16:03 <bdmurray> #startmeeting Weekly Ubuntu Foundations team
16:03 <meetingology> Meeting started Thu Feb 20 16:03:06 2020 UTC.  The chair is bdmurray. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
16:03 <meetingology> 
16:03 <meetingology> Available commands: action commands idea info link nick
16:03 <bdmurray> #topic Lightning rounds
16:03 <bdmurray> echo $(shuf -e vorlon bdmurray xnox tdaitx doko sil2100 rbalint infinity mwhudson juliank waveform)
16:03 <bdmurray> bdmurray mwhudson tdaitx rbalint sil2100 doko xnox infinity vorlon waveform juliank
16:03 <bdmurray> submitted RT regarding ET mojo spec jenkins run failing
16:03 <bdmurray> reading / posting on Raspberry Pi forums re Ubuntu
16:03 <bdmurray> writing / discussing ubuntu-patch-status specification
16:03 <bdmurray> uploaded u-r-u fixing LP: #1834501
16:03 <bdmurray> testing updated linux-firmware-raspi2 and 802.11ac
16:03 <bdmurray> sru-review of containerd for mwhudson
16:03 <bdmurray> short week due to a holiday
16:03 <ubottu> Launchpad bug 1834501 in lxd (Ubuntu) "do-release-upgrade from bionic->any disables lxd without snapstore access" [Undecided,Confirmed] https://launchpad.net/bugs/1834501
16:03 <bdmurray> ✔ done
16:04 <bdmurray> tdaitx:
16:04 <bdmurray> rbalint: how about you?
16:04 <rbalint> * partner work
16:04 <rbalint> * proposed migration: task
16:05 <rbalint> * ftbfs fix: flatbuffers
16:05 <rbalint> * SRUs: unattended-upgrades
16:05 <rbalint> * speeding up unattended-upgrades in progress
16:05 <rbalint> (done)
16:05 <bdmurray> rbalint: Do those SRUs need review?
16:05 <bdmurray> sil2100:
16:05 <rbalint> bdmurray, sil2100 kindly did them
16:05 <sil2100> Skip
16:05 <sil2100> plz
16:05 <doko> - finally sent the python2 removal email
16:05 <doko> - python3.7 removal: https://people.canonical.com/~ubuntu-archive/transitions/html/python3.8-only.html
16:05 <doko> - icu transition
16:05 <doko> - working on a gcc-8 upstream issue, armhf, for the 8.4 release
16:05 <doko> - gcc-8 update
16:05 <doko> - gcc-9 update, add the fix for -O3 ppc64el endless loop
16:05 <doko> - some unrelated uploads/transitions
16:05 <doko> (done)
16:06 <bdmurray> xnox:
16:06 <bdmurray> vorlon:
16:06 <vorlon> * short week, out last Friday and this Monday
16:06 <vorlon> * proposed-migration and merges
16:06 <vorlon> * a few more i386 autopkgtests passing
16:06 <vorlon> (done)
16:07 <bdmurray> waveform is ill
16:07 <bdmurray> juliank:
16:07 <juliank> you guys are too fast
16:07 <sil2100> I'm ready
16:07 <bdmurray> go for it
16:07 <sil2100> - SRU reviews and releases
16:07 <sil2100> - Some kernel SRUs
16:07 <sil2100> - Looking into the quirk check-free-space update-manager crash bug - already fixed
16:07 <sil2100> - Looked into bubblewrap/libcap2 ADT issues, filled bugs in LP and Debian
16:07 <sil2100> - Refreshed core16 images
16:07 <xnox> bdmurray:  i was still in IBM call sorry.
