Repository navigation
Conversation
… fix) into the scroll lock's base
…xposes no ontouchstart is taken once Chromium fixes ontouchstart when the page is created, so turning touch emulation on later sends Touch Events without it: the host reads Pointer Events for the finger and must not read its Touch Events too. A one-finger long press fails here, the finger tracked twice.
…from their pointer events install() reads a finger from its pointer events when the window has no ontouchstart, but the four touch handlers still took the Touch Events a browser sends anyway (Chromium after touch emulation turns on), tracking the finger twice, so a one-finger long press never recognized. Each touch handler now returns while usesTouchEvents is false, after noting the fingers down, as install()'s comment promised.
…ter events) into meta#8320
…ter events) into meta#8293
…ter events) into meta#8356
…ter events) into meta#8358
…n before the click guard internal/motion.ts (meta#8562) re-exports UIKIT_CLICK from the barrel, so a tree carrying both fails typecheck and the Storybook build since this pull request dropped it. The barrel keeps it, with the click guard's window, and the table of every gesture group still leaves it out. Fails until the barrel exports it.
…'s window This pull request dropped UIKIT_CLICK when the click guard took the stray-click window over, but importers written before it read the barrel's UIKIT_CLICK (internal/motion.ts in meta#8562 re-exports it), so a tree carrying both failed typecheck and the Storybook build. UIKIT_CLICK stays in the barrel, deprecated, with the guard's 400 ms (CLICK_GUARD_WINDOW_MS), and UIKIT_GESTURES still leaves it out.
…ter proofs
spec:AST-094 FR43 has gestures attach for every pointer, so each FR43 proof needs a pen and a wheel as real engine input, not only a finger. webInputDriver(page) adds wheel({x, y, dx, dy, ctrl}) and pen(points): in Chromium through CDP Input.dispatchMouseEvent, with a wheel step's deltas scaled back so the page receives exactly its CSS px (Chromium divides them by the device scale factor times the visual viewport's scale: a raw 120 px step reached a phone page as 40), and a pen stroke with pointerType pen, pressure, tilt and scripted timestamps; in WebKit through page.mouse.wheel in a desktop context. WebKit's phone emulation has no wheel and Playwright has no pen for WebKit: the driver's wheels and pens flags say so, with WHEEL_SKIP and PEN_SKIP. The input log also records wheel steps and pointer pressure.
…ter events) into meta#8360
…ter events) into the every-pointer host # Conflicts: # packages/core/src/gestures/__tests__/GestureHost.test.ts
…k) into the FR43 harness helper Keeps the helper's harness self-test green in WebKit too: meta#8535 fixes the backward case's flake under load.
…Core's internal entries Main's internal/touch, its test, the touch-support bundle test, the internal-entries test, sync-exports and the exports map are kept exactly; this branch adds only what main lacks: internal/motion, the named re-exports Lab builds on (the pan recognizer, the gesture types, the touch presentation queries), their tests and the COVERED list. # Conflicts: # packages/core/package.json # packages/core/src/internal/touch.test.ts # packages/core/src/internal/touch.ts # scripts/core-internal-entries.test.mjs # scripts/sync-exports.js # scripts/touch-support-bundle.test.mjs
On a page whose wheel listeners are all passive (every page with the input log), Chromium answers Input.dispatchMouseEvent before the page has the wheel step and delivers it afterwards, so a log read right after wheel() could miss it. wheel() now arms a one-shot passive capture listener before the dispatch and waits until the page has the step (2 s limit), in WebKit too. The ctrl-only note was wrong: a ctrl step reaches an all-passive page as well, and nothing zooms. The docs cite spec:AST-100 FR4, which is current.
… into the FR43 harness helper
Bundled with the size report's esbuild options, none of Core's public entries carries a byte of gestures/ until a component imports it (architecture:react-component-runtime INV13). The click guard in utils/ imports the haptic switch's attribute from gestures/, so the root and hooks entries carry gestures/hapticSwitch.ts: both cases fail until it moves.
…blic entries carry no gesture code The click guard (utils/clickGuard.ts) imported HAPTIC_SWITCH_ATTRIBUTE from gestures/hapticSwitch.ts, which put gesture code in Core's root and hooks entries. The attribute moves to utils/hapticSwitch.ts, as the coarse-pointer query did; the click guard, the gesture host and haptics read it there, and gestures/hapticSwitch.ts re-exports it, deprecated, for importers written before the move. The module-graph case follows the new path.
