside of the actual rendering so they can be moved around and swapped easily (unsure about this but seems to work good for now)
+
+under heavy initial development so not gonna try to explain status, maybe check TODO for that;
+sizable chance it gets a rewrite once I know everything I need and what seems to work best
+
+it's called iris because it's the structure around what you actually want to display and colorful

iris/run-headless.sh

@@ -0,0 +1,108 @@
+#!/bin/sh
+# Run an iris example on this machine, which has no display.
+#
+#   ./run-headless.sh tabs [-- cargo args]
+#   ./run-headless.sh tabs --shot /tmp/tabs.png --seconds 4
+#
+# The VM has a virtio-gpu render node (Vulkan 1.4 through Venus, GL 4.6
+# through virgl), so wgpu runs on the host's real GPU -- what is missing is
+# only a compositor to give winit a surface. So: a headless sway, the same
+# trick `emu` uses for the Android emulator, and `grim` to see the result.
+#
+# It is deliberately *not* `emu`'s compositor. sway tiles, so adding a window
+# to the one an emulator is sitting in resizes that emulator's window, and a
+# peer session's `emu up` could join at any moment. This one has its own
+# socket and its own runtime directory and goes away with the machine.
+set -eu
+
+here=$(cd "$(dirname "$0")" && pwd)
+run="${XDG_RUNTIME_DIR:-/tmp}/iris-headless"
+seconds=3
+shot=""
+example=""
+
+while [ $# -gt 0 ]; do
+    case "$1" in
+        --shot) shot=$2; shift 2 ;;
+        --seconds) seconds=$2; shift 2 ;;
+        --) shift; break ;;
+        *) example=$1; shift ;;
+    esac
+done
+[ -n "$example" ] || { echo "usage: $0 EXAMPLE [--shot PNG] [--seconds N] [-- cargo args]" >&2; exit 2; }
+
+mkdir -p "$run"
+export SWAYSOCK="$run/sway.sock"
+
+# Named rather than left to sway's pid-based default, so a second run reuses
+# this compositor instead of starting another beside it.
+if ! swaymsg -t get_version >/dev/null 2>&1; then
+    rm -f "$SWAYSOCK"
+    WLR_BACKENDS=headless WLR_LIBINPUT_NO_DEVICES=1 LIBSEAT_BACKEND=noop \
+        setsid sway -c "$here/headless.conf" >"$run/sway.log" 2>&1 &
+    i=0
+    while [ $i -lt 20 ]; do
+        swaymsg -t get_version >/dev/null 2>&1 && break
+        i=$((i + 1)); sleep 0.5
+    done
+    swaymsg -t get_version >/dev/null 2>&1 || {
+        echo "run-headless: compositor did not start; see $run/sway.log" >&2
+        exit 1
+    }
+fi
+
+# Asked of the compositor rather than guessed: sway takes the first free
+# wayland-N, and this machine may already have one.
+rm -f "$run/display"
+swaymsg exec -- "sh -c 'printf %s \"\$WAYLAND_DISPLAY\" > $run/display'" >/dev/null
+i=0
+while [ $i -lt 20 ]; do
+    [ -s "$run/display" ] && break
+    i=$((i + 1)); sleep 0.5
+done
+[ -s "$run/display" ] || { echo "run-headless: could not read WAYLAND_DISPLAY" >&2; exit 1; }
+WAYLAND_DISPLAY=$(cat "$run/display")
+export WAYLAND_DISPLAY
+
+echo "run-headless: $WAYLAND_DISPLAY (sway $(swaymsg -t get_version --raw | sed -n 's/.*"human_readable":"\([^"]*\)".*/\1/p'))" >&2
+
+cd "$here"
+cargo build --example "$example" "$@" >&2
+bin="$here/target/debug/examples/$example"
+
+"$bin" >"$run/$example.log" 2>&1 &
+pid=$!
+trap 'kill "$pid" 2>/dev/null || true' EXIT INT TERM
+
+# Wait for the window to be mapped rather than for a number of seconds. A
+# fixed sleep took an all-black screenshot the first time this ran, when sway
+# had started in the same invocation and had not composited its output yet --
+# which is indistinguishable from an app that draws nothing.
+i=0
+while [ $i -lt 40 ]; do
+    kill -0 "$pid" 2>/dev/null || break
+    swaymsg -t get_tree --raw 2>/dev/null | grep -q "\"pid\":$pid," && break
+    i=$((i + 1)); sleep 0.25
+done
+
+# Then settle, for whatever the example does after its first frame.
+i=0
+while [ $i -lt "$((seconds * 2))" ]; do
+    kill -0 "$pid" 2>/dev/null || break
+    i=$((i + 1)); sleep 0.5
+done
+
+if kill -0 "$pid" 2>/dev/null; then
+    [ -n "$shot" ] && grim "$shot" && echo "run-headless: wrote $shot" >&2
+    kill "$pid" 2>/dev/null || true
+    wait "$pid" 2>/dev/null || true
+    status=0
+else
+    wait "$pid" 2>/dev/null || status=$?
+    echo "run-headless: $example exited early (status ${status:-0})" >&2
+    status=${status:-1}
+fi
+
+echo "--- $example output ---" >&2
+cat "$run/$example.log" >&2
+exit "$status"

iris/rust-toolchain.toml

@@ -0,0 +1,11 @@
+# iris needs nightly (see the #![feature] list in core/src/lib.rs and src/lib.rs).
+# The pin is dated rather than "nightly" because the const-traits feature set
+# changes shape between nightlies: on 2026-09-04 the vendored January tree would
+# not parse at all, because `impl const Trait for T` had become
+# `const impl Trait for T`. A rolling channel turns that into a build that
+# breaks unattended on whatever machine Dev Updater happens to build on.
+# Advance this deliberately, with the feature list in RUST.md's I0b.
+[toolchain]
+channel = "nightly-2026-09-03"
+components = ["clippy", "rustfmt"]
+targets = ["aarch64-linux-android", "x86_64-linux-android"]

iris/src/default/app.rs

@@ -0,0 +1,60 @@
+use winit::{
+    application::ApplicationHandler,
+    event::WindowEvent,
+    event_loop::{ActiveEventLoop, EventLoop, EventLoopProxy},
+    window::WindowId,
+};
+
+pub trait AppState {
+    type Event: 'static;
+    fn new(event_loop: &ActiveEventLoop, proxy: EventLoopProxy<Self::Event>) -> Self;
+    fn window_event(&mut self, event: WindowEvent, event_loop: &ActiveEventLoop);
+    fn event(&mut self, event: Self::Event, event_loop: &ActiveEventLoop);
+    fn exit(&mut self);
+
+    fn run()
+    where
+        Self: Sized,
+    {
+        App::<Self>::run();
+    }
+}
+
+pub struct App<State: AppState> {
+    state: Option<State>,
+    proxy: EventLoopProxy<State::Event>,
+}
+
+impl<State: AppState> App<State> {
+    pub fn run() {
+        let event_loop = EventLoop::with_user_event().build().unwrap();
+        let proxy = event_loop.create_proxy();
+        event_loop
+            .run_app(&mut App::<State> { state: None, proxy })
+            .unwrap();
+    }
+}
+
+impl<State: AppState> ApplicationHandler<State::Event> for App<State> {
+    fn resumed(&mut self, event_loop: &ActiveEventLoop) {
+        if self.state.is_none() {
+            let state = State::new(event_loop, self.proxy.clone());
+            self.state = Some(state);
+        }
+    }
+
+    fn window_event(&mut self, event_loop: &ActiveEventLoop, _id: WindowId, event: WindowEvent) {
+        let state = self.state.as_mut().unwrap();
+        state.window_event(event, event_loop);
+    }
+
+    fn user_event(&mut self, event_loop: &ActiveEventLoop, event: State::Event) {
+        let state = self.state.as_mut().unwrap();
+        state.event(event, event_loop);
+    }
+
+    fn exiting(&mut self, _: &ActiveEventLoop) {
+        let state = self.state.as_mut().unwrap();
+        state.exit();
+    }
+}

iris/src/default/attr.rs

@@ -0,0 +1,83 @@
+use crate::prelude::*;
+use std::time::{Duration, Instant};
+use winit::dpi::{LogicalPosition, LogicalSize};
+
+pub struct Selector;
+
+impl<Rsc: HasEvents, W: Widget + 'static> WidgetAttr<Rsc, W> for Selector
+where
+    Rsc::State: HasDefaultUiState,
+{
+    type Input = WeakWidget<TextEdit>;
+
+    fn run(rsc: &mut Rsc, container: WeakWidget<W>, id: Self::Input) {
+        rsc.register_event(container, CursorSense::click_or_drag(), move |ctx, rsc| {
+            let region = ctx.data.render.window_region(&id, &*rsc).unwrap();
+            let id_pos = region.top_left;
+            let container_pos = ctx
+                .data
+                .render
+                .window_region(&container, &*rsc)
+                .unwrap()
+                .top_left;
+            let pos = ctx.data.pos + container_pos - id_pos;
+            let size = region.size();
+            select(
+                rsc,
+                ctx.data.render,
+                ctx.state,
+                id,
+                pos,
+                size,
+                ctx.data.sense.is_dragging(),
+            );
+        });
+    }
+}
+
+pub struct Selectable;
+
+impl<Rsc: HasEvents> WidgetAttr<Rsc, TextEdit> for Selectable
+where
+    Rsc::State: HasDefaultUiState,
+{
+    type Input = ();
+
+    fn run(rsc: &mut Rsc, id: WeakWidget<TextEdit>, _: Self::Input) {
+        rsc.register_event(id, CursorSense::click_or_drag(), move |ctx, rsc| {
+            select(
+                rsc,
+                ctx.data.render,
+                ctx.state,
+                id,
+                ctx.data.pos,
+                ctx.data.size,
+                ctx.data.sense.is_dragging(),
+            );
+        });
+    }
+}
+
+fn select(
+    rsc: &mut impl UiRsc,
+    render: &UiRenderState,
+    state: &mut impl HasDefaultUiState,
+    id: WeakWidget<TextEdit>,
+    pos: Vec2,
+    size: Vec2,
+    dragging: bool,
+) {
+    let state = state.default_state_mut();
+    let now = Instant::now();
+    let recent = (now - state.last_click) < Duration::from_millis(300);
+    state.last_click = now;
+    id.edit(rsc).select(pos, size, dragging, recent);
+    if let Some(region) = render.window_region(&id, &*rsc) {
+        state.window.set_ime_allowed(true);
+        state.window.set_ime_cursor_area(
+            LogicalPosition::<f32>::from(region.top_left.tuple()),
+            LogicalSize::<f32>::from(region.size().tuple()),
+        );
+    }
+    state.focus = Some(id);
+}

iris/src/default/event.rs

@@ -0,0 +1,9 @@
+use iris_core::Event;
+
+#[derive(Eq, PartialEq, Hash, Clone)]
+pub struct Submit;
+impl Event for Submit {}
+
+#[derive(Eq, PartialEq, Hash, Clone)]
+pub struct Edited;
+impl Event for Edited {}

iris/src/default/input.rs

@@ -0,0 +1,78 @@
+use crate::prelude::*;
+use winit::{
+    event::{MouseButton, MouseScrollDelta, WindowEvent},
+    keyboard::{Key, NamedKey},
+};
+
+#[derive(Default)]
+pub struct Input {
+    cursor: CursorState,
+    pub modifiers: Modifiers,
+}
+
+impl Input {
+    pub fn event(&mut self, event: &WindowEvent) -> bool {
+        match event {
+            WindowEvent::CursorMoved { position, .. } => {
+                self.cursor.pos = Vec2::new(position.x as f32, position.y as f32);
+                self.cursor.exists = true;
+            }
+            WindowEvent::MouseInput { state, button, .. } => {
+                let buttons = &mut self.cursor.buttons;
+                let pressed = state.is_pressed();
+                match button {
+                    MouseButton::Left => buttons.left.update(pressed),
+                    MouseButton::Right => buttons.right.update(pressed),
+                    MouseButton::Middle => buttons.middle.update(pressed),
+                    _ => (),
+                }
+            }
+            WindowEvent::MouseWheel { delta, .. } => {
+                let mut delta = match *delta {
+                    MouseScrollDelta::LineDelta(x, y) => Vec2::new(x, y),
+                    MouseScrollDelta::PixelDelta(pos) => Vec2::new(pos.x as f32, pos.y as f32),
+                };
+                if delta.x == 0.0 && self.modifiers.shift {
+                    delta.x = delta.y;
+                    delta.y = 0.0;
+                }
+                self.cursor.scroll_delta = delta;
+            }
+            WindowEvent::CursorLeft { .. } => {
+                self.cursor.exists = false;
+                self.modifiers.clear();
+            }
+            WindowEvent::KeyboardInput { event, .. } => {
+                if let Key::Named(named) = event.logical_key {
+                    let pressed = event.state.is_pressed();
+                    match named {
+                        NamedKey::Control => {
+                            self.modifiers.control = pressed;
+                        }
+                        NamedKey::Shift => {
+                            self.modifiers.shift = pressed;
+                        }
+                        _ => (),
+                    }
+                }
+            }
+            _ => return false,
+        }
+        true
+    }
+
+    pub fn end_frame(&mut self) {
+        self.cursor.end_frame();
+    }
+}
+
+impl DefaultUiState {
+    pub fn window_size(&self) -> Vec2 {
+        let size = self.renderer.window().inner_size();
+        (size.width, size.height).into()
+    }
+
+    pub fn cursor_state(&self) -> &CursorState {
+        &self.input.cursor
+    }
+}

