jbichavorlon: https://ubuntu-archive-team.ubuntu.com/madison.cgi is displaying code instead of the webpage00:52
vorlonjbicha: well that's fun, I didn't know we had any cgi scripts under there. :|02:00
vorlonjbicha: does that break rmadison?02:00
jbichayes, that's how I noticed02:01
vorlonthat's annoying.  I'll file an RT about it shortly but first I'm looking at the fact that I broke all image builds by having archive-team.internal DNS changed to a machine that's not running apache02:01
vorlonok filing an RT but also trying to figure out if I can magic up an .htaccess in the meantime that redirects everything NOT madison.cgi02:07
vorlonso the answer to the latter is that I definitely can and it's just a question of how magic02:20
vorlonjbicha: hacky hacky02:27
vorlon(it's working)02:27
arraybolt3Hacky is how it's done :)02:28
jbichathe rmadison output seems a little out of date. gnome-pkg-tools is a test case02:32
vorlonyes it will be02:34
vorlonI need to turn a cron job back on on snakefruit to make it work02:34
vorlonbut first I need to figure out which one02:35
vorlon(probably the big one which means I need to neuter it)02:35
vorlonjbicha: fixed02:44
=== chris14_ is now known as chris14
-queuebot:#ubuntu-release- Packageset: Added oem-sutton-banagher-meta to canonical-oem-metapackages in bionic04:24
-queuebot:#ubuntu-release- Packageset: Added oem-sutton-cappy-meta to canonical-oem-metapackages in bionic04:24
-queuebot:#ubuntu-release- Packageset: Added oem-sutton-caressa-meta to canonical-oem-metapackages in bionic04:24
-queuebot:#ubuntu-release- Packageset: Added oem-sutton-banagher-meta to canonical-oem-metapackages in focal04:24
-queuebot:#ubuntu-release- Packageset: Added oem-sutton-cappy-meta to canonical-oem-metapackages in focal04:24
-queuebot:#ubuntu-release- Packageset: Added oem-sutton-caressa-meta to canonical-oem-metapackages in focal04:24
-queuebot:#ubuntu-release- Packageset: Added oem-sutton-banagher-meta to canonical-oem-metapackages in jammy04:24
-queuebot:#ubuntu-release- Packageset: Added oem-sutton-cappy-meta to canonical-oem-metapackages in jammy04:24
-queuebot:#ubuntu-release- Packageset: Added oem-sutton-caressa-meta to canonical-oem-metapackages in jammy04:24
-queuebot:#ubuntu-release- New sync: oem-sutton-banagher-meta (jammy-proposed/primary) [22.04~ubuntu1]06:39
-queuebot:#ubuntu-release- New sync: oem-sutton-cappy-meta (jammy-proposed/primary) [22.04~ubuntu1]06:40
-queuebot:#ubuntu-release- New sync: oem-sutton-caressa-meta (jammy-proposed/primary) [22.04~ubuntu1]06:42
LocutusOfBorgsergiodj, I uploaded emacs to Ubuntu with gcc-13 on riscv64, lets cross fingers (and also merged from Debian)06:48
LocutusOfBorgwe had security fixes there06:49
schopinvorlon: ack on the FFe status field. I'm honestly annoyed at the context-sensitive nature of the status field in LP, though.09:55
-queuebot:#ubuntu-release- Unapproved: linux-firmware (jammy-proposed/main) [20220329.git681281e4-0ubuntu3.10 => 20220329.git681281e4-0ubuntu3.11] (core, kernel)10:53
-queuebot:#ubuntu-release- Unapproved: linux-firmware (focal-proposed/main) [1.187.36 => 1.187.37] (core, kernel)11:04
juliankWho can rescore-ppa-builds ubuntu-uefi-team ppa for me?11:57
juliankThere's a riscv64 build in there that's queuing a bit long11:58
* juliank is waiting on that to continue with the gnu-efi reverse depends rebuilds for kinetic to ensure SRU doesn't regress (or we at least know what regresses)11:59
-queuebot:#ubuntu-release- Unapproved: fwupd (bionic-proposed/main) [1.2.14-0~18.04.2 => 1.2.14-0~18.04.3] (desktop-core)12:00
cjwatsonjuliank: Done, but I think I'd probably better do some builder repair which will be more useful12:36
kanashiro[m]slyon: hi, not sure if you saw my comment from yesterday, but could you update the status of the mosh MIR bug? (LP: #1997106) It is already in the component mismatches page12:36
-ubottu:#ubuntu-release- Launchpad bug 1997106 in mosh (Ubuntu) "[MIR] mosh" [Undecided, In Progress] https://launchpad.net/bugs/199710612:36
cjwatsonriscv64 has a hard enough time keeping up as it is without half the builders disabled12:37
-queuebot:#ubuntu-release- Builds: Ubuntu Server Subiquity amd64 [Focal 20.04.6] has been marked as ready12:38
juliankcjwatson: thanks. basically I just need to ensure it's built before the other PPA starts its 4 test rebuilds12:39
juliankI don't think we have riscv64 docker containers or I could test that with qemu-user-static emulation locally12:40
juliankoh we do12:42
cjwatsonIt's building now, anyway12:42
juliankyeah but this means I can run the test rebuilds locally12:42
juliankof the reverse deps12:42
juliankThe PPA build is needed anyhow as it's binary copied to the main archive12:43