This pull request carried meta#8666 only up to b246458; meta#8666's head adds its main merge and its later harness text. The two conflicts are meta#8666's harness files, resolved with meta#8666's author keeping both sides: - internal/interaction-harness/README.md: meta#8360's module and engine-limits tables with meta#8666's webInput row and its wheel and pen rows, both engine columns (the same text as meta#8666 merged onto meta#8360); - harness.interaction.browser.spec.ts: meta#8360's header with meta#8666's wheel and pen clause, and every test from both sides; this pull request's own changes to the backward-run timing stay.
…esture host's subscriptions The gesture host subscribes no click: the one cancelling click subscriber is the click guard's, added at a window's first arm and kept. The release note, the stream's header and the gestures README listed click with the host's touchend and contextmenu; they now name the guard's own.
… lift and the browser's menu, keeps a kept press's finger, and goes when the page hides From the review of this pull request, three cases fail on its head: - an armed press installed the page listeners without the claiming ones, so a claimed finger's touchend stayed passive and the browser's menu opened during a claimed long press (spec:AST-094 FR21); - a kept press got its blocking touchmove during an armed press, but its moves were cancelled only where the primary pointer is coarse, so a held stepper on a touchscreen laptop lost its finger (FR11, DEC-21); - a page hidden mid-press cancelled the press but kept the page's listeners until the next lift anywhere (FR24).
…lift and the browser's menu, keeps a kept press's finger, and goes when the page hides - A finger's or a pen's armed press now brings the claiming listeners (the cancelable touchend and the contextmenu canceller) with the page listeners, as wherever the host installs, so a claimed lift clicks nothing and a claimed hold opens no browser menu (spec:AST-094 FR21). A mouse's press needs neither and adds neither. - A kept press keeps its finger under any pointer: its target blocks and its moves are cancelled once its press is delivered, not only where the primary pointer is coarse (FR11, DEC-21). - A page hidden mid-press releases the armed press as a blur does, so the page's listeners go with the cancelled press (FR24).
…'s selection An action group's press is a long press with no minimum duration, so it claims at the press's pointerdown, before any movement. The drag selection hold then cleared the page's selection, so a plain click on a group's button left its Copy or Cut with nothing to act on. The case fails on this head: the selection is gone after the pointerdown.
…age's selection The drag selection hold cleared the page's selection when any mouse or pen contact was claimed, but an action group's press claims at its pointerdown, before it has moved or selected anything, so a plain click on a group's button left its Copy or Cut with nothing to act on. A claim before any movement now keeps the page's selection: the hold still refuses a new one, and puts the page's back over a selection that appears while it runs. A claim after movement clears what its drag selected, as before (spec:AST-094 FR41). The case before the new one now waits for its hold to end a task after the lift, so no hold outlives that test.
…t every pointer does - GestureHost's header, the gestures README and coarsePointer.ts no longer describe the coarse-only host: a finger or a pen takes the touch-down lock under any primary pointer, a kept press's finger is kept during a fine pointer's press too, and recognizers attach for every pointer while the page listeners last a press (spec:AST-094 FR19, FR24, FR41, FR43). - The fine-pointer selection case asserted the replaced rule and passed only because it read iOS WebKit's inline lock without iOS WebKit: it now asserts the lock a fine pointer's finger and pen take, and that a mouse takes none. - FR41's sentence and its verification row say that a press claimed at its own pointerdown, before it moves, keeps the page's selection, while a drag that moved before its claim loses what it selected.
…ion, in the present tense The sentence said a drag that moved before its claim loses the selection it started, but the hold clears the page's selection then, whatever made it; and an unclaimed mouse drag kept "today's" behavior, a phrase that ages. It now says a claim after movement clears the page's selection, a press claimed at its own pointerdown keeps it, and an unclaimed mouse drag keeps its released behavior.
meta#8665's review fixes: an armed press under a fine pointer brings the claiming listeners, keeps a kept press's finger and goes when the page hides; a press claimed at its own pointerdown keeps the page's selection; FR41 says what a claimed drag does with the selection; meta#8360's input-stream notes. The merge is clean.
…he page presented Two proofs failed on a loaded machine with no defect in the control: - the busy-main-thread settle counted the screencast's delivered frames against its cadence. It now hands the compositor's own frame record to busyWindowProof, whose verdict alone judges the settle; a settle paused before the block fails it. - the catch read the settle's position before the touch, which lands later under load. A capture listener now records where the settle stood as the finger landed; a catch that restarts from zero travel fails it.
vjeux
marked this pull request as draft
October 29, 2026 22:39
![astracat-bot[bot]](https://avatars.githubusercontent.com/in/4796231?s=60&v=4){kind=small}
There was a problem hiding this comment.
