[Webkit-unassigned] [Bug 103771] [GStreamer] support in-band text tracks

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=103771





--- Comment #13 from Brendan Long <b.long at cablelabs.com>  2013-04-24 15:37:20 PST ---
I'm about to upload a patch to uses uridecodebin and playsink to implement this. I tried using playbin, but I kept coming back to how uridecodebin is a perfect match for what we're doing (handling streams individually).

One thing that would be nice is if playsink had multiple inputs and outputs and had:

  * An adder on the audio streams (with some way to enable and disable them).
  * An inputselector on the video streams

The changes I made are:

  * m_playbin becomes m_pipeline, which has a uridecodebin and a playsink
  * We automatically connect the first audio and video stream to the playsink
  * We connect all text streams to appsinks
  * Add InbandTextTrackPrivateGStreamer
  * InbandTextTrackPrivateClient gets setLabel() and setLanguage()
  * HTMLMediaElement::updateActiveTextTrackCues() ignores disabled tracks
  * Made GRefPtr<GstTagList> work
  * Simplify position query to just use gst_element_query_position()
  * Add test track-in-band-no-forced.html and update track-in-band-cues-added-once.html to work with this implementation (You'll need gst-plugins-bad from git to run these)

Some outstanding issues (in progress):

  * On GStreamer 0.10, restarting the video after it ends doesn't work.
  * When two cues are displayed at the same time, they get rendered on top of eachother.
  * I'm probably not handling all of the GRefPtr's correctly at the moment.
  * Katedec doesn't seem to be giving me any tags right now on 1.0

It's obviously not done, but I wanted to get this out here to see if I'm doing anything obviously wrong at the moment. It should build on Qt with GStreamer 0.10 and 1.0, and hopefully GTK. Do any other ports use GStreamer?

I'll add the audio and video tracks next.

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