juliank(because this is building against -security only)12:43
cjwatsonRight, riscv64 builders all healthy again now12:53
-queuebot:#ubuntu-release- Unapproved: accepted linux-firmware [source] (jammy-proposed) [20220329.git681281e4-0ubuntu3.11]12:54
-queuebot:#ubuntu-release- Unapproved: accepted linux-firmware [source] (focal-proposed) [1.187.37]12:57
slyonkanashiro[m]: thanks for the reminder, I missed it yesterday. let me have a look13:00
-queuebot:#ubuntu-release- Unapproved: livecd-rootfs (focal-proposed/main) [2.664.45 => 2.664.46] (desktop-core, i386-whitelist)13:05
slyonkanashiro[m]: it seems the mandadory MIR requirement #3 and #4 are not yet addressed. Is there anything planned to fix this?13:05
-queuebot:#ubuntu-release- Unapproved: accepted livecd-rootfs [source] (focal-proposed) [2.664.46]13:07
slyoncan we at least run the build-time unittests against the installed package during autopkgtests for example?13:07
sil2100Hello everyone! We'll have to respin 20.04.6 images sadly! But the changeset should be small, so I wouldn't require retesting everything13:10
ricotzsil2100, hello :), would you be so kind to take a look at https://bugs.launchpad.net/ubuntu/+source/libreoffice/+bug/200935413:30
-ubottu:#ubuntu-release- Launchpad bug 2009354 in libreoffice (Ubuntu Kinetic) "[SRU] libreoffice 7.4.6 for kinetic" [High, Incomplete]13:30
ahasenacksil2100: ginggs: hi, about the focal soft freeze, is that still ongoing? I have a rabbitmq sru for f, j, k, and was wondering if I should hold the focal release14:07
sil2100ahasenack: please hold for now14:07
ahasenacksil2100: ack14:07
sil2100ahasenack: we'll be respinning soon...14:07
tillkamppeterHi, I need an exception for a failing autopkg test. It is the test of the foo2zjs package which is triggered by several printing-related packages, especially cups-filters: https://ubuntu-archive-team.ubuntu.com/proposed-migration/lunar/update_excuses.html#cups-filters14:15
tillkamppeterThe test is triggered because foo2zjs depends on cups-filters.14:16
kanashiro[m]slyon: let me take a look and I will get back to you, thanks for checking!14:16
tillkamppeterThe relevance of the failure is very low for the distro, as it foo2zjs is a driver package for ~100, usually older, printer models, rather cheap printers. Once, the package is not maintained upstream any more, second, it will be removed in Ubuntu 23.10, as printer drivers will get replaced by Printer Application Snaps. and third, it is not very probable that someone connects such a printer to a ppc64el system (server?).14:19
tillkamppeterSo having the build of the whole distro be held by a bug in foo2zjs does not make sense. Also there are drivers for other printers, ~10000 models where the autopkg tests pass with the current cups-filters (2.0b3).14:21
tillkamppeterDo I want to ask you to let failures of the autopkgtests of foo2zjs not be treated as regression. Could this be done?14:22
zhsjtillkamppeter: the regression looks strange, it only fails on ppc64el. it makes me think if it's related to -O3 option. and if it's broken on ppc64el, is it better to drop this arch?14:51
sergiodjLocutusOfBorg: ah, thanks.  I'm working with doko to determine what's going on, but I see that your upload's riscv64 build already passed the stage where it was failing before14:53
sergiodjso I think it'll work14:53
LocutusOfBorgsergiodj, I wanted to unblock some emacs stuff stuck on riscv6414:58
LocutusOfBorgfeel free to override my upload once you get the gcc fix :D14:58
tillkamppeterzhsj, you mean to not build it for that arch at all any more? We could do so. What do I have to add to the package so that its build on ppc64el gets excluded? Or will you in the release team make it getting not built for ppc64el any more?14:59
sergiodjLocutusOfBorg: thanks a lot!  :)14:59
zhsjtillkamppeter: i'm not the release team, lol. just my 2 cent :P14:59
-queuebot:#ubuntu-release- Unapproved: linux-firmware (kinetic-proposed/main) [20220923.gitf09bebf3-0ubuntu1.4 => 20220923.gitf09bebf3-0ubuntu1.5] (core, kernel)15:01
icey[m]hey vorlon : the fixes for our Ceph 17.2.5 SRU in Kinetic and Jammy are in the unapproved queues again, and I've run our distro-regression testing on that upload now15:04
kanashiro[m]slyon: as Robie was working on the mosh MIR and I got assigned just to finish the promotion (surprisingly it is not ready), I will not address your comments right now (it will require some package changes) and wait for Christian to get back from PTO.  I have another unfinished roadmap item requiring my attention (release is coming :)15:19
kanashiro[m]again, thanks for checking!15:19
dokoLocutusOfBorg: it's probably better to disable the jit in that case, same like for s390x (although I don't know the reasons for that=15:25
