16:02 #startmeeting Weekly Ubuntu Foundations team 16:02 Meeting started Thu Dec 12 16:02:05 2019 UTC. The chair is bdmurray. Information about MeetBot at http://wiki.ubuntu.com/meetingology. 16:02 16:02 Available commands: action commands idea info link nick 16:02 #topic Lightning rounds 16:02 echo $(shuf -e vorlon bdmurray xnox tdaitx doko sil2100 rbalint infinity cyphermox mwhudson juliank waveform) 16:03 mwhudson cyphermox bdmurray infinity waveform doko vorlon tdaitx rbalint juliank xnox sil2100 16:03 winnitude! 16:03 cyphermox: 16:03 MIR reviews 16:03 updating doc for UEFI/ SecureBoot for juliank's benefit 16:03 I have a hangout with sil2100 later for netplan hand-off 16:04 discussing shim review process with potential contributors 16:04 other hand-offy things prior to my departure 16:04 netplan NM plugin still in -progress by progressing nicely 16:04 (done) 16:04 +1 16:04 I've got nothing to report 16:05 infinity is at a conference 16:05 waveform: 16:05 * Finished LP: #1311056 16:05 * Answering lots of forum posts and questions about the Eoan re-spin 16:05 * Added one issue to release notes with workaround (attached hardware interfering with serial console, preventing boot) 16:05 * Started work on Bionic Pi 4 support; investigating whether the u-boot script changes that broke in Eoan can be added back in safely 16:05 * Work on merging core and classic boot scripts 16:05 Launchpad bug 1311056 in python-apt (Ubuntu) "apt-add-repository adds duplicate commented/disabled source lines" [High,Triaged] https://launchpad.net/bugs/1311056 16:05 * Brief experiment with ZFS on Pi (it works ...?) 16:05 (done) 16:05 doko: 16:06 gcc-10 packages are available in a frist version now, native and cross 16:07 looked at component mismatches, and python2 getting pulled into main again, maybe some i386 breajage? 16:07 some MIR work 16:07 (done) 16:07 vorlon: 16:07 bdmurray: skip? need another minute 16:07 tdaitx: 16:07 * apport O_PATH fix (LP: #1851806) 16:07 - backported LP: #1854237 to xenial, bionic, disco, eoan 16:07 - found and fixed some issues 16:07 - found more issues with the previous security update regarding python2 compatibility 16:07 - got autopkgtest to run using python2, but too many stuff fails for it to be useful right now; some bits of code are not python2 compatible; should we improve this so future sru's are python2 compatible? 16:07 Launchpad bug 1851806 in apport (Ubuntu) "'module' object has no attribute 'O_PATH'" [Undecided,Confirmed] https://launchpad.net/bugs/1851806 16:07 - testing alternatives but seems there's no "secure" way to open() /proc/pid in python2 16:07 Launchpad bug 1854237 in apport (Ubuntu Eoan) "autopkgtests fail after security fixes" [High,New] https://launchpad.net/bugs/1854237 16:07 * setup environment to build adoptopenjdk with debug and extra verbosity (currently fails to build) to compare libjawt.so config/flags 16:07 Other: 16:08 - setting up travel for the sprint 16:08 (done) 16:08 rbalint: go! =) 16:08 * uploaded vmdk-stream-converter to unstable finally 16:08 * uploads via Debian: wireshark 3.0.7-1 16:08 * fixed LP: #1855520 with a revert, then properly (please accept the SRUs) 16:08 Launchpad bug 1855520 in wslu (Ubuntu) "Issue with wsl-integration.sh script causing slow start of Ubuntu shell with WSL2" [High,Fix released] https://launchpad.net/bugs/1855520 16:08 * all uploads & srus to show motd from /etc/profile.d when show-motd is installed 16:08 * merged systemd 244-3, tests tests are running under bileto ticket 3801 16:08 * fixed and SRU-d (please accept it): LP: #1855522 16:08 Launchpad bug 1855522 in landscape-client (Ubuntu Eoan) "/etc/update-motd.d/50-landscape-sysinfo shows error in WSL" [Undecided,New] https://launchpad.net/bugs/1855522 16:08 * several rounds of upstream binutils code reviews for LP: #1843479 16:08 Launchpad bug 1843479 in gzip (Ubuntu) "gzip in Ubuntu Eoan results in Exec format error on WSL1" [High,In progress] https://launchpad.net/bugs/1843479 16:08 (done) 16:09 juliank: 16:09 bdmurray: ready now 16:09 * ubiquity: When removing packages [for minimal install], also remove automatically installed packages that are no longer required (LP: #1798992) 