iris/src/default/mod.rs

@@ -0,0 +1,354 @@
+use crate::prelude::*;
+use arboard::Clipboard;
+use std::{
+    marker::{PhantomData, Sized},
+    sync::Arc,
+    time::Instant,
+};
+use winit::{
+    event::{Ime, WindowEvent},
+    event_loop::{ActiveEventLoop, EventLoopProxy},
+    window::{Window, WindowAttributes},
+};
+
+mod app;
+mod attr;
+mod event;
+mod input;
+mod render;
+mod sense;
+mod state;
+mod task;
+
+pub use app::*;
+pub use attr::*;
+pub use event::*;
+pub use input::*;
+pub use render::*;
+pub use sense::*;
+pub use state::*;
+pub use task::*;
+
+pub type Proxy<Event> = EventLoopProxy<Event>;
+
+pub struct DefaultUiState {
+    pub root: Option<StrongWidget>,
+    pub renderer: UiRenderer,
+    pub input: Input,
+    pub focus: Option<WeakWidget<TextEdit>>,
+    pub clipboard: Clipboard,
+    pub window: Arc<Window>,
+    pub ime: usize,
+    pub last_click: Instant,
+}
+
+impl HasRoot for DefaultUiState {
+    fn set_root(&mut self, root: StrongWidget) {
+        self.root = Some(root);
+    }
+}
+
+impl DefaultUiState {
+    pub fn new(window: impl Into<Arc<Window>>) -> Self {
+        let window = window.into();
+        Self {
+            root: None,
+            renderer: UiRenderer::new(window.clone()),
+            window,
+            input: Input::default(),
+            clipboard: Clipboard::new().unwrap(),
+            ime: 0,
+            last_click: Instant::now(),
+            focus: None,
+        }
+    }
+}
+
+pub trait HasDefaultUiState: Sized + 'static {
+    fn default_state(&self) -> &DefaultUiState;
+    fn default_state_mut(&mut self) -> &mut DefaultUiState;
+}
+
+pub trait DefaultAppState: HasDefaultUiState {
+    type Event = ();
+    fn new(ui_state: DefaultUiState, rsc: &mut DefaultRsc<Self>, proxy: Proxy<Self::Event>)
+    -> Self;
+    #[allow(unused_variables)]
+    fn event(
+        &mut self,
+        event: Self::Event,
+        rsc: &mut DefaultRsc<Self>,
+        render: &mut UiRenderState,
+    ) {
+    }
+    #[allow(unused_variables)]
+    fn exit(&mut self, rsc: &mut DefaultRsc<Self>, render: &mut UiRenderState) {}
+    #[allow(unused_variables)]
+    fn window_event(
+        &mut self,
+        event: WindowEvent,
+        rsc: &mut DefaultRsc<Self>,
+        render: &mut UiRenderState,
+    ) {
+    }
+    fn window_attributes() -> WindowAttributes {
+        Default::default()
+    }
+}
+
+pub struct DefaultRsc<State: 'static> {
+    pub ui: UiData,
+    pub events: EventManager<Self>,
+    pub tasks: Tasks<Self>,
+    pub state: WidgetState,
+    _state: PhantomData<State>,
+}
+
+impl<State> DefaultRsc<State> {
+    fn init(window: Arc<Window>) -> (Self, TaskMsgReceiver<Self>) {
+        let (tasks, recv) = Tasks::init(window);
+        (
+            Self {
+                ui: Default::default(),
+                events: Default::default(),
+                tasks,
+                state: Default::default(),
+                _state: Default::default(),
+            },
+            recv,
+        )
+    }
+
+    pub fn create_state<T: 'static>(&mut self, id: impl IdLike, data: T) -> WeakState<T> {
+        self.state.add(id.id(), data)
+    }
+}
+
+impl<State> UiRsc for DefaultRsc<State> {
+    fn ui(&self) -> &UiData {
+        &self.ui
+    }
+
+    fn ui_mut(&mut self) -> &mut UiData {
+        &mut self.ui
+    }
+
+    fn on_draw(&mut self, active: &ActiveData) {
+        self.events.draw(active);
+    }
+
+    fn on_undraw(&mut self, active: &ActiveData) {
+        self.events.undraw(active);
+    }
+
+    fn on_remove(&mut self, id: WidgetId) {
+        self.events.remove(id);
+        self.state.remove(id);
+    }
+}
+
+impl<State: 'static> HasState for DefaultRsc<State> {
+    type State = State;
+}
+
+impl<State: 'static> HasEvents for DefaultRsc<State> {
+    fn events(&self) -> &EventManager<Self> {
+        &self.events
+    }
+
+    fn events_mut(&mut self) -> &mut EventManager<Self> {
+        &mut self.events
+    }
+}
+
+impl<State: 'static> HasTasks for DefaultRsc<State> {
+    fn tasks_mut(&mut self) -> &mut Tasks<Self> {
+        &mut self.tasks
+    }
+}
+
+impl<State: 'static> HasWidgetState for DefaultRsc<State> {
+    fn widget_state(&self) -> &WidgetState {
+        &self.state
+    }
+
+    fn widget_state_mut(&mut self) -> &mut WidgetState {
+        &mut self.state
+    }
+}
+
+pub struct DefaultApp<State: DefaultAppState> {
+    rsc: DefaultRsc<State>,
+    render: UiRenderState,
+    state: State,
+    task_recv: TaskMsgReceiver<DefaultRsc<State>>,
+}
+
+impl<State: DefaultAppState> AppState for DefaultApp<State> {
+    type Event = State::Event;
+
+    fn new(event_loop: &ActiveEventLoop, proxy: EventLoopProxy<Self::Event>) -> Self {
+        let window = event_loop
+            .create_window(State::window_attributes())
+            .unwrap();
+        let default_state = DefaultUiState::new(window);
+        let (mut rsc, task_recv) = DefaultRsc::init(default_state.window.clone());
+        let state = State::new(default_state, &mut rsc, proxy);
+        let render = UiRenderState::new();
+        Self {
+            rsc,
+            state,
+            render,
+            task_recv,
+        }
+    }
+
+    fn event(&mut self, event: Self::Event, _: &ActiveEventLoop) {
+        self.state.event(event, &mut self.rsc, &mut self.render);
+    }
+
+    fn window_event(&mut self, event: WindowEvent, event_loop: &ActiveEventLoop) {
+        let Self {
+            rsc,
+            render,
+            state,
+            task_recv,
+        } = self;
+
+        for update in task_recv.try_iter() {
+            update(state, rsc);
+        }
+
+        let ui_state = state.default_state_mut();
+        let input_changed = ui_state.input.event(&event);
+        let cursor_state = ui_state.cursor_state().clone();
+        let old = ui_state.focus;
+        if cursor_state.buttons.left.is_start() {
+            ui_state.focus = None;
+        }
+        if input_changed {
+            let window_size = ui_state.window_size();
+            render.run_sensors(rsc, state, cursor_state, window_size);
+        }
+        let ui_state = state.default_state_mut();
+        if old != ui_state.focus
+            && let Some(old) = old
+        {
+            old.edit(rsc).deselect();
+        }
+        match &event {
+            WindowEvent::CloseRequested => event_loop.exit(),
+            WindowEvent::RedrawRequested => {
+                render.update(&ui_state.root, rsc);
+                ui_state.renderer.update(&mut rsc.ui, render);
+                ui_state.renderer.draw();
+            }
+            WindowEvent::Resized(size) => {
+                render.resize((size.width, size.height));
+                ui_state.renderer.resize(size)
+            }
+            WindowEvent::KeyboardInput { event, .. } => {
+                if let Some(sel) = ui_state.focus
+                    && event.state.is_pressed()
+                {
+                    let mut text = sel.edit(rsc);
+                    match text.apply_event(event, &ui_state.input.modifiers) {
+                        TextInputResult::Unfocus => {
+                            ui_state.focus = None;
+                            ui_state.window.set_ime_allowed(false);
+                        }
+                        TextInputResult::Submit => {
+                            rsc.run_event::<Submit>(sel, (), state);
+                        }
+                        TextInputResult::Paste => {
+                            if let Ok(t) = ui_state.clipboard.get_text() {
+                                text.insert(&t);
+                            }
+                            rsc.run_event::<Edited>(sel, (), state);
+                        }
+                        TextInputResult::Copy(text) => {
+                            if let Err(err) = ui_state.clipboard.set_text(text) {
+                                eprintln!("failed to copy text to clipboard: {err}")
+                            }
+                        }
+                        TextInputResult::Used => {
+                            rsc.run_event::<Edited>(sel, (), state);
+                        }
+                        TextInputResult::Unused => {}
+                    }
+                }
+            }
+            WindowEvent::Ime(ime) => {
+                if let Some(sel) = ui_state.focus {
+                    let mut text = sel.edit(rsc);
+                    match ime {
+                        Ime::Enabled | Ime::Disabled => (),
+                        Ime::Preedit(content, _pos) => {
+                            // TODO: highlight once that's real
+                            text.replace(ui_state.ime, content);
+                            ui_state.ime = content.chars().count();
+                        }
+                        Ime::Commit(content) => {
+                            text.insert(content);
+                        }
+                    }
+                }
+            }
+            _ => (),
+        }
+        state.window_event(event, rsc, render);
+        let ui_state = self.state.default_state_mut();
+        if render.needs_redraw(&ui_state.root, rsc.widgets()) {
+            ui_state.renderer.window().request_redraw();
+        }
+        ui_state.input.end_frame();
+    }
+
+    fn exit(&mut self) {
+        self.state.exit(&mut self.rsc, &mut self.render);
+    }
+}
+
+pub trait RscIdx<Rsc> {
+    type Output;
+    fn get(self, rsc: &Rsc) -> &Self::Output;
+    fn get_mut(self, rsc: &mut Rsc) -> &mut Self::Output;
+}
+
+impl<State: 'static, I: RscIdx<DefaultRsc<State>>> std::ops::Index<I> for DefaultRsc<State> {
+    type Output = I::Output;
+
+    fn index(&self, index: I) -> &Self::Output {
+        index.get(self)
+    }
+}
+
+impl<State: 'static, I: RscIdx<DefaultRsc<State>>> std::ops::IndexMut<I> for DefaultRsc<State> {
+    fn index_mut(&mut self, index: I) -> &mut Self::Output {
+        index.get_mut(self)
+    }
+}
+
+impl<W: Widget, Rsc: UiRsc> RscIdx<Rsc> for WeakWidget<W> {
+    type Output = W;
+
+    fn get(self, rsc: &Rsc) -> &Self::Output {
+        &rsc.ui().widgets[self]
+    }
+
+    fn get_mut(self, rsc: &mut Rsc) -> &mut Self::Output {
+        &mut rsc.ui_mut().widgets[self]
+    }
+}
+
+impl<T: 'static, Rsc: HasWidgetState> RscIdx<Rsc> for WeakState<T> {
+    type Output = T;
+
+    fn get(self, rsc: &Rsc) -> &Self::Output {
+        rsc.widget_state().get(self)
+    }
+
+    fn get_mut(self, rsc: &mut Rsc) -> &mut Self::Output {
+        rsc.widget_state_mut().get_mut(self)
+    }
+}

iris/src/default/render.rs

@@ -0,0 +1,154 @@
+use iris_core::{UiData, UiRenderNode, UiRenderState};
+use pollster::FutureExt;
+use std::sync::Arc;
+use wgpu::*;
+use winit::{dpi::PhysicalSize, window::Window};
+
+pub const CLEAR_COLOR: Color = Color::BLACK;
+
+pub struct UiRenderer {
+    window: Arc<Window>,
+    surface: Surface<'static>,
+    device: Device,
+    queue: Queue,
+    config: SurfaceConfiguration,
+    encoder: CommandEncoder,
+    pub ui: UiRenderNode,
+}
+
+impl UiRenderer {
+    pub fn update(&mut self, ui: &mut UiData, render: &mut UiRenderState) {
+        self.ui.update(&self.device, &self.queue, ui, render);
+    }
+
+    pub fn draw(&mut self) {
+        let output = self.surface.get_current_texture().unwrap();
+        let view = output
+            .texture
+            .create_view(&TextureViewDescriptor::default());
+
+        let mut encoder = std::mem::replace(&mut self.encoder, Self::create_encoder(&self.device));
+        {
+            let render_pass = &mut encoder.begin_render_pass(&RenderPassDescriptor {
+                color_attachments: &[Some(RenderPassColorAttachment {
+                    view: &view,
+                    resolve_target: None,
+                    ops: Operations {
+                        load: LoadOp::Clear(CLEAR_COLOR),
+                        store: StoreOp::Store,
+                    },
+                    depth_slice: None,
+                })],
+                ..Default::default()
+            });
+            self.ui.draw(render_pass);
+        }
+
+        self.queue.submit(std::iter::once(encoder.finish()));
+        // Immediately before presenting, so the windowing system can schedule
+        // the frame. On Wayland this is what ties the commit to the surface's
+        // frame callback; without it a frame drawn when nothing else follows
+        // could sit unpresented, and the window kept the layout it had before
+        // the compositor's first resize -- intermittently, on about a fifth of
+        // starts, with nothing left to flush it.
+        self.window.pre_present_notify();
+        output.present();
+    }
+
+    pub fn resize(&mut self, size: &PhysicalSize<u32>) {
+        self.config.width = size.width;
+        self.config.height = size.height;
+        self.surface.configure(&self.device, &self.config);
+        self.ui.resize((size.width, size.height), &self.queue);
+    }
+
+    fn create_encoder(device: &Device) -> CommandEncoder {
+        device.create_command_encoder(&CommandEncoderDescriptor {
+            label: Some("Render Encoder"),
+        })
+    }
+
+    pub fn new(window: Arc<Window>) -> Self {
+        let size = window.inner_size();
+
+        let instance = Instance::new(&InstanceDescriptor {
+            backends: Backends::PRIMARY,
+            ..Default::default()
+        });
+
+        let surface = instance
+            .create_surface(window.clone())
+            .expect("Could not create window surface!");
+
+        let adapter = instance
+            .request_adapter(&RequestAdapterOptions {
+                power_preference: PowerPreference::default(),
+                compatible_surface: Some(&surface),
+                force_fallback_adapter: false,
+            })
+            .block_on()
+            .expect("Could not get adapter!");
+
+        // No features beyond what wgpu asks for by default, and no
+        // binding-array limits: the atlas is one texture_2d_array and a
+        // standalone image is its own ordinary bind group, neither of which
+        // needs descriptor indexing. See TEXTURES.md's "Recommended shape"
+        // for why the old binding array asked for
+        // VK_EXT_descriptor_indexing unconditionally and did not survive a
+        // real share of Android GPUs.
+        let (device, queue) = adapter
+            .request_device(&DeviceDescriptor {
+                required_limits: Limits {
+                    max_buffer_size: 1 << 30,
+                    ..Default::default()
+                },
+                ..Default::default()
+            })
+            .block_on()
+            .expect("Could not get device!");
+
+        let surface_caps = surface.get_capabilities(&adapter);
+        let surface_format = surface_caps
+            .formats
+            .iter()
+            .copied()
+            .find(|f| f.is_srgb())
+            .unwrap_or(surface_caps.formats[0]);
+
+        let config = SurfaceConfiguration {
+            usage: TextureUsages::RENDER_ATTACHMENT,
+            format: surface_format,
+            width: size.width,
+            height: size.height,
+            // Vsync, because a toolkit aiming at battery life must not present
+            // frames a display will never show: AutoNoVsync accepts them as
+            // fast as the GPU will take them, so a redraw burst costs whatever
+            // the hardware can be made to do rather than one frame.
+            // AutoVsync picks Fifo, which every backend supports.
+            present_mode: PresentMode::AutoVsync,
+            alpha_mode: surface_caps.alpha_modes[0],
+            desired_maximum_frame_latency: 2,
+            view_formats: vec![],
+        };
+
+        surface.configure(&device, &config);
+
+        let encoder = Self::create_encoder(&device);
+
+        let ui = UiRenderNode::new(&device, &queue, &config);
+
+        Self {
+            surface,
+            device,
+            queue,
+            config,
+            encoder,
+            ui,
+            window,
+        }
+    }
+
+    pub fn window(&self) -> &Window {
+        self.window.as_ref()
+    }
+}

