[00:36] valorie: hey! Do you remember if you saw LP: #1752453 in previous point-release testing? [00:36] Launchpad bug 1752453 in ubiquity (Ubuntu) "on a test of xenial .4 release, selecting "connect to wireless" freezes the process" [Undecided,New] https://launchpad.net/bugs/1752453 [00:36] valorie: is it reproducible? [00:36] over and over [00:36] however, I've given up because I get no logs of the failures [00:37] on to 64 bit [00:37] I don't recall ever seeing it before [00:37] that said, I've not tested point releases much [00:37] nor do I recall the choice of "connect to wireless" being available [00:38] once ubiquity freezes, I can't drop to the terminal to get logs [00:39] overwriting that daily right now, so no more from me [00:39] :( [00:39] -queuebot:#ubuntu-release- New source: lxc-templates (bionic-proposed/primary) [3.0.0~beta1-0ubuntu1] [00:43] -queuebot:#ubuntu-release- New binary: libkf5pimcommon [amd64] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [00:43] -queuebot:#ubuntu-release- New binary: libkf5pimcommon [i386] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [00:44] sil2100: you can be sure I will test the 64 for the same bug [00:45] -queuebot:#ubuntu-release- New: accepted lxc-templates [source] (bionic-proposed) [3.0.0~beta1-0ubuntu1] [00:48] -queuebot:#ubuntu-release- New binary: libkf5pimcommon [arm64] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [00:49] -queuebot:#ubuntu-release- New binary: libkf5pimcommon [armhf] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [00:50] valorie: thanks [00:51] valorie: as said, on the VM it all works as intended - at best I'd like someone else with real hardware trying to reproduce the issue [00:52] which is what I'm using [01:01] -queuebot:#ubuntu-release- Builds: Ubuntu GNOME Desktop amd64 [Xenial 16.04.4] has been marked as ready [01:01] -queuebot:#ubuntu-release- Builds: Ubuntu GNOME Desktop i386 [Xenial 16.04.4] has been marked as ready [01:10] jbicha: thanks for re-testing! [01:11] slangasek: can you mark docker-libkv/0.2.1-1 as a badtest? http://autopkgtest.ubuntu.com/packages/d/docker-libkv/bionic/amd64 <- broken in release [01:11] final scheduled Ubuntu GNOME iso release, right? [01:12] jbicha: Nope, there's still .5 [01:12] oh ok, I was looking at http://old-releases.ubuntu.com/releases/trusty/ which is why I got confused [01:18] mwhudson: done [01:19] slangasek: thanks [01:19] slangasek: did you see my wibble about fpc above? [01:19] mwhudson: reading now [01:20] mwhudson: so the test is broken and needs fixing (possibly in the neutering sense)? [01:20] slangasek: well the test with glibc release segfaults, the test with glibc fails to compile, which apparently results in different enough log output to break the autopkgtest [01:21] slangasek: the reason for the failure to compile does suggest that fpc might be slightly incompatible with glibc 2.27 [01:21] but no other tests fail to compile because of this so... [01:24] * sil2100 EODs now [01:24] o/ [01:27] slangasek: i also have to run now ttyl [01:47] -queuebot:#ubuntu-release- Unapproved: snapcraft (xenial-proposed/universe) [2.39.2 => 2.39.3] (no packageset) [01:49] -queuebot:#ubuntu-release- Unapproved: accepted snapcraft [source] (xenial-proposed) [2.39.3] [02:01] -queuebot:#ubuntu-release- New source: python3-lxc (bionic-proposed/primary) [3.0.0~beta1-0ubuntu1] [02:05] -queuebot:#ubuntu-release- New: accepted python3-lxc [source] (bionic-proposed) [3.0.0~beta1-0ubuntu1] [03:10] -queuebot:#ubuntu-release- Builds: Lubuntu Desktop amd64 [Xenial 16.04.4] has been marked as ready [03:10] -queuebot:#ubuntu-release- Builds: Lubuntu Desktop i386 [Xenial 16.04.4] has been marked as ready [03:18] * tsimonq2 crosses fingers for no more respins... [04:11] -queuebot:#ubuntu-release- Builds: Kubuntu Desktop amd64 [Xenial 