RikMillsseed change I made has not propagated to https://ubuntu-archive-team.ubuntu.com/seeds/kubuntu.lunar/desktop15:26
slyonkanashiro[m]: OK wfm. Thanks.15:29
LocutusOfBorgdoko, what is the problem of using gcc-13 on riscv64?15:35
LocutusOfBorgit works...15:35
vorlonRikMills: how long ago?15:40
vorlonicey[m]: I won't look at the SRU queue until tomorrow; ahasenack and sil2100 are on for today15:41
icey[m]tomorrow's ok too, I'm just hoping for a less-than-two-week acceptance :-)15:41
RikMillsvorlon: 2hrs for me. I also not that a change push to the ubuntu-mate seed at 10pm UTC yesterday has also not been updated there15:50
RikMills*also note15:50
vorlonRikMills: thanks, digging15:51
vorlonRikMills: we run update-seeds every 5 minutes, but it will fall over at the first entry that has failures updating and then the rest silently don't get processed15:54
bdrungcan we build pycurl for i386 as well (to get rid of the devscripts diff)? Especially since we have now unit tests for the code that uses pycurl.15:55
RikMillsyeah, I recall this from a while back15:55
vorlonbdrung: do you know why devscripts isn't Arch: all?15:56
vorlon/usr/bin/debpkg:             ELF 64-bit LSB pie executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, BuildID[sha1]=b2c596013974032c717e71cf9d6c1d0f33b1254d, for GNU/Linux 3.2.0, stripped15:57
vorlona command I've never heard of before15:57
vorlonand looks like an all-around bad idea15:57
bdrungtime to rewrite some scripts in rust to complete the language mix :D15:57
vorlonkill that upstream and then we don't have to worry about an i386 build ;)15:58
schopinThe question is why do we have an i386 build in the first place? Otherwise you're just moving the issue.15:59
-queuebot:#ubuntu-release- Builds: Ubuntu Desktop (Legacy) amd64 [Focal 20.04.6] has been updated (20230316)15:59
schopinunless pycurl is *only* used for the unit tests?16:00
vorlonbdrung: pycurl build-depends on vsftpd.  I'm not going down that road16:00
vorlonschopin: gem2deb build-depends on devscripts16:00
vorlonRikMills: running update-seeds manually has succeeded16:04
vorlonRikMills: rerunning with the proxy variables from cron set to see what's up16:05
RikMillsvorlon: thanks!16:05
bdrunggem2deb uses at least dch and wrap-and-sort16:06
vorlonRikMills: ok, 'bzr pull' on kubuntu-active.trusty is failing with the proxy variables set16:09
vorlonI'll dig in16:09
vorlon(the git pulls work fine)16:09
tillkamppeterAnyone of the release team could have a look at my problem with foo2zjs and ppc64el?16:09
tillkamppeter@vorlon? ^^16:10
zhsjbdrung: i want to learn the history of debpkg, why it exists in the first place. but salsa downs...16:10
vorlontillkamppeter: I agree with zhsj; test failures like this are a hint that there are broken binaries, and it's better for us to stop making those binaries available on archs where they are not useful and we're not going to support them, than to hint away failures.  I'm going to take a look on ppc64el at what's actually broken16:11
vorlonRikMills: oh this is the crash of bzr I've already seen elsewhere, that focal bzr + urllib are incompatible with http_proxy env var16:12
vorlonif we just got everyone off of bzr, we'd be golden16:12
vorlonas it is, I'll shove in a fix to unset http_proxy16:13
bdrungzhsj, having a local checkout is useful for this case.16:14
tillkamppeterThanks, @vorlon. if you have some solution, please also ping me on Mattermost, as I am not always on IRC.16:14
bdrungzhsj, https://manpages.debian.org/testing/devscripts/debpkg.1.en.html gives some explanation, but it is not clear to me what having debpkg benefits over making dpkg setuid.16:15
vorlonRikMills: deployed. seed updates should work fine now16:15
ItzSwirlzHi release-team, I just booted the Cinnamon image in a VM, and the.. I guess, Ubiquity stage? (The try or install ubuntu) - the background is still showing a debian logo, and I don't think it's using the Yaru theme. Where should I look? I'm willing to attempt a PR16:16
zhsjbdrung: that's why i want to check the git commits, the manpage really hurts my head.16:16
jbichaItzSwirlz: is it possible for you to not have desktop-base pre-installed since that shouldn't be needed on an Ubuntu system?16:18
ItzSwirlzjbicha: can you send a link of what desktop-base installs?16:18
jbichaItzSwirlz: you can find the file list at the bottom of https://launchpadlibrarian.net/647812491/buildlog_ubuntu-lunar-amd64.desktop-base_12.0.2ubuntu1_BUILDING.txt.gz16:19
jbichaItzSwirlz: or run dpkg -L desktop-base  or my favorite is  apt-file list16:20
ItzSwirlzYeah, that would explain why there's so much debian lying around.16:20
jbichaI don't think that would solve your problem but it would get rid of the Debian wallpaper16:20
