[00:12] cool [00:13] * cjwatson looks: I ported most of it in 2012, and a few more bits in 2013/2014 [00:14] In fact I think I did the shell-to-Python port with an eye to py3 compat in the first place, and just had to fix a couple of minor mistakes in that plan [00:28] nice [00:28] uploaded nlopt which fixes alternative python3 extension name. The default one was renamed correctly, but the alternative pythons were not. [02:03] vorlon: please ignore the missing sagemath/armhf build, not built anymore [02:09] vorlon: is the vim/i386 autopkg test supposed to fail? [04:52] why doesn't assimp show up in update_excuses? v5 is in proposed [05:06] force-synced odil to get it rebuilt with boost1.71. hope that's all to let boost-defaults migrate [05:40] doko: the vim/i386 autopkgtest has been passing, I don't know why it started to fail [05:42] doko: sagemath doesn't need ignored, it needs the binaries removed from focal-proposed on armhf and s390x (why they're in focal-proposed for a source version that's in the release pocket, I don't know) [05:42] doko: (sagemath removals done) [05:49] ta [07:17] doko: it looks like we're down to 4 packages (libreoffice, kopanocore, assimp, link-grammar) so I'm going to shut off autosync for the moment just in case [07:18] vorlon: ugh, lo and odil already picked up icu, and I now uploaded to first batch of icu rebuilds [07:19] I don't see odil tied to the python3-defaults transition [07:20] libreoffice, we could roll back to the previous version and force breaking of the libreoffice-gtk stuff [07:20] (but you'd need someone else on the release team to do it, I'm not going to be up late enough to) [07:21] I definitely don't think we should wait for icu [07:22] odil is tight to ros [07:22] ok, would a rollback be sufficient there also? [07:23] I don't see a ros transition, only an assimp one [07:23] let me look, it should up blocking boost-dedaults [07:24] doko: so I'll do the libreoffice rollback and then maybe Laney can force break libreoffice-gtk during your day [07:26] vorlon: apparmor/ppc64el needs to be addressed or ignored, triggered by kopanocore [07:27] and kopanocore fails on armhf [07:28] and arm64 [07:28] I love failing tests without any output [07:33] vorlon: if you have time today, could you add matplotlib to i386 blacklist please? [08:01] vorlon, I can do link-grammar and kopanocore [08:03] link-grammar is uploaded [08:04] why does this need to stay like this for 5 days? [08:05] doko, I was busy with some other 20 entangled packages. I spent yesterday fixing something else... [08:05] btw with update excuses showing ton of stuff, I didn't notice they were broken! [08:05] I mean, they were blocking [08:06] (I still don't get what regressed kopanocore and now magically fixed it btw) [08:53] -queuebot:#ubuntu-release- Builds: 32 entries have been added, updated or disabled [09:03] * Laney peers in [09:05] what does force break mean in this context? === andrewc is now known as Guest70315 [09:19] apparmor ppc64el test should autoheal in a publisher run apparmor/focal/ppc64el because of evice just built on ppc64el [09:20] vorlon, do you think libextutils-pkgconfig-perl [focal/i386] is elibigle for an hint? [09:20] I see you also tried it... [09:53] Laney: it's build with the missing libreoffice-gtk2 binary package [09:57] ah [09:57] if we had a newer britney there's an allow-uninst hint for this :( [09:59] Laney, sigh [10:02] doko: ok, so when libreoffice-gtk is the only thing listed we can do that I guess [10:02] did someone tell the LO people already? [10:12] they are aware, see #u-d [10:12] also gnucash 1:3.8b-1build3 needs to be removed, restored by -1build2 [10:13] oh ok that's what ubuntu5 was fixing [10:14] yeah, you can do that can't you? [10:14] * Laney can't [10:14] I can remove -build3, will -build2 still be available? [10:14] copy it back, I think that should work [10:15] not sure if you have to wait for the removal to publish first [10:15] I'll wait [10:16] they are separate publishing records, so i would think there is no need to wait [10:16] yeah I'd try it anyway, worst case it doesn't work, best case you save a cycle [10:17] * doko looks up the syntax again for the restore [10:21] copied back [10:21] zsh: /home/laney/bin/ubuntu-archive-tools/copy-package: bad interpreter: /usr/bin/python: no such file or directory [10:21] doh [10:24] so it looks