Hi colleagues, Timo,
On 24-09-2025 16:13, Simon McVittie wrote:
On Wed, 24 Sep 2025 at 14:39:45 +0200, Timo Röhling wrote:
this is a heads-up that I'm planning to upload the next minor release
of CMake 4 to unstable. The release is scheduled about two weeks from
now, around October 8th. Once CMake 4 hits unstable (again), I will
upgrade the remaining open CMake 4 FTBFS bugs to serious.
The affected bugs are https://udd.debian.org/cgi-bin/bts-usertags.cgi?
user=cmake@packages.debian.org&tag=cmake-4 of which (at the time of
writing) 794 out of 974 are still open. Is
that correct?
(cc release team since they might have input on how this relates to
transitions)
I guess we could/should set up a tracker page to judge the impact. I'm
also seeing that 76 of the bugs [1] are against key packages, so I agree
there's substantial risk of interference with more work than usual. I'm
normally not handling transitions, so maybe those that do have a more
nuanced opinion.
Next to the concern raised by Simon I'd like to raise the following. The
recent gcc-default migration cause quite some fall-out in testing
because it migrated before packages were fixed or autoremoved. I'm
wondering if we should hold off migrating cmake 4 while we're waiting a
bit for autoremoval to do its job (including sending notifications). Or
should we be more aggressive with manual removal? What do you think?
Timo, if I see correctly, most of those bugs were filed about one month
ago. I'm wondering if raising the severity even before the upload
happens makes sense to start the autoremoval clock (but please wait for
objections to be raised before acting on this suggestion).