ItzSwirlzIt's possible desktop-base overrides what is placed16:22
ItzSwirlzI also need to check GRUB, I didn't but I don't think the oem-config-gtk task is added either16:24
vorlontillkamppeter: cups-filters isn't moving anyway, the MIRs aren't finished16:28
tillkamppeter@vorlon, @zhsj by the way, -O3 cannot be the problem, as the package uses only O2 for building.16:28
vorlontillkamppeter: incorrect. dpkg overrides the default compiler flags for all builds on ppc64el16:28
vorlonyou can see it in the build log16:28
jbichaItzSwirlz: does Ubuntu Cinnamon use the correct wallpaper by default if you complete the Lunar install?16:29
tillkamppeter@vorlon, I know, but @seb128 reminded me that the foo2zjs test on ppc64el was still hanging, and for migration both have be solved. That was also a reason why I had prioritized other tasks before looking into the foo2zjs problem again.16:29
ItzSwirlzjbicha: move to dms16:30
vorlontillkamppeter: ok16:30
jbichaItzSwirlz: other people here or in #ubuntu-devel may have helpful advice too 🤷16:31
tillkamppeter@vorlon, OK, so it is O2 anyway, what I have seen now that even on amd64 it only uses O2, but this is probably due to the upstream build system. So we know it is not the O2/O3 problem here, so best is to take it away from ppc64el.16:32
vorlontillkamppeter: it is NOT -O2, look at the build log16:33
vorlonanyway I'm having an annoying time reproducing the failure because ppc64el uses too much memory so I use chroots instead of containers for my ppc64el debugging and the test case assumes cupsd is running. :|16:35
vorlonright manually starting it works around that16:36
vorlonbut the foo2zjs autopkgtest is buggy to not declare the isolation it needs16:36
tillkamppeter@vorlon, this I did not know, I have created and applied the test years ago to practically all printer drivers and it worked, up to now. So declaring isolation on it could perhaps help?16:38
vorlonit could be made compatible with chroots by stopping the service and manually starting cupsd but that's not really all that interesting16:40
vorlon(nobody runs tests in chroot in production)16:40
vorlonyeah now I'm getting completely different test failures16:41
tillkamppeter@vorlon, I only want to know now whether simply uploading foo2zjs with isolation declared for the test could already make the test passing and everything is OK, or whether we should now right away remove this package from ppc64el?16:57
vorlontillkamppeter: no, that has no bearing on the way autopkgtests are run on our infrastructure17:03
tillkamppeter@vorlon, OK, thanks, then it seems we should remove it from ppc64el. How to proceed for that? Do I need to mark it somehow in the package and upload or does the release team mark something internally?17:05
vorlonfoo2zjs isn't what regressed17:06
vorlonthis binary is unchanged and its tests now fail17:06
vorlonso what changed that is making foo2zjs not work?  that's the broken binary17:06
vorlonI'm working on getting a sane reproducer environment now (deploying a new larger ppc64el VM so I can launch containers)17:08
tillkamppeter@vorlon, OK, assuming a change in cups-filters (or the new dependencies libcupsfilters and libppd) would cause it, is that all the other printer drivers are no affected.17:10
sil2100FYI: 20.04.6 images respinning, confirmed with the desktop image that the workaround helped18:12
-queuebot:#ubuntu-release- Builds: Ubuntu Kylin Desktop amd64 [Focal 20.04.6] has been updated (20230316)18:33
-queuebot:#ubuntu-release- Builds: Xubuntu Desktop amd64 [Focal 20.04.6] has been updated (20230316)18:36
-queuebot:#ubuntu-release- Builds: Kubuntu Desktop amd64 [Focal 20.04.6] has been updated (20230316)18:37
-queuebot:#ubuntu-release- Builds: Ubuntu Budgie Desktop amd64 [Focal 20.04.6] has been updated (20230316)18:38
-queuebot:#ubuntu-release- Builds: Ubuntu MATE Desktop amd64 [Focal 20.04.6] has been updated (20230316)18:38
Eickmeyervorlon: ubuntustudio seeds not getting picked-up by update-seeds? I replaced the gnome-3-38-2004 snap with the gnome-42-2204 snap as required and latest build attempt doesn't reflect it happened.18:46
EickmeyerI made the change hours ago, fwiw.18:47
vorlonEickmeyer: siggggghhhh let me check again what's going on19:01
vorlonI did do a complete update-seeds run that succeeded when RikMills asked19:01
Eickmeyervorlon: I saw, but the last cron job build for Studio failed.19:02
Eickmeyer(merely 1/2 hour ago-ish)19:02
vorlonoh I know what it is19:02
vorlonI forgot to fix this last night19:02
vorlonI fixed the update of seeds on the new host, but livecd-rootfs hits archive-team.internal for the seeds; and that points to snakefruit still because ubuntu-archive-toolbox isn't running apache; and I turned off all the cronjobs on snakefruit19:03