iris/src/default/sense.rs

@@ -0,0 +1,359 @@
+use crate::prelude::*;
+use std::{
+    ops::{BitOr, Deref, DerefMut},
+    rc::Rc,
+};
+
+#[derive(Clone, Copy, PartialEq)]
+pub enum CursorButton {
+    Left,
+    Right,
+    Middle,
+}
+
+#[derive(Clone, Copy, PartialEq)]
+pub enum CursorSense {
+    PressStart(CursorButton),
+    Pressing(CursorButton),
+    PressEnd(CursorButton),
+    HoverStart,
+    Hovering,
+    HoverEnd,
+    Scroll,
+}
+
+#[derive(Clone)]
+pub struct CursorSenses(Vec<CursorSense>);
+
+impl Event for CursorSenses {
+    type Data<'a> = CursorData<'a>;
+    type State = SensorState;
+    fn should_run<'a>(&self, data: &Self::Data<'a>) -> Option<Self::Data<'a>> {
+        if let Some(sense) = should_run(self, &data.cursor, data.hover) {
+            let mut data = data.clone();
+            data.sense = sense;
+            Some(data)
+        } else {
+            None
+        }
+    }
+}
+
+impl CursorSense {
+    pub fn click() -> Self {
+        Self::PressStart(CursorButton::Left)
+    }
+    pub fn click_or_drag() -> CursorSenses {
+        Self::click() | Self::Pressing(CursorButton::Left)
+    }
+    pub fn unclick() -> Self {
+        Self::PressEnd(CursorButton::Left)
+    }
+    pub fn is_dragging(&self) -> bool {
+        matches!(self, CursorSense::Pressing(CursorButton::Left))
+    }
+
+    /// True for a sense that names a specific thing happening this frame
+    /// (a button transitioning, a scroll) as opposed to the ambient,
+    /// always-on-while-over `Hover*` family. Used to decide whether a
+    /// widget actually *consumes* an input for fall-through purposes: a
+    /// widget that merely highlights on hover must not be able to block a
+    /// scroll or a click meant for whatever is behind it, the way it
+    /// currently could when "the cursor is over this widget" and
+    /// "this widget handled the event" were the same check. See
+    /// `SensorUi::run_sensors`.
+    pub fn is_momentary(&self) -> bool {
+        !matches!(
+            self,
+            CursorSense::HoverStart | CursorSense::Hovering | CursorSense::HoverEnd
+        )
+    }
+}
+
+#[derive(Default, Clone)]
+pub struct CursorState {
+    pub pos: Vec2,
+    pub exists: bool,
+    pub buttons: CursorButtons,
+    pub scroll_delta: Vec2,
+}
+
+#[derive(Default, Clone)]
+pub struct CursorButtons {
+    pub left: ActivationState,
+    pub middle: ActivationState,
+    pub right: ActivationState,
+}
+
+impl CursorButtons {
+    pub fn select(&self, button: &CursorButton) -> &ActivationState {
+        match button {
+            CursorButton::Left => &self.left,
+            CursorButton::Right => &self.right,
+            CursorButton::Middle => &self.middle,
+        }
+    }
+
+    pub fn end_frame(&mut self) {
+        self.left.end_frame();
+        self.middle.end_frame();
+        self.right.end_frame();
+    }
+
+    pub fn iter(&self) -> impl Iterator<Item = (CursorButton, &ActivationState)> {
+        [
+            CursorButton::Left,
+            CursorButton::Middle,
+            CursorButton::Right,
+        ]
+        .into_iter()
+        .map(|b| (b, self.select(&b)))
+    }
+}
+
+impl CursorState {
+    pub fn end_frame(&mut self) {
+        self.buttons.end_frame();
+        self.scroll_delta = Vec2::ZERO;
+    }
+}
+
+#[derive(Debug, Clone, Copy, Default, PartialEq)]
+pub enum ActivationState {
+    Start,
+    On,
+    End,
+    #[default]
+    Off,
+}
+
+/// this and other similar stuff has a generic
+/// because I kind of want to make CursorModule generic
+/// or basically have some way to have custom senses
+/// that depend on active widget positions
+/// but I'm not sure how or if worth it
+pub struct Sensor<Ctx: HasEvents, Data> {
+    pub senses: CursorSenses,
+    pub f: Rc<dyn EventFn<Ctx, Data>>,
+}
+
+pub type SenseShape = UiRegion;
+
+#[derive(Default, Debug)]
+pub struct SensorState {
+    pub hover: ActivationState,
+}
+
+#[derive(Clone)]
+pub struct CursorData<'a> {
+    /// where this widget was hit
+    pub pos: Vec2,
+    pub size: Vec2,
+    pub scroll_delta: Vec2,
+    pub hover: ActivationState,
+    pub cursor: CursorState,
+    /// the first sense that triggered this
+    pub sense: CursorSense,
+    pub render: &'a UiRenderState,
+}
+
+pub trait SensorUi {
+    fn run_sensors<Rsc: HasEvents>(
+        &self,
+        rsc: &mut Rsc,
+        state: &mut Rsc::State,
+        cursor: CursorState,
+        window_size: Vec2,
+    );
+}
+
+impl SensorUi for UiRenderState {
+    fn run_sensors<Rsc: HasEvents>(
+        &self,
+        rsc: &mut Rsc,
+        state: &mut Rsc::State,
+        cursor: CursorState,
+        window_size: Vec2,
+    ) {
+        // in order to remove this take, need to store active list in UiRenderState somehow
+        // this would probably be done through a generic parameter that adds yet another rsc /
+        // state like thing, but local to render state, and is passed to UiRsc events so you can
+        // update it there?
+        // Whether *something specific* is happening this frame (a button
+        // transitioning, a scroll) as opposed to the cursor merely resting
+        // over widgets. Only this decides whether a widget can block a
+        // lower layer from also seeing the event -- see the `consumed`
+        // comment below.
+        let momentary_active =
+            cursor.scroll_delta != Vec2::ZERO || cursor.buttons.iter().any(|(_, a)| !a.is_off());
+
+        let mut active = std::mem::take(&mut rsc.events_mut().get_type::<CursorSense>().active);
+        for layer in self.layers.indices().rev() {
+            let mut sensed = false;
+            for (id, sensor) in active.get_mut(&layer).into_flat_iter() {
+                let shape = self.resolved_region(id, rsc).unwrap();
+                let region = shape.to_px(window_size);
+                let in_shape = cursor.exists && region.contains(cursor.pos);
+                sensor.hover.update(in_shape);
+                if sensor.hover == ActivationState::Off {
+                    continue;
+                }
+
+                // A widget in shape always still runs (a hover-only
+                // highlight must fire on the topmost thing under the
+                // cursor even while a scroll or click passes through it),
+                // but whether it *consumes* the input -- stopping a lower
+                // layer from also seeing it -- is judged per input kind
+                // (LAYOUT.md's coordinator asked for this alongside the
+                // hit-test rewrite, since both are about `resolved_region`
+                // and what "under the pointer" means): with nothing
+                // momentary happening, "in shape" is consumption, same as
+                // before (the topmost widget wins an idle hover). With a
+                // scroll or a press/release actually happening, only a
+                // widget that registered a matching non-hover sense
+                // consumes it -- a button that only registered `click()`
+                // must not block a scroll meant for the list behind it.
+                let consumed = if momentary_active {
+                    rsc.events_mut()
+                        .get_type::<CursorSense>()
+                        .registered(*id)
+                        .any(|senses| {
+                            matches!(should_run(senses, &cursor, sensor.hover), Some(s) if s.is_momentary())
+                        })
+                } else {
+                    true
+                };
+                if consumed {
+                    sensed = true;
+                }
+
+                let cursor = cursor.clone();
+
+                let data = CursorData {
+                    pos: cursor.pos - region.top_left,
+                    size: region.bot_right - region.top_left,
+                    scroll_delta: cursor.scroll_delta,
+                    hover: sensor.hover,
+                    cursor,
+                    // this does not have any meaning;
+                    // might wanna set up Event to have a prepare stage
+                    sense: CursorSense::Hovering,
+                    render: self,
+                };
+                rsc.run_event::<CursorSense>(*id, data, state);
+            }
+            if sensed {
+                break;
+            }
+        }
+        rsc.events_mut().get_type::<CursorSense>().active = active;
+    }
+}
+
+pub fn should_run(
+    senses: &CursorSenses,
+    cursor: &CursorState,
+    hover: ActivationState,
+) -> Option<CursorSense> {
+    for sense in senses.iter() {
+        if match sense {
+            CursorSense::PressStart(button) => cursor.buttons.select(button).is_start(),
+            CursorSense::Pressing(button) => cursor.buttons.select(button).is_on(),
+            CursorSense::PressEnd(button) => cursor.buttons.select(button).is_end(),
+            CursorSense::HoverStart => hover.is_start(),
+            CursorSense::Hovering => hover.is_on(),
+            CursorSense::HoverEnd => hover.is_end(),
+            CursorSense::Scroll => cursor.scroll_delta != Vec2::ZERO,
+        } {
+            return Some(*sense);
+        }
+    }
+    None
+}
+
+impl ActivationState {
+    pub fn is_start(&self) -> bool {
+        *self == Self::Start
+    }
+    pub fn is_on(&self) -> bool {
+        *self == Self::Start || *self == Self::On
+    }
+    pub fn is_end(&self) -> bool {
+        *self == Self::End
+    }
+    pub fn is_off(&self) -> bool {
+        *self == Self::End || *self == Self::Off
+    }
+    pub fn update(&mut self, on: bool) {
+        *self = match *self {
+            Self::Start => match on {
+                true => Self::On,
+                false => Self::End,
+            },
+            Self::On => match on {
+                true => Self::On,
+                false => Self::End,
+            },
+            Self::End => match on {
+                true => Self::Start,
+                false => Self::Off,
+            },
+            Self::Off => match on {
+                true => Self::Start,
+                false => Self::Off,
+            },
+        }
+    }
+
+    pub fn end_frame(&mut self) {
+        match self {
+            Self::Start => *self = Self::On,
+            Self::End => *self = Self::Off,
+            _ => (),
+        }
+    }
+}
+
+impl EventLike for CursorSense {
+    type Event = CursorSenses;
+    fn into_event(self) -> Self::Event {
+        self.into()
+    }
+}
+
+impl Deref for CursorSenses {
+    type Target = Vec<CursorSense>;
+
+    fn deref(&self) -> &Self::Target {
+        &self.0
+    }
+}
+
+impl DerefMut for CursorSenses {
+    fn deref_mut(&mut self) -> &mut Self::Target {
+        &mut self.0
+    }
+}
+
+impl From<CursorSense> for CursorSenses {
+    fn from(val: CursorSense) -> Self {
+        CursorSenses(vec![val])
+    }
+}
+
+impl BitOr for CursorSense {
+    type Output = CursorSenses;
+
+    fn bitor(self, rhs: Self) -> Self::Output {
+        CursorSenses(vec![self, rhs])
+    }
+}
+
+impl BitOr<CursorSense> for CursorSenses {
+    type Output = Self;
+
+    fn bitor(mut self, rhs: CursorSense) -> Self::Output {
+        self.0.push(rhs);
+        self
+    }
+}

iris/src/default/state.rs

@@ -0,0 +1,75 @@
+use iris_core::{
+    WidgetId,
+    util::{HashMap, HashSet},
+};
+use std::{
+    any::{Any, TypeId},
+    marker::PhantomData,
+};
+
+#[derive(Clone, Copy, PartialEq, Eq, Hash)]
+struct Key {
+    id: WidgetId,
+    ty: TypeId,
+    i: usize,
+}
+
+#[derive(Default)]
+pub struct WidgetState {
+    widgets: HashMap<WidgetId, HashSet<(TypeId, usize)>>,
+    counts: HashMap<(WidgetId, TypeId), usize>,
+    map: HashMap<Key, Box<dyn Any>>,
+}
+
+impl WidgetState {
+    pub fn new() -> Self {
+        Self::default()
+    }
+    pub fn add<T: 'static>(&mut self, id: WidgetId, data: T) -> WeakState<T> {
+        let ty = TypeId::of::<T>();
+        let count = self.counts.entry((id, ty)).or_default();
+        let i = *count;
+        let key = Key { ty, i, id };
+        self.map.insert(key, Box::new(data));
+        self.widgets.entry(id).or_default().insert((ty, i));
+        *count += 1;
+        WeakState {
+            key,
+            _pd: PhantomData,
+        }
+    }
+    pub fn remove(&mut self, id: WidgetId) {
+        for &(ty, i) in self.widgets.get(&id).into_iter().flatten() {
+            self.map.remove(&Key { id, ty, i });
+        }
+    }
+    pub fn get<T: 'static>(&self, state: WeakState<T>) -> &T {
+        self.map.get(&state.key).unwrap().downcast_ref().unwrap()
+    }
+    pub fn get_mut<T: 'static>(&mut self, state: WeakState<T>) -> &mut T {
+        self.map
+            .get_mut(&state.key)
+            .unwrap()
+            .downcast_mut()
+            .unwrap()
+    }
+}
+
+#[derive(Clone, Copy)]
+pub struct WeakState<T> {
+    key: Key,
+    _pd: PhantomData<T>,
+}
+
+pub trait HasWidgetState {
+    fn widget_state(&self) -> &WidgetState;
+    fn widget_state_mut(&mut self) -> &mut WidgetState;
+}
+
+impl<'a, T: 'static> FnOnce<(&'a mut WidgetState,)> for WeakState<T> {
+    type Output = &'a mut T;
+
+    extern "rust-call" fn call_once(self, (state,): (&'a mut WidgetState,)) -> Self::Output {
+        state.get_mut(self)
+    }
+}