Approval waits on an engineering owner's decision about new spec:AST-094 FR41 text this branch carries ("A mouse or pen drag that a recognizer claims takes no new text selection for that drag only…"), which is not on main. An engineering owner records that decision by commenting /approve-spec with this pull request's full head SHA, which the author may do as an engineering owner, or by landing the text on main through its own pull request. The trackpad and wheel pull itself matches FR43, which is current on main.
- Please rebase once the carried pull requests land, so the diff holds only the RefreshControl change (639 of its 38,912 authored lines).
scripts/touch-support-bundle.test.mjsfails on this branch becauseuseScrollLock.tsimportsgestures/pageHold.tsinto Core's root and hooks entries; it needs #8671 first, or that import kept out of those entries.
[Automated review]
Comment on lines
+471
to
+475
| (`spec:AST-017` FR59; DEC-9). A mouse or pen drag that a recognizer claims | ||
| takes no new text selection for that drag only. A claim made after the | ||
| drag has moved clears the page's selection; a press claimed at its own | ||
| pointerdown, before it moves, keeps it. An unclaimed mouse drag keeps its | ||
| released behavior. |
There was a problem hiding this comment.
This FR41 text is not on main, and no engineering owner has approved it at this head. Approval of this pull request waits on that decision: an engineering owner comments /approve-spec with the full head SHA, or the text lands on main through its own pull request first.
…dy part The proof failed on a loaded machine with one long task: it judged long tasks over the whole trace, the pull's trigger and the release's render included, while Layout and Paint were judged only in the steady part. Under a 4x CPU throttle the release's last commit frame also landed past the fixed 50 ms start. The steady part now starts once the page has rendered three frames after the settle began, counted in frames, and long tasks are judged there too. It passes 4 of 4 under the throttle; a paint on every frame of the settle, or an 80 ms block inside it, still fails it.
There was a problem hiding this comment.
Approval waits on an engineering owner's decision about the new spec:AST-094 FR41 text this branch carries ("A mouse or pen drag that a recognizer claims takes no new text selection for that drag only…"), which is not on main; the same decision settles whether RefreshControl's mouse pull, a watcher that never claims, may clear the drag's selection. An engineering owner records it by commenting /approve-spec with this pull request's full head SHA, which the author may do as an engineering owner, or by landing the text on main through its own pull request. The trackpad and wheel pull itself matches FR43, which is current on main, and the settle proof's new window looks sound.
- Please rebase once the carried pull requests land, so the diff holds only the RefreshControl change (639 of its 38,952 authored lines).
scripts/touch-support-bundle.test.mjsfails on this branch becauseuseScrollLock.tsimportsgestures/pageHold.tsinto Core's root and hooks entries; it needs #8671 first, or that import kept out of those entries.
[Automated review]
This branch had an error being deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Depends on #8688 (the pen and mouse pull) and #8676 (wheel and trackpad input). Carries #8562 (Core's internal entries) as a dependency; if the entries land differently, this follows her decision. Review only this PR's change: vjeux/astryx@933330e...e6dbcd1; later heads only merge #8688 and #8676.
Why
On a laptop, a trackpad's two-finger scroll or a mouse wheel at the top of a refreshable list only rubber-bands the page: without a touch screen or pen, nothing pulls to refresh.
What led to it
Part of the touch-presentations program (
spec:AST-094), which gives Astryx native-feeling touch behavior. Its FR43, now onmain, has a pan take wheel and trackpad scrolling. #8688 made a pen's and a mouse's drag pull; this adds a trackpad's and a wheel's scroll.Options and tradeoffs
Why this choice
The second pan, with notches in, follows FR43 as written. It claims the scrolls it takes, so its blocking
wheellistener is a claimer's (FR20); the touch pull stays a watcher. It drives the same pull as a finger and keeps the page's text selection. Whether a notch should pull less on short lists is asked on #8646.How to check it
In Chromium or WebKit, on the RefreshControl story "On the page":
Evidence
touch-support-bundlefails from fix(useScrollLock): keep the page where it is while an overlay is open #8098's scroll lock, which this branch carries; test(core): a touch component's import carries its gesture code, and nothing else does #8671's per-component check replaces it and passes here.