Hi Lukas, Lukas Märdian <slyon@debian.org> (2024-04-25): > Turns out d-i was unable to finish the installation, due to > installability issues of packages in the target system, especially > systemd-sysv vs libssl3 in this case. The archive is still very much > in flux and this is probably also why we still see d-i daily build > failures [0] (but its looking much better already!) and Salsa-CI > failures for d-i [1]. What I've reported a while back, and what I ended up fixing last week were all udeb relationship issues making it impossible to build d-i or to run it (I don't have the details but I'm pretty sure all runtime issues would already show up at build time anyway, and yes that explains a lot of red on the d-i daily build graphs). And yeah, at the moment we end up with a debootstrap problem when installing trixie, since libssl3 is gone from trixie (except on 32-bit arms) while coreutils still Pre-Depends on it. That's one blocker I've just confirmed again, but didn't look for possible others. I haven't followed recent progress on the 64-bit time_t front, maybe rebuilding packages within testing might help get rid of such issues, that might make some other things harder… Seeing how the package in unstable has DEP-17 changes on top, I'm not sure I want to get involved in the intersection of two gigantic transitions I'm not knowledgeable about. :/ Installing unstable is more likely to succeed though (as at least a plain debootstrap sid works, as opposed to a plain debootstrap trixie). Thanks for the write-up, I'll check it when time permits. Cheers, -- Cyril Brulebois (kibi@debian.org) <https://debamax.com/> D-I release manager -- Release team member -- Freelance Consultant
Attachment:
signature.asc
Description: PGP signature