On Sat, Mar 16, 2019 at 6:54 PM Andreas Beckmann <[email protected]> wrote: > Source: anjuta > Version: 2:3.28.0-6 > Severity: serious > Justification: fails to build from source (but built successfully in the past) > > Hi, > > anjuta/experimental fails to build twice in a row. The first build > succeeds, but the subsequent clean deletes > libanjuta/interfaces/Makefile.am.gir which is not regenerated during the > second build, fausing this to fail with:
My opinion is that I would not consider it a RC bug for a package to FTBFS twice in a row. For a variety of reasons, I think it makes sense to build packages from a clean chroot. (By the way, source-only uploads are now required for updated packages for bullseye.) Therefore, unless you can point to a clear statement in Debian Policy or by the Release Team about this particular type of bug, I suggest we lower this severity and stop the autoremoval of anjuta from Testing. Or if you come up with a working patch, I would be happy to apply that instead. Thanks, Jeremy Bicha