EickmeyerOh nooooo....19:03
vorlonsorry.  I totally knew about this issue and then it slipped my mind19:04
vorlonrefreshing the seeds now19:04
EickmeyerIt's alright. I'm not unbelievably as stressed about this as I am other things.19:04
vorlonI can kick off another ubuntustudio build after this19:05
EickmeyerThat would be really helpful.19:05
vorlonI'll do kubuntu, ubuntu-mate, ubuntustudio which are the ones that have been mentioned here19:06
EickmeyerFWIW, edubuntu is looking sunshine and rainbows.19:06
vorlonah ubuntu-mate still failing with pipewire vs pulse19:21
EickmeyerI've been trying to guide them through that mess, but it seems there's a lot of desktop deps on pulse they need to navigate through.19:23
* vorlon nods19:24
-queuebot:#ubuntu-release- Unapproved: openldap (kinetic-proposed/main) [2.5.14+dfsg-0ubuntu0.22.10.1 => 2.5.14+dfsg-0ubuntu0.22.10.2] (i386-whitelist, ubuntu-server)19:24
seb128vorlon, could you trigger an ubuntu lunar rebuild as well? the iso tracker seems to think it's already rebuilding when it's not19:47
vorlonseb128: looking19:47
vorlonseb128: to be clear, you're looking for a rebuild in addition to the daily that ran at 7:41?19:48
vorlon(the cowboy is still in place for 20.04.6)19:49
seb128vorlon, yes, firefox is broken on the current ISO and we want to verify that the store assertion update just issue fixes the issue19:49
arraybolt3Out of curiosity, how long does it take for seed changes to propagate? I'm impatiently waiting for something I changed just a few minutes ago to go through so I can rebuild Lubuntu, and would like to know when I should check next.19:50
arraybolt3(I guess that's only partially out of curiosity, more out of wanting to not keep re-checking needlessly :P)19:50
vorlonarraybolt3: update-seeds is run on a 5-minute schedule.  The actual implementation of the command is has to walk through all the source repositories one-by-one which is annoyingly slow19:53
vorlonI haven't timed it but should still be sub-20 minutes for the whole thing19:53
-queuebot:#ubuntu-release- New: accepted gcr4 [source] (lunar-proposed) [4.1.0-0ubuntu1]19:55
-queuebot:#ubuntu-release- New binary: gcr4 [amd64] (lunar-proposed/none) [4.1.0-0ubuntu1] (no packageset)19:59
-queuebot:#ubuntu-release- New binary: gcr4 [s390x] (lunar-proposed/none) [4.1.0-0ubuntu1] (no packageset)20:00
RikMillsvorlon: thank you for the image respins. kubuntu livefs succeeded \o/20:00
RikMillsubuntu-mate did not /o\20:01
-queuebot:#ubuntu-release- New binary: gcr4 [arm64] (lunar-proposed/none) [4.1.0-0ubuntu1] (no packageset)20:04
vorlontillkamppeter: I can reproduce the ppc64el failure.  How can I manually invoke the filters *without* going through cupsd, to have a minimal test case?20:16
vorlontillkamppeter: I've tried rebuilding all of libcupsfilters, libppd, and cups-filters with -O2 instead of -O3 and these still fail.  and it happens sooner for me on my test VM (with an earlier driver and test file in the loop) than on the autopkgtest infra.  So I am left doubtful about the reliability of the whole cups-filters stack on ppc64el as a whole20:20
-queuebot:#ubuntu-release- New binary: gcr4 [riscv64] (lunar-proposed/universe) [4.1.0-0ubuntu1] (no packageset)20:29
vorlonxnox: if these linux-* packages in -proposed are never going to migrate should I just remove them?20:33
tillkamppeter@vorlon CUPS has a tool named "cupsfilter" (binary package "cups") which allows to run filter chains without using cupsd. It has also a man page. You supply the PPD file, input and output MIME types, options, an input file ... Then the same methods are applied to find the needed filter chain.20:36
-queuebot:#ubuntu-release- New binary: gcr4 [armhf] (lunar-proposed/universe) [4.1.0-0ubuntu1] (no packageset)20:43
-queuebot:#ubuntu-release- New binary: gcr4 [ppc64el] (lunar-proposed/universe) [4.1.0-0ubuntu1] (no packageset)20:44
tewardnotes for anyone watching GNOME and archive: Studio is having an issue with Germinate and requested (with agreement from Wimpy and others) to swap the ordering of dependencies in libcanberra-pulse and asked me with my coredev to upload.  Apparently, GNOME packaging uses an annoying d/control.in to populate d/control which i missed in -10ubuntu3 so I reuploaded WITH the fix as -10ubuntu4.  In case someone wonders why i did two20:48
teward... nearly at the same time give or take a few20:49
tewardPOC for the request came from arraybolt3 and Eickmeyer but Wimpy and RikMills both requested as well20:49
arraybolt3teward: MATE, not Studio.20:50
Eickmeyerteward: Nope MATE, not Studio.20:50
tewardoops already put in chagnelog from Studio team but oops20:50
tewarddoesn't matter as long as it's fixed20:50
