Bug 106723

Summary: Styling disappears from the cue that's being styled by ::cue pseudo element
Product: WebKit Reporter: Dima Gorbik <dgorbik>
Component: MediaAssignee: Dima Gorbik <dgorbik>
Status: RESOLVED FIXED    
Severity: Normal CC: dgorbik, eric.carlson, feature-media-reviews, koivisto, ojan.autocc, rniwa, webkit-bug-importer, webkit.review.bot
Priority: P2 Keywords: InRadar
Version: 528+ (Nightly build)   
Hardware: Unspecified   
OS: Unspecified   
Attachments:
Description Flags
Proposed fix 0.1 none

Description Dima Gorbik 2013-01-11 18:20:42 PST
When the cue is first shown it's styled correctly, but then (in few fractions of a second) it reverts to the display of the the cue like it had no styles.
Comment 1 Dima Gorbik 2013-01-12 20:12:46 PST
Created attachment 182479 [details]
Proposed fix 0.1
Comment 2 Radar WebKit Bug Importer 2013-01-12 22:55:47 PST
<rdar://problem/13004246>
Comment 3 Antti Koivisto 2013-01-13 03:52:42 PST
Comment on attachment 182479 [details]
Proposed fix 0.1

View in context: https://bugs.webkit.org/attachment.cgi?id=182479&action=review

> Source/WebCore/html/track/TextTrackCue.cpp:661
> +void TextTrackCue::markFutureAndPastNodes(Node* root, double previousTimestamp, double movieTime)

Could still be ContainerNode
Comment 4 Dima Gorbik 2013-01-13 03:55:44 PST
(In reply to comment #3)
> (From update of attachment 182479 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=182479&action=review
> 
> > Source/WebCore/html/track/TextTrackCue.cpp:661
> > +void TextTrackCue::markFutureAndPastNodes(Node* root, double previousTimestamp, double movieTime)
> 
> Could still be ContainerNode

I will tighten this with the next patch.
Thanks for a review!
Comment 5 WebKit Review Bot 2013-01-13 04:26:47 PST
Comment on attachment 182479 [details]
Proposed fix 0.1

Clearing flags on attachment: 182479

Committed r139562: <http://trac.webkit.org/changeset/139562>
Comment 6 WebKit Review Bot 2013-01-13 04:26:50 PST
All reviewed patches have been landed.  Closing bug.
Comment 7 Ryosuke Niwa 2013-01-14 12:17:44 PST
It appears that this test has been failing on WK2 bots since it was initially added:
http://test-results.appspot.com/dashboards/flakiness_dashboard.html#group=%40ToT%20-%20webkit.org&tests=media%2Ftrack%2Ftrack-css-cue-lifetime.html
Comment 8 Ryosuke Niwa 2013-01-14 12:18:11 PST
I guess not necessarily on all WK2 bots but some.
Comment 9 Dima Gorbik 2013-01-14 15:38:39 PST
(In reply to comment #7)
> It appears that this test has been failing on WK2 bots since it was initially added:
> http://test-results.appspot.com/dashboards/flakiness_dashboard.html#group=%40ToT%20-%20webkit.org&tests=media%2Ftrack%2Ftrack-css-cue-lifetime.html

Looks like this is happening because video seeking is much slower in WK2 for some reason.