iris/src/default/task.rs

@@ -0,0 +1,82 @@
+use iris_core::HasState;
+use std::{
+    pin::Pin,
+    sync::{
+        Arc,
+        mpsc::{Receiver as SyncReceiver, Sender as SyncSender, channel as sync_channel},
+    },
+};
+use tokio::{
+    runtime::Runtime,
+    sync::mpsc::{
+        UnboundedReceiver as AsyncReceiver, UnboundedSender as AsyncSender,
+        unbounded_channel as async_channel,
+    },
+};
+use winit::window::Window;
+
+pub type TaskMsgSender<Rsc> = SyncSender<Box<dyn TaskUpdate<Rsc>>>;
+pub type TaskMsgReceiver<Rsc> = SyncReceiver<Box<dyn TaskUpdate<Rsc>>>;
+
+pub trait TaskUpdate<Rsc: HasState>: FnOnce(&mut Rsc::State, &mut Rsc) + Send {}
+impl<F: FnOnce(&mut Rsc::State, &mut Rsc) + Send, Rsc: HasState> TaskUpdate<Rsc> for F {}
+
+pub struct Tasks<Rsc: HasState> {
+    start: AsyncSender<BoxTask>,
+    window: Arc<Window>,
+    msg_send: SyncSender<Box<dyn TaskUpdate<Rsc>>>,
+}
+
+pub struct TaskCtx<Rsc: HasState> {
+    send: TaskMsgSender<Rsc>,
+}
+
+impl<Rsc: HasState> TaskCtx<Rsc> {
+    pub fn update(&mut self, f: impl TaskUpdate<Rsc> + 'static) {
+        let _ = self.send.send(Box::new(f));
+    }
+}
+impl<Rsc: HasState + 'static> TaskCtx<Rsc> {
+    fn new(send: TaskMsgSender<Rsc>) -> Self {
+        Self { send }
+    }
+}
+
+type BoxTask = Pin<Box<dyn Future<Output = ()> + Send>>;
+
+impl<Rsc: HasState> Tasks<Rsc> {
+    pub fn init(window: Arc<Window>) -> (Self, TaskMsgReceiver<Rsc>) {
+        let (start, start_recv) = async_channel();
+        let (msgs, msgs_recv) = sync_channel();
+        std::thread::spawn(|| {
+            let rt = Runtime::new().unwrap();
+            rt.block_on(listen(start_recv))
+        });
+        (
+            Self {
+                start,
+                msg_send: msgs,
+                window,
+            },
+            msgs_recv,
+        )
+    }
+
+    pub fn spawn<F: AsyncFnOnce(TaskCtx<Rsc>) + 'static + std::marker::Send>(&mut self, task: F)
+    where
+        F::CallOnceFuture: Send,
+    {
+        let send = self.msg_send.clone();
+        let window = self.window.clone();
+        let _ = self.start.send(Box::pin(async move {
+            task(TaskCtx::new(send)).await;
+            window.request_redraw();
+        }));
+    }
+}
+
+async fn listen(mut recv: AsyncReceiver<BoxTask>) {
+    while let Some(task) = recv.recv().await {
+        tokio::spawn(task);
+    }
+}

iris/src/event.rs

@@ -0,0 +1,87 @@
+use iris_core::*;
+use iris_macro::*;
+use std::sync::Arc;
+
+use crate::default::{TaskCtx, TaskUpdate, Tasks};
+
+pub trait Eventable<Rsc: HasEvents, Tag>: WidgetLike<Rsc, Tag> {
+    fn on<E: EventLike>(
+        self,
+        event: E,
+        f: impl for<'a> WidgetEventFn<Rsc, <E::Event as Event>::Data<'a>, Self::Widget>,
+    ) -> impl WidgetIdFn<Rsc, Self::Widget> {
+        move |rsc| {
+            let id = self.add(rsc);
+            rsc.register_event(id, event.into_event(), move |ctx, rsc| {
+                f(
+                    EventIdCtx {
+                        widget: id,
+                        state: ctx.state,
+                        data: ctx.data,
+                    },
+                    rsc,
+                );
+            });
+            id
+        }
+    }
+}
+impl<WL: WidgetLike<Rsc, Tag>, Rsc: HasEvents, Tag> Eventable<Rsc, Tag> for WL {}
+
+widget_trait! {
+    pub trait TaskEventable<Rsc: HasEvents + HasTasks>;
+    fn task_on<'a, E: EventLike, F: AsyncWidgetEventFn<Rsc, WL::Widget>>(
+        self,
+        event: E,
+        f: F,
+    ) -> impl WidgetIdFn<Rsc, WL::Widget>
+        where <E::Event as Event>::Data<'a>: Send,
+        for<'b> F::CallRefFuture<'b>: Send,
+    {
+        let f = Arc::new(f);
+        move |rsc| {
+            let id = self.add(rsc);
+            rsc.register_event(id, event.into_event(), move |_, rsc| {
+                let f = f.clone();
+                rsc.tasks_mut().spawn(async move |task| {
+                    f(AsyncEventIdCtx {
+                        widget: id,
+                        task,
+                    }).await;
+                });
+            });
+            id
+        }
+    }
+}
+
+pub trait HasTasks: Sized + HasState + HasEvents {
+    fn tasks_mut(&mut self) -> &mut Tasks<Self>;
+
+    fn spawn_task<F: AsyncFnOnce(TaskCtx<Self>) + 'static + std::marker::Send>(&mut self, task: F)
+    where
+        F::CallOnceFuture: Send,
+    {
+        self.tasks_mut().spawn(task);
+    }
+}
+
+pub trait AsyncWidgetEventFn<Rsc: HasEvents, W: ?Sized>:
+    AsyncFn(AsyncEventIdCtx<Rsc, W>) + Send + Sync + 'static
+{
+}
+impl<Rsc: HasEvents, F: AsyncFn(AsyncEventIdCtx<Rsc, W>) + Send + Sync + 'static, W: ?Sized>
+    AsyncWidgetEventFn<Rsc, W> for F
+{
+}
+
+pub struct AsyncEventIdCtx<Rsc: HasEvents, W: ?Sized> {
+    pub widget: WeakWidget<W>,
+    task: TaskCtx<Rsc>,
+}
+
+impl<Rsc: HasEvents, W: ?Sized> AsyncEventIdCtx<Rsc, W> {
+    pub fn update(&mut self, f: impl TaskUpdate<Rsc> + 'static) {
+        self.task.update(f);
+    }
+}

iris/src/layout_tests.rs

@@ -0,0 +1,183 @@
+//! Pass conditions for LAYOUT.md section 8, exercised as plain unit tests
+//! rather than through `run-headless.sh`: `UiRenderState` and `Widgets` do
+//! not touch a GPU or a window, so a tree can be built and driven directly.
+//! No GPU-backed rendering (`UiRenderNode`) is exercised here -- only the
+//! CPU-side layout/move machinery LAYOUT.md is about.
+
+use crate::prelude::*;
+
+/// The minimal `UiRsc` a test needs: just the shared `UiData`, none of the
+/// event/window/state plumbing `DefaultRsc` carries.
+struct TestRsc {
+    ui: UiData,
+}
+
+impl UiRsc for TestRsc {
+    fn ui(&self) -> &UiData {
+        &self.ui
+    }
+    fn ui_mut(&mut self) -> &mut UiData {
+        &mut self.ui
+    }
+}
+
+/// A `Scroll` over a `Span` of `n` fixed-height rects -- N primitives large
+/// enough that an O(N) regression in the move path would show up as a
+/// non-trivial counter rather than being lost in noise (LAYOUT.md section
+/// 8, condition 3, using rects rather than glyphs to avoid pulling the font
+/// stack into a plain unit test). Returns the scroll widget (weak, for
+/// mutating it later), the erased root to draw, and the rows (weak, for
+/// hit-testing one of them).
+fn scrolled_rects(
+    rsc: &mut TestRsc,
+    n: usize,
+) -> (WeakWidget<Scroll>, StrongWidget, Vec<WeakWidget<Rect>>) {
+    let mut span = Span::empty(Dir::DOWN);
+    let mut rects = Vec::with_capacity(n);
+    for _ in 0..n {
+        let rect = rsc.ui.widgets.add_strong(Rect::new(UiColor::WHITE));
+        rects.push(rect.weak());
+        // Each row gets a fixed height so the span's total content is
+        // genuinely taller than the viewport -- rest-sized rows would just
+        // divide whatever space is offered and never need scrolling.
+        let row = rsc.ui.widgets.add_strong(Sized {
+            inner: rect.any(),
+            x: None,
+            y: Some(Len::abs(10.0)),
+        });
+        span.push(row.any());
+    }
+    let span = rsc.ui.widgets.add_strong(span);
+    let scroll = rsc.ui.widgets.add_strong(Scroll::new(span.any(), Axis::Y));
+    let weak = scroll.weak();
+    (weak, scroll.any(), rects)
+}
+
+#[test]
+fn an_unchanged_frame_draws_and_rewrites_nothing() {
+    let mut rsc = TestRsc {
+        ui: UiData::default(),
+    };
+    let (_scroll, root, _rects) = scrolled_rects(&mut rsc, 500);
+    let mut render = UiRenderState::new();
+    render.resize((800.0, 20000.0));
+
+    render.update(&root, &mut rsc);
+    render.take_counters(); // discard the first, real draw
+
+    render.update(&root, &mut rsc);
+    let (draws, rewrites, moves) = render.take_counters();
+    assert_eq!((draws, rewrites, moves), (0, 0, 0));
+}
+
+#[test]
+fn scrolling_moves_in_o1_without_a_redraw() {
+    let mut rsc = TestRsc {
+        ui: UiData::default(),
+    };
+    let (scroll, root, _rects) = scrolled_rects(&mut rsc, 500);
+    let mut render = UiRenderState::new();
+    render.resize((800.0, 600.0));
+
+    // The first draw offers `Scroll`'s content a zero-height region
+    // (nothing has been measured yet) and learns the real content length
+    // from what comes back; `update()` only redraws widgets actually
+    // marked dirty, so that corrected length is not reflected in the
+    // content's own *active* region until something -- here a no-op
+    // scroll tick -- actually asks `Scroll` to redraw again. Only after
+    // that warm-up does the content's offered size stop changing between
+    // draws, which is what makes a further, real scroll tick a same-size
+    // move instead of a resize. See scroll.rs.
+    render.update(&root, &mut rsc);
+    rsc.ui.widgets.get_mut(&scroll).unwrap().scroll(0.0);
+    render.update(&root, &mut rsc);
+    render.take_counters();
+
+    // Negative: `scroll`'s sign convention subtracts from `amt`, and
+    // `amt` starts at (and is clamped to) 0 at the top of the content, so
+    // a *positive* argument here would be scrolling further up (a no-op,
+    // already clamped) rather than actually moving anything.
+    rsc.ui.widgets.get_mut(&scroll).unwrap().scroll(-40.0);
+    render.update(&root, &mut rsc);
+    let (draws, _rewrites, moves) = render.take_counters();
+
+    // The pass condition (LAYOUT.md section 8, condition 3) is 0 draws and
+    // 1 move_offsets write, independent of how many rects are in the
+    // scrolled subtree. `draws` here is exactly 1: `Scroll` itself is
+    // marked dirty by `scroll()` and its own body is cheap arithmetic with
+    // no primitives of its own, so it is the one real `Widget::draw` this
+    // counts -- the 500 rects underneath move via the O(1) chain and are
+    // never revisited.
+    assert_eq!(draws, 1, "only Scroll itself should redraw");
+    assert_eq!(moves, 1, "the scrolled subtree should move in one write");
+}
+
+#[test]
+fn hit_testing_follows_a_scrolled_widget() {
+    let mut rsc = TestRsc {
+        ui: UiData::default(),
+    };
+    let (scroll, root, rects) = scrolled_rects(&mut rsc, 500);
+    let mut render = UiRenderState::new();
+    render.resize((800.0, 600.0));
+    render.update(&root, &mut rsc);
+    rsc.ui.widgets.get_mut(&scroll).unwrap().scroll(0.0);
+    render.update(&root, &mut rsc);
+
+    let target = &rects[2];
+    let before = render.resolved_region(target, &rsc).unwrap();
+
+    rsc.ui.widgets.get_mut(&scroll).unwrap().scroll(-37.0);
+    render.update(&root, &mut rsc);
+
+    let after = render.resolved_region(target, &rsc).unwrap();
+    let before_px = before.to_px((800.0, 600.0).into());
+    let after_px = after.to_px((800.0, 600.0).into());
+
+    // Scrolling by -37 moves `amt` from 0 to 37, sliding the content's
+    // top-left up by 37px -- `resolved_region` (the CPU twin of the vertex
+    // shader's chain walk) must reflect that immediately, not the
+    // pre-scroll position, or a tap routed through it would land on
+    // whatever is now at the old coordinates instead of this widget.
+    assert!(
+        (after_px.top_left.y - (before_px.top_left.y - 37.0)).abs() < 0.01,
+        "before={before_px:?} after={after_px:?}"
+    );
+}
+
+#[test]
+fn a_mask_stays_put_while_its_scrolled_content_moves() {
+    let mut rsc = TestRsc {
+        ui: UiData::default(),
+    };
+    let (scroll, inner_root, _rects) = scrolled_rects(&mut rsc, 500);
+    let masked = rsc.ui.widgets.add_strong(Masked { inner: inner_root });
+    let masked_id = masked.id();
+    let root = masked.any();
+    let mut render = UiRenderState::new();
+    render.resize((800.0, 600.0));
+
+    render.update(&root, &mut rsc);
+    rsc.ui.widgets.get_mut(&scroll).unwrap().scroll(0.0);
+    render.update(&root, &mut rsc);
+
+    let masked_slot_before = render.active.get(&masked_id).unwrap().move_slot;
+    let mask_delta_before = rsc.ui.move_offsets[masked_slot_before.idx()].delta;
+
+    rsc.ui.widgets.get_mut(&scroll).unwrap().scroll(-40.0);
+    render.update(&root, &mut rsc);
+
+    let masked_slot_after = render.active.get(&masked_id).unwrap().move_slot;
+    let mask_delta_after = rsc.ui.move_offsets[masked_slot_after.idx()].delta;
+
+    // `Masked` itself is never the target of a `mov`/`reposition` here --
+    // only its scrolled child is -- so the slot its own mask references
+    // (`Painter::set_mask` bakes in `self.move_slot`, i.e. this one) must
+    // still read zero after the scroll. The visible counterpart of this
+    // (the clipped edge follows the scroll while the viewport border does
+    // not) is `iris/run-headless.sh`'s job to catch in a real frame; this
+    // is the numeric half, on the same data the fragment shader's
+    // `resolve_move` reads. See LAYOUT.md section 2b.
+    assert_eq!(mask_delta_before, [0.0, 0.0]);
+    assert_eq!(mask_delta_after, [0.0, 0.0]);
+}