like there's sagemath (ipython -> dask failures), ros-rviz (assimp -> assimp failures) apart from the known ones? [10:30] dask tests are still running, the scikit-learn in -proposed is removed [10:31] assimp is a real issue [10:32] is the sync needed for the transition or can it be removed? [10:34] looks like it is [10:35] or could it be replaced by a rebuild of the version in focal release? [10:36] games-python2-dev still depends on python-pyassimp [10:37] and we have lot of dependencies on the new assimp in -proposed [10:38] mmmmmmmmmmmeh [10:38] so somebody needs to have a look ... [10:38] ok I'll hint it, Debian report would be appreciated [10:39] "it's possible your package is broken on many arches" [10:39] pyassimp.errors.AssimpError: Could not import file! [10:39] trying: /usr/share/assimp/models/invalid/empty.smd [10:39] Error encountered while loading '/usr/share/assimp/models/invalid/empty.smd' [10:40] why are the trying to load invalid files? [10:43] Laney, ipython is fixed [10:43] good [10:43] I mean, with doko fixing scikit sadness [10:44] doko, TBH, I spent something like two weeks on assimp, reporting reopening reporting reopening debian rc bugs [10:44] all the tests that fails are new tests, at some point we can just ignore [10:44] https://ci.debian.net/data/autopkgtest/unstable/ppc64el/a/assimp/4032912/log.gz [10:44] LocutusOfBorg: please can you write that in a bug report, and then point the release team to it? [10:44] pyassimp.errors.AssimpError: Could not import file! [10:44] ** Loaded 264 models, got 54 assimp errors and 4 other errors [10:44] Error:Invalid token "�" expected "{" [10:45] command2 PASS [10:45] afk now, errand [10:45] meh [10:45] doko, http://bugs.debian.org/944742 [10:45] Debian bug 944742 in src:assimp "assimp: testsuite regression" [Serious,Fixed] [10:45] I did that [10:45] many, many times [10:46] also Debian for some reasons have the same failures with "pass" result [10:51] I did hint it, this seems like not the best use of time to me [11:22] sil2100, could you please review the ec2-instance-connect srus in the unapproved queues? [11:58] kopanocore green [11:59] so with gnucash publish and libreoffice builds... migration=? [12:43] new britney run in 5 minutes === andrewc is now known as Guest43475 [14:35] let's see now [14:56] rbalint: looking o/ [15:20] * doko has 220 [ubuntu/focal] messsages in the inbox \o/ [15:26] interesting [15:26] don't really undetstand how that worked [15:27] it looked like it failed due to libreoffice-gtk as expected, but then they were all accepted anyway [15:27] but you overrode that? [15:27] no that happened before my hint [15:27] I didn't notice that it worked [15:28] if you look for "final:" in update_output.txt [15:29] it accepted everything on the basis of an "easy" hint from the auto hinter that looks like it failed [15:29] * Laney is confused [15:29] why did I get these emails? [15:30] because it did work [15:30] I just don't know why! [15:30] perhaps uninstallability trading? [16:04] marcustomlinson: I don't think libreoffice 1:6.4.0-0ubuntu6 was needed, 0ubuntu5 had already been built against the new icu and I was going to bring that version back into -proposed now that python3-defaults is through [16:05] vorlon: not on ppc64el [16:05] ah ok [16:05] marcustomlinson: ^^ then nevermind, all good [16:05] :|) [16:10] autosyncs reenabled, thanks Laney doko et al for getting that transition through [16:10] np [16:10] still unsure how it worked in the end [16:10] but maybe I should let that be [16:19] Laney: ah did you not add a hint? :-) with so many packages migrating and a non-zero number of prior uninstallables, it's possible britney made a trade [16:22] vorlon: Nope. Trading's the only thing I can think of, but not sure what the situation was like before [16:22] 🤷 [16:23] it was probably a trade that fixed one of the packages doko had made uninstallable recently with removals :) [16:23] heh [16:23] someone needs to set infinity loose [16:24] * doko doesn't feel guilty about debian-med removals [16:24] $ python [16:24] Command 'python' not found, but can be installed with: [16:24] sudo apt install python3 # version 3.7.5-1ubuntu1, or [16:24] sudo apt install python # version 2.7.17-1 [16:24] sudo apt install python-minimal # version 2.7.17-1 [16:24] You also have python3 installed, you can run 'python3' instead. [16:24] to