tewardeither way, MATE team requested ^^20:50
tewardnot vgoing ot reup -10ubuntu5 for the changelog20:50
tewardbut either way, was requested, was uploaded20:50
tewardi'm going to find more coffee20:50
tewardarraybolt3: you owe me $2020:50
arraybolt3I just sent you more coffee!20:51
arraybolt3vorlon: Hate to bother you about seeds again, but Lubuntu's seed still hasn't updated on ubuntu-archive-team.ubuntu.com and it's been way more than 20 minutes since I pushed things.20:54
vorlonarraybolt3: I have to step out for a half hour, but I'll look at it when I get back.  I'm now realizing that even hitting archive-team.internal gets redirected to ubuntu-archive-team.ubuntu.com, which has a different, rsync delay21:02
-queuebot:#ubuntu-release- New: accepted gcr4 [amd64] (lunar-proposed) [4.1.0-0ubuntu1]21:06
-queuebot:#ubuntu-release- New: accepted gcr4 [armhf] (lunar-proposed) [4.1.0-0ubuntu1]21:06
-queuebot:#ubuntu-release- New: accepted gcr4 [riscv64] (lunar-proposed) [4.1.0-0ubuntu1]21:06
-queuebot:#ubuntu-release- New: accepted gcr4 [arm64] (lunar-proposed) [4.1.0-0ubuntu1]21:06
-queuebot:#ubuntu-release- New: accepted gcr4 [s390x] (lunar-proposed) [4.1.0-0ubuntu1]21:06
-queuebot:#ubuntu-release- New: accepted gcr4 [ppc64el] (lunar-proposed) [4.1.0-0ubuntu1]21:06
utkarsh2102vorlon: hi! I have few questions - mostly just to make sure I understand everything clearly :)21:40
vorlonutkarsh2102: hi21:41
vorlonarraybolt3: https://ubuntu-archive-team.ubuntu.com/seeds/lubuntu.lunar/desktop is up-to-date right now; did it fix itself?21:41
utkarsh2102first, every seed in ubuntu-seeds is germniated via the ./update script in ubuntu-meta, no? and then the output is uploaded to https://ubuntu-archive-team.ubuntu.com/germinate-output/ubuntu.lunar/21:41
utkarsh2102is that correct?21:41
vorlonutkarsh2102: ubuntu-archive-team.ubuntu.com is a front-end that pulls from ubuntu-archive-toolbox, which runs germinate.  This is not done using the update script, which is specific to metapackages21:42
utkarsh2102aaaaaaah, got it! noting it down21:43
arraybolt3vorlon: Looks like it, thanks you!21:43
* utkarsh2102 is looking for what and where is ubuntu-archive-toolbox :)21:46
vorlonarraybolt3: ok.  Seems like it took longer than it ought to, but the cycle there is that we run update-seeds every 5 minutes on ubuntu-archive-toolbox, but ubuntu-archive-team.ubuntu.com currently only pulls from that machine to the web frontend hourly.  (There's an open RT to make that a push trigger instead.)21:46
vorlonutkarsh2102: the replacement for snakefruit.  Privileged machine, you don't have access21:46
utkarsh2102now that makes sense; I remember your RT around that21:47
vorlonwe run a bunch of stuff on there that *isn't* privileged, should be split out, and should be wrapped in juju+mojo.  But it isn't yet.21:47
* utkarsh2102 nods21:48
utkarsh2102if only I was given some spare cycles :)21:48
utkarsh2102out of curiosity, is germinate on ububtu-archive-toolbox different from the germinate call (germinate-update-metapackage) in ubuntu-meta?21:49
utkarsh2102and how so?21:49
vorlonarraybolt3: so anyway, in the worst case it might take around ~2h5m before a change is reflected on the frontend21:50
vorlonif everything's working as expected right now21:50
vorlonwait that's not right. 1h5m21:50
arraybolt3Makes sense.21:51
vorlonutkarsh2102: lp:ubuntu-archive-scripts, update-germinate. $HOME/germinate in turn is a checkout of lp:germinate21:52
utkarsh2102I see, so it's pure germinate21:55
utkarsh2102and the ubuntu-meta one works only for metapackages21:55
utkarsh2102vorlon: second, not every seed in ubuntu-seeds in turned into a metapackage. Even though the ./update script (germinate-update-metapackage) is called and they're germinated, but only the ones mentioned in d/control gets created, no?22:00
utkarsh2102that is, the usual packaging process :)22:01
utkarsh2102so the ./update script in ubuntu-meta only germinates the seeds that we want the meta-packages for (that is, by mentioning it somewhere? d/control? idk :/)22:03
vorlonutkarsh2102: debian/control is itself generated22:03
vorlonthe seed pods contain the information about which seeds should be turned into metapackages22:04
utkarsh2102oh wow! TIL!22:05
utkarsh2102where's this "seed pod"?22:05
vorlonutkarsh2102: that's the branch that contains the seeds22:05
vorlon(not a frequently used name but we have no other name for it)22:05
utkarsh2102https://git.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/+git/ubuntu/tree/?h=lunar I suppose?22:07
utkarsh2102where exactly do we store that information (to turn seeds to metapackages)22:07
vorloncjwatson: huh do you remember why you changed run-proposed-migration (...10 years ago) to checkout lp:auto-package-testing? I don't find anything using this checkout22:08