16.04.4] has been marked as ready [05:04] slangasek and/or infinity: I know the feature freeze is coming up. I am curious if this would count as a feature. I want to make a change to Kubuntu for editing default shortcuts for some KWin stuff. What that be blocked by the feature freeze? [05:12] infinity: ^ [05:54] -queuebot:#ubuntu-release- New binary: polymake [s390x] (bionic-proposed/universe) [3.2r2-3] (no packageset) [06:00] -queuebot:#ubuntu-release- New binary: polymake [amd64] (bionic-proposed/universe) [3.2r2-3] (no packageset) [06:38] -queuebot:#ubuntu-release- New binary: polymake [ppc64el] (bionic-proposed/universe) [3.2r2-3] (no packageset) [06:40] -queuebot:#ubuntu-release- New binary: polymake [i386] (bionic-proposed/universe) [3.2r2-3] (no packageset) [06:54] acheronuk: what's wrong with all the kde autopkgtests failing, again? [06:55] doing a trivial mesa upload always results in hand-helding all these racy/buggy tests [06:55] *holding [07:15] tjaalton: the PIM ones, only half of new PIM stack is built yet. issues should go away mostly when that is done. [07:24] ok then [07:26] they've really taken the unix philosophy to the limit.. [07:27] 50+ pkgs, seriously [07:28] acheronuk: shouldn't there be more strict dependencies if the tests fail every time the versions don't match? [07:32] it's usually not such an issue. but there are library version bumps here, so like always with those, things have to build against the new one to remain installable. [07:34] but yeah, PIM really ticks me off sometimes. used to be just several large source packages, then was split, and split again, and split again..... grrr [07:35] tjaalton: http://gpul.grupos.udc.es/ka-iron-hand_reports/applications_archive/17.12.2_bionic_retry_builds.pdf [07:36] ^^^ one giant flying spaghetti monster of build deps [07:36] think I've seen that before ;) [07:37] I've seen it too much! [07:38] ok, so I'll just wait to see it sort itself [07:40] tjaalton: hopefully if AAs are kind today and let the new binaries through as they appear for PIM, then later once all built most tests should re-run ok. [07:40] if any don't, I will try to sort [07:41] new binaries too? ok [07:42] tjaalton: just library bumps [07:42] but they still end up in the new queue [07:43] like libkf5pimcommon in there at the moment [07:45] PIM devs really went to town breaking their ABI on this release :( [07:59] -queuebot:#ubuntu-release- New binary: polymake [armhf] (bionic-proposed/universe) [3.2r2-3] (no packageset) [08:03] so it seems [08:33] -queuebot:#ubuntu-release- New: accepted libkf5pimcommon [amd64] (bionic-proposed) [4:17.12.2-0ubuntu1] [08:33] -queuebot:#ubuntu-release- New: accepted libkf5pimcommon [armhf] (bionic-proposed) [4:17.12.2-0ubuntu1] [08:33] -queuebot:#ubuntu-release- New: accepted libkf5pimcommon [arm64] (bionic-proposed) [4:17.12.2-0ubuntu1] [08:33] -queuebot:#ubuntu-release- New: accepted libkf5pimcommon [i386] (bionic-proposed) [4:17.12.2-0ubuntu1] [08:34] -queuebot:#ubuntu-release- Builds: Netboot amd64 [Xenial 16.04.4] has been marked as ready [08:34] -queuebot:#ubuntu-release- Builds: Netboot i386 [Xenial 16.04.4] has been marked as ready [08:40] hi. could we force badtest plasma-framework/5.43.0-0ubuntu1 against mesa for amd64 please? [08:41] that seems to consistently fail on ubuntu's tests, but passes fine here in a lxd autopkgtest run [08:44] -queuebot:#ubuntu-release- Builds: Ubuntu Kylin Desktop amd64 [Xenial 16.04.4] has been marked as ready [08:44] -queuebot:#ubuntu-release- Builds: Ubuntu Kylin Desktop i386 [Xenial 16.04.4] has been marked as ready [09:08] -queuebot:#ubuntu-release- New: accepted leatherman [amd64] (bionic-proposed) [1.4.0+dfsg-1] [09:08] -queuebot:#ubuntu-release- New: accepted leatherman [armhf] (bionic-proposed) [1.4.0+dfsg-1] [09:08] -queuebot:#ubuntu-release- New: