On Sun, Jan 22, 2012 at 7:19 PM, Anton Khirnov <an...@khirnov.net> wrote: > > On Sun, 22 Jan 2012 18:19:58 +0100, Reinhard Tartler <siret...@tauware.de> > wrote: >> On So, Jan 22, 2012 at 10:28:46 (CET), Anton Khirnov wrote: >> >> > 0.8 is finally out so we can start preparing the next major bump. Before >> > we can actually bump, there are still some issues to resolve: >> >> I guess you want to bump the soname of libavcodec only. Is that correct? >> > > No, libavformat too. > >> Could you please summarize the non-cosmetic benefits of this? >> >> From my last Debian archive rebuilds, I've seen that there are still *a >> lot* of packages using deprecated APIs, and I fear that dropping them >> will cause a serious amount of work to get all upstreams updated. I'd >> therefore prefer to have some solid arguments ready before starting to >> argue with fellow package maintainers. >> > > For libavformat this means that protocol private options get enabled. > > In libavcodec, this will fix some libx264 options problems (resulting > from the fact that they were previously implemented as flags, so it's > impossible to tell if the user set them or no). > > And of course much less overall mess and clutter (the diffstat so far is > over 2000 lines removed and I'm not finished with lavf yet), which is > quite important for us, though admittedly less so for our users. > > Anyway, I'd expect bumping early to make stuff easier for you. We > would bump at some point before 0.9 in any case, so bumping now means > you have more time to convince people to update their packages.
Yes, that makes sense. Thanks for the explanation. > Meanwhile you can ship the bumped libs in experimental installable > concurrently with 0.8. Indeed, that makes sense. I know have also much better means to recompile packages against 0.9 test packages. Let's to a quick 0.9 alpha release shortly after the bump, which I can upload to experimental. -- regards, Reinhard _______________________________________________ libav-devel mailing list libav-devel@libav.org https://lists.libav.org/mailman/listinfo/libav-devel
