correct: gst-plugins-bad1.0 and gst-plugins-bad1.0-contrib are the same code. Only one library, containing 2 different GStreamer elements (fdkaacdec and fdkaacenc) are packaged from gst-plugins-bad1.0-contrib into the binary package gstreamer1.0-fdkaac and uploaded to contrib. So yes, the debian/ directory defines this.
All the rest of the code (over 99%) is packaged and uploaded to main as part of gst-plugins-bad1.0. The (default) debdiff takes all the changes the GStreamer bad plugins release 1.20.0-1 to 1.26.1-1; which is over 3 years where steady releases and packages were made of gst-plugins-bad1.0. We did not upload gst-plugins-bad1.0-contrib/gstreamer1.0-fdkaac; so that one is still the copy of 3 years ago. I'll try to find something to exclude all but the relevant changes On Thu, 15 May 2025 at 17:28, Paul Gevers <elb...@debian.org> wrote: > > Control: tags -1 moreinfo > > Hi, > > On 07-05-2025 14:28, Marc Leeman wrote: > > Please unblock package gst-plugins-bad1.0-contrib > > > At this moment, my feeling says no, it's too late in the cycle. > > > The debdiff is not attached because this is a request for uploading of a > > single library that is uploaded to contrib, while the bulk of the > > libraries from that code base are in the regular gst-plugins-bad1.0 > > packages in main. > > > I'm not sure I fully understand this remark, can you elaborate? Are you > perhaps saying that the source of gst-plugins-bad1.0 and > gst-plugins-bad1.0-contib is normally equal, except for the files in > ./debian/ ? We really like to see a debdiff before acking something like > this, but you can filter it to exclude irrelevant changes if you tell us > the filter you applied and why that filter hides irrelevant changes. > > Paul > -- g. Marc GPG: 827C FD74 BA46 8152 A041 F3A0 7A6A 4F17 5995 A65B
