15:01 #startmeeting Weekly Ubuntu Foundations team 15:01 Meeting started Thu Jul 16 15:01:55 2020 UTC. The chair is bdmurray. Information about MeetBot at http://wiki.ubuntu.com/meetingology. 15:01 15:01 Available commands: action commands idea info link nick 15:02 #topic Lightning rounds 15:02 echo $(shuf -e vorlon bdmurray xnox tdaitx doko sil2100 rbalint infinity mwhudson juliank waveform slyon) 15:02 tdaitx bdmurray waveform doko mwhudson juliank vorlon slyon infinity xnox rbalint sil2100 15:02 WON 15:02 tdaitx: 15:02 \o/ 15:02 uh oh, I don't have it ready T_T 15:03 sil2100: so I will go last instead =P 15:03 SRU verified LP: #1884148 15:03 investigation into update-motd autopkgtest failure on Focal (it strangely passed for Groovy) 15:03 Launchpad bug 1884148 in ubuntu-release-upgrader (Ubuntu Focal) "dist upgrade process passes a number along with package to apport" [Critical,Fix released] https://launchpad.net/bugs/1884148 15:03 updated the fix for LP: #1873545 not to use $EUID as #! is /bin/sh 15:03 Launchpad bug 1873545 in Ubuntu WSL "Permission denied error from show-motd when updates available" [High,Fix released] https://launchpad.net/bugs/1873545 15:03 review / testing of ~slyon/ubuntu-release-upgrader:slyon/focal-python2-fix 15:03 reported autopkgtest bug LP: #1887183 15:03 uploaded u-r-u for the 18.04.5 point release 15:03 +1 maintenance 15:03 Launchpad bug 1887183 in autopkgtest (Ubuntu) "--associate-ip does not work" [Undecided,New] https://launchpad.net/bugs/1887183 15:03 short week with some holidays 15:03 waveform: 15:03 * Fixed issue in pending raspberrypi-userland (inclusion of EGL/GL headers clashing with libegl-dev) (LP: #1883111) 15:03 * Addressed review of rpi-eeprom (LP: #1884748) 15:03 * Prepared split linux-firmware-raspi2 using multi-upstream tarball style (LP: #1886581) 15:03 * Bumped wifi firmware in linux-firmware-raspi2 to latest version 15:03 * Worked on removal of u-boot from classic (postinst transition in flash-kernel and adjustment of custom "pi" method in flash-kernel to avoid u-boot tools) 15:03 (done) 15:03 Launchpad bug 1883111 in Ubuntu "[needs-packaging] raspberrypi-userland" [Wishlist,New] https://launchpad.net/bugs/1883111 15:03 Launchpad bug 1884748 in Ubuntu "[needs-packaging] rpi-eeprom" [Wishlist,New] https://launchpad.net/bugs/1884748 15:03 Launchpad bug 1886581 in linux-firmware-raspi2 (Ubuntu) "Ubuntu 18.04 64-bit RasPi 3 and 4 image fails to boot on RasPi 4B" [Undecided,New] https://launchpad.net/bugs/1886581 15:04 doko: 15:04 how about juliank! 15:05 Not sure it's complete, but here we go: 15:05 jul 09: 15:05 * short week, off sick mon & tue 15:05 * apt 2.1.7 15:05 * fixed make-dfsg on s390x by disabling use of posix_spawn, still needs investigating whether posix_spawn is broken on s390x or the make code uses it in a way that only fails on s390x (and only in flatpak-builder tests) 15:05 * retry regressions 15:05 jul 16: 15:05 * initial work of freeing apt of masters, and slaves 15:05 * cleaned up apt-key net-update spec 15:05 * dug around in os-prober and snapshots 15:05 * private stuff fba872e3c2d435b7a69acbb618fb888d0437d7b471758b2bb9e74de341b070fe 15:05 * opened bugs for google chrome and keybase to stop using apt-key add, and use trusted.gpg.d 15:05 * reviewed rpi eeprom package thingy for waveform 15:05 (done) 15:06 vorlon: 15:06 * haskell transition work (mostly deleting broken packages) 15:06 * pandoc on riscv64 takes foreeeeever 15:06 * reviewing current net-apt key-update spec 15:06 * working with server team on some changes to the motd-news implementation 15:06 * partner meetings 15:06 * discussions around an embargoed issue 15:06 (done) 15:06 slyon: 15:06 - Fixed