[Bug 794911] Implement support for ULP Forward Error Correction

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=794911

--- Comment #6 from Mathieu Duponchelle <mduponchelle1 at gmail.com> ---
(In reply to Sebastian Dröge (slomo) from comment #5)
> (In reply to Mathieu Duponchelle from comment #4)
> Yeah, does that work for PLAY already then? Probably not as we don't add the
> relevant pieces to the SDP?

No, it doesn't do anything for PLAY yet.


> For RTX this is configurable though, it's the same thing there right?

Sort of, for rtxsend you're safe enough setting it to be strictly equal to the
client-side latency, as any request beyond that will be too late, for the
rtpstorage it's a bit different as we request packets from the storage just as
we get packet lost events, which is why we must have some tolerance on top of
that, hope I'm explaining this well enough :)

> 
 w> > > ::: gst/rtsp-server/rtsp-stream.c
> > > @@ +5019,3 @@
> > > +
> > > +
> > > +/**
> > > 
> > > Double new-line above
> > > 
> > > ::: gst/rtsp-server/rtsp-stream.h
> > > @@ +310,3 @@
> > > +/* ULP Forward Error Correction (RFC 5109) */
> > > +GST_RTSP_SERVER_API
> > > +gboolean           gst_rtsp_stream_get_ulpfec_enabled (GstRTSPStream
> > > *stream);
> > > 
> > > I wonder if some more generic API could be added here, similar to what is in
> > > rtpbin. So that we won't have to add yet another set of functions for
> > > flexfec later.
> > 
> > I wondered about that too, then opted for the less generic choice, because
> > I'm not really sure what the API will have to look like for flex,
> > specifically because of the fact multiple streams can be protected by the
> > same flex protection stream.
> 
> Ok then :)

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.