16:07 <sil2100> - Archive administration work
16:08 <sil2100> - Poked people to get Dave added to the right teams for pi-gadget commit access
16:08 * xnox starts typing my own notes
16:08 <sil2100> - netplan:
16:08 <sil2100> * Played around with the code
16:08 <sil2100> * Looked at the NetworkManager read-write spec, answered some questions, asked a few more
16:08 <sil2100> * Prepared packaging for the required libnetplan library, did some test builds
16:08 <sil2100> - Prepared some symlinks to pi4 images on nusakan
16:08 <sil2100> - Interviews
16:08 <sil2100> (done)
16:08 * juliank ready
16:08 <juliank> * learning more about grub scripting and dmi, and modaliases
16:08 <juliank> * apt 1.9.10
16:08 <juliank> * started breaking the APT ABI in master
16:08 <juliank> * learning how to build images locally with ubuntu-cdimage and friends
16:08 <juliank> * sponsoring maas deb2snap uploads for ack and co
16:08 <juliank> * fixing apt fuzz segfaults in tagfile parser
16:08 <juliank> * python-apt:
16:08 <juliank> - ported to pybuild
16:08 <juliank> - fixed segmentation fault in python-apt
16:08 <juliank> - added set_priority calls for rbalint
16:08 <juliank> - made it build against apt master
16:08 <juliank> * fixed aptitude
16:08 <juliank> * broke shim? it can't load fwupd on focal, investigating did not yield useful results so far
16:09 <juliank> * found i915 patches
16:09 <juliank> * "testing" mesa 20 from proposed :D
16:09 <juliank> (done)
16:09 <bdmurray> xnox: are you ready?
16:09 <juliank> thanks xnox for ubuntu-cdimage tutorial
16:09 <xnox> Good progress on icu transition, about just a dozen packages left
16:09 <xnox> Ubuntu-core-initramfs did upload with better cryptsetup, still needs work as post-initrd boot is still borked up a bit.
16:09 <xnox> Subiquity luks is done in edge channel now, it’s nice.
16:09 <xnox> Done
16:09 <xnox> juliank:  your welcome =)
16:09 <xnox> bdmurray:  let's call that my status
16:09 <bdmurray> xnox: works for me
16:10 <bdmurray> #topic Release incoming bugs
16:10 <bdmurray> #link http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-ff-incoming-bug-tasks.html#foundations-bugs
16:10 <bdmurray> bug 1863532
16:10 <ubottu> bug 1863532 in python3-defaults (Ubuntu) "Invoking "python" brings inappropriate response from command-not-found when python3 installed" [Undecided,Confirmed] https://launchpad.net/bugs/1863532
16:11 <bdmurray> this seems worth taking
16:12 <bdmurray> I'll target and card it
16:12 <bdmurray> bug 1863195
16:12 <ubottu> bug 1863195 in dh-python (Ubuntu) "py3compile should not emit python runtime warnings during dist-upgrade" [Undecided,New] https://launchpad.net/bugs/1863195
16:13 <bdmurray> I also saw this during an upgrade
16:14 <bdmurray> This would add a lot of noise to debugging dist-upgrades
16:18 <doko> I don't care, do what you have to do
16:23 <rbalint> +1 for carding it
16:24 <bdmurray> We'll talk about this at the sprint in Frankfurt but will also target and card it
16:25 <bdmurray> bug 1862846 is already targetted and in progress
16:25 <ubottu> bug 1862846 in util-linux (Ubuntu Focal) "Crash and failure installing focal" [Medium,In progress] https://launchpad.net/bugs/1862846
16:26 <bdmurray> bug 1861655
16:26 <ubottu> bug 1861655 in The Ubuntu-power-systems project "[Ubu 20.04][kernel-5.4.0-12-generic][WSP-DD2.3] Unable to install Ubu 20.4 on NVMe disk" [High,New] https://launchpad.net/bugs/1861655
16:27 <bdmurray> we'll card that
16:28 <bdmurray> It looks like bryce has made progress with bug 1861472 but has a lot of philosophical questions in there
16:28 <ubottu> bug 1861472 in openssh (Ubuntu) "upgrade from fresh bionic to focal needlessly prompts user" [High,Triaged] https://launchpad.net/bugs/1861472
16:29 <bdmurray> vorlon will follow up on the questions
16:30 <bdmurray> bug 1861470 sounds familiar did we discuss it?
