[Webkit-unassigned] [Bug 79167] Expose a setting to exempt media playback from user gesture requirement after a user gesture is initiated on loading/playing a media

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=79167





--- Comment #28 from Philippe Normand <pnormand at igalia.com>  2012-02-27 11:37:02 PST ---
(In reply to comment #27)
> Sorry, I made a mistake in my previous comment.
> There are some tests that are using eventSender. But this change should not affect them.
> Here is a list of media tests that rely on eventSender:
> 
> These tests do not click the play button, so nothing should happen after user gesture:
> audio-delete-while-slider-thumb-clicked.html.
> controls-drag-timebar.html
> controls-right-click-on-timebar.html
> video-volume-slider.html
> video-mouse-focus.html
> crash-closing-page-with-media-as-plugin-fallback.html
> context-menu-actions.html
> audio-delete-while-step-button-clicked.html
> media-fullscreen-inline.html
> media-fullscreen-not-in-document.html
> 
> These tests clicks on the play button, however, they just check whether the media is not paused afterwards:
> video-controls-transformed.html
> video-controls-visible-audio-only.html
> video-controls-zoomed.html

You can see the list of flaky tests there:
http://build.webkit.org/results/GTK%20Linux%2064-bit%20Release/r109009%20(18855)/results.html

I wonder, though. Could the new test affect the test environment so that the tests executed after it run differently? That would explain the flakiness.
For instance I see the new test does:

window.internals.setMediaPlaybackRequiresUserGesture(document, true);

but doesn't seem to revert that before ending.

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