On Tue, Jan 12, 2016 at 02:53:03PM +0100, Michael Stefaniuc wrote:
On 01/12/2016 02:42 PM, Andrew Eikum wrote:
On Tue, Jan 12, 2016 at 11:25:53AM +0100, Michael Stefaniuc wrote:
Hello Andrew,
On 01/11/2016 09:13 PM, Andrew Eikum wrote:
This gets winegstreamer with gstreamer-0.10 working for me again, fixing Bug 30557. However, gstreamer-0.10 is no longer developed and some distros don't ship it anymore.
which distros are those? gstreamer-0.10 is still available in Fedora 23.
Yes, Debian most notably.
I intend to pull this patch (or a version of it) into Stable. But I'd rather stick with the old gstreamer-0.10 then go to the new API; it is the smaller change.
I'm okay with that if you want to, but you will be enabling a DLL that was not enabled before, which can obviously cause regressions. See the "quartz" component in bugzilla (though not all of those are related to winegstreamer).
Uh? winegstreamer is already enabled, just prone to crashing. By cherry picking this change I'm actually fixing it.
I guess in theory. In practice, Arch, Debian, and Fedora all disable winegstreamer at configure-time. Picking across support for it into the stable branch might encourage them to build it. Like I said, I'm not opposed, just be warned.
https://projects.archlinux.org/svntogit/community.git/tree/trunk/PKGBUILD?h=...
https://packages.debian.org/stretch/wine
http://koji.fedoraproject.org/koji/buildinfo?buildID=710361
Andrew