[Bug 764733] qtdemux: Regression in YouTube TV tests in WebKit MSE after fix for #760779

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=764733

--- Comment #10 from Seungha Yang <sh.yang at lge.com> ---
(In reply to Xabier Rodríguez Calvar from comment #9)
> (In reply to Seungha Yang from comment #6)
> > Could you kindly explain more please?
> > 
> > What I'm understanding about this issue is as follows
> > - bug #760779 caused delay in expose stream, if moov have no sample
> > information.
> > - MSS mode case, however, exposing stream causes no problem because upstream
> > element provides caps information (as far as I understand)...
> > ==> bug #760779 did not fully consider MSS case so, delay also happened for
> > MSS mode.
>  
> > - Attached patch expects exposing streams with the first moov although the
> > moov has no sample information.
> 
> Phil is right, your patch does not fix the issue. I uploaded a couple of
> logs, the right one is run after reverting your patch and the wrong one is
> run before reverting your patch.
> 
> I hope this can provide some useful information to you.

Thanks for providing more information. I perfectly misunderstood this issue....
Uploaded "Wrong run log" shows invalid framerate is set in caps. (the purpose
of bug #760779 is exactly fixing it). However, because the delayed exposing
stream causes regression, I think we have no choice but to revert bug #760779.

Actually, there is no available information to calculate framerate at that
moment.

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