vorloncjwatson: includes references to jenkins :)22:09
vorlonutkarsh2102: maybe I'm a liar about debian/control being autogenerated hmm22:12
utkarsh2102yes, I was reallllyyyyy surprised about that!22:12
-queuebot:#ubuntu-release- New binary: ubuntustudio-look [amd64] (lunar-proposed/universe) [23.04.1] (ubuntustudio)22:12
EickmeyerHey look! Wallpapers! ^ :)22:13
Eickmeyer(pun slightly intended)22:13
vorlonutkarsh2102: yeah looks like we have to manually update debian/control and update.cfg22:14
utkarsh2102so we update the two things there, run the update script and there, we have the new meta-package, correct?22:15
utkarsh2102is there a git repository for ubuntu-meta where you push to?22:17
utkarsh2102or is it just something you have locally?22:17
utkarsh2102because https://git.launchpad.net/ubuntu/+source/ubuntu-meta/log/?h=ubuntu/lunar is just imports of the uploads22:17
utkarsh2102I'd like to see git log if that's available somewhere22:17
vorlonit's not pushed. the uploads are the history22:19
vorlonthere's not going to be any complicated history there.  upload-level granularity is pretty complete22:20
utkarsh2102noted, thanks! \o/22:21
utkarsh2102vorlon: third, I see a lot of seeds referencing the meta-packages themselves. How does that really work? I mean, I am super curious to understand what's happening behind the scenes?22:24
utkarsh2102how can the seed depend on the metapackage when there isn't one yet :)22:25
utkarsh2102for eg: https://git.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/+git/ubuntu/tree/desktop-minimal#n17622:25
utkarsh2102 * ubuntu-desktop-minimal # metapackage for everything here22:25
vorlonutkarsh2102: everything that uses germinate ignores missing packages22:26
utkarsh2102then why do we even mention it there? what's the purpose of that? for the subsequent runs?22:27
vorlonyou'll find plenty of packages in the seeds that no longer exist; nothing enforces their presence at build time and we have no process for auditing the seeds to remove them22:27
vorlonutkarsh2102: it's used to set the task header on the metapackage itself22:27
cjwatsonvorlon: I very much do not, sorry!22:27
vorlonit would be silly if livecd-rootfs used task_install to install the ubuntu-desktop task and this failed to include the metapackage22:28
vorloncjwatson: ok then I'll try deleting it!22:28
vorloncjwatson: last commit was by pitti in 201522:29
vorlonutkarsh2102: (otoh there's growing consensus that it's silly to use the task field at all for installing packages in livecd-rootfs and we are angling to stop doing that next cycle)22:29
cjwatsonvorlon: I think that was probably the first attempt to integrate autopkgtest into p-m22:30
cjwatsonAnd I'm not sure I properly understood it at the time either22:31
utkarsh2102vorlon: so IOW, having meta-packages mentioned in the seeds themselves is going to be moot from next cycle onward?22:32
vorlonutkarsh2102: that's my expectation yes22:32
utkarsh2102okeydoke, thank you!22:33
utkarsh2102vorlon: next, are you OK with calling the common seed as server-cloud-minimal-common? :)22:33
vorlonoh god22:33
utkarsh2102(you know the history behind that request :))22:33
utkarsh2102or I'll let it be, when you review the spec, please add a comment around that :)22:34
vorlonutkarsh2102: my understanding is that the 'common' there refers to the fact that it's common to both baremetal server and cloud images.  Let it be either server-minimal-common or server-cloud-minimal, not server-cloud-hyphen-minimal-common-ampersand22:36
utkarsh2102noted, I'll bring this up within CPC. Hopefully we can agree on server-cloud-minimal.22:38
utkarsh2102vorlon: lastly, and also very stupid, some people suggested adding comments in the Task-* fields; for eg: `Task-Seeds: server-cloud-minimal  # this is a comment`. Do you know, at the top of your head, if those comments are allowed like that in the Tasks-* fields?22:47
vorlonutkarsh2102: not a clue22:51
utkarsh2102hehe, danke22:51
utkarsh2102vorlon: this is the last, for now :P23:04
utkarsh2102if my seed has23:04
utkarsh2102 Task-Seeds: server-cloud-minimal23:04
utkarsh2102and   * feature: no-follow-recommends23:04
utkarsh2102will it at least have the Recommends which are explicitly mentioned in the Task-Seeds seed?23:05
vorlonI do not know23:05
utkarsh2102a hunch?23:05
vorlonutkarsh2102: useless for me to guess in this context23:06
vorlonrun it and see23:06
utkarsh2102okeydoke, noted. Thank youuuu!23:07
utkarsh2102for answering all the questions :)23:07
vorlonbdmurray: so why do we have a checkout of lp:ubuntu-bugpatterns under public_html https://ubuntu-archive-team.ubuntu.com/bugpatterns23:12
