On Tue, 2020-11-24 at 22:17 +0200, Otto Kekäläinen wrote: > > Well, my point was more that we already have buster > unstable, and > > that I was assuming you'd fix that by uploading a package to > > unstable, rather than an even newer package to buster. :-) > > As MariaDB 10.5 is already in unstable, it is in practice impossible > to do any 10.3 uploads to unstable anymore. Once the 10.5->10.3 is > safely over, 10.3 will be completely removed from unstable/testing, > but not yet.
OK... In that case, the result will be that the binary packages mariadb*10.3 in unstable after the point release will be from 10.3.27-0+deb10u1. That's basically unavoidable if you want to update the package in stable at this point, given the above. > > If there isn't an upload to unstable before the point release, is > > pushing the package from p-u to unstable and testing likely to > > cause any issues? We can't have the result of the point release be > > that stable has a newer version of the package than unstable. > > We already have 10.3.25 in Buster (via security update) and 10.3.24 > in testing/unstable. It will not be any worse by increasing the "gap" > to 10.3.27 vs 10.3.24. It's not actually in buster, though. Yes, 10.3.25 is in buster- security, but right now buster itself still has 10.3.23, which is what matters. Regards, Adam