16:09 Launchpad bug 1798992 in ubiquity (Ubuntu Focal) "Fresh minimal desktop installation has packages pending autoremoval, pending updates" [High,Fix committed] https://launchpad.net/bugs/1798992 16:09 * apt: fix segmentation fault when passing empty argument to 'satisfy' 16:09 * python-apt 1.9.1 (soon 1.9.2): 16:09 - merged waveform's "Don't duplicate disabled sources during add()" (LP: #1311056) 16:09 Launchpad bug 1311056 in python-apt (Ubuntu) "apt-add-repository adds duplicate commented/disabled source lines" [High,Triaged] https://launchpad.net/bugs/1311056 16:09 - migration from md5 to HashStringList (debian bug 944696) 16:09 Debian bug 944696 in python-apt "python-apt: relies on MD5 internally to download packages" [Serious,Fixed] http://bugs.debian.org/944696 16:09 - migration to PY_SSIZE_T_CLEAN (debian bug 944091) 16:09 Debian bug 944091 in src:python-apt "deprecation warnings with 3.8 in the autopkg tests" [Serious,Fixed] http://bugs.debian.org/944091 16:09 - fix for the HashStringList migration (debian bug 946597) 16:09 Debian bug 946597 in python-apt "python-apt: security regression in 1.9.1" [Critical,Open] http://bugs.debian.org/946597 16:09 * initial look at the poppler thing I botched up in eoan (LP: #1849773) 16:09 Launchpad bug 1849773 in evince (Ubuntu) "/usr/bin/evince:11:strstr:TextSelectionPainter::hasGlyphLessFont:TextSelectionPainter::endPage:TextPage::drawSelection:poppler_page_render_selection" [High,Confirmed] https://launchpad.net/bugs/1849773 16:09 * preparing some stuff 16:09 (done) 16:09 vorlon: 16:10 * partially killing architectures is tedious 16:10 * found and worked around a bug in the germinate handling that was wrongly pruning build-deps that we still needed 16:10 * in response to the above ballooning the package set again, pruned opencv, gdcm, etc by changing gst-plugins-bad1.0 to not build a binary on i386 that isn't required for us 16:10 * still eyeing the numpy package being in the set and considering more source changes to get rid of it 16:10 * working through the list of autopkgtest regressions on i386 now that we're cross-testing (total test count reduced from 311 to 299 after the last germinate change) 16:10 * discussions around handling of /usr/bin/python in 20.04 16:10 (done) 16:10 xnox: 16:10 12th of December 16:10 Proposed migration: 16:10 Sheparded autopkgtest regressions blocking glibc and other uploads 16:10 Doing a few removals in debian to drop python2 things 16:10 This should make boost migratable soon* (debian removals pending before end of year) 16:10 Working on subiquity encryption UX with mwhudson 16:10 Working with snapd team on uc20 firstboot install 16:10 Working with desktopy people on oem stuff 16:10 done 16:10 - SRU reviews and releases 16:10 * Work on getting the toolchain SRUs reviewed and accepted as well 16:10 - Finished manifest differ, deployed it on a canonistack instance 16:10 - Finalizing 19.10.1 raspi stuff 16:10 - Work on enabling uc20 daily image builds, changes to livecd-rootfs and cdimage 16:11 * A bit back and forth with snapd 16:11 * Now still a small u-i change is needed to get the builds going 16:11 - Some digging into netplan 16:11 - Meeting with Iain about the new autopkgtest deployment plan 16:11 (done) 16:11 #topic Release incoming bugs 16:11 link http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-ff-incoming-bug-tasks.html#foundations-bugs 16:12 bug 1856059 is very new 16:12 bug 1856059 in jellyfish (Ubuntu) "jellyfish: please remove arm64, ppc64el binaries" [Undecided,New] https://launchpad.net/bugs/1856059 16:12 that doesn't seem like it belongs on ff-incoming 16:12 make sure ubuntu-archive is subscribed, move on? 16:13 Do we really need to card this or can somebody JFDI it? 16:13 ubuntu-archive is subscribed 16:13 I'll untag it 16:13 rls-ee is empty 16:14 rls-bb has nothing new 16:14 #topic Team proposed-migration report 16:14 urgh 16:14 https://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses_by_team.html#foundations-bugs 16:15 fwiw cyrus-imapd will probably disappear shortly, I have just been looking at it 16:19 claiming libmoo-perl 16:20 I will work on glibc with xnox 16:21 I'll pick python-packaging, 16:21 cyphermox for libmoo-perl (really cyrus-imapd) 16:21 I'll take apport 16:21 bdmurray gets apport 16:21 waveform: can you look at polib? 