Hi again, Cyril Brulebois <kibi@debian.org> (2021-05-21): > Paul Gevers <elbrus@debian.org> (2021-05-20): > > Ok from my side. As this upload is to fix the d-i issue I'm pretty > > sure that debian-boot is also fine, but I promised kibi this morning > > that I'll follow the process and wait for an explicit ACK from their > > side. > > Yes, please don't rush it into testing. > > I'm currently debugging a regression from bullseye that's seen with > the combination of updated udebs from both cdebconf and gtk+2.0. After > a few attempts, I'd say that's an issue with cdebconf, but I'd rather > have that confirmed before we start getting packages into bullseye. > > (Basically we have the obvious benefits from no longer hanging, but > also a buggy focus; I'll track this in a separate bug report once I > know a little bit more, and loop Simon back in.) OK, I think we're good now, the regression was detailed in: https://bugs.debian.org/988951 and that was addressed thanks to a different approach in cdebconf regarding the original bugfix (trying not to run into gtk+2.0's infinite loop, which itself is avoided thanks to the gtk+2.0 patch, what a maze), as discussed in: https://salsa.debian.org/installer-team/cdebconf/-/merge_requests/12 Once cdebconf was updated with this new approach, I've tested d-i built against unstable's udebs, and it seems to be working quite nicely, see tests listed in this comment: https://salsa.debian.org/installer-team/cdebconf/-/merge_requests/12#note_240698 I'm happy to have gtk+2.0 migrate to testing as soon as seems reasonable from the release team point of view. Ditto for cdebconf, but I can file a separate request for that, as is customary for unblock requests. Cheers, -- Cyril Brulebois (kibi@debian.org) <https://debamax.com/> D-I release manager -- Release team member -- Freelance Consultant
Attachment:
signature.asc
Description: PGP signature