Skip to content

Capture html5 clips from a freshly loaded document - #218

Merged
dazzatronus merged 1 commit into
mainfrom
derk/eng-1094-capture-html5-from-fresh-document
Oct 8, 2026
Merged

dazzatronus merged 1 commit into
mainfrom
derk/eng-1094-capture-html5-from-fresh-document

Conversation

@dazzatronus

Copy link
Copy Markdown
Contributor

The html5 player captured frames from the same iframe the editor seeks while you scrub. A seek can leave state behind that a freshly loaded page doesn't have. For example, once a GSAP tween has rendered and been rewound, its start values stay inline. The captured frames therefore depended on where the playhead had been, and could differ from a render that starts from a fresh page. One case we measured: an identity transform left inline changed text antialiasing on about 5% of pixels.

runCapture now reloads the iframe's srcdoc and waits for it to load before capturing. Cached captures skip the reload. The player is already in capturing mode with the iframe parked, so the reload isn't visible.

Verify: npm run verify:ci. The new test scrubs the iframe, starts a capture, and checks that the first seek happens only after a fresh load of the composed document.

@dazzatronus
dazzatronus merged commit c8378e7 into main Oct 8, 2026
1 check passed
github-actions Bot pushed a commit that referenced this pull request Oct 8, 2026
## [2.24.1](v2.24.0...v2.24.1) (2026-10-08)

### Bug Fixes

* capture html5 clips from a freshly loaded document ([#218](#218)) ([c8378e7](c8378e7))
* leave scripts out of html5 frame capture ([#219](#219)) ([b296049](b296049))
@github-actions

github-actions Bot commented Oct 8, 2026

Copy link
Copy Markdown

🎉 This PR is included in version 2.24.1 🎉

The release is available on:

Your semantic-release bot 📦🚀

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant