* fast/mediastream/media-element-current-time.html [ Failure ] --- /home/igalia/dpino/workspace/webkit-gtk/layout-test-results/retries/fast/mediastream/media-element-current-time-expected.txt +++ /home/igalia/dpino/workspace/webkit-gtk/layout-test-results/retries/fast/mediastream/media-element-current-time-actual.txt @@ -1,4 +1,4 @@ -PASS Check video.currentTime behavior +FAIL Check video.currentTime behavior assert_not_equals: Playback has started, currentTime must not be zero got disallowed value 0
I think the way this test was written kind of makes it prone to flakiness. Changing the playback state and waiting 10ms until performing a test is not great for the gst engine. It would be better to rely on events.
Current failure is: -PASS Check video.currentTime behavior +FAIL Check video.currentTime behavior assert_greater_than: video is playing, time should advance expected a number greater than 0.284381 but got 0.284381 Which happens because the position is queried immediately after the play() call. Waiting 10ms gives time to the engine to have the pipeline position going forward.
Created attachment 448027 [details] Patch
Committed r287477 (245612@main): <https://commits.webkit.org/245612@main> All reviewed patches have been landed. Closing bug and clearing flags on attachment 448027 [details].
<rdar://problem/86967699>
This is still very flaky. (In reply to Philippe Normand from comment #1) > I think the way this test was written kind of makes it prone to flakiness. > Changing the playback state and waiting 10ms until performing a test is not > great for the gst engine. > > It would be better to rely on events. Patch incoming...
Pull request: https://github.com/WebKit/WebKit/pull/3085
Committed 253220@main (3e994a2e3142): <https://commits.webkit.org/253220@main> Reviewed commits have been landed. Closing PR #3085 and removing active labels.