python-mkdocs & streamlink for fonts-font-awesome migration 15:06 * sponsored by Dimiri & merged in Debian Salsa 15:06 - Fixed cider for fonts-font-awesome migration 15:06 * sponsoed via Debian NMU by Dimitri 15:06 - Discovered XSBC- field bug in dh-make-perl, fixed upstream 15:06 * Fix released, bug #1886461 15:06 - Debugged 20.04 upgrade, if legacy python[-minimal] installed 15:06 * Updated Focal release notes about 'python-is-python2' 15:06 * MP awaiting review, bug #1875523 15:06 - SRU verification for apport snap handling 15:06 * Fix released, bug #1861082 15:06 - Improved IWD vs WPA Supplicant evaluation document 15:06 - Collected netplan.io community engagement metrics 15:06 === Netplan === 15:07 - Fixed non-OVS bonds in OVS netplan setup, with James & Lukasz 15:07 - Improve OVS integration tests 15:07 - Review Lukasz' implicit OVS fake VLAN bridges PR#151 15:07 - Checked upstream NM's refusal of our netplan plugin 15:07 * considering different solutions, like a netplan-nm-bridge daemon 15:07 * needs team discussion 15:07 - Reviewed Lukasz' SR-IOV first boot PR #141 15:07 - Prepared OVS interface cleanup PR #153 15:07 * In review 15:07 - Started working on Wireguard PR #113 15:07 (done) 15:07 xnox: 15:08 * mid-cycle product sprint + working on all the .1 release things & proposed-migration. Done. 15:08 rbalint: 15:08 * +1 maintenance 15:08 * uploads via Debian: wireshark 3.2.5-1, libevent 2.1.12-stable-1 15:08 * merges: systemd as 245.6-2ubuntu1, then as 245.6-3ubuntu1 with additional fixes, mtd-utils 15:08 * partner work 15:08 * +1 maintenance (ongoing) 15:08 (done) 15:08 sil2100: 15:09 - Shorter week, had half-day on Friday, status only from this week 15:09 - SRU reviews and releases 15:09 - Kernel SRUs 15:09 - +1 maintenance on Monday/Tuesday-ish 15:09 * Got slightly side-tracked fixing britney in bileto 15:09 - Phasing out new core18 into stable 15:09 - 20.04.1 and 18.04.5 point-release preparation 15:09 * Prepared base-files 15:09 * Pushed updated language-packs (base for focal, updates for bionic) 15:09 * Coordination with various teams 15:09 * General preparation per the checklists 15:09 - netplan: 15:09 * Looking into the OVS config clean-up PR from Lukas 15:09 * Started looking at SR-IOV improvement PR review comments 15:09 * NM-plugin discussions 15:09 - Pushed appliance-build livecd-rootfs changes to bionic Unapproved 15:09 - Asked for upload of the new lxd appliance model assertion to the store 15:09 - Various meetings 15:09 (done) 15:09 * improved jck snap documentation 15:09 * working on regression detection for the jck testsuite in jenkins 15:09 - grabbing/copying artifacts in jenkins 15:09 * openjdk-11 and openjdk-8 security updates 15:09 - packaged openjdk-11, build failure on bionic, investigating 15:09 - more CET investigations 15:09 - GCC 10.2 release candidate 15:09 - binutils 2.35 snapshots 15:09 - Python 3.9 beta4 15:09 - two MIRs, NEW processing 15:09 - packaging openjdk-8 15:09 - get lintian again installable on i386 15:09 * fixed jtreg regression 15:09 - started libgc and mpfr4 transitions 15:09 (done) 15:09 - two +1 days 15:09 (done) 15:10 any questions on status? 15:10 None here! 15:10 #topic Release incoming bugs 15:11 #link http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-gg-incoming-bug-tasks.html#foundations-bugs 15:11 bug 1886314 - is this for desktop or us? 15:11 bug 1886314 in ubuntu-meta (Ubuntu) "preinstall `earlyoom` by default" [Wishlist,New] https://launchpad.net/bugs/1886314 15:12 vorlon: ? 15:12 uh 15:13 I would want the kernel team to weigh in on this 15:13 before we install a package that changes the default oom handling 15:13 okay 15:14 huh, why does it need a userspace sercice? 