Philipp Albrecht

I put together a hacky solution passing [video_size_preserved_after_ended.html](https://searchfox.org/mozilla-central/source/testing/web-platform/tests/html/semantics/embedded-content/the-video-element/video_size_preserved_after_ended.html) locally to get some feedback via `./mach try wpt`. Turns out there are *a lot* of `TIMEOUT`s now (with the occasional `FAIL`...

`/html/semantics/embedded-content/media-elements/track/track-element/track-mode-not-changed-by-new-track.html` times out locally as well (however without the ALSA errors). Output with --headless ``` ~/dev/personal/servo/servo שׂ ready-states % ./mach test-wpt --headless tests/wpt/tests/html/semantics/embedded-content/media-elements/track/track-element/track-mode-not-changed-by-new-track.html No build type specified so assuming `--dev`....

Just for the record: I only "verified" flakiness with this test: [video_size_preserved_after_ended.html](https://searchfox.org/mozilla-central/source/testing/web-platform/tests/html/semantics/embedded-content/the-video-element/video_size_preserved_after_ended.html), as I was always running it to get feedback on my changes. I haven't tried any other test...

I ran some of the tests against `upstream/main` and they are still timing out; usually with the following: ``` UNSUPPORTED (log once): POSSIBLE ISSUE: unit 1 GLD_TEXTURE_INDEX_2D is unloadable and...

Most of the `TIMEOUT`s I am experiencing locally seem to be just flaky. The tests in the "Stable unexpected results" section have been showing consistent results locally. (Thanks again @mrobinson...

I spend the last days studying the spec and code more. Now I think I *really* understand what this statement means: > So the problem is the interaction between the...

I have a foundational assumption that I would like to confirm: As `HTMLMediaElementFetchListener` is only responsible for fetching the media resource, it should not change the ready state ever. Instead...

I realized I was chasing too many moving parts at once, so I will continue as follows: 1. Make [video_size_preserved_after_ended.html](https://searchfox.org/mozilla-central/source/testing/web-platform/tests/html/semantics/embedded-content/the-video-element/video_size_preserved_after_ended.html) pass 2. Deal with stable unexpected test results one at...

While removing the currently selected video track from the track list makes [video_size_preserved_after_ended.html](https://searchfox.org/mozilla-central/source/testing/web-platform/tests/html/semantics/embedded-content/the-video-element/video_size_preserved_after_ended.html) pass, I think that's not precise enough when taking a close look at the spec: > If...

I'm having this exact issue with my [Scylla](https://bastardkb.com/scylla/) with a [Splinky v3](https://github.com/Bastardkb/Splinky). Tested with a onekey firmware on a T480 (Arch Linux) and a P14s G2 (Windows 11). On both...