16:21 I'll take gawk given that it's cyrus-imapd 16:21 i take curl 16:22 vorlon, okay 16:22 lets hold on there and start from the top 16:22 you still get what you took though 16:22 doko: can you also take python-fixtures? 16:23 oh, the apt blocked by ... stuff I guess I can do some retries 16:23 bdmurray: why is apport even blocking stuff? update excuses points to 2.20.11-0ubuntu9 but we have 2.20.11-0ubuntu13 on focal for a while now 16:23 if a package is a 'candidate' but stuck, look at https://people.canonical.com/~ubuntu-archive/proposed-migration/update_output.txt 16:23 vorlon: no, this is all autopilot 16:24 doko: and because of that you can't look at it? 16:25 I'll take python-testtools 16:25 that's autopilot as well 16:25 doko: so? 16:25 it still needs tdoing 16:25 vorlon: do you work on autopilot removal? 16:25 I retried the python-apt triggered failures, seems they should have been fixed in the meantime 16:26 doko: two archive admins are better than one for working on this 16:26 bdmurray: I requested autopktest retrys for apt and gdb so new apport will unblock them 16:26 I expect it'll take several rounds 16:26 tdaitx: steal my work! 16:26 doko: I'll let you have python-fixtures and python-testtools then 16:26 tdaitx: me too, for apt 16:27 gawk, curl are claimed 16:27 tdaitx: can you look at cmake? 16:27 makes me wonder if we have a fix for dropping duplicate requests in autopkgtest queue yet 16:27 vorlon: yeah 16:27 juliank: psycopg2? 16:27 juliank: no, Simon prepped a branch one time, I requested changes and I don't think it moved since then 16:28 isn't that better handled by server team, psycopg2? 16:28 it would be, I guess 16:28 it's listed as a foundations package 16:28 we can retry 16:29 let me try that 16:29 juliank: ok you're on point for that 16:29 apport/gdb is taken by bdmurray 16:29 systemd/libseccomp: sil2100 ? 16:29 cyphermox: can you also look at sudo? (cyrus-imapd, but also mediawiki) 16:29 i take libseccomp 16:29 vorlon: ok o/ 16:30 i think systemd 244 fixed it 16:30 sil2100: ah, rbalint is fighting you for it 16:30 sil2100: you can have libwww-perl isntead 16:30 vorlon: aww, ok, fine as well ;) 16:30 Let me card it then 16:30 so, cmake being blocked by ros-robot-state-publisher seems to be done, seb got it to pass 2 hours ago 16:30 juliank: are you taking apt? 16:30 vorlon: ack, sudo 16:30 vorlon: Yeah 16:30 ok 16:31 netcat-openbsd: waveform: ? 16:31 vorlon: Those were temporary failures I think, so I just retried them too :) 16:31 vorlon, I can take a look 16:31 doko: can you also take requests, or are you full? 16:32 tdaitx: libdate-manip-perl 16:32 I don't want to commit to that, but again, the oslo stuff is openstack ... 16:32 vorlon: ack 16:32 oh, if only we could grep all failures for postgres 11 and retry them 16:32 anyone think they need more for this week? 16:32 otherwise I would stop there 16:32 juliank: download the artifacts tarballs and look for postgres? 16:32 +1 for stopping :-) 16:32 I'll look at python-meshio 16:32 well, yes it'd work 16:33 vorlon: what about the packages with missing i386 builds, like binutils? 16:33 doko: binutils should not be missing any i386 builds, it's in the whitelist 16:33 what exactly are you referring to? 16:33 ahh, no, failing atopkg test 16:34 can we remove packages from release, and allow new version to be stuck in proposed? ie. some debian packages are like that, due to e.g. automatic removal from testing 16:34 xnox: demoting to proposed, instead of removing from proposed and release, has a non-negligible cost of clogging up proposed-migration reports 16:34 if you can convince rm to override autopkg tests too ... 16:35 doko: yes, I had overridden binutils autopkgtest for the version in release, I'll do the proposed version now as well 16:35 vorlon: ack 16:37 I think we're done with this topic 16:37 bdmurray: ? 16:38 #topic AOB 16:39 Anybody? 16:39 nothing here 16:39 I won't be at the next meeting, it's past my EOY :) 16:40 Okay then. Thanks everybody! 16:40 Thanks! 16:40 thanks! 16:40 #endmeeting