Ralph suggested an initial strategy, which I've expanded somewhat:
* # the nsVideoFrame changes to add the display div. Flowing from this will be other layout considerations, like text positioning, CSS, and other display aspects of the WebVTT spec (not all of which we'll finish this term).* # the import and build support for the parser library in media/webvtt. This will include massaging libwebvtt's public API to work well with Mozilla.* # the WebVTTDecoder (see [https://bugzilla.mozilla.org/show_bug.cgi?id=629350#c29 discussion]) implementation in content/media/webvtt* # DOM bindings for HTMLTrackElement, TextTrack, TextTrackCue, TextTrackCueList. Flowing from this will be the DOM API implementation for everything these objects can do.* # DOM Tests to test the new types (HTMLTrackElement, TextTrack, TextTrackCue, TextTrackCueList) and also any aspects of the WebVTT spec related to rendering.
Files that this [https://bug629350.bugzilla.mozilla.org/attachment.cgi?id=704716 initial patch] touches:
