+1 maintenance report

Sat Jan 8 02:36:55 UTC 2022

I've been on +1 maintenance part of this week (from Tuesday to Friday).
Unfortunately, due to a combination of meetings + other work priorities
+ some personal stuff I wasn't able to dedicate as much time as I
initially planned.

Retriggers that worked
======================

sqlite> SELECT DISTINCT test.package, result.version, result.triggers FROM result INNER JOIN test ON test.id = result.test_id WHERE result.requester = 'sergiodj' AND result.exitcode = 0 AND result.run_id LIKE '2022010%';
lebiniou-data|3.64.0-1|lebiniou-data/3.64.0-1
asterisk|1:16.16.1~dfsg-4build1|systemd/249.5-2ubuntu3 dpdk/21.11-1
cluster-glue|1.0.12-20ubuntu2|curl/7.81.0-1
ceph|16.2.7-0ubuntu1|binutils/2.37.50.20220106-2ubuntu1
asterisk|1:16.16.1~dfsg-4build1|curl/7.81.0-1
tang|11-1|curl/7.81.0-1
apt|2.3.13|dpkg/1.21.1ubuntu1
casper|1.465|finalrd/9
pkg-js-tools|0.10.3|dpkg/1.21.1ubuntu1
node-redis|3.1.2+~cs6.15.1-1|pkg-js-tools/0.10.5
python-parsel|1.6.0+dfsg-3|lxml/4.6.4-1ubuntu1
python-oslo.vmware|3.10.0-0ubuntu1|lxml/4.6.4-1ubuntu1
dvisvgm|2.12-4|lxml/4.6.4-1ubuntu1
node-labeled-stream-splicer|2.0.2+~2.0.0-1|node-labeled-stream-splicer/2.0.2+~2.0.0-1
node-grunt-contrib-copy|1.0.0-4|pkg-js-tools/0.10.5
python3.9|3.9.9-2|readline/8.1.2-1
macaulay2|1.19.1+ds-4|readline/8.1.2-1

Retriggers that didn't work
===========================

sqlite> SELECT DISTINCT test.package, result.version, result.triggers FROM result INNER JOIN test ON test.id = result.test_id WHERE result.requester = 'sergiodj' AND result.exitcode != 0 AND result.run_id LIKE '2022010%';
exim4|4.95-2ubuntu2|systemd/249.5-2ubuntu3 dpdk/21.11-1
pg-fact-loader|1.6.0-5|pglogical/2.4.1-1
golang-github-komkom-toml|0.0~git20211215.3c8ee9d-1|golang-github-komkom-toml/0.0~git20211215.3c8ee9d-1
mediawiki|1:1.35.5-1|mediawiki/1:1.35.5-1 postgresql-14/14.1-5 curl/7.81.0-1 php8.1/8.1.1-4 openssl/3.0.1-0ubuntu1 php-defaults/91 ocamlnet/4.1.8-3
python-pypowervm|1.1.24-0ubuntu1|lxml/4.6.4-1ubuntu1
node-chokidar|3.4.3-3|node-is-binary-path/2.1.0-4
gtk4|4.4.1+ds1-3|mesa/21.3.3-1
dpdk|20.11.3-0ubuntu3|hwdata/0.355-1
dgit|9.15|sqlite3/3.37.2-1 diffoscope/198 php-doctrine-cache/2.1.1-2 supysonic/0.6.2+ds-4
pyopencl|2021.2.9-1|mesa/21.3.3-1

Investigations
==============

kma & kmerresistance
--------------------

Debian dropped support for 32-bit architectures on kma.  This required
manual intervention from the AAs in order to deleted the old armhf
binaries and make the package migrate.  I requested this, and cjwatson
pointed out that kmerresistance was actually a rdep that needed to be
fixed first.  I noticed the fix was available in the salsa repo but
never got uploaded.  Got in touch with the author, he had just forgotten
about it.  Uploaded the fix to Debian, which got sync'ed to Ubuntu, but
didn't solve the problem because we still had kma's armhf binary
available in -release.  I decided to patch kmerresistance and just use
the same Architecture line as kma in d/control.  This finally solved the
problem, and the two packages finally migrated (kma had been stuck in
-proposed for more than 80 days, IIRC).

dnsdbq
------

Package is FTBFS'ing due to changes in glibc 2.34.  I filed
https://bugs.launchpad.net/dnsdbq/+bug/1956639 and
https://github.com/dnsdb/dnsdbq/issues/198.  Upstream is willing to
accept a temporary solution to fix the issue; I will file a PR soon.

libsdl & osk-sdl
-----------------

osk-sdl's autopkgtest is failing with libsdl >= 2.0.18.  This is also
impacting the Debian package, and there's an upstream bug where a
discussion is happening.  I filed
https://bugs.launchpad.net/debian/+source/osk-sdl/+bug/1956534.  I've
just got notified that the bug has been fixed in Debian, which means
that it will likely be fixed in Ubuntu as well.  I will monitor the
situation and close the Ubuntu bug if everything is OK.

guile-3.0
---------

guile-3.0 is FTBFS'ing with glibc 2.34.  There's an upstream bug which
has apparently been fixed, but is not present in any release yet.  I
filed https://bugs.launchpad.net/ubuntu/+source/guile-3.0/+bug/1956636.
I chose not to backport the upstream patch that fixed the problem
because it's a really long diff (gnulib update, for those who are
curious), so I will just wait until the fix is released, packaged into
Debian, and then the bug can be safely closed.

python-aioamqp
--------------

autopkgtest fails due to python-pamqp 2->3 transition.  Upstream seems a
bit abandoned, but there's a PR to fix the issue and it seems OK.  I
backported the patch to the Debian package and did the upload.
Eventually this will be fixed in Ubuntu.

Thanks,

-- 
Sergio
GPG key ID: E92F D0B3 6B14 F1F4 D8E0  EB2F 106D A1C8 C3CB BF14