From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Seiderer Date: Wed, 13 Mar 2019 20:17:33 +0100 Subject: [Buildroot] [RFC v1] package/gst1-plugins-bad: add webrtcbin option In-Reply-To: <71e0a8c0-0809-e02c-cb29-25b2771202bd@mind.be> References: <20190310175109.13444-1-ps.report@gmx.net> <20190312222745.06cf1b34@windsurf> <71e0a8c0-0809-e02c-cb29-25b2771202bd@mind.be> Message-ID: <20190313201733.5ac6fb18@gmx.net> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Hello Arnout, On Wed, 13 Mar 2019 10:05:19 +0100, Arnout Vandecappelle wrote: > On 12/03/2019 22:27, Thomas Petazzoni wrote: > > Hello, > > > > +Arnout in Cc, our Config.in.legacy/backward-compatibility guru. > > > > On Sun, 10 Mar 2019 18:51:09 +0100 > > Peter Seiderer wrote: > > > >> Signed-off-by: Peter Seiderer > >> --- > >> Notes: > >> - just compile tested yet > >> - named webrtcbin option (instead of webrtc), because of the > >> BR2_PACKAGE_GST1_PLUGINS_BAD_PLUGIN_WEBRTC entry in Config.in.legacy > >> (see [1] for history) > > > > Would it be a big problem to reuse the > > BR2_PACKAGE_GST1_PLUGINS_BAD_PLUGIN_WEBRTC option ? > > I believe the plugin got a different name, and is used in a different way. So > reusing the same option could lead to things that stop working - if it's a > different name, at least there is still the legacy warning. > > That said, in -bad API changes sometimes happen anyway, so maybe we shouldn't > worry too much about that. The history is there is a webrtcdsp plugin in gst1-plugins-bad which was (false) named simple webrtc in buildroot and was renamed (with a legacy entry) to webrtcdsp. Since gstreamer-1.12.x there is an 'real' additional/new webrtc plugin in gst1-plugins-bad (nothing to do with -bad API changes, just a buildroot history problem)... Two possibilities: A) add a new buildroot option with the right name (and delete the legacy entry?) B) add a new buildroot option with the wrong name webrtcbin instead of webrtc, but avoid the legacy entry problem (as the suggested RFC patch does) Regards, Peter > > Regards, > Arnout >