media/media-element-play-after-eos.html failed 12 out of 30 times on SnowLeopard Intel Debug (Tests) between r97742 and r98022 (inclusive). Failures: http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r98022%20(2662)/results.html http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r98020%20(2661)/results.html http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r97946%20(2655)/results.html http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r97886%20(2646)/results.html http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r97873%20(2644)/results.html http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r97856%20(2640)/results.html http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r97851%20(2639)/results.html http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r97843%20(2637)/results.html http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r97827%20(2633)/results.html http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r97821%20(2632)/results.html http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r97798%20(2630)/results.html http://build.webkit.org/results/SnowLeopard%20Intel%20Debug%20(Tests)/r97769%20(2625)/results.html
Flaky on GTK as well, bug 83790.
I wonder if this flakyness is related with the duration not being determined yet when mediaPlaying() is called. media/video-duration-known-after-eos.html could be affected by this issue as well. The solution would be to wait for the loadedmetadata event, I think.