Hi,
On 16/05/2013 04:04, Kevin Anthony wrote:
Chow, where can I find that patch?
Is gstreamer 1.0 on the slate for 3.0?
I don't have a clue.
dnielsen/knocte, don't you have anything to say about this? Both of you seem to
be somewhat informed about the situation with Gst(#)1.0, yet every time I bring
up the issue, the both of you just stfu and disappear. Very nice.
The patch in the bug report has been sitting there since last year, and there
isn't so much as even a reply on the bug report saying if you'll accept or
reject it, let alone why.
I'm sorry about the lack of communication on this issue. I wanted to
reply on your previous remark about this, but never got to it.
The patch in the above bug ports libbanshee, our C GStreamer backend,
to GStreamer 1.0, and seems OK.
But it doesn't address our GStreamerSharp C# backend, which uses the
gstreamer-sharp backend. This backend is currently only used on
Windows, but our long term goal is to use it by default on all
platforms.
Olivier did some work to get gstreamer-sharp ready:
But he had to remove the bundled copy of glib-sharp, and add a
dependency to glib-sharp 2.99.x. I don't remember the exact reasons,
but I think it's the only way he could get video playback to work.
Mixing glib-sharp 2.12.x and glib-sharp 2.99.x in the same app is not
possible, so it means we can't have a GStreamerSharp backend targeting
GStreamer 1.0 before Banshee is ported to gtk/glib-sharp 2.99.
So we have 2 ways of going forward:
1/ Release Banshee 3.0 depending on gtk+ 3.x, and at the same time, or
right after that, port the 2 backends to GStreamer 1.0, applying the
above patch in the process.
2/ Apply the libbanshee patch, release Banshee 2.8 from git master in
its current state, still depending on GTK+ 2.x. This would be a
release mainly for Linux and Mac, because the Windows version would
not see any GStreamer changes.
In my previous "long term strategy" e-mail [1], I proposed going with
option 1/ :
At the time, I thought the overall transition (other apps, distros,
etc.) to GStreamer 1.0 would go at the same pace as the one to GTK +
3, leaving us with enough time to do it this way. But apparently, the
GStreamer transition is going much faster.
So I'm open to reconsidering this, especially since we (purposefully)
don't have a timeline for Banshee 3.0.
