[01:16] <bryceh> LocutusOfBorg, sorry Utkarsh and I were both out sick today.  The 'Unable to load dynamic library '<whatever>' is occuring among a lot of the php packages right now due to the transition still under way.  I don't think the error is actually due to php-apcu*, but rather a larger issue with autoloading that Utkarsh and I have been actively investigating.
[01:18] <bryceh> LocutusOfBorg, I'm keeping notes on current status of things (and next actions for some packages) at https://wiki.ubuntu.com/ServerTeam/Transition/Php8.1
[08:12] <LocutusOfBorg> thanks bryceh I'm happy it wasn't my fault
[08:12] <LocutusOfBorg> I guess no need to do anything and situation will auto fix then :D
[09:18] <icey> I'm taking a look at the pythonmagick build failures this morning
[09:39] <ginggs> icey: great, feel free to ping if you need a sponsor
[09:40] <icey> thanks ginggs
[10:33] <icey> ginggs: I just built it fine locally with a jammy pbuilder, could you just poke a retry on it please?
[10:42] <ginggs> icey: given back
[10:51] <ginggs> icey: no luck :(
[10:51] <icey> ginggs: frustrating
[10:51] <icey> thanks!
[11:01] <ginggs> icey: https://mail.gnu.org/archive/html/bug-autoconf/2021-10/msg00000.html
[11:05] <icey> ginggs: that'd do it :-/
[11:08] <icey> ginggs: I get the feeling that this is going to end up hitting us a lot :-/
[11:11] <ginggs> icey: https://codesearch.debian.net/search?q=AX_PYTHON_DEVEL&literal=1&perpkg=1
[11:12] <ginggs> icey: i'll file a bug against autoconf-archive in debian
[13:15] <icey> ginggs: how do we handle things like python 3.10 being incompatible with things right now (took a peek at one I can upload, saw that it's failing on 3.10 because of referring to a `collections.abc.$THING` indirectly, which is deprecated in 3.9 and gone in 3.10
[13:15] <icey> so there's upstream work to get them updated, but how do we handle that blocking the 3.10 migration?
[13:30] <icey> ginggs: in essence: `cannot import name 'Mapping' from 'collections' `; the correct update is `import collections.abc.Mapping` (in this case) but seems to hit a lot of packages
[13:31] <icey> although I'm seeing some common bits hitting unittest2
[13:40] <ginggs> icey: adding python3.10 is a non-blocking transition; python3.10 is already in -release, so things can migrate as soon as they're ready.  for things requiring upstream work, we'll file bugs against the packages in debian. i believe doko is already looking at that
[13:41] <icey> ginggs: I was wondering when looking at that python3-defaults autopkgtest failure list
[13:56] <ginggs> icey: many of those should pass on retry soon, e.g. when assimp and cvxopt migrate
[13:57] <mfo> juliank, hi o/ sorry to bother. when you have a chance, may i ask you to review this grub2 mr in bug 1840560 ? it's not high priority, fyi. but just wanted to check if this is still on your radar, or would it better fit someone else at this moment?  thanks! :)
[13:57] <icey> ginggs: except for the unes using unittest2: https://bugs.launchpad.net/ubuntu/+source/unittest2/+bug/1949778 ;-)
[14:32] <Odd_Bloke> rbasak: o/ If I've `git ubuntu clone`d a package, and I want to pull everything new from the LP repo, what command(s) would I run?
[14:35] <rbasak> Odd_Bloke: git fetch pkg
[14:55] <Odd_Bloke> rbasak: Thanks!
[17:15] <juliank> mfo: I'll try to have a look at it tomorrow
[17:15] <mfo> juliank, cool, thanks!
[22:42] <bdmurray> sergiodj: Did this end up getting merged? I don't see it in impish https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=2bf3b79d05bf85e41cbdcb020bd1cc424f59dd9a
[22:46] <sergiodj> bdmurray: upstream, you mean?  yeah, it was merged
[22:47] <sergiodj> it should be part of impish because we have gdb 11.1 there
[22:48] <bdmurray> I strangely don't see it in there
[22:48] <sergiodj> bdmurray: http://paste.debian.net/1218273/
[22:48] <sergiodj> it's there
[22:50] <sergiodj> bdmurray: are you seeing this bug manifest in Impish?
[22:50] <bdmurray> sergiodj: ah the move to dwz.c and having a ChangeLog-2020 explains why I didn't find it
[22:51] <sergiodj> yeah.  GDB has recently dropped ChangeLogs altogether, FWIW
[22:51] <bdmurray> ?
[22:51] <sergiodj> sorry, just an unrelated-but-somewhat-related comment
[22:51] <sergiodj> :)
[22:52] <bdmurray> sure that just seems weird
[22:52] <sergiodj> why?
[22:52] <bdmurray> what is the alternative to a ChangeLog?
[22:52] <sergiodj> git commit, basically
[22:53] <bdmurray> that seems hard for downstream consumers
[22:54] <sergiodj> yeah, it's something being discussed.  the glibc folks came up with a script that goes through the commits and generates an equivalent to ChangeLogs
[22:54] <sergiodj> but because GDB has switched to C++ we have to adjust that script, which is non-trivial to do
[22:54] <sergiodj> so for now people are just not writing ChangeLogs anymore
[22:55] <sergiodj> most of the community wanted to get rid of them for quite a while (I remember seeing discussions about this almost a decade ago), and now they finally pulled the trigger
[22:56] <sergiodj> either way, I'm sure that some input from downstream consumers would be most welcome.  I don't see us going back to ChangeLogs, but maybe the automation script I mentioned could gain some momentum
[23:04] <cjwatson> There's a thing in gnulib (thus passed around a lot of GNU and GNU-adjacent projects) that just generates a ChangeLog from git log at tarball release time
[23:05] <cjwatson> Various projects I'm involved in have been using it for years without it occasioning much comment