be fixed too ... [16:37] xnox: reminder that rdkit is on you to fix [16:38] xnox: also cegui-mk2 IIRC [16:38] yeap. [16:39] i ignored them until after python3-defaults migrates, which it did now [16:39] taking them [17:10] sil2100, how do you like it? :-) [17:25] rbalint: uh, my terminal window got lost somewhere and I forgot about it! Sorry, too much going on again ;) [17:25] Finding it now [17:41] -queuebot:#ubuntu-release- New binary: ibus [amd64] (focal-proposed/main) [1.5.21-5ubuntu1] (desktop-core, i386-whitelist, input-methods) [17:44] -queuebot:#ubuntu-release- New binary: blis [s390x] (focal-proposed/universe) [0.6.1-2] (no packageset) [17:48] -queuebot:#ubuntu-release- New binary: blis [ppc64el] (focal-proposed/universe) [0.6.1-2] (no packageset) [17:49] -queuebot:#ubuntu-release- New binary: blis [armhf] (focal-proposed/universe) [0.6.1-2] (no packageset) [17:56] -queuebot:#ubuntu-release- New binary: blis [amd64] (focal-proposed/universe) [0.6.1-2] (no packageset) [17:56] vorlon: did you complete the icu uploads? i don't think i see any for virtual provides/depends [17:56] otherwise i'll do them once i'm back [18:18] -queuebot:#ubuntu-release- New binary: blis [arm64] (focal-proposed/universe) [0.6.1-2] (no packageset) [18:22] -queuebot:#ubuntu-release- New: accepted blis [amd64] (focal-proposed) [0.6.1-2] [18:22] -queuebot:#ubuntu-release- New: accepted blis [armhf] (focal-proposed) [0.6.1-2] [18:22] -queuebot:#ubuntu-release- New: accepted blis [s390x] (focal-proposed) [0.6.1-2] [18:22] -queuebot:#ubuntu-release- New: accepted blis [arm64] (focal-proposed) [0.6.1-2] [18:22] -queuebot:#ubuntu-release- New: accepted blis [ppc64el] (focal-proposed) [0.6.1-2] [18:25] -queuebot:#ubuntu-release- New: accepted ibus [amd64] (focal-proposed) [1.5.21-5ubuntu1] [18:30] xnox: I ^C'ed because it was wrongly trying to pick up libreoffice, I'm continuing now that there's been a publisher cycle so that I'm not re-uploading all the things I already uploaded [18:30] sil2100: I've update the meta-release files for 18.04.4 [18:31] bdmurray: thank you o/ [18:37] doko: why did you remove the gnucash that you uploaded for icu? [18:58] how often does the rdepends service update nowadays? [19:03] twice a day: 7 9,21 * * * [19:05] xnox: bombono-dvd> lol failed rebuild against boost because of python3-incompatible scons, enjoy [19:07] sil2100, vorlon could you please merge this for systemd? https://code.launchpad.net/~rbalint/britney/hints-ubuntu/+merge/379168 [19:11] rbalint: done [19:24] tumbleweed: thanks! [19:28] can I sync ocaml? [19:28] doko, your delta seems to be not applicable anymore... :/ [19:34] after a ton of bothering to remove hint htslib/s390x, I got an upstream fix, and uploaded it back :/ [19:39] ocaml syncd, lets see what happens [19:40] locutus_: that wasn't very long to wait for objections [19:40] we're in the middle of rebuilds for the icu transition, might've been nice to let that finish first [20:01] actually I asked months ago for ocaml... [20:01] and pinged here from time to time, and FFe is approaching... [20:01] in any case I don't want to issue rebuilds today or tomorrow [20:01] for sure not before icu [20:01] yes, but the icu rebuilds are still /in progress/, so ocaml revdeps can get entangled [20:02] anyway, we'll deal with that if it happen [20:02] s [20:02] I wasn't aware of entangling between icu and ocaml... [20:03] if you look e.g. here https://release.debian.org/transitions/html/ocaml.html [20:03] there is no icu mention [20:03] -queuebot:#ubuntu-release- New binary: ocaml [amd64] (focal-proposed/universe) [4.08.1-8] (i386-whitelist, kubuntu) [20:03] I wouldn't expect 'icu' to be in the names of the revdeps there [20:03] anyway it's possible there's no entanglement [20:04] btw I'm already doing the haskell icu part... [20:04] something you did ftbfs [20:05] well, those were no-change rebuild [20:05] s [20:22] btw please libextutils-pkgconfig-perl/i386 hint if possible? [20:34] locutus_: done [20:50] ta [20:58] Laney, juliank: fwiw I was trying to launch an adt vm on lgw01 to debug vim/i386 and was hitting memory quota; nova list showed 6 adt-prepare VMs running for 12days+ which I've now nuked [21:01] Thanks vorlon [21:02] Laney, juliank: also, the thing I'm debugging is that vim/i386 autopkgtest failed, then started to pass, now fails again due to wanting to install