15:14 yeah exactly 15:15 So is anybody interested in this enough to drive it? 15:15 not really 15:15 nope 15:15 bdmurray: not fixing, possibly punt to like cking maybe 15:15 it sounds interesting but also dubious 15:15 ack 15:15 bug 1887098 15:15 bug 1887098 in what-is-python (Ubuntu) "Installing python-is-python3 should affect pip as well" [Undecided,Confirmed] https://launchpad.net/bugs/1887098 15:17 is there a pip2? 15:17 this bug actually has a duplicate 15:18 doko: we have python2 pip and python3 pip3, we don't have pip2 15:18 hmm, is python2 based pip really available in focal? 15:18 let me check. 15:18 afaicr, that was only re-added later in unstable after the focal release. 15:18 doesn't appear to be (at least there's no python-pip or python2-pip?) 15:19 $ dpkg -S bin/pip 15:19 python3-pip: /usr/bin/pip3 15:19 python3-pip: /usr/bin/pip 15:19 in groovy 15:19 probably should be packported 15:19 if we want to have a python2 based pip like debian will ship with bullseye 15:20 but I'm -1 on pip-is-* 15:21 doko: the request is for existing "python-is-python3" to ship pip -> pip3 symlink, like it does for `python` `python-config` etc. 15:21 python-is-python3 should add pip->pip3 symlink 15:21 not a new package 15:21 however 15:21 and presumably then python-is-python2 should do likewise for pip? 15:21 juliank: no way, please don't fuck around with extra dependencies 15:21 doko: no extra dependencies 15:21 dangling symlink 15:21 groovy ships it in python3-pip 15:21 so dangling symlink? 15:22 note that in groovy /bin/pip is python3 shebang 15:22 if we add dangling symlink to python-is-python3, we also need replaces/breaks in python3-pip in groovy 15:22 but there is no python2 pip... so python-is-python2 -> pip does not work. Futhermore, there would be file conflicts if python2 pip is re-introduced 15:23 right 15:27 slyon: we will not do that in the python-is-python2 package 15:27 so /bin/pip will be a dangling symlink in python-is-python3 package only 15:28 we can make it a file trigger 15:28 I think this warrants some further discussion via email with a clear design. 15:28 ok 15:28 to be continued 15:29 xnox: will you draft a plan? 15:29 ok 15:29 bug 1878225 already has a card so skipping 15:29 bug 1878225 in autopkgtest (Ubuntu) "Please remove lxd.snap from lxd images, as it fails to seed thus failing the first boot - snapd.seeded.service waits forever (?) to have snaps seeded in LXD on s390x and arm64" [Undecided,Triaged] https://launchpad.net/bugs/1878225 15:30 #link http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-ff-incoming-bug-tasks.html#foundations-bugs 15:30 bug 1871641 the desktop team is looking for help from "a systemd expert" 15:30 bug 1871641 in plymouth (Ubuntu) "Ubuntu never finishes booting: A start job is running for Hold until boot process finishes up (3min 7s / no limit) -- removing 'splash' kernel parm fixes it" [High,Confirmed] https://launchpad.net/bugs/1871641 15:31 i guess that's a +1 for me :-) 15:31 could you reach out to daniel and see what he needs? 15:32 sure, will do so 15:32 NB he's probably not very timezone-compatible 15:33 bdmurray, should i card it for myself instead for the team? 15:33 rbalint: that sounds good to me 15:34 are we waiting for hardware for bug 1883229? 15:34 done 15:34 bug 1883229 in subiquity "Ubuntu server installation on S3260M4 server fails" [High,Confirmed] https://launchpad.net/bugs/1883229 15:37 alright, that's that for new bugs 15:37 #topic Team proposed-migration report 15:37 vorlon: 15:38 https://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses_by_team.html#foundations-bugs 15:38 41 packages needing attention, that doesn't sound good 15:38 vim is still mine 15:38 well, build-essential triggered quite a few I suppose 15:38 lintian is with me 15:39 I don't see why it should be 15:39 bdmurray: ok thanks 15:39 python3.8 vs sqlite 15:39 libio-async-perl is still mine 15:39 i'm fixing systemd 15:40 sil2100: are you still on the python3.8 vs sqlite regression? 