[Webkit-unassigned] [Bug 54910] media/video-replaces-poster.html fails on all platforms

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=54910


Eric Carlson <eric.carlson at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |eric.carlson at apple.com




--- Comment #24 from Eric Carlson <eric.carlson at apple.com>  2011-02-28 09:38:33 PST ---
(In reply to comment #23)
> Seeking to time 0 will replace the poster with the first frame of the video, so this should still test the right thing.
> 
  Oops, you are correct!


> But, I think you are right, seeking to time 1 should have the same effect as seeking to time 0.  I had just changed that because I was trying to figure out why some tests showed 00:00.24 in the image and some showed 00:02.03.  Pretty sure that is due to the pause() after 'autoplay' landing on different times for different platforms combined with the asynchronous response to "seeked" (or any other event) resulting in unpredictable pixel repainting.
> 
  I wonder if the problem is that the video pipeline hasn't necessarily pushed out the new frame of video yet when the 'seeked' event fires?

> Do we really have to play() or 'autoplay' the video for this test?
>
  No, seeking like you have done should also replace the poster. 

  I am afraid that the only way to check one method or another is to land a change because the results are inconsistent. 

  Maybe it is worth seeking to time 1 and adding "checkExpected(video.currentTime, 1)" in the 'seeked' function so we will at least have more information if the test continues to fail.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.