vim-tiny:i386 and remove vim-tiny:amd64 + ubuntu-minimal. do you know of anything that would've changed on autopkgtest infra to explain why apt would flip-flop on this? [21:03] No idea sorry [21:32] rbalint: ok, I got completely side-tracked again [21:32] vorlon: hey, so, iptables 1.8.4-2 is... let's just say problematic [21:33] vorlon: you should see autopkgtests for ufw fail [21:33] rbalint: seeing that vorlon did the previous review, I would anyway opt for you to reach out to him and see if he's okay with the new version now [21:33] rbalint: if not, I can take a look at it on Monday [21:34] vorlon: I tested the patch added to 1.8.4-3 this morning and all ufw tests pass with it. Please merge or cherrypick https://git.netfilter.org/iptables/commit/?id=8e76391096f12212985c401ee83a67990aa27a29 [21:38] vorlon: actually, that is only with the nft backend and the ufw autopkgtests don't try both nft and legacy, so the upload will probably make it. but as soon as someone uses iptables-restore-nft, they are in a world of hurt with 1.8.4-2 [21:39] * jdstrand makes a note to adjust the autopkgtests to do both [21:40] vorlon: though, 1.8.4-1 drops the /sbin -> /usr/sbin compat symlinks (on fresh installs iirc) so ufw will die there as well [21:41] vorlon: I have a ufw upload read for that, but was waiting for all the upstream fixes to flow into iptables. that happened this morning with 1.8.4-3 (also, check your email if you want to help a maintainer out ;) [21:50] vorlon: if you need me to do the iptables merge with 1.8.4-3, let me know [22:14] jdstrand: please go ahead and do the iptables 1.8.4-3 merge, since you know what's up [22:20] vorlon: could you please action LP: #1757815 [22:20] Launchpad bug 1757815 in qt-assistant-compat (Ubuntu) "Please remove from focal archive" [Undecided,New] https://launchpad.net/bugs/1757815 [22:21] thansk [22:21] *thanks [22:23] also #1757816 #1757600 [22:23] LP: #1757816 [22:24] Launchpad bug 1757816 in qarecord (Ubuntu) "Please remove from focal archive" [Undecided,New] https://launchpad.net/bugs/1757816 [22:24] LP: #1757600 [22:24] Launchpad bug 1757600 in autopilot-qt (Ubuntu) "RM: Please remove from focal (Qt4 removal)" [Undecided,Triaged] https://launchpad.net/bugs/1757600 [22:37] "'Dead upstream for ~10 years' but the last upload was in karmic, karmic wasn't 10 years ag.... oh." [22:38] vorlon: called 'flogging a dead horse' I think [22:39] RikMills: all done [22:42] vorlon: thanks! I will go though more qt4 things left tomorrow if no-one else does. I guess if we get down to just a few that 'might' get an upstream solution on next 2 months, we could bump those to proposed? [22:43] they should be removed from release pocket and from -proposed, and if an upstream solution becomes available they can be reuploaded [22:43] (or resynced) [22:43] even better [23:20] vorlon: hey, i think i have done waf conversion to python3 before. Unpacked/repacked waf (btw still my favourite build system) and then they had custom plugins i'm like "if 2to3 -w works, i'm uploading it" [23:22] -queuebot:#ubuntu-release- New binary: golang-ginkgo [s390x] (focal-proposed/universe) [1.12.0-7] (no packageset) [23:24] -queuebot:#ubuntu-release- New binary: golang-ginkgo [ppc64el] (focal-proposed/universe) [1.12.0-7] (no packageset) [23:25] -queuebot:#ubuntu-release- New binary: efl [i386] (focal-proposed/universe) [1.23.3-7] (i386-whitelist, ubuntustudio) [23:25] -queuebot:#ubuntu-release- New binary: golang-ginkgo [amd64] (focal-proposed/universe) [1.12.0-7] (no packageset) [23:26] -queuebot:#ubuntu-release- New binary: efl [s390x] (focal-proposed/universe) [1.23.3-7] (i386-whitelist, ubuntustudio) [23:28] -queuebot:#ubuntu-release- New binary: golang-ginkgo [armhf] (focal-proposed/universe) [1.12.0-7] (no packageset) [23:29] -queuebot:#ubuntu-release- New binary: golang-ginkgo [arm64] (focal-proposed/universe) [1.12.0-7] (no packageset) [23:35] -queuebot:#ubuntu-release- New binary: python-cassandra-driver [amd64] (focal-proposed/universe) [3.20.2-1] (no packageset) [23:47] vorlon: i hope you enjoy my justification on https://bugs.launchpad.net/ubuntu/+source/bombono-dvd/+bug/1863185 [23:47] Ubuntu bug 1863185 in bombono-dvd (Ubuntu) "RM RoM bombono-dvd" [High,Triaged] [23:49] -queuebot:#ubuntu-release- New binary: efl [ppc64el] (focal-proposed/universe) [1.23.3-7] (i386-whitelist, ubuntustudio)