iris/src/lib.rs

@@ -0,0 +1,30 @@
+#![feature(unboxed_closures)]
+#![feature(fn_traits)]
+#![feature(associated_type_defaults)]
+#![feature(unsize)]
+#![feature(option_into_flat_iter)]
+#![feature(async_fn_traits)]
+
+pub mod default;
+pub mod event;
+pub mod widget;
+
+#[cfg(test)]
+mod layout_tests;
+#[cfg(test)]
+mod sense_tests;
+
+pub use iris_core as core;
+pub use iris_macro as macros;
+
+pub mod prelude {
+    use super::*;
+    pub use default::*;
+    pub use event::*;
+    pub use iris_core::*;
+    pub use iris_macro::*;
+    pub use widget::*;
+
+    pub use iris_core::util::Vec2;
+    pub use len_fns::*;
+}

iris/src/sense_tests.rs

@@ -0,0 +1,124 @@
+//! IRIS_TODO.md's "Input does not fall through by input type": a widget
+//! that only registered `click()` used to also block a `Scroll` meant for
+//! whatever is behind it, because `run_sensors` decided "consumed, stop
+//! looking at lower layers" from mere hover, not from anything actually
+//! matching. Exercised as a plain unit test for the same reason
+//! `layout_tests.rs` is one: `UiRenderState` and a minimal `HasEvents`
+//! impl need no GPU or window.
+
+use crate::prelude::*;
+use std::{cell::Cell, rc::Rc};
+
+struct SenseRsc {
+    ui: UiData,
+    events: EventManager<SenseRsc>,
+}
+
+impl UiRsc for SenseRsc {
+    fn ui(&self) -> &UiData {
+        &self.ui
+    }
+    fn ui_mut(&mut self) -> &mut UiData {
+        &mut self.ui
+    }
+    fn on_draw(&mut self, active: &ActiveData) {
+        self.events.draw(active);
+    }
+    fn on_undraw(&mut self, active: &ActiveData) {
+        self.events.undraw(active);
+    }
+    fn on_remove(&mut self, id: WidgetId) {
+        self.events.remove(id);
+    }
+}
+
+impl HasState for SenseRsc {
+    type State = ();
+}
+
+impl HasEvents for SenseRsc {
+    fn events(&self) -> &EventManager<Self> {
+        &self.events
+    }
+    fn events_mut(&mut self) -> &mut EventManager<Self> {
+        &mut self.events
+    }
+}
+
+fn cursor_at(pos: Vec2) -> CursorState {
+    CursorState {
+        pos,
+        exists: true,
+        buttons: Default::default(),
+        scroll_delta: Vec2::ZERO,
+    }
+}
+
+#[test]
+fn a_button_over_a_list_scrolls_the_list_and_still_clicks() {
+    let mut rsc = SenseRsc {
+        ui: UiData::default(),
+        events: EventManager::default(),
+    };
+
+    // Both cover the whole window -- the button "sitting over" the list,
+    // the case in IRIS_TODO.md's report.
+    let list = rsc.ui.widgets.add_strong(Rect::new(UiColor::WHITE));
+    let list_weak = list.weak();
+    let button = rsc.ui.widgets.add_strong(Rect::new(UiColor::RED));
+    let button_weak = button.weak();
+
+    let scrolled = Rc::new(Cell::new(false));
+    let clicked = Rc::new(Cell::new(false));
+    {
+        let scrolled = scrolled.clone();
+        rsc.register_event(list_weak, CursorSense::Scroll, move |_ctx, _rsc| {
+            scrolled.set(true);
+        });
+    }
+    {
+        let clicked = clicked.clone();
+        rsc.register_event(button_weak, CursorSense::click(), move |_ctx, _rsc| {
+            clicked.set(true);
+        });
+    }
+
+    // A Stack draws its children on separate layers in order, which is
+    // exactly the "one thing drawn over another" shape `run_sensors`
+    // walks top layer first.
+    let root = rsc
+        .ui
+        .widgets
+        .add_strong(Stack {
+            children: vec![list.any(), button.any()],
+            size: StackSize::default(),
+        })
+        .any();
+
+    let mut render = UiRenderState::new();
+    render.resize((100.0, 100.0));
+    render.update(&root, &mut rsc);
+
+    let mut state = ();
+    let mut scroll_cursor = cursor_at((50.0, 50.0).into());
+    scroll_cursor.scroll_delta = (0.0, 10.0).into();
+    render.run_sensors(&mut rsc, &mut state, scroll_cursor, (100.0, 100.0).into());
+
+    assert!(
+        scrolled.get(),
+        "a scroll over the button must still reach the list underneath it"
+    );
+    assert!(
+        !clicked.get(),
+        "a scroll is not a click; the button must not have fired"
+    );
+
+    let mut click_cursor = cursor_at((50.0, 50.0).into());
+    click_cursor.buttons.left = ActivationState::Start;
+    render.run_sensors(&mut rsc, &mut state, click_cursor, (100.0, 100.0).into());
+
+    assert!(
+        clicked.get(),
+        "the button on top must still receive an actual click"
+    );
+}

iris/src/widget/image.rs

@@ -0,0 +1,58 @@
+use crate::prelude::*;
+use image::DynamicImage;
+
+pub struct Image {
+    handle: TextureHandle,
+}
+
+impl Widget for Image {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        // Drawn at its own natural size, anchored top-left of whatever it
+        // was offered, not stretched to fill it -- its primitive is
+        // independent of the offered region, matching `is_size_independent`
+        // below. A caller that wants it placed differently wraps it (e.g.
+        // `.center()`, `.align(...)`).
+        let size = self.handle.size();
+        painter.texture_within(&self.handle, size.align(Align::TOP_LEFT));
+        Size::abs(size)
+    }
+
+    fn is_size_independent(&self) -> bool {
+        true // a decoded image's primitive never depends on the region it is offered
+    }
+}
+
+pub fn image<State: UiRsc>(image: impl LoadableImage) -> impl WidgetFn<State, Image> {
+    let image = image.get_image().expect("Failed to load image");
+    move |state| Image {
+        handle: state.ui_mut().textures.add(image),
+    }
+}
+
+pub trait LoadableImage {
+    fn get_image(self) -> Result<DynamicImage, String>;
+}
+
+impl LoadableImage for &str {
+    fn get_image(self) -> Result<DynamicImage, String> {
+        image::open(self).map_err(|e| format!("{e:?}"))
+    }
+}
+
+impl LoadableImage for String {
+    fn get_image(self) -> Result<DynamicImage, String> {
+        image::open(self).map_err(|e| format!("{e:?}"))
+    }
+}
+
+impl<const LEN: usize> LoadableImage for &[u8; LEN] {
+    fn get_image(self) -> Result<DynamicImage, String> {
+        image::load_from_memory(self).map_err(|e| format!("{e:?}"))
+    }
+}
+
+impl LoadableImage for DynamicImage {
+    fn get_image(self) -> Result<DynamicImage, String> {
+        Ok(self)
+    }
+}

iris/src/widget/mask.rs

@@ -0,0 +1,12 @@
+use crate::prelude::*;
+
+pub struct Masked {
+    pub inner: StrongWidget,
+}
+
+impl Widget for Masked {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        painter.set_mask(painter.region());
+        painter.widget(&self.inner)
+    }
+}

iris/src/widget/mod.rs

@@ -0,0 +1,15 @@
+mod image;
+mod mask;
+mod position;
+mod ptr;
+mod rect;
+mod text;
+mod trait_fns;
+
+pub use image::*;
+pub use mask::*;
+pub use position::*;
+pub use ptr::*;
+pub use rect::*;
+pub use text::*;
+pub use trait_fns::*;

iris/src/widget/position/align.rs

@@ -0,0 +1,35 @@
+use crate::prelude::*;
+
+pub struct Aligned {
+    pub inner: StrongWidget,
+    pub align: Align,
+}
+
+impl Widget for Aligned {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        // Draw once at the whole region this widget was offered to learn
+        // the child's real size -- this placement is provisional and
+        // corrected below without a second draw. `painter.widget` (not
+        // `widget_within(..., painter.region())`) is what "my whole,
+        // already-resolved region, unmodified" means: `widget_within`
+        // composes its argument as a *local*, `UiRegion::FULL`-relative
+        // box against `painter.region()`, so handing it the
+        // already-resolved region double-applies that composition and is
+        // wrong for any widget nested below the root.
+        let used = painter.widget(&self.inner);
+        let region = match self.align.tuple() {
+            (Some(x), Some(y)) => used.to_uivec2().align(RegionAlign { x, y }),
+            (Some(x), None) => {
+                let x = used.x.apply_rest().align(x);
+                UiRegion::new(x, UiSpan::FULL)
+            }
+            (None, Some(y)) => {
+                let y = used.y.apply_rest().align(y);
+                UiRegion::new(UiSpan::FULL, y)
+            }
+            (None, None) => UiRegion::FULL,
+        };
+        painter.reposition(&self.inner, region); // O(1): one offset write, no second draw
+        used
+    }
+}

iris/src/widget/position/layer.rs

@@ -0,0 +1,15 @@
+use crate::prelude::*;
+
+pub struct LayerOffset {
+    pub inner: StrongWidget,
+    pub offset: usize,
+}
+
+impl Widget for LayerOffset {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        for _ in 0..self.offset {
+            painter.next_layer();
+        }
+        painter.widget(&self.inner)
+    }
+}

iris/src/widget/position/max_size.rs

@@ -0,0 +1,55 @@
+use crate::prelude::*;
+
+pub struct MaxSize {
+    pub inner: StrongWidget,
+    pub x: Option<Len>,
+    pub y: Option<Len>,
+}
+
+impl MaxSize {
+    /// Caps a reported length at `max`, comparing in pixels since `Len`'s
+    /// rel/abs/rest components are not otherwise comparable.
+    fn clamp(len: Len, max: Option<Len>, output: f32) -> Len {
+        let Some(max) = max else {
+            return len;
+        };
+        let len_px = len.apply_rest().to_abs(output);
+        let max_px = max.apply_rest().to_abs(output);
+        if len_px > max_px { max } else { len }
+    }
+
+    /// The span (in this widget's own local, `UiRegion::FULL`-relative
+    /// terms) to actually offer the child: unconstrained if it already fits
+    /// within `max`, or a box of exactly `max`, anchored at this axis's
+    /// start, if it does not. Needed so the child is never painted bigger
+    /// than the size this widget reports for it -- see the identical
+    /// requirement noted on `Sized::draw`.
+    fn clamp_region(offered_px: f32, max: Option<Len>, output: f32) -> UiSpan {
+        let Some(max) = max else {
+            return UiSpan::FULL;
+        };
+        let max_scalar = max.apply_rest();
+        let max_px = max_scalar.to_abs(output);
+        if offered_px > max_px {
+            max_scalar.align(AxisAlign::Neg)
+        } else {
+            UiSpan::FULL
+        }
+    }
+}
+
+impl Widget for MaxSize {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        let output = painter.output_size();
+        let offered = painter.px_size();
+        let region = UiRegion {
+            x: Self::clamp_region(offered.x, self.x, output.x),
+            y: Self::clamp_region(offered.y, self.y, output.y),
+        };
+        let used = painter.widget_within(&self.inner, region);
+        Size {
+            x: Self::clamp(used.x, self.x, output.x),
+            y: Self::clamp(used.y, self.y, output.y),
+        }
+    }
+}

iris/src/widget/position/mod.rs

@@ -0,0 +1,19 @@
+mod align;
+mod layer;
+mod max_size;
+mod offset;
+mod pad;
+mod scroll;
+mod sized;
+mod span;
+mod stack;
+
+pub use align::*;
+pub use layer::*;
+pub use max_size::*;
+pub use offset::*;
+pub use pad::*;
+pub use scroll::*;
+pub use sized::*;
+pub use span::*;
+pub use stack::*;

iris/src/widget/position/offset.rs

@@ -0,0 +1,13 @@
+use crate::prelude::*;
+
+pub struct Offset {
+    pub inner: StrongWidget,
+    pub amt: UiVec2,
+}
+
+impl Widget for Offset {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        let region = UiRegion::FULL.offset(self.amt);
+        painter.widget_within(&self.inner, region)
+    }
+}

