Lost microphone, some news

[David Woodhouse]

On Fri, 2018-06-29 at 13:58 +1000, Jan Schmidt wrote:
> 
> > https://bugzilla.gnome.org/show_bug.cgi?id=795374
> > 
> > I tried a hack to set the pipeline into PLAYING later, in an attempt to
> > make it go away. It made it happen 100% of the time instead. Which I
> > suppose is a step towards diagnosing it...
> 
> What does 'later' mean in that context? It must mean that the pipeline
> transition is now reliably happening later than some other event - but
> what else is happening that might be related?

My hack, which was attached to the last email, made it happen on
receiving farstream-codecs-changed instead of when the src element is
created. I think I also tried transitioning to PAUSED and then PLAYING
again even later, which didn't help to fix it IIRC. 

If it's helpful I can reconstitute my test case which makes a dummy FS
conference in Pidgin using audiotestsrc/filesink instead of any actual
network I/O. I think it should happen there too (although it's really
bloody obvious when it happens on a Chime call, because we end up
sending payloadless audio frames from a timer and thus are clearly
shown as 'muted' in the list of participants.)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 5213 bytes
Desc: not available
URL: <https://pidgin.im/pipermail/devel/attachments/20180629/9795e7e2/attachment.bin>