Re: Pidgin 2.13.0 has been released!!

On Thu, 2018-03-08 at 23:39 -0600, Gary Kramlich wrote:
> Greetings Programs,
> 
> Pidgin 2.13.0 has been released.  You can find source tarballs on
> Sourceforge[1] or Bitbucket [2].  Windows versions are still in the
> works, so hang tight and we'll announce when they're ready as well.


When building with GStreamer 0.10:

mediamanager.c: In function ‘purple_media_manager_remove_output_window’:
mediamanager.c:1705:47: warning: passing argument 1 of ‘gst_plugin_feature_get_name’ from incompatible pointer type [-Wincompatible-pointer-types]
    factory_name = gst_plugin_feature_get_name(factory);
                                               ^
In file included from /usr/include/gstreamer-0.10/gst/gstelementfactory.h:34:0,
                 from /usr/include/gstreamer-0.10/gst/gstelement.h:60,
                 from /usr/include/gstreamer-0.10/gst/gstbin.h:27,
                 from /usr/include/gstreamer-0.10/gst/gst.h:35,
                 from media-gst.h:33,
                 from mediamanager.c:36:
/usr/include/gstreamer-0.10/gst/gstpluginfeature.h:143:14: note: expected ‘GstPluginFeature * {aka struct _GstPluginFeature *}’ but argument is of type ‘GstElementFactory * {aka struct _GstElementFactory *}’
 const gchar    *gst_plugin_feature_get_name             (GstPluginFeature *feature);
              ^

This was introduced in PR#52 with a backport of a GStreamer-1.0-only
patch from 3.0.

Do we actually care about GStreamer 0.10 any more? Should we stop
supporting it for future builds even in the 2.x branch? It's not
actually supposed to be a user-visible ABI, is it? It's all wrapped in
Purple APIs, so we're perfectly compatible even if libpurple changes
which implementation it uses behind the scenes?

Having fixed the device selection for GST1.0 in PR#322 I did also do a
test build with GST0.10 and make sure it's doing the right thing just
because I'm a good boy, which is how I found this... but I confess I'm
finding it hard to care... (and my toy PRPL is hard-coded to GST1.0
anyway so I couldn't do a *real* test of an actual call).

_______________________________________________
Devel mailing list
[email protected]
https://pidgin.im/cgi-bin/mailman/listinfo/devel