iris/src/widget/position/pad.rs

@@ -0,0 +1,120 @@
+use crate::prelude::*;
+
+pub struct Pad {
+    pub padding: Padding,
+    pub inner: StrongWidget,
+}
+
+impl Widget for Pad {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        let used = painter.widget_within(&self.inner, self.padding.region());
+        let width = self.padding.left + self.padding.right;
+        let height = self.padding.top + self.padding.bottom;
+        Size {
+            x: used.x + Len::abs(width),
+            y: used.y + Len::abs(height),
+        }
+    }
+}
+
+pub struct Padding {
+    pub left: f32,
+    pub right: f32,
+    pub top: f32,
+    pub bottom: f32,
+}
+
+impl Padding {
+    pub const ZERO: Self = Self {
+        left: 0.0,
+        right: 0.0,
+        top: 0.0,
+        bottom: 0.0,
+    };
+
+    pub fn uniform(amt: impl UiNum) -> Self {
+        let amt = amt.to_f32();
+        Self {
+            left: amt,
+            right: amt,
+            top: amt,
+            bottom: amt,
+        }
+    }
+    pub fn region(&self) -> UiRegion {
+        let mut region = UiRegion::FULL;
+        region.x.start.abs += self.left;
+        region.y.start.abs += self.top;
+        region.x.end.abs -= self.right;
+        region.y.end.abs -= self.bottom;
+        region
+    }
+    pub fn x(amt: impl UiNum) -> Self {
+        let amt = amt.to_f32();
+        Self {
+            left: amt,
+            right: amt,
+            top: 0.0,
+            bottom: 0.0,
+        }
+    }
+    pub fn y(amt: impl UiNum) -> Self {
+        let amt = amt.to_f32();
+        Self {
+            left: 0.0,
+            right: 0.0,
+            top: amt,
+            bottom: amt,
+        }
+    }
+
+    pub fn top(amt: impl UiNum) -> Self {
+        let mut s = Self::ZERO;
+        s.top = amt.to_f32();
+        s
+    }
+
+    pub fn bottom(amt: impl UiNum) -> Self {
+        let mut s = Self::ZERO;
+        s.bottom = amt.to_f32();
+        s
+    }
+
+    pub fn left(amt: impl UiNum) -> Self {
+        let mut s = Self::ZERO;
+        s.left = amt.to_f32();
+        s
+    }
+
+    pub fn right(amt: impl UiNum) -> Self {
+        let mut s = Self::ZERO;
+        s.right = amt.to_f32();
+        s
+    }
+
+    pub fn with_top(mut self, amt: impl UiNum) -> Self {
+        self.top = amt.to_f32();
+        self
+    }
+
+    pub fn with_bottom(mut self, amt: impl UiNum) -> Self {
+        self.bottom = amt.to_f32();
+        self
+    }
+
+    pub fn with_left(mut self, amt: impl UiNum) -> Self {
+        self.left = amt.to_f32();
+        self
+    }
+
+    pub fn with_right(mut self, amt: impl UiNum) -> Self {
+        self.right = amt.to_f32();
+        self
+    }
+}
+
+impl<T: UiNum> From<T> for Padding {
+    fn from(amt: T) -> Self {
+        Self::uniform(amt.to_f32())
+    }
+}

iris/src/widget/position/scroll.rs

@@ -0,0 +1,77 @@
+use crate::prelude::*;
+
+pub struct Scroll {
+    inner: StrongWidget,
+    axis: Axis,
+    amt: f32,
+    snap_end: bool,
+    container_len: f32,
+    content_len: f32,
+}
+
+impl Widget for Scroll {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        // The region offered to the child is sized using *last* frame's
+        // content length, not a fresh measurement -- deliberately, so that
+        // an ordinary scroll tick (`amt` changes, content does not) offers
+        // the child the exact same size it was last drawn with, only
+        // shifted. That is what lets `draw_inner` dispatch this as an O(1)
+        // move (LAYOUT.md section 2) instead of a redraw: sizing the region
+        // to a *fresh* measurement would require drawing the child first to
+        // learn it, and a provisional draw almost never matches the
+        // previously active size, forcing a real redraw on every tick. A
+        // genuine content-size change (not just a scroll) therefore lags
+        // one frame before the container's clamp reflects it; the content
+        // length itself (read below from what was actually drawn) is never
+        // stale, so this self-corrects the next frame and never leaves the
+        // scroll range wrong for long. See LAYOUT.md section 4.
+        let axis = self.axis;
+        let output_len = painter.output_size().axis(axis);
+        let container_len = painter.region().axis(axis).len();
+        self.container_len = container_len.to_abs(output_len);
+
+        if self.snap_end {
+            self.amt = self.content_len - self.container_len;
+        }
+        self.update_amt();
+
+        let mut region = UiRegion::FULL;
+        region.axis_mut(axis).end = region.axis(axis).start.offset(self.content_len);
+        let region = region.offset(Vec2::from_axis(axis, -self.amt, 0.0));
+
+        let used = painter.widget_within(&self.inner, region);
+
+        self.content_len = used
+            .axis(axis)
+            .apply_rest()
+            .within_len(container_len)
+            .to_abs(output_len);
+
+        used
+    }
+}
+
+impl Scroll {
+    pub fn new(inner: StrongWidget, axis: Axis) -> Self {
+        Self {
+            inner,
+            axis,
+            amt: 0.0,
+            snap_end: true,
+            container_len: 0.0,
+            content_len: 0.0,
+        }
+    }
+
+    pub fn update_amt(&mut self) {
+        self.amt = self.amt.max(0.0);
+        let len = (self.content_len - self.container_len).max(0.0);
+        self.amt = self.amt.min(len);
+        self.snap_end = self.amt == len;
+    }
+
+    pub fn scroll(&mut self, amt: f32) {
+        self.amt -= amt;
+        self.update_amt();
+    }
+}

iris/src/widget/position/sized.rs

@@ -0,0 +1,33 @@
+use crate::prelude::*;
+
+pub struct Sized {
+    pub inner: StrongWidget,
+    pub x: Option<Len>,
+    pub y: Option<Len>,
+}
+
+impl Widget for Sized {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        // The child is drawn within a region that actually carves out the
+        // fixed axes, not whatever region this widget itself happened to
+        // be offered -- needed so the painted geometry matches the
+        // declared size returned below regardless of how much room a
+        // parent offers. `Aligned`'s single-draw pattern (LAYOUT.md
+        // section 6) draws its child once at its own *full* region to
+        // learn its size, then moves it into place with a pure
+        // translation; that translation is only valid if what got painted
+        // is already the reported size, anchored the same way both times.
+        let mut region = UiRegion::FULL;
+        if let Some(x) = self.x {
+            region.x = x.apply_rest().align(AxisAlign::Neg);
+        }
+        if let Some(y) = self.y {
+            region.y = y.apply_rest().align(AxisAlign::Neg);
+        }
+        let used = painter.widget_within(&self.inner, region);
+        Size {
+            x: self.x.unwrap_or(used.x),
+            y: self.y.unwrap_or(used.y),
+        }
+    }
+}

iris/src/widget/position/span.rs

@@ -0,0 +1,155 @@
+use crate::prelude::*;
+use std::marker::PhantomData;
+
+pub struct Span {
+    pub children: Vec<StrongWidget>,
+    pub dir: Dir,
+    pub gap: f32,
+}
+
+impl Widget for Span {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        let axis = self.dir.axis;
+
+        // Phase 1: draw each child once, at the ambient (unmodified, full)
+        // region a size-only query used to see before this migration, to
+        // learn its length along the layout axis. This paints real
+        // primitives at a provisional slot; phase 2 below places each
+        // child for real via the normal `widget_within` dispatch, which
+        // only actually redraws it when that slot's *size* differs from
+        // this provisional one (most children: a resize, since the
+        // provisional slot is the whole span, not this child's share).
+        let lens: Vec<Len> = self
+            .children
+            .iter()
+            .map(|child| painter.widget(child).axis(axis))
+            .collect();
+
+        let gap_total = self.gap * self.children.len().saturating_sub(1) as f32;
+        let total = lens.iter().fold(Len::abs(gap_total), |s, &l| s + l);
+
+        // Phase 2: place each child for real, using the lengths just
+        // learned -- the same arithmetic this loop always used. The cross-
+        // axis length of *this* draw (used for `Span`'s own reported size
+        // below) falls out of each child's real, resolved-width `used`
+        // here for free -- this is what replaces `desired_ortho`'s former
+        // duplicate simulation of this same loop (see LAYOUT.md section 4).
+        let mut start = UiScalar::rel_min();
+        let mut ortho_len = Len::ZERO;
+        let mut ortho_mixed = false;
+        for (child, &len) in self.children.iter().zip(&lens) {
+            let mut span = UiSpan::FULL;
+            span.start = start;
+            if len.rest > 0.0 {
+                let offset = UiScalar::new(total.rel, total.abs);
+                let rel_end = UiScalar::rel(len.rest / total.rest);
+                let end = (UiScalar::rel_max() + start) - offset;
+                start = rel_end.within(&start.to(end));
+            }
+            start.abs += len.abs;
+            start.rel += len.rel;
+            span.end = start;
+            let mut child_region = UiRegion::from_axis(axis, span, UiSpan::FULL);
+            if self.dir.sign == Sign::Neg {
+                child_region.flip(axis);
+            }
+            let used = painter.widget_within(child, child_region);
+            start.abs += self.gap;
+
+            let ortho = used.axis(!axis);
+            if ortho.rel > 0.0 || ortho.rest > 0.0 {
+                ortho_mixed = true;
+            } else {
+                ortho_len.abs = ortho_len.abs.max(ortho.abs);
+            }
+        }
+        if ortho_mixed {
+            ortho_len = Len::default();
+        }
+
+        let along = if total.rest == 0.0 && total.rel == 0.0 {
+            total
+        } else {
+            Len::default()
+        };
+
+        Size::from_axis(axis, along, ortho_len)
+    }
+}
+
+impl Span {
+    pub fn empty(dir: Dir) -> Self {
+        Self {
+            children: Vec::new(),
+            dir,
+            gap: 0.0,
+        }
+    }
+
+    pub fn gap(mut self, gap: impl UiNum) -> Self {
+        self.gap = gap.to_f32();
+        self
+    }
+
+    pub fn push(&mut self, w: StrongWidget) {
+        self.children.push(w);
+    }
+
+    pub fn pop(&mut self) -> Option<StrongWidget> {
+        self.children.pop()
+    }
+}
+
+pub struct SpanBuilder<State, const LEN: usize, Wa: WidgetArrLike<State, LEN, Tag>, Tag> {
+    pub children: Wa,
+    pub dir: Dir,
+    pub gap: f32,
+    _pd: PhantomData<(State, Tag)>,
+}
+
+impl<Rsc, const LEN: usize, Wa: WidgetArrLike<Rsc, LEN, Tag>, Tag> WidgetFnTrait<Rsc>
+    for SpanBuilder<Rsc, LEN, Wa, Tag>
+{
+    type Widget = Span;
+
+    #[track_caller]
+    fn run(self, rsc: &mut Rsc) -> Self::Widget {
+        Span {
+            children: self.children.add(rsc).arr.into_iter().collect(),
+            dir: self.dir,
+            gap: self.gap,
+        }
+    }
+}
+
+impl<State, const LEN: usize, Wa: WidgetArrLike<State, LEN, Tag>, Tag>
+    SpanBuilder<State, LEN, Wa, Tag>
+{
+    pub fn new(children: Wa, dir: Dir) -> Self {
+        Self {
+            children,
+            dir,
+            gap: 0.0,
+            _pd: PhantomData,
+        }
+    }
+
+    pub fn gap(mut self, gap: impl UiNum) -> Self {
+        self.gap = gap.to_f32();
+        self
+    }
+}
+
+impl std::ops::Deref for Span {
+    type Target = Vec<StrongWidget>;
+
+    fn deref(&self) -> &Self::Target {
+        &self.children
+    }
+}
+
+impl std::ops::DerefMut for Span {
+    fn deref_mut(&mut self) -> &mut Self::Target {
+        &mut self.children
+    }
+}

iris/src/widget/position/stack.rs

@@ -0,0 +1,77 @@
+use std::marker::PhantomData;
+
+use crate::prelude::*;
+
+pub struct Stack {
+    pub children: Vec<StrongWidget>,
+    pub size: StackSize,
+}
+
+impl Widget for Stack {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        let mut picked = None;
+        let mut iter = self.children.iter().enumerate();
+        if let Some((i, child)) = iter.next() {
+            painter.child_layer();
+            let used = painter.widget(child);
+            if matches!(self.size, StackSize::Child(j) if j == i) {
+                picked = Some(used);
+            }
+        }
+        for (i, child) in iter {
+            painter.next_layer();
+            let used = painter.widget(child);
+            if matches!(self.size, StackSize::Child(j) if j == i) {
+                picked = Some(used);
+            }
+        }
+        match self.size {
+            StackSize::Default => Size::default(),
+            StackSize::Child(_) => picked.unwrap_or_default(),
+        }
+    }
+}
+
+#[derive(Default, Debug)]
+pub enum StackSize {
+    #[default]
+    Default,
+    Child(usize),
+}
+
+pub struct StackBuilder<State, const LEN: usize, Wa: WidgetArrLike<State, LEN, Tag>, Tag> {
+    pub children: Wa,
+    pub size: StackSize,
+    _pd: PhantomData<(State, Tag)>,
+}
+
+impl<Rsc, const LEN: usize, Wa: WidgetArrLike<Rsc, LEN, Tag>, Tag> WidgetFnTrait<Rsc>
+    for StackBuilder<Rsc, LEN, Wa, Tag>
+{
+    type Widget = Stack;
+
+    #[track_caller]
+    fn run(self, rsc: &mut Rsc) -> Self::Widget {
+        Stack {
+            children: self.children.add(rsc).arr.into_iter().collect(),
+            size: self.size,
+        }
+    }
+}
+
+impl<State, const LEN: usize, Wa: WidgetArrLike<State, LEN, Tag>, Tag>
+    StackBuilder<State, LEN, Wa, Tag>
+{
+    pub fn new(children: Wa) -> Self {
+        Self {
+            children,
+            size: StackSize::default(),
+            _pd: PhantomData,
+        }
+    }
+
+    pub fn size(mut self, size: StackSize) -> Self {
+        self.size = size;
+        self
+    }
+}

iris/src/widget/ptr.rs

@@ -0,0 +1,47 @@
+use crate::prelude::*;
+use std::marker::{Sized, Unsize};
+
+pub struct WidgetPtr {
+    pub inner: Option<StrongWidget>,
+}
+
+impl Widget for WidgetPtr {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        if let Some(id) = &self.inner {
+            painter.widget(id)
+        } else {
+            Size::ZERO
+        }
+    }
+
+    fn is_size_independent(&self) -> bool {
+        self.inner.is_none()
+    }
+}
+
+impl WidgetPtr {
+    pub fn new() -> Self {
+        Self::default()
+    }
+    pub fn empty() -> Self {
+        Self {
+            inner: Default::default(),
+        }
+    }
+    pub fn set<W: ?Sized + Unsize<dyn Widget>>(&mut self, to: StrongWidget<W>) {
+        self.inner = Some(to)
+    }
+
+    pub fn replace<W: ?Sized + Unsize<dyn Widget>>(
+        &mut self,
+        to: StrongWidget<W>,
+    ) -> Option<StrongWidget> {
+        self.inner.replace(to)
+    }
+}
+
+impl Default for WidgetPtr {
+    fn default() -> Self {
+        Self::empty()
+    }
+}