accepted leatherman [ppc64el] (bionic-proposed) [1.4.0+dfsg-1] [09:08] -queuebot:#ubuntu-release- New: accepted leatherman [arm64] (bionic-proposed) [1.4.0+dfsg-1] [09:08] -queuebot:#ubuntu-release- New: accepted leatherman [s390x] (bionic-proposed) [1.4.0+dfsg-1] [09:08] -queuebot:#ubuntu-release- New: accepted leatherman [i386] (bionic-proposed) [1.4.0+dfsg-1] [09:16] -queuebot:#ubuntu-release- New: accepted polymake [amd64] (bionic-proposed) [3.2r2-2] [09:16] -queuebot:#ubuntu-release- New: accepted polymake [armhf] (bionic-proposed) [3.2r2-2] [09:16] -queuebot:#ubuntu-release- New: accepted polymake [ppc64el] (bionic-proposed) [3.2r2-2] [09:16] -queuebot:#ubuntu-release- New: accepted polymake [arm64] (bionic-proposed) [3.2r2-2] [09:16] -queuebot:#ubuntu-release- New: accepted polymake [s390x] (bionic-proposed) [3.2r2-2] [09:16] -queuebot:#ubuntu-release- New: accepted polymake [i386] (bionic-proposed) [3.2r2-2] [09:19] -queuebot:#ubuntu-release- New binary: polymake [arm64] (bionic-proposed/universe) [3.2r2-3] (no packageset) [09:20] -queuebot:#ubuntu-release- Builds: Ubuntu Desktop amd64 [Xenial 16.04.4] has been marked as ready [09:21] -queuebot:#ubuntu-release- Builds: Ubuntu Desktop i386 [Xenial 16.04.4] has been marked as ready [09:43] acheronuk: fails for me on my machine (qemu not lxd, that's more like what the autopkgtet machines are running for amd64) [09:44] 4 retries might have been a bit much [09:44] and tjaalton has a fair point, it's not very good to have a half broken stack be installable - I think you should try to find something to depend on that prevents things moving forward until everything is consisitent [09:45] it creates problems for other people trying to do work in the dev release as you can see here :( [09:47] Laney, weird request -> could you please, re-kick, bionic ubuntu-base, for all arches? I know it was built today, but it was 3 hours too early =) [09:48] i.e. http://cdimage.ubuntu.com/ubuntu-base/daily/pending/ these things [09:54] it's not installable. that's why tests are failing. if I made it less installable still, they would still fail [09:55] It is a problem we are thinking about. just solutions move the problem somewhere else. :/ [09:57] xnox: okey dokey [10:06] -queuebot:#ubuntu-release- New binary: libkf5calendarsupport [amd64] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [10:12] -queuebot:#ubuntu-release- New binary: uwsgi [amd64] (bionic-proposed/universe) [2.0.15-10.2ubuntu2] (no packageset) [10:12] -queuebot:#ubuntu-release- New binary: uwsgi [i386] (bionic-proposed/universe) [2.0.15-10.2ubuntu2] (no packageset) [10:14] -queuebot:#ubuntu-release- New binary: uwsgi [s390x] (bionic-proposed/universe) [2.0.15-10.2ubuntu2] (no packageset) [10:25] -queuebot:#ubuntu-release- New binary: libkf5calendarsupport [i386] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [10:27] -queuebot:#ubuntu-release- New binary: kf5-messagelib [amd64] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [10:33] -queuebot:#ubuntu-release- New binary: kf5-messagelib [i386] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [10:33] -queuebot:#ubuntu-release- New binary: uwsgi [ppc64el] (bionic-proposed/universe) [2.0.15-10.2ubuntu2] (no packageset) [10:37] -queuebot:#ubuntu-release- Unapproved: kbuild (xenial-proposed/universe) [1:0.1.9998svn2813+dfsg-1 => 1:0.1.9998svn2814+dfsg-2~ubuntu16.04.1] (no packageset) [10:38] -queuebot:#ubuntu-release- Unapproved: virtualbox (artful-proposed/multiverse) [5.1.30-dfsg-1 => 5.1.34-dfsg-0ubuntu1.17.10.1~17.10.1] (ubuntu-cloud) [10:39] -queuebot:#ubuntu-release- Unapproved: virtualbox-ext-pack (artful-proposed/multiverse) [5.1.30-2 => 5.1.34-0ubuntu1.17.10.1] (no packageset) [10:39] -queuebot:#ubuntu-release- Unapproved: virtualbox (xenial-proposed/multiverse) [5.0.40-dfsg-0ubuntu1.16.04.2 => 5.1.34-dfsg-0ubuntu1.16.04.1~16.04.1] (ubuntu-cloud) [10:39] -queuebot:#ubuntu-release- Unapproved: virtualbox-ext-pack (xenial-proposed/multiverse) [5.0.40-0ubuntu1.16.04.1 => 5.1.34-0ubuntu1.16.04.1] (no packageset) [10:41] -queuebot:#ubuntu-release- Unapproved: virtualbox-guest-additions-iso (xenial-proposed/multiverse) [5.0.40-0ubuntu1.16.04.1 => 5.1.34-0ubuntu1.16.04.1] (no packageset) [10:42] -queuebot:#ubuntu-release- Unapproved: virtualbox-hwe (xenial-proposed/multiverse) [5.0.40-dfsg-0ubuntu1.16.04.1~16.04.4 => 5.1.34-dfsg-0ubuntu1.16.04.1~16.04.1] (no packageset) [10:42] -queuebot:#ubuntu-release- Unapproved: virtualbox-guest-additions-iso (artful-proposed/multiverse) [5.1.30-1 => 5.1.34-0ubuntu1.17.10.1] (no packageset) [10:45] apw or another AA. would someone be kind enough to remove blogilo from bionic? [10:45] https://bugs.launchpad.net/ubuntu/+source/blogilo/+bug/1752544 [10:45] Ubuntu bug 1752544 in blogilo (Ubuntu) "Please remove blogilo from Bionic 18.04" [Undecided,New] [10:46] ^^^ that is also a test fail against mesa that I can't fix [10:47] acheronuk, having a look [10:47] ty [10:48] sil2100, the virtualbox* stack has been uploaded for xenial/artful, it is needed if you want to test the new xenial kernel 4.13... [10:48] I gave all my time to this, and I finally managed to fix it, so maybe accepting them would be nice [10:48] kbuild needs a minor update fox xenial, I have uploaded it too [10:48] -queuebot:#ubuntu-release- New binary: libkf5calendarsupport [arm64] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [10:50] -queuebot:#ubuntu-release- New binary: libkf5calendarsupport [armhf] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [10:50] LocutusOfBorg: hey! I'll look into those in a few minutes, thanks! [10:50] <3 [10:57] sil2100, I think we will need to close our eyes for vbox 5.0 -> 5.1 jump, and wait for community testing, It is unfortunate, but 5.0 is dead beyond repair (and even completely broken wrt new kernels), 5.2 can't build because of old qt, so 5.1 is the way to go, even if... it changed a lot [10:58] -queuebot:#ubuntu-release- New: accepted polymake [amd64] (bionic-proposed) [3.2r2-3] [10:58] -queuebot:#ubuntu-release- New: accepted polymake [armhf] (bionic-proposed) [3.2r2-3] [10:58] -queuebot:#ubuntu-release- New: accepted polymake [ppc64el] (bionic-proposed) [3.2r2-3] [10:58] -queuebot:#ubuntu-release- New: accepted polymake [arm64] (bionic-proposed) [3.2r2-3] [10:58] -queuebot:#ubuntu-release- New: accepted polymake [s390x] (bionic-proposed) [3.2r2-3] [10:58] -queuebot:#ubuntu-release- New: accepted polymake [i386] (bionic-proposed) [3.2r2-3] [11:00] -queuebot:#ubuntu-release- New: accepted libkf5calendarsupport [amd64] (bionic-proposed) [4:17.12.2-0ubuntu1] [11:00] -queuebot:#ubuntu-release- New: accepted libkf5calendarsupport [armhf] (bionic-proposed) [4:17.12.2-0ubuntu1] [11:00] -queuebot:#ubuntu-release- New: accepted libkf5calendarsupport [arm64] (bionic-proposed) [4:17.12.2-0ubuntu1] [11:00] -queuebot:#ubuntu-release- New: accepted libkf5calendarsupport [i386] (bionic-proposed) [4:17.12.2-0ubuntu1] [11:00] -queuebot:#ubuntu-release- New binary: uwsgi [armhf] (bionic-proposed/universe) [2.0.15-10.2ubuntu2] (no packageset) [11:01] -queuebot:#ubuntu-release- New binary: uwsgi [arm64] (bionic-proposed/universe) [2.0.15-10.2ubuntu2] (no packageset) [11:06] -queuebot:#ubuntu-release- New binary: kf5-messagelib [arm64] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [11:07] -queuebot:#ubuntu-release- New binary: kf5-messagelib [armhf] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [11:11] -queuebot:#ubuntu-release- New: accepted kf5-messagelib [amd64] (bionic-proposed) [4:17.12.2-0ubuntu1] [11:11] -queuebot:#ubuntu-release- New: accepted kf5-messagelib [armhf] (bionic-proposed) [4:17.12.2-0ubuntu1] [11:11] -queuebot:#ubuntu-release- New: