Mathias Gibbens <gibmat@debian.org> writes: > On Sun, 2025-09-28 at 23:03 +0200, Simon Josefsson wrote: >> All, >> >> I have uploaded golang-go.crypto 0.42.0 to experimental. >> >> Does anyone see any reason not to try to get this into sid? > > Nope, I'd say let's go for it. We're early in the forky development > cycle so it's a much better time to perform the update. Agreed. >> https://salsa.debian.org/jas/golang-go.crypto/-/pipelines/946302 >> >> Doing a reverse build check is the only concern that I'm aware of. >> >> Could anyone share a recipe for doing a reverse build of all >> dependencies? Debusine? Ratt? Or maybe we could try a Salsa revbuild, >> if we get a pre-approval for doing this, although I'm not sure how salsa >> CI handles ~1000 jobs. > > For all of the "x-whatever" packages, I think there are so many > reverse dependencies that trying to run ratt locally would take too > long. If uploaded to experimental, we can at least see the autopkgtest > results to check if anything obviously breaks. I see a few failures at > the moment, but it looks like test failures (not compilation failures); > I haven't investigated if the failures are due to the updated golang- > go.crypto or some other reason (flaky, broken due to other update, > etc). I don't think autopkgtest runs for experimental, or is there a way to trigger it to rebuild reverse dependencies? Anyway, I think a 100 job Salsa CI pipeline will give us sufficient indication of any problems before doing the unstable upload. I'm sure some packages will break, but that will be for good reasons (algorithm deprecation etc) and they need to be taken care of. /Simon
Attachment:
signature.asc
Description: PGP signature