[Bug 795330] pad: Handle changing sticky events in pad probes

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=795330

--- Comment #18 from Thibault Saunier <tsaunier at gnome.org> ---
(In reply to Sebastian Dröge (slomo) from comment #15)
> (In reply to Thibault Saunier from comment #13)
> 
> > In the testcase yes, in the real world example it was not the case as the
> > probe was removed,
> 
> Nothing is preventing upstream from sending the same event again, you can't
> just remove the pad probe and expect your changes to be there forever :)

Well, in my case I know when a stream-start should happen, and if I know that
the sticky events are updated with the probe, then yes I can do that without
any problem.

Tbh I am not sure what is conceptually expected in that case, ie. sticked even
are before probes or after probes? To me both option make sense and have their
pros and cons :-)


> > and without that patch I believe that the STREAM_START
> > was unsticked before each buffer or something like that (I do not have that
> > code anymore so my memory might be wrong and note that I do not need that
> > patch anymore as I reworked the whole thing)
> 
> That sounds like a bug we should fix then

Well, this was one of the point of that patch ;)

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