[ While your message is dated earlier, looking at email headers it wasn't actually sent until 20:41 UTC, and I'm only seeing it now after being away from my computer most of the day. Apologies if it seemed like I was blindly ignoring your response when I modified the bug's severity at 15:22 UTC. ]
On Sun, 2025-03-09 at 07:26 +0100, Helmut Grohne wrote: > Why would you knowingly let breakage slip through when there is a very > simple path for mitigation (upgrading Replaces to Conflicts)? What is it > that you gain in not just uploading the suggested change? Incus hasn't yet been part of a stable release. If it were, I would be more concerned about this edge case bug. As it is, people installing Incus via bookworm-backports (including myself on several machines), are by definition tracking a package version from the next eventual release. Reviewing Policy sections 7.4 & 7.6.1 make it pretty clear that Breaks+Replaces is the preferred way to move files between packages, and using just Conflicts seems hacky for an issue that doesn't affect trixie or sid. Updating the backported version of Incus will also resolve this edge case. I expect users to have fully updated their systems prior to upgrading once trixie is released, as outlined in the release notes for every Debian release. For the case of users jumping to testing _right now_, I think we've already spent more time and effort worrying about "what ifs" than is justified. > Generally speaking, I reached an agreement with the release team that > /usr-move issues should prevent testing migration via filing rc bugs. > This bug really is meant as a migration blocker. In case you lower > severity, I'm going to let a release team member judge the severity. Prior to seeing your reply, I did lower the severity of the bug and it migrated during the day, sorry. Mathias
signature.asc
Description: This is a digitally signed message part