iris/src/widget/rect.rs

@@ -0,0 +1,48 @@
+use crate::prelude::*;
+
+#[derive(Clone, Copy)]
+pub struct Rect {
+    pub color: UiColor,
+    pub radius: f32,
+    pub thickness: f32,
+    pub inner_radius: f32,
+}
+
+impl Rect {
+    pub fn new(color: UiColor) -> Self {
+        Self {
+            color,
+            radius: 0.0,
+            inner_radius: 0.0,
+            thickness: 0.0,
+        }
+    }
+    pub fn color(mut self, color: UiColor) -> Self {
+        self.color = color;
+        self
+    }
+    pub fn radius(mut self, radius: impl UiNum) -> Self {
+        self.radius = radius.to_f32();
+        self
+    }
+}
+
+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
+    }
+}
+
+pub fn rect(color: UiColor) -> Rect {
+    Rect::new(color)
+}

iris/src/widget/text/build.rs

@@ -0,0 +1,132 @@
+use crate::prelude::*;
+use std::marker::{PhantomData, Sized};
+
+pub struct TextBuilder<State, O = TextOutput, H: WidgetOption<State> = ()> {
+    pub content: String,
+    pub attrs: TextAttrs,
+    pub hint: H,
+    pub output: O,
+    state: PhantomData<State>,
+}
+
+impl<State, O, H: WidgetOption<State>> TextBuilder<State, O, H> {
+    pub fn size(mut self, size: impl UiNum) -> Self {
+        self.attrs.font_size = size.to_f32();
+        self.attrs.line_height = self.attrs.font_size * LINE_HEIGHT_MULT;
+        self
+    }
+    pub fn color(mut self, color: UiColor) -> Self {
+        self.attrs.color = color;
+        self
+    }
+    pub fn family(mut self, family: Family) -> Self {
+        self.attrs.family = family;
+        self
+    }
+    pub fn line_height(mut self, height: f32) -> Self {
+        self.attrs.line_height = height;
+        self
+    }
+    pub fn text_align(mut self, align: impl Into<RegionAlign>) -> Self {
+        self.attrs.align = align.into();
+        self
+    }
+    pub fn center_text(mut self) -> Self {
+        self.attrs.align = Align::CENTER;
+        self
+    }
+    pub fn wrap(mut self, wrap: bool) -> Self {
+        self.attrs.wrap = wrap;
+        self
+    }
+    pub fn editable(self, mode: EditMode) -> TextBuilder<State, TextEditOutput, H> {
+        TextBuilder {
+            content: self.content,
+            attrs: self.attrs,
+            hint: self.hint,
+            output: TextEditOutput { mode },
+            state: PhantomData,
+        }
+    }
+}
+
+impl<Rsc: UiRsc, O> TextBuilder<Rsc, O> {
+    pub fn hint<W: WidgetLike<Rsc, Tag>, Tag>(
+        self,
+        hint: W,
+    ) -> TextBuilder<Rsc, O, impl WidgetOption<Rsc>> {
+        TextBuilder {
+            content: self.content,
+            attrs: self.attrs,
+            hint: move |rsc: &mut Rsc| Some(hint.add_strong(rsc).any()),
+            output: self.output,
+            state: PhantomData,
+        }
+    }
+}
+
+pub trait TextBuilderOutput<State>: Sized {
+    type Output;
+    fn run<H: WidgetOption<State>>(
+        state: &mut State,
+        builder: TextBuilder<State, Self, H>,
+    ) -> Self::Output;
+}
+
+pub struct TextOutput;
+impl<Rsc: UiRsc> TextBuilderOutput<Rsc> for TextOutput {
+    type Output = Text;
+
+    fn run<H: WidgetOption<Rsc>>(
+        state: &mut Rsc,
+        builder: TextBuilder<Rsc, Self, H>,
+    ) -> Self::Output {
+        let buf = TextBuffer::new(&builder.content);
+        let hint = builder.hint.get(state);
+        let mut text = Text {
+            content: builder.content.into(),
+            view: TextView::new(buf, builder.attrs, hint),
+        };
+        text.content.changed = false;
+        text
+    }
+}
+
+pub struct TextEditOutput {
+    mode: EditMode,
+}
+
+impl<State: UiRsc> TextBuilderOutput<State> for TextEditOutput {
+    type Output = TextEdit;
+
+    fn run<H: WidgetOption<State>>(
+        state: &mut State,
+        builder: TextBuilder<State, Self, H>,
+    ) -> Self::Output {
+        let buf = TextBuffer::new(&builder.content);
+        TextEdit::new(
+            TextView::new(buf, builder.attrs, builder.hint.get(state)),
+            builder.output.mode,
+        )
+    }
+}
+
+impl<State, O: TextBuilderOutput<State>, H: WidgetOption<State>> FnOnce<(&mut State,)>
+    for TextBuilder<State, O, H>
+{
+    type Output = O::Output;
+
+    extern "rust-call" fn call_once(self, args: (&mut State,)) -> Self::Output {
+        O::run(args.0, self)
+    }
+}
+
+pub fn wtext<State>(content: impl Into<String>) -> TextBuilder<State> {
+    TextBuilder {
+        content: content.into(),
+        attrs: TextAttrs::default(),
+        hint: (),
+        output: TextOutput,
+        state: PhantomData,
+    }
+}