15:40 (and sqlite3 generally) 15:41 vorlon: yes, sorry, still on it... 15:41 ok 15:41 libio-async-perl is just waiting on its dep which is part of the linitan stuff 15:41 rbalint mentioned being on systemd 15:41 juliank: xz-utils? 15:41 yeah sounds fine 15:42 juliank: I think I tried to assign this to you last time but I'm not sure you saw 15:43 casper blocking qemu, what's that about 15:43 looks like we should leave this with the server team 15:44 curl is blocked on nettle, which is the ghc transition 15:44 freetds, gnutls28, libarchive same 15:44 vorlon: I see my mention in the backlog, but I'm not sure I actively noticed 15:45 open-iscsi also tied to qemu, so skipping 15:45 rtmpdump, nettle again; then systemd and ubuntu-image for qemu again 15:45 libzstd 15:45 I can look into libzstd vs badger, as I have been dealing with badger before 15:45 slyon: thanks 15:45 (at least for the 64bit failure, 32bit problems need upstream fixing it) 15:46 build-essential, I've actually been picking at these already this morning, strangely there are some i386 autopkgtest false-positives that were never seen before 15:46 so I'll follow through on that one 15:46 https://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html#build-essential 15:47 glib2.0, might just need some autopkgtest retriggers. xnox can you take? 15:48 vorlon: seb128 was doing some today, and some of them depend on the qemu fix previously mentioned. 15:48 vorlon: but sure, can try to help poking it. 15:48 well the qemu in release pocket should be ok surely 15:48 python-tz needs oops to drop python2 support, should be easy 15:49 waveform: ^^ can you take? 15:49 vorlon, will do 15:49 s390-tools waiting on AA new to accepted signing 15:49 ok I'll look at that shortly 15:49 dosfstools regressing livecd-rootfs, doesn't sound good 15:50 tdaitx: ^^ can you look at this? 15:50 vorlon: ack 15:50 rbalint: is systemd/amd64 regressing for kbd related to your ongoing work? 15:50 tdaitx: maybe livecd-rootfs broken by itself? " python3-secretstorage : Depends: python3-jeepney but it is not installable" 15:50 could be just components missmatches 15:50 that's a -proposed-only issue 15:51 python3-jeepney is (was?) an MIR 15:51 vorlon, i take kbd it then i'll see 15:51 ok 15:51 bdmurray: can you take dnspython vs python-idna? 15:52 libsereal-decoder-perl, rather looks like a Debian issue, we might just ignore this for now and see if something better comes in via sync in the next week 15:52 and libevent vs mysql-8.0 15:52 who wants this last one 15:52 vorlon, it is mysql 15:53 server team! 15:53 vorlon: okay 15:53 rbalint: meaning? 15:53 vorlon, i take it, it regressed in release 15:53 (mysql regressed) 15:54 Failing test(s): main.ssl_crl main.ssl_crl_crlpath main.ssl_crl_clients_valid 15:54 rbalint: ok, you'll follow through with hint mp? 15:54 sounds like dates / times / timing valid/from valid/until 15:54 vorlon, yes 15:54 thanks 15:54 and that's done 15:54 bdmurray: 15:54 #topic AOB 15:55 just a heads up, GCC 10.2 will be released next, and once that's built the defaults change will happen 15:55 next week even 15:56 nice 15:56 okay, thanks for the notification 15:56 doko: horay! 15:56 doko: can you communicate this to ubuntu-devel also? 15:56 vorlon: no, it's in the release schedule ;o 15:57 more notifications don't hurt! 15:58 thanks everyone 15:58 #endmeeting