16:30 <ubottu> bug 1861470 in linux-signed-5.4 (Ubuntu) "BIOS logo reappears mid-boot after the purple screen" [Medium,Confirmed] https://launchpad.net/bugs/1861470
16:30 <vorlon> is this the one I've commented on?
16:30 <vorlon> yeah
16:33 <bdmurray> This sounds like something we should coordinate across teams so a spec would be appropriate to get agreement and buy in.
16:33 <bdmurray> Or start via an email which xnox will send.
16:34 <bdmurray> bug 1861082 is a duplicate and think is a won't fix from us
16:34 <ubottu> bug 1861082 in apport (Ubuntu) "ubuntu-bug doesn't know how to file bugs against snaps" [Undecided,Confirmed] https://launchpad.net/bugs/1861082
16:35 <bdmurray> because there isn't any information provided by snaps about where bugs should go
16:37 <bdmurray> I seem to recall some previous work about apport telling you something is a snap e.g. if you do ubuntu-bug git-ubuntu
16:38 <bdmurray> and that does tell you something
16:39 <bdmurray> xnox will do some stuff with that bug
16:40 <bdmurray> #link http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-ee-incoming-bug-tasks.html
16:41 <bdmurray> bug 1842439
16:41 <ubottu> bug 1842439 in apport (Ubuntu) "apport-gtk crashed with SIGSEGV in _gtk_settings_get_screen(settings=0x0)" [Medium,Confirmed] https://launchpad.net/bugs/1842439
16:41 <bdmurray> That errors bucket is somewhat large so I think its worth having a look. Agreed?
16:42 <sil2100> +1
16:42 <bdmurray> rls-bb-incoming is fine
16:43 <bdmurray> #topic Team proposed-migration report
16:43 <bdmurray> vorlon:
16:43 <vorlon> [LINK] https://people.canonical.com/~ubuntu-archive/proposed-migration/focal/update_excuses_by_team.html#foundations-bugs
16:44 <vorlon> the top one is libyaml-libyaml-perl, which is stuck in both Debian unstable and -proposed because of some upstream incompatibilities
16:44 <vorlon> I took this one last time but didn't make any progress on it
16:44 <vorlon> I'm inclined to deprioritize it, unless someone else is keen?
16:44 <xnox> kick it out of -proposed?
16:45 <vorlon> xnox: why?  it is likely to be fixed eventually in Debian
16:45 <xnox> if it can't migrate, and we can't update the package if we need to = bad
16:45 <xnox> cause /this/ version will never migrate
16:45 <vorlon> when have we ever updated this package
16:45 <xnox> =)
16:45 <xnox> lolz ok
16:45 <vorlon> xnox: no, it's a /related/ package whose autopkgtests are failing
16:45 <xnox> ah
16:45 <vorlon> and that package might get fixed so that they can go in together
16:45 <xnox> ok
16:45 <vorlon> so I would leave it there and ignore unless something else comes up that makes it a priority
16:45 <vorlon> gmp
16:45 <xnox> open a bug report with update-execuse tag explaining this?
16:46 <vorlon> rbalint: you took this one last week, do you need any help?
16:46 <vorlon> xnox: ok
16:46 <rbalint> gmp is mine, behavour changes
16:46 <vorlon> I'll do that
16:46 <rbalint> need revdep fixes, on it
16:46 <vorlon> ok
16:46 <vorlon> boost-defaults
16:46 <vorlon> xnox: need any help here for this transition?
16:46 <xnox> can we just migrate it?
16:47 <vorlon> sounds like no (britney is right!)
16:47 <vorlon> and sounds like boost-defaults+icu is going to be ready to go soon
16:48 <vorlon> pyserial/mu-editor
16:48 <vorlon> waveform took this
16:48 <vorlon> waveform: (when you're around) any progress on pyserial, mu-editor on armhf?
