Fixing race condition #6551
Open
Fixing race condition #6551
+6
−2
Conversation
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
dosubot
bot
added
the
size:XS
🎭 Playwright Test Results⏰ Completed at: 11/03/2025, 05:02:18 AM UTC 📈 Summary
📊 Test Reports by Browser
🎉 Click on the links above to view detailed test results for each browser configuration. |
🎨 Storybook Build Status
⏰ Started at: 11/03/2025, 03:38:59 AM UTC 🚀 Building Storybook
⏱️ Please wait while the Storybook build is in progress... |
Bundle Size ReportSummary
Category Glance Per-category breakdownApp Entry Points — 3.27 MB (baseline 3.27 MB) • 🟢 -15 BMain entry bundles and manifests
Status: 3 added / 3 removed Graph Workspace — 725 kB (baseline 725 kB) • 🟢 -24 BGraph editor runtime, canvas, workflow orchestration
Status: 1 added / 1 removed Views & Navigation — 8.18 kB (baseline 8.18 kB) • ⚪ 0 BTop-level views, pages, and routed surfaces
Status: 1 added / 1 removed Panels & Settings — 293 kB (baseline 293 kB) • ⚪ 0 BConfiguration panels, inspectors, and settings screens
Status: 6 added / 6 removed UI Components — 12.5 kB (baseline 12.6 kB) • 🟢 -111 BReusable component library chunks
Status: 1 added / 1 removed Data & Services — 11.4 kB (baseline 10.4 kB) • 🔴 +1 kBStores, services, APIs, and repositories
Status: 1 added / 1 removed Utilities & Hooks — 1.07 kB (baseline 1.07 kB) • ⚪ 0 BHelpers, composables, and utility bundles
Vendor & Third-Party — 5.32 MB (baseline 5.32 MB) • ⚪ 0 BExternal libraries and shared vendor chunks
Other — 2.55 MB (baseline 2.55 MB) • ⚪ 0 BBundles that do not match a named category
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.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.
Summary
Original PR and discussion: Comfy-Org/litegraph.js#1082
Original Example HTML to visually describe the issue: https://github.com/Comfy-Org/litegraph.js/blob/8e5c4452cd67ccd91146ba79514bb001c470900a/public/mousemoving_example_with_blur.html
The ordering of mouseup and mousemove events is not defined. So there may be a mouseup event before the next mousemove event where buttons=0, but there may also be a mouseup immediately after.
The original code was designed to handle the case where a user click and drags, then tabs out. It turns out, you still get mousemove events in that case, so we can reset the CanvasPointer the moment it happens, as opposed to attempting to detect if it did happen in the past. The original code was designed under the assumption that if there was no mouseup, and then there was a mousemove with buttons=0, then we must have lost the focus at some point. Unfortunately, on some computers (notably my Macbook M2) this has the side effect of making it so that if your mouse pointer was even slightly moving when releasing the mouse button, it would reset and not connect the graph nodes. Extremely frustrating.
Changes
Review Focus
To replicate this, use a high pixel density display and a sensitive mouse device, such as a Retina Display and a macbook touchpad. This also happens in low resolution displays with less sensitive mouse devices, but the race condition is much harder to create, because you need to move at least 1 pixel during the release of the mouse button. If you want to replicate the bug, and you're on a non-Retina display or you have a basic USB mouse, you'll need to move your mouse rather vigorously. If you're on a macbook M2, just use the computer normally.
┆Issue is synchronized with this Notion page by Unito