iris/src/widget/text/edit.rs

@@ -0,0 +1,671 @@
+use crate::prelude::*;
+use iris_core::{TextData, UiColor};
+use parley::{Affinity, Layout, Selection};
+use std::ops::{Deref, DerefMut};
+use winit::{
+    event::KeyEvent,
+    keyboard::{Key, NamedKey},
+};
+
+/// Which way a cursor movement goes. Named here rather than taken from the text
+/// stack so that the key handling below does not have to change when the stack
+/// does; the mapping onto parley lives in one place, in `apply_motion`.
+#[derive(Clone, Copy, PartialEq, Eq, Debug)]
+pub enum Motion {
+    Left,
+    Right,
+    LeftWord,
+    RightWord,
+    Up,
+    Down,
+    LineStart,
+    LineEnd,
+}
+
+pub struct TextEdit {
+    view: TextView,
+    /// `None` when the field is not focused -- which parley's `Selection` has no
+    /// way to say, since it always denotes some position in the text. A
+    /// collapsed selection is a caret; an uncollapsed one is a span.
+    selection: Option<Selection>,
+    history: Vec<(String, Option<Selection>)>,
+    double_hit: Option<usize>,
+    pub mode: EditMode,
+}
+
+#[derive(Clone, Copy, PartialEq)]
+pub enum EditMode {
+    SingleLine,
+    MultiLine,
+}
+
+impl TextEdit {
+    pub fn new(view: TextView, mode: EditMode) -> Self {
+        Self {
+            view,
+            selection: None,
+            history: Default::default(),
+            double_hit: None,
+            mode,
+        }
+    }
+
+    pub fn selected_text(&self) -> Option<String> {
+        let sel = self.selection?;
+        if sel.is_collapsed() {
+            return None;
+        }
+        Some(self.buf.text()[sel.text_range()].to_string())
+    }
+}
+
+impl Widget for TextEdit {
+    fn draw(&mut self, painter: &mut Painter) -> Size {
+        let base = painter.layer;
+        painter.child_layer();
+        let used = self.view.draw(painter);
+        painter.layer = base;
+        let region = self.region();
+
+        let Some(selection) = self.selection else {
+            return used;
+        };
+        let layout = self.view.buf.layout();
+
+        // parley reports selection as boxes in layout space, so bidi and
+        // wrapped lines come out right without this code knowing about either.
+        for (rect, _) in selection.geometry(layout) {
+            let size = vec2(rect.width() as f32, rect.height() as f32);
+            let top_left = vec2(rect.x0 as f32, rect.y0 as f32);
+            painter.primitive_within(
+                RectPrimitive::color(Color::SKY),
+                size.align(Align::TOP_LEFT).offset(top_left).within(&region),
+            );
+        }
+
+        let caret = selection.focus().geometry(layout, CARET_WIDTH);
+        let size = vec2(caret.width() as f32, caret.height() as f32);
+        let top_left = vec2(caret.x0 as f32, caret.y0 as f32);
+        painter.primitive_within(
+            RectPrimitive::color(Color::WHITE),
+            size.align(Align::TOP_LEFT).offset(top_left).within(&region),
+        );
+        used
+    }
+}
+
+const CARET_WIDTH: f32 = 1.0;
+
+pub struct TextEditCtx<'a> {
+    pub text: &'a mut TextEdit,
+    pub data: &'a mut TextData,
+}
+
+impl<'a> TextEditCtx<'a> {
+    /// The layout, brought up to date with the text first.
+    ///
+    /// Every cursor movement and hit test goes through parley's layout, so an
+    /// edit that left it stale would move the caret against the previous text.
+    /// Shaping is skipped when nothing changed, so calling this freely is fine.
+    fn layout(&mut self) -> &Layout<UiColor> {
+        let attrs = self.text.view.attrs.clone();
+        let width = self.text.view.wrap_width();
+        self.text.view.buf.shape(self.data, &attrs, width);
+        self.text.view.buf.layout()
+    }
+
+    /// Keep the selection valid after the text underneath it changed.
+    fn refresh(&mut self) {
+        if let Some(sel) = self.text.selection {
+            let layout = self.layout();
+            self.text.selection = Some(sel.refresh(layout));
+        }
+    }
+
+    pub fn take(&mut self) -> String {
+        let text = self.text.view.buf.text().to_string();
+        self.set("");
+        text
+    }
+
+    pub fn set(&mut self, text: &str) {
+        let text = self.string(text);
+        self.text.view.buf.set_text(text);
+        self.text.view.buf.changed = true;
+        self.text.selection = None;
+    }
+
+    pub fn motion(&mut self, motion: Motion, select: bool) {
+        let Some(sel) = self.text.selection else {
+            return;
+        };
+        let layout = self.layout();
+        // Collapsing a span with an unshifted left/right puts the caret at the
+        // near end rather than moving one character from the focus, which is
+        // what every other editor does.
+        let sel = if !select && !sel.is_collapsed() {
+            match motion {
+                Motion::Left | Motion::LeftWord => {
+                    Selection::from(sel.text_range().start_cursor(layout))
+                }
+                Motion::Right | Motion::RightWord => {
+                    Selection::from(sel.text_range().end_cursor(layout))
+                }
+                _ => apply_motion(sel, layout, motion, false),
+            }
+        } else {
+            apply_motion(sel, layout, motion, select)
+        };
+        self.text.selection = Some(sel);
+    }
+
+    /// Replace the `len` characters before the caret. This is the IME's
+    /// preedit path: it re-sends the whole composition each time.
+    pub fn replace(&mut self, len: usize, text: &str) {
+        let text = self.string(text);
+        for _ in 0..len {
+            self.backspace(false);
+        }
+        self.insert_str(&text);
+    }
+
+    fn string(&self, text: &str) -> String {
+        if self.text.mode == EditMode::SingleLine {
+            text.replace('\n', "")
+        } else {
+            text.to_string()
+        }
+    }
+
+    pub fn insert(&mut self, text: &str) {
+        let text = self.string(text);
+        self.insert_str(&text);
+    }
+
+    fn insert_str(&mut self, text: &str) {
+        if text.is_empty() {
+            return;
+        }
+        self.clear_span();
+        let at = match self.text.selection {
+            Some(sel) => sel.focus().index(),
+            None => return,
+        };
+        let at = at.min(self.text.view.buf.text().len());
+        self.text.view.buf.edit().insert_str(at, text);
+        self.text.view.buf.changed = true;
+        self.set_caret(at + text.len());
+    }
+
+    /// True when there was a span to remove.
+    pub fn clear_span(&mut self) -> bool {
+        let Some(sel) = self.text.selection else {
+            return false;
+        };
+        if sel.is_collapsed() {
+            return false;
+        }
+        let range = sel.text_range();
+        self.text.view.buf.edit().replace_range(range.clone(), "");
+        self.text.view.buf.changed = true;
+        self.set_caret(range.start);
+        true
+    }
+
+    fn set_caret(&mut self, index: usize) {
+        let index = index.min(self.text.view.buf.text().len());
+        let layout = self.layout();
+        self.text.selection = Some(Selection::from_byte_index(
+            layout,
+            index,
+            Affinity::default(),
+        ));
+    }
+
+    pub fn newline(&mut self) {
+        if self.text.mode == EditMode::MultiLine {
+            self.insert_str("\n");
+        }
+    }
+
+    pub fn backspace(&mut self, word: bool) {
+        if self.clear_span() {
+            return;
+        }
+        let Some(sel) = self.text.selection else {
+            return;
+        };
+        let end = sel.focus().index();
+        if end == 0 {
+            return;
+        }
+        let layout = self.layout();
+        let start = if word {
+            sel.focus().previous_logical_word(layout).index()
+        } else {
+            sel.focus().previous_visual(layout).index()
+        };
+        self.delete_range(start, end);
+    }
+
+    pub fn delete(&mut self, word: bool) {
+        if self.clear_span() {
+            return;
+        }
+        let Some(sel) = self.text.selection else {
+            return;
+        };
+        let start = sel.focus().index();
+        if start >= self.text.view.buf.text().len() {
+            return;
+        }
+        let layout = self.layout();
+        let end = if word {
+            sel.focus().next_logical_word(layout).index()
+        } else {
+            sel.focus().next_visual(layout).index()
+        };
+        self.delete_range(start, end);
+    }
+
+    fn delete_range(&mut self, start: usize, end: usize) {
+        let len = self.text.view.buf.text().len();
+        let (start, end) = (start.min(end).min(len), start.max(end).min(len));
+        if start == end {
+            return;
+        }
+        self.text.view.buf.edit().replace_range(start..end, "");
+        self.text.view.buf.changed = true;
+        self.set_caret(start);
+    }
+
+    pub fn select_all(&mut self) {
+        let len = self.text.view.buf.text().len();
+        if len == 0 {
+            return;
+        }
+        let layout = self.layout();
+        let anchor = parley::Cursor::from_byte_index(layout, 0, Affinity::default());
+        let focus = parley::Cursor::from_byte_index(layout, len, Affinity::default());
+        self.text.selection = Some(Selection::new(anchor, focus));
+    }
+
+    pub fn select(&mut self, pos: Vec2, size: Vec2, drag: bool, recent: bool) {
+        let pos = pos - self.text.region().top_left().to_abs(size);
+        let prev_sel = self.text.selection;
+        let prev_hit = self.text.double_hit;
+
+        // The layout borrows `self`, so the whole decision is made in here and
+        // only the answer escapes.
+        let outcome = {
+            let layout = self.layout();
+            let inside =
+                pos.x >= 0.0 && pos.y >= 0.0 && pos.x <= layout.width() && pos.y <= layout.height();
+
+            if !inside {
+                if drag { None } else { Some((None, None)) }
+            } else if drag {
+                prev_sel.map(|sel| (Some(sel.extend_to_point(layout, pos.x, pos.y)), prev_hit))
+            } else {
+                let hit = Selection::from_point(layout, pos.x, pos.y);
+                let index = hit.focus().index();
+                // A second click in the same place takes the word and a third
+                // the line; `double_hit` is what remembers that the previous
+                // click had already grown to a word.
+                Some(if recent && prev_hit == Some(index) {
+                    (Some(Selection::line_from_point(layout, pos.x, pos.y)), None)
+                } else if recent && prev_sel.map(|s| s.focus().index()) == Some(index) {
+                    (
+                        Some(Selection::word_from_point(layout, pos.x, pos.y)),
+                        Some(index),
+                    )
+                } else {
+                    (Some(hit), None)
+                })
+            }
+        };
+
+        if let Some((selection, double_hit)) = outcome {
+            self.text.selection = selection;
+            self.text.double_hit = double_hit;
+        }
+    }
+
+    pub fn deselect(&mut self) {
+        self.text.selection = None;
+        self.text.double_hit = None;
+    }
+
+    pub fn apply_event(&mut self, event: &KeyEvent, modifiers: &Modifiers) -> TextInputResult {
+        let old = (self.text.view.buf.text().to_string(), self.text.selection);
+        let mut undo = false;
+        let res = self.apply_event_inner(event, modifiers, &mut undo);
+        if undo {
+            if let Some((old, selection)) = self.text.history.pop() {
+                self.set(&old);
+                self.text.selection = selection;
+                self.refresh();
+            }
+        } else if self.text.view.buf.text() != old.0 {
+            self.text.history.push(old);
+        }
+        res
+    }
+
+    fn apply_event_inner(
+        &mut self,
+        event: &KeyEvent,
+        modifiers: &Modifiers,
+        undo: &mut bool,
+    ) -> TextInputResult {
+        match &event.logical_key {
+            Key::Named(named) => match named {
+                NamedKey::Backspace => self.backspace(modifiers.control),
+                NamedKey::Delete => self.delete(modifiers.control),
+                NamedKey::Space => self.insert(" "),
+                NamedKey::Enter => {
+                    if modifiers.shift {
+                        self.newline();
+                    } else {
+                        return TextInputResult::Submit;
+                    }
+                }
+                NamedKey::ArrowRight => {
+                    let motion = if modifiers.control {
+                        Motion::RightWord
+                    } else {
+                        Motion::Right
+                    };
+                    self.motion(motion, modifiers.shift);
+                }
+                NamedKey::ArrowLeft => {
+                    let motion = if modifiers.control {
+                        Motion::LeftWord
+                    } else {
+                        Motion::Left
+                    };
+                    self.motion(motion, modifiers.shift);
+                }
+                NamedKey::ArrowUp => self.motion(Motion::Up, modifiers.shift),
+                NamedKey::ArrowDown => self.motion(Motion::Down, modifiers.shift),
+                NamedKey::Home => self.motion(Motion::LineStart, modifiers.shift),
+                NamedKey::End => self.motion(Motion::LineEnd, modifiers.shift),
+                NamedKey::Escape => {
+                    self.deselect();
+                    return TextInputResult::Unfocus;
+                }
+                _ => return TextInputResult::Unused,
+            },
+            Key::Character(text) => {
+                if modifiers.control {
+                    match text.as_str() {
+                        "v" => return TextInputResult::Paste,
+                        "c" => {
+                            if let Some(content) = self.text.selected_text() {
+                                return TextInputResult::Copy(content);
+                            }
+                        }
+                        "x" => {
+                            if let Some(content) = self.text.selected_text() {
+                                self.clear_span();
+                                return TextInputResult::Copy(content);
+                            }
+                        }
+                        "a" => self.select_all(),
+                        "z" => *undo = true,
+                        _ => self.insert(text),
+                    }
+                } else {
+                    self.insert(text);
+                }
+            }
+            _ => return TextInputResult::Unused,
+        }
+        TextInputResult::Used
+    }
+}
+
+fn apply_motion(
+    sel: Selection,
+    layout: &Layout<UiColor>,
+    motion: Motion,
+    extend: bool,
+) -> Selection {
+    match motion {
+        Motion::Left => sel.previous_visual(layout, extend),
+        Motion::Right => sel.next_visual(layout, extend),
+        Motion::LeftWord => sel.previous_visual_word(layout, extend),
+        Motion::RightWord => sel.next_visual_word(layout, extend),
+        Motion::Up => sel.previous_line(layout, extend),
+        Motion::Down => sel.next_line(layout, extend),
+        Motion::LineStart => sel.line_start(layout, extend),
+        Motion::LineEnd => sel.line_end(layout, extend),
+    }
+}
+
+/// The ends of a byte range as cursors, so collapsing a selection can put the
+/// caret at whichever end the movement asked for.
+trait RangeCursors {
+    fn start_cursor(&self, layout: &Layout<UiColor>) -> parley::Cursor;
+    fn end_cursor(&self, layout: &Layout<UiColor>) -> parley::Cursor;
+}
+
+impl RangeCursors for std::ops::Range<usize> {
+    fn start_cursor(&self, layout: &Layout<UiColor>) -> parley::Cursor {
+        parley::Cursor::from_byte_index(layout, self.start, Affinity::default())
+    }
+    fn end_cursor(&self, layout: &Layout<UiColor>) -> parley::Cursor {
+        parley::Cursor::from_byte_index(layout, self.end, Affinity::default())
+    }
+}
+
+#[derive(Default)]
+pub struct Modifiers {
+    pub shift: bool,
+    pub control: bool,
+}
+
+impl Modifiers {
+    pub fn clear(&mut self) {
+        self.shift = false;
+        self.control = false;
+    }
+}
+
+pub enum TextInputResult {
+    Used,
+    Unused,
+    Unfocus,
+    Submit,
+    Copy(String),
+    Paste,
+}
+
+impl TextInputResult {
+    pub fn unfocus(&self) -> bool {
+        matches!(self, TextInputResult::Unfocus)
+    }
+}
+
+impl Deref for TextEdit {
+    type Target = TextView;
+
+    fn deref(&self) -> &Self::Target {
+        &self.view
+    }
+}
+
+impl DerefMut for TextEdit {
+    fn deref_mut(&mut self) -> &mut Self::Target {
+        &mut self.view
+    }
+}
+
+pub trait TextEditable {
+    fn edit<'a>(&self, ui: &'a mut impl UiRsc) -> TextEditCtx<'a>;
+}
+
+impl<I: IdLike<Widget = TextEdit>> TextEditable for I {
+    fn edit<'a>(&self, ui: &'a mut impl UiRsc) -> TextEditCtx<'a> {
+        let ui = ui.ui_mut();
+        TextEditCtx {
+            text: ui.widgets.get_mut(self).unwrap(),
+            data: &mut ui.text,
+        }
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use iris_core::{TextAttrs, TextBuffer};
+
+    /// The editor is the one part of iris that is pure logic over a string and
+    /// a layout, and it was rewritten wholesale when the text stack changed --
+    /// so it is the one part worth testing directly. Everything else here
+    /// needs a GPU and a window.
+    fn edit(text: &str, mode: EditMode) -> (TextEdit, TextData) {
+        let view = TextView::new(TextBuffer::new(text), TextAttrs::default(), None);
+        (TextEdit::new(view, mode), TextData::default())
+    }
+
+    fn ctx<'a>(text: &'a mut TextEdit, data: &'a mut TextData) -> TextEditCtx<'a> {
+        TextEditCtx { text, data }
+    }
+
+    fn content(text: &TextEdit) -> String {
+        text.buf.text().to_string()
+    }
+
+    #[test]
+    fn insert_at_the_caret() {
+        let (mut t, mut d) = edit("ac", EditMode::SingleLine);
+        ctx(&mut t, &mut d).set_caret(1);
+        ctx(&mut t, &mut d).insert("b");
+        assert_eq!(content(&t), "abc");
+        assert_eq!(t.selection.unwrap().focus().index(), 2);
+    }
+
+    #[test]
+    fn backspace_removes_the_character_before_the_caret() {
+        let (mut t, mut d) = edit("abc", EditMode::SingleLine);
+        ctx(&mut t, &mut d).set_caret(2);
+        ctx(&mut t, &mut d).backspace(false);
+        assert_eq!(content(&t), "ac");
+    }
+
+    #[test]
+    fn backspace_at_the_start_does_nothing() {
+        let (mut t, mut d) = edit("abc", EditMode::SingleLine);
+        ctx(&mut t, &mut d).set_caret(0);
+        ctx(&mut t, &mut d).backspace(false);
+        assert_eq!(content(&t), "abc");
+    }
+
+    #[test]
+    fn delete_removes_the_character_after_the_caret() {
+        let (mut t, mut d) = edit("abc", EditMode::SingleLine);
+        ctx(&mut t, &mut d).set_caret(1);
+        ctx(&mut t, &mut d).delete(false);
+        assert_eq!(content(&t), "ac");
+    }
+
+    #[test]
+    fn delete_at_the_end_does_nothing() {
+        let (mut t, mut d) = edit("abc", EditMode::SingleLine);
+        ctx(&mut t, &mut d).set_caret(3);
+        ctx(&mut t, &mut d).delete(false);
+        assert_eq!(content(&t), "abc");
+    }
+
+    #[test]
+    fn select_all_then_typing_replaces_everything() {
+        let (mut t, mut d) = edit("hello", EditMode::SingleLine);
+        ctx(&mut t, &mut d).select_all();
+        assert_eq!(t.selected_text().as_deref(), Some("hello"));
+        ctx(&mut t, &mut d).insert("x");
+        assert_eq!(content(&t), "x");
+    }
+
+    #[test]
+    fn clearing_a_span_leaves_the_caret_at_its_start() {
+        let (mut t, mut d) = edit("abcdef", EditMode::SingleLine);
+        ctx(&mut t, &mut d).select_all();
+        assert!(ctx(&mut t, &mut d).clear_span());
+        assert_eq!(content(&t), "");
+        assert_eq!(t.selection.unwrap().focus().index(), 0);
+    }
+
+    #[test]
+    fn a_single_line_field_refuses_newlines() {
+        let (mut t, mut d) = edit("", EditMode::SingleLine);
+        ctx(&mut t, &mut d).set_caret(0);
+        ctx(&mut t, &mut d).insert("a\nb");
+        assert_eq!(content(&t), "ab");
+        ctx(&mut t, &mut d).newline();
+        assert_eq!(content(&t), "ab");
+    }
+
+    #[test]
+    fn a_multi_line_field_keeps_newlines() {
+        let (mut t, mut d) = edit("", EditMode::MultiLine);
+        ctx(&mut t, &mut d).set_caret(0);
+        ctx(&mut t, &mut d).insert("a\nb");
+        assert_eq!(content(&t), "a\nb");
+    }
+
+    #[test]
+    fn take_empties_the_field_and_hands_back_what_was_there() {
+        let (mut t, mut d) = edit("some text", EditMode::SingleLine);
+        assert_eq!(ctx(&mut t, &mut d).take(), "some text");
+        assert_eq!(content(&t), "");
+    }
+
+    /// The IME's preedit path: each keystroke resends the whole composition,
+    /// so `replace` has to remove exactly what it added last time.
+    #[test]
+    fn ime_preedit_replaces_its_own_previous_text() {
+        let (mut t, mut d) = edit("", EditMode::SingleLine);
+        ctx(&mut t, &mut d).set_caret(0);
+        ctx(&mut t, &mut d).replace(0, "n");
+        assert_eq!(content(&t), "n");
+        ctx(&mut t, &mut d).replace(1, "ni");
+        assert_eq!(content(&t), "ni");
+        ctx(&mut t, &mut d).replace(2, "に");
+        assert_eq!(content(&t), "に");
+    }
+
+    #[test]
+    fn motion_moves_the_caret_and_shift_extends_a_span() {
+        let (mut t, mut d) = edit("abc", EditMode::SingleLine);
+        ctx(&mut t, &mut d).set_caret(0);
+        ctx(&mut t, &mut d).motion(Motion::Right, false);
+        assert_eq!(t.selection.unwrap().focus().index(), 1);
+        ctx(&mut t, &mut d).motion(Motion::Right, true);
+        assert_eq!(t.selected_text().as_deref(), Some("b"));
+    }
+
+    /// Collapsing a span with an unshifted arrow goes to the near end rather
+    /// than stepping one character from the focus.
+    #[test]
+    fn an_unshifted_arrow_collapses_a_span_to_its_edge() {
+        let (mut t, mut d) = edit("abcdef", EditMode::SingleLine);
+        ctx(&mut t, &mut d).select_all();
+        ctx(&mut t, &mut d).motion(Motion::Left, false);
+        assert_eq!(t.selection.unwrap().focus().index(), 0);
+
+        ctx(&mut t, &mut d).select_all();
+        ctx(&mut t, &mut d).motion(Motion::Right, false);
+        assert_eq!(t.selection.unwrap().focus().index(), 6);
+    }
+
+    /// Byte offsets, not character counts: a caret placed after a multi-byte
+    /// character must not split it.
+    #[test]
+    fn multibyte_text_is_edited_by_byte_offset() {
+        let (mut t, mut d) = edit("aé", EditMode::SingleLine);
+        ctx(&mut t, &mut d).set_caret(3);
+        ctx(&mut t, &mut d).backspace(false);
+        assert_eq!(content(&t), "a");
+    }
+}