16:49 <vorlon> libsereal-decoder-perl is tdaitx, who is also not around, carry over
16:49 <vorlon> icu, as discussed
16:49 <vorlon> openjdk-lts is blocked on the i386 autopkgtest regression
16:49 <vorlon> well, "regression"
16:49 <vorlon> doko: do I need to dig up my patch for i386 autopkgtests and send it to you?
16:50 <vorlon> or have you talked to tdaitx about this?
16:51 <doko> vorlon: not yet, he's still lunching :-/
16:51 <vorlon> doko: so will you follow up with him or should I dig up my link?
16:52 <doko> maybe digging it up is safer
16:53 <vorlon> ok
16:53 <vorlon> apt + libzstd
16:53 <vorlon> juliank: ^^ any insight?  do you want to take this and dig into it?
16:53 <vorlon> might just need a retry
16:53 <juliank> ill loo
16:53 <vorlon> ok
16:53 <juliank> k
16:53 <vorlon> xfsprogs, entangled with icu
16:54 <vorlon> procps: this was my upload, I've pinged the server team about figuring out why postgresql-common doesn't like the new version, I'll re-ping
16:54 <vorlon> (the postgresql-comomn tests require a VM rather than a chroot, so I can't be bothered to reproduce it locally ;p)
16:54 <vorlon> lintian
16:54 <vorlon> ftbfs
16:55 <vorlon> who wants this one?
16:55 <vorlon> bdmurray: ?
16:56 <vorlon> and I'll take netbase, which is munin, which is flaky
16:56 <vorlon> bdmurray: can you take lintian?
16:56 <bdmurray> okay
16:56 <vorlon> thanks
16:56 <vorlon> and that's the list for today
16:57 <bdmurray> #topic AOB
16:57 <bdmurray> I'm out starting the 25th
16:57 <juliank> I'll have to land libapt-pkg6.0 next week before FF. I might still want a FFe to add more patterns later when I'm less busy
16:57 <ahasenack> can I bring something up? (don't know what aob is, any other business?)
16:57 <doko> ahh, I wanted to point out the two pending transitions:
16:57 <juliank> ahasenack: yes it is
16:58 <doko> https://people.canonical.com/~ubuntu-archive/transitions/html/python3.8-only.html
16:58 <bdmurray> its not ahasenack other business though
16:58 <juliank> bdmurray: that's true
16:58 <doko> https://people.canonical.com/~ubuntu-archive/transitions/html/icu.html
16:58 <doko> help on those is welcome
16:58 <vorlon> doko: from what xnox says, icu should be "done soon", are there specific things here that you and xnox want help with?
16:58 <xnox> doko:  are unknown actually bad? or unknown actually good?
16:58 <vorlon> I can pitch in but don't want to duplicate effort
16:59 <juliank> bdmurray: we could invite him to share, though, let's be friendly
16:59 <xnox> vorlon:  there are like 12 packages left, i'm not stuck yet. if i do get stuck i'll ask for help.
16:59 <vorlon> ok
16:59 <doko> xnox: I'm ignoring unknown
16:59 <xnox> so far it was fallout from new SDL2 and new CMAKE nothing related to icu itself, just generic bitrot FTBFS
16:59 <bdmurray> ahasenack: what do you want to bring up?
16:59 <ahasenack> my topic is bind 9.16, upstream's "lts" stable version. isc-dhcp, bind-dyndb-ldap (universe) don't build with it, and debian has come up with https://salsa.debian.org/dns-team/bind9-libs/commit/40cab7029da3680cb9e58fa849018d942a6013b1
16:59 <xnox> ahasenack:  i asked them to release LTS stable in time for ubuntu LTS.
17:00 <ahasenack> which is package 9.16.x as just libraries, to keep isc-dhcp happy, and ship 9.16 with the full blown server
17:00 <xnox> ahasenack:  my hope was that we will take it.