vorlonmake that https://ubuntu-archive-team.ubuntu.com/bugpatterns/23:12
vorlonbdmurray: I guess the answer on this is bugpatterns.xml?23:28
tillkamppeter@vorlon, I have uploaded a cups-browsed package (cups-browsed SOURCE package should be still in Universe) and get this rejection: https://paste.ubuntu.com/p/kjb9q5dfFF/ The version before I could upload without problem. MIR is bug 200325923:37
tillkamppeterIt says "Signer is not permitted to upload to the component 'main'.".23:38
vorlontillkamppeter: do you have motu rights or PPU rights?23:38
tillkamppeterI have MOTU rights, so I can upload arbitrary packages into Universe and I have PPU rights to upload printing/scanning-related packages into Main.23:39
vorlontillkamppeter: then perhaps this is a package that's not in your PPU set?23:40
vorlontillkamppeter: confirmed that it doesn't show up for you in https://ubuntu-archive-team.ubuntu.com/archive-permissions/individuals23:41
tillkamppeter@vorlon, the cups-browsed SOURCE package should still be in Universe (MIR is not yet "Fix Released"), so I should have upload rights via MOTU, I could upload the earlier versions of the very same source package (2.0~b2-0ubuntu4).23:41
vorlontillkamppeter: I see cups-browsed having been promoted to main in -proposed23:42
tillkamppeter@vorlon, source package or binary package?23:43
vorlontillkamppeter: possibly because it was a source split and therefore a source-only promotion so I promoted it without worrying about an MIR23:43
vorlonit quite clearly should be added to your ppu set but I have to remember how to do that23:43
tillkamppeterOK, Meaning that cups-browsed is fully in Main now. Who can update my PPU list to reflect this source split (add libcupsfilters, libppd, and cups-browsed)?23:44
vorlonI *can* update it but I'm not sure if the DMB cares about being consulted before I d23:44
tillkamppeter@vorlon and once on it, pappl, pappl-retrofit, cpdb-libs, cpdb-backend-cups, and cpdb-backend-file need to get added to my list, too.23:45
vorlonutkarsh2102, kanashiro[m], rbasak, teward, bdmurray ^^ do you need any sort of process before I add packages to tillkamppeter's PPU acl that have been split out of source packages he already had PPU on?23:46
vorlontillkamppeter: I'm definitely not adding packages to the PPU list that *aren't* a result of a source split, without DMB approval23:46
utkarsh2102vorlon: I don’t think so. A mail to devel-permissions would be good I think.23:47
tillkamppeterOK, so for the pappl* and cpdb* we wait for the answer of DMB then ...23:47
utkarsh2102tillkamppeter: you have that :)23:47
utkarsh2102but feel free to just do that and send an fyi on the list.23:47
tillkamppeter@utkarsh2102, what do you mean now?23:48
utkarsh2102I meant, vorlon can add those packages to your PPU ACL and then send an FYI on the devel-permissions mailing list.23:49
utkarsh2102or you could request it there and someone can do it and reply. Either works.23:49
tillkamppeterOK, @vorlon tell me when you have done it.23:49
vorlontillkamppeter: AIUI libppd was not split out of an existing source package you had ppu on but is a replacement for the old libppd source.  I think that one needs to go through the DMB process23:49
utkarsh2102But I think the former is faster.23:49
vorlonI will add cups-browsed and libcupsfilters to your ACL and email the DMB23:50
tillkamppeter@vorlon, libppd is split off from CUPS. It contains all PPD support functionality of CUPS.23:50
vorlontillkamppeter: libppd> how am I to know this?23:52
vorlontillkamppeter: cups-browsed, libcupsfilters are added to your acl23:52
vorlontillkamppeter: for the others, please email the DMB and I can add when they've signed off23:52
tillkamppeter@vorlon I have written about this in the MIR, bug 200325923:53
tillkamppeter@vorlon,.cups-browsed got accepted now.23:55
tillkamppeter@vorlon, so I have to mail to devel-permissions@ubuntu.com for asking for the PPU permissions?23:56
teward*appears with DMB hat on*23:57
tewardi have no issue with Vorlon assigning rights to source packages broken out of source packages you *already* had PPU access on23:57
tewardbut if there's things that are now split that you *didn't* have upload rights to previously, then no, you have to go through the PPU application process again for the new stuff23:57
teward(case in point for me if I didn't have my coredev rights, nginx now has a dozen libnginxmod source packages incoming from Debian, they were part of src:nginx, but my PPU in theory could be extended to all the things that were now split out of it)23:58
teward(i had nginx PPU before I had coredev)23:58
tewardthis is my current understanding of our policies23:58
tewardso reading scrollback:23:59

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!