At the time of writing, the three tests below are skipped. fast/mediastream/MediaStream-video-element-displays-buffer.html fast/mediastream/MediaStream-video-element-video-tracks-disabled-then-enabled.html fast/mediastream/MediaStream-video-element-video-tracks-disabled.html
Updated list webkit.org/b/160996 fast/mediastream/MediaStream-video-element-video-tracks-disabled-then-enabled.html [ Skip ] webkit.org/b/160996 fast/mediastream/MediaStream-video-element-video-tracks-disabled.html [ Skip ] webkit.org/b/160996 fast/mediastream/apply-constraints-advanced.html [ Skip ] webkit.org/b/160996 fast/mediastream/apply-constraints-audio.html [ Skip ] webkit.org/b/160996 fast/mediastream/apply-constraints-video.html [ Skip ]
I think is better to mark tests as failing or timing out rather than just skip them. Otherwise when any of the following happens: * Some revision introduces a regression that makes the test crash * Some revision makes the test pass It goes completely undetected. A real example.. I'm checking that now, of the 5 mediastream tests skipped assigned to this bug, 2 are working now. And nobody was aware of that. If those 5 test were marked as failing, the bots would notice the new passes. But with skipped tests nobody notices.
Current status: Regressions: Unexpected image-only failures (1) fast/mediastream/MediaStream-video-element-video-tracks-disabled.html [ ImageOnlyFailure ] Regressions: Unexpected timeouts (1) fast/mediastream/MediaStream-video-element-video-tracks-disabled-then-enabled.html [ Timeout ]
Done: unskipped all the MediaStream tests and instead marked the ones still failing or timing out as such in r210838: <http://trac.webkit.org/changeset/210838>