17:00 <doko> anyway, if anybody is picking up a python3.8 issue, just mention it on irc
17:00 <ahasenack> yeah, I've been working on 9.15.8 for the past few days, trying to pacakge it, check the differences, etc
17:00 <ahasenack> this affects also freeipa, but it's my understanding we removed it from focal already, due to java issues, or something like that
17:00 <xnox> ahasenack:  we do want to keep isc-dhcp working
17:01 <xnox> ahasenack:  so i guess we do need to do this mess, and take it into focal
17:01 <juliank> :/
17:01 <xnox> don't care about bind-dyndb-ldap, but we do care about isc-dhcp to be working
17:01 <ahasenack> #security also prefers to have 9.16 as the server
17:01 <vorlon> xnox: do you want to create a card for the isc-dhcp work?
17:01 <xnox> can we fix isc-dhcp or is it hard?
17:01 <ahasenack> and is ok with having 9.11 as libs only, even though they are used (for what?) in isc-dhcp
17:01 <xnox> or like make isc-dhcp static link?
17:01 <xnox> cause i'm not sure we want to expose old libs for everyone
17:02 <xnox> ahasenack:  it is ok for having 9.11 as libs only, yes.
17:02 <ahasenack> there are also the udebs which I don't know how they are used
17:02 <ahasenack> debian's 9.16 package doesn't build them
17:02 <vorlon> only in d-i, so let's just break those deliberately
17:02 <vorlon> ;p
17:02 <xnox> ahasenack:  it's isc-dhcp which is used in d-i.
17:02 <ahasenack> I mean, I know it's in d-i, just not how, or how to test it
17:02 <xnox> ahasenack:  hence you'll probably need the old bind libs as udebs
17:03 <xnox> but not the new ones
17:03 <ahasenack> xnox: the bind9-libs src pkg from that salsa link above builds the udebs
17:03 <ahasenack> the 9.11 ones
17:03 <xnox> yes
17:03 <ahasenack> so it looks like a working plan
17:03 <xnox> yeap
17:03 <ahasenack> but wanted to check with you
17:03 <ahasenack> specially because we will have two major versions of bind 9 in main then
17:03 <ahasenack> even though one only builds libraries
17:03 <vorlon> :/
17:03 <xnox> yes
17:04 <xnox> ahasenack:  alternative is to build systemd-networkd netplan python3.8 into initramfs and udebs
17:04 <ahasenack> I don't know about building isc-dhcp with static bind9.11 libraries
17:04 <xnox> ahasenack:  drop / break isc-dhcp
17:04 <xnox> and thus drop / break ifupdown
17:04 <juliank> nooooooooo
17:04 <ahasenack> I tried merging isc-dhcp once from debian, couldn't do it. Too much undocumented history
17:04 <xnox> which is not going to happen
17:04 <xnox> ahasenack:  i think people at this point depend on the bugs in isc-dhcp as stable ABI
17:05 <juliank> but .. can't it go to universe?
17:05 <xnox> juliank:  we use it in the initrd
17:05 <juliank> ifupdown is in universe
17:05 <xnox> juliank:  isc-dhcp
17:05 <juliank> hmm
17:05 <vorlon> xnox: isc-dhcp is also used in initramfs
17:05 <ahasenack> I was told by upstream that isc-dhcp is on life support
17:05 <vorlon> sure, but in favor of what
17:06 <ahasenack> isc-kea probably
17:06 <juliank> systemd dhcp client?
17:06 <ahasenack> ah, mixed streams, sorry
17:06 <vorlon> juliank: in initramfs
17:06 <vorlon> that's not a change to make for 20.04
17:06 <juliank> yes
17:07 <juliank> :D
17:07 <vorlon> did we manage to close out this meeting, or is it going to go forever because bdmurray is afk? :)
17:07 <xnox> juliank:  i don't think we can use networkd dhcp client stand alone, as it requests different set of DHCP options / not enough for like ISCSI root mount on Oracle
17:07 <bdmurray> I'll close it out
17:07 <bdmurray> #endmeeting