2.12.1 Release

[David Woodhouse]

On Mon, 2017-10-02 at 00:43 -0500, Gary Kramlich wrote:
> On Fri, Sep 29, 2017 at 5:53 AM, David Woodhouse <dwmw2 at infradead.org> wrote:
> > 
> > I'd love to have https://developer.pidgin.im/ticket/17246 resolved.
> > Does the patch I attached there make sense?
>
> In the future please submit these as pull requests via
> https://bitbucket.org/pidgin/main they will get address much faster
> than trying to sift through trac.

But then somebody might pull it! :)

That one definitely wasn't a pull request; it was very much a request
for comments. I haven't even got it working yet; that just stops the
immediate symptom. I'd happily accept "your PRPL shouldn't do it like
that, it should do <this>..." as an answer.

> > We still haven't got full audio working — I've been able to save it to
> > a .wav file for weeks, but persuading Pidgin to actually *play* the
> > stream is distinctly non-trivial. But I think the above is all we need
> > in Pidgin itself, to make it stop crashing when we use anything other
> > than fsrtpconference. I'd love someone with a clue to take a look
> > though.
>
> Might be tough as the main person familiar with it has been out of
> touch for a bit now.

Hm, OK. I'll keep playing with it. 

Jakub, can you offer any insight? In short, my new PRPL can spew hex
bytes of audio data to the terminal but I can't see how to actually
turn that into a *call* in Pidgin. Short of setting up an RTP proxy on
localhost for Pidgin to talk to.

> > It's also been pointed out to me that since I brought the 2.x branch up
> > to date with everything we needed for Lync, other things have been
> > added to 3.x and not backported. This one in particular, which *mostly*
> > looks like it can go in without changing ABI, with a few tweaks:
> >  https://bitbucket.org/pidgin/main/pull-requests/86/gstreamer-10-media-device-management/diff
>
> Our intent is not to backport stuff to 2.x.y.  If it needs to go in
> 2.x.y it does and is then forward ported into 3.0.0 when we release
> the 2.x.y version.

I've updated https://github.com/tieto/sipe/issues/157 with that
comment.

> ABI was broken in that patch.  I'm not qualified enough in
> gstreamer/farstream to present any solutions, so what did you have in
> mind?

I had a quick look through it, and only spotted one ABI change (as
opposed to addition), which was the addition of a new argument to a
certain callback function. I'm fairly sure we can come up with a
backward-compatibility story for that. I'll take another look...
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 4938 bytes
Desc: not available
URL: <https://pidgin.im/pipermail/devel/attachments/20171002/13c19f64/attachment-0001.bin>