accepted kf5-messagelib [arm64] (bionic-proposed) [4:17.12.2-0ubuntu1] [11:11] -queuebot:#ubuntu-release- New: accepted kf5-messagelib [i386] (bionic-proposed) [4:17.12.2-0ubuntu1] [11:13] -queuebot:#ubuntu-release- New: accepted bolt [source] (bionic-proposed) [0.1-0ubuntu1] [11:15] -queuebot:#ubuntu-release- New binary: bolt [s390x] (bionic-proposed/none) [0.1-0ubuntu1] (no packageset) [11:16] -queuebot:#ubuntu-release- New binary: bolt [armhf] (bionic-proposed/none) [0.1-0ubuntu1] (no packageset) [11:16] -queuebot:#ubuntu-release- New binary: bolt [ppc64el] (bionic-proposed/none) [0.1-0ubuntu1] (no packageset) [11:16] -queuebot:#ubuntu-release- New binary: bolt [i386] (bionic-proposed/none) [0.1-0ubuntu1] (no packageset) [11:17] -queuebot:#ubuntu-release- New binary: bolt [arm64] (bionic-proposed/none) [0.1-0ubuntu1] (no packageset) [11:24] seb128, bolt failed its tests on amd64 [11:25] (and therefore ftbfs there) [11:45] -queuebot:#ubuntu-release- New: accepted bolt [arm64] (bionic-proposed) [0.1-0ubuntu1] [11:45] -queuebot:#ubuntu-release- New: accepted bolt [i386] (bionic-proposed) [0.1-0ubuntu1] [11:45] -queuebot:#ubuntu-release- New: accepted bolt [s390x] (bionic-proposed) [0.1-0ubuntu1] [11:45] -queuebot:#ubuntu-release- New: accepted bolt [armhf] (bionic-proposed) [0.1-0ubuntu1] [11:45] -queuebot:#ubuntu-release- New: accepted bolt [ppc64el] (bionic-proposed) [0.1-0ubuntu1] [11:45] ^ that is lacking the amd64 bits .... which failed to build [11:57] -queuebot:#ubuntu-release- New binary: python-diskimage-builder [amd64] (bionic-proposed/universe) [2.11.0-0ubuntu1] (no packageset) [12:05] -queuebot:#ubuntu-release- New binary: bolt [amd64] (bionic-proposed/universe) [0.1-0ubuntu1] (no packageset) [12:09] -queuebot:#ubuntu-release- New: accepted uwsgi [amd64] (bionic-proposed) [2.0.15-10.2ubuntu1] [12:09] -queuebot:#ubuntu-release- New: accepted uwsgi [armhf] (bionic-proposed) [2.0.15-10.2ubuntu1] [12:09] -queuebot:#ubuntu-release- New: accepted uwsgi [ppc64el] (bionic-proposed) [2.0.15-10.2ubuntu1] [12:10] -queuebot:#ubuntu-release- New: accepted uwsgi [arm64] (bionic-proposed) [2.0.15-10.2ubuntu1] [12:10] -queuebot:#ubuntu-release- New: accepted uwsgi [s390x] (bionic-proposed) [2.0.15-10.2ubuntu1] [12:10] -queuebot:#ubuntu-release- New: accepted uwsgi [i386] (bionic-proposed) [2.0.15-10.2ubuntu1] [12:10] -queuebot:#ubuntu-release- New: accepted uwsgi [amd64] (bionic-proposed) [2.0.15-10.2ubuntu2] [12:10] -queuebot:#ubuntu-release- New: accepted uwsgi [armhf] (bionic-proposed) [2.0.15-10.2ubuntu2] [12:10] -queuebot:#ubuntu-release- New: accepted uwsgi [ppc64el] (bionic-proposed) [2.0.15-10.2ubuntu2] [12:10] -queuebot:#ubuntu-release- New: accepted uwsgi [arm64] (bionic-proposed) [2.0.15-10.2ubuntu2] [12:10] -queuebot:#ubuntu-release- New: accepted uwsgi [s390x] (bionic-proposed) [2.0.15-10.2ubuntu2] [12:10] -queuebot:#ubuntu-release- New: accepted uwsgi [i386] (bionic-proposed) [2.0.15-10.2ubuntu2] [12:20] -queuebot:#ubuntu-release- New binary: libkf5mailcommon [amd64] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [12:21] apw, thx [12:21] seb128, though i now see it has built, so i am assuming someone stabbed it in retry [12:22] apw, seems it worked now? [12:22] so i assume you have a nice racy test, yay for poo tests [12:22] apw, do you remember what was the failure? I don't have the log now [12:22] it was an assertion failure [12:22] but i didn't record what it was, something which was compared != "" which was "" [12:22] and it was in the first test if that helps [12:22] k, thanks [12:24] 1/3 test-common [12:24] seb128, ^ yeah it was definatly that test which cried [12:24] apw, ok, I'm going to do some loop testing here see if I hit it [12:26] -queuebot:#ubuntu-release- New binary: libkf5mailcommon [armhf] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [12:29] is anyone looking at unblocking the glibc transition from bionic-proposed? [12:32] -queuebot:#ubuntu-release- New binary: libkf5mailcommon [arm64] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [12:34] -queuebot:#ubuntu-release- New binary: libkf5mailcommon [i386] (bionic-proposed/universe) [4:17.12.2-0ubuntu1] (kubuntu) [12:38] so, we have some extra desktop transitions to land, unsure what to do [12:39] it would be annoying to pile them on the currently non sorted out ones [12:39] but if we wait for those to clear out we miss ff and then we have extra paperwork to do [12:40] what are the chances that glibc and libglvnd clears out of proposed this afternoon? [12:43] seb128: depends on the kde pim stuff to get sorted first.. [12:43] shrug [12:43] that's new ? [12:43] yes [12:43] :-( [12:43] *uck [12:44] we should lock down new transitions to be accepted when current ones are being sorted out [12:54] seb128: I was theone that retried bolt, log excerpt at https://paste.debian.net/1012554/ [12:56] could an AA accept binaries for libkf5mailcommon please [12:56] jbicha, thx [12:57] acheronuk: here's an idea; put the pim tarballs in one big source package which builds them all? :) [12:57] "ERROR: ld.so: object 'libumockdev-preload.so.0' from LD_PRELOAD cannot be [12:57] preloaded (cannot open shared object file): ignored." [12:57] weird error, doesn't seem to be due to the code tested itself then [13:00] tjaalton: you could call it something like — I dunno — kdepim ;) [13:00] lol [13:01] we should delete that new transition from proposed [13:01] clear out the ones that were ready to migrate [13:01] and add it back [13:11] Can someone sponsor an upload of the brlaser (printer driver) package? It seems not to be in my main upload package list (by the way, how can I see this list?). [13:12] Needs to be done before FF as it introduces a new upstream versions. [13:13] tkamppeter: did you file a LP bug and subscribe ubuntu-sponsors? I recommend doing that first and then ask in #ubuntu-devel [13:20] -queuebot:#ubuntu-release- New binary: libkf5incidenceeditor [i386] (bionic-proposed/universe) [17.12.2-0ubuntu1] (kubuntu) [13:21] -queuebot:#ubuntu-release- New: accepted libkf5mailcommon [amd64] (bionic-proposed) [4:17.12.2-0ubuntu1] [13:21] -queuebot:#ubuntu-release- New: accepted libkf5mailcommon [armhf] (bionic-proposed) [4:17.12.2-0ubuntu1] [13:21] -queuebot:#ubuntu-release- New: accepted libkf5mailcommon [arm64] (bionic-proposed) [4:17.12.2-0ubuntu1] [13:21] -queuebot:#ubuntu-release- New: accepted libkf5mailcommon [i386] (bionic-proposed) [4:17.12.2-0ubuntu1] [13:22] -queuebot:#ubuntu-release- New binary: libkf5incidenceeditor [amd64] (bionic-proposed/universe) [17.12.2-0ubuntu1] (kubuntu) [13:23] seb128, what all are these transitions you would like to start on the day of FF :) [13:23] -queuebot:#ubuntu-release- New binary: libkf5incidenceeditor [arm64] (bionic-proposed/universe) [17.12.2-0ubuntu1] (kubuntu) [13:23] apw, it's not "on the day of FF", we have just been waiting for the current ones to clear out for a while [13:24] apw, but mostly GNOME 3.28 ones [13:24] evolution-data-server then gnome-settings-daemon [13:24] but we wanted to migrate gnome-desktop3 first [13:24] which got stucked due to libglvnd [13:25] well logic would dictate with the system as it is you should just dump them in -proposed and make [13:25] the transitions a total mess, or we could somehow record that you [13:25] did want to do them beforehand and are only holding them to be nice a [13:25] and not make you do FFe's for everytyhing, because you are doing it to be nice [13:26] yeah, which is basically what I'm asking [13:26] well i would be down with that, but i would like someone else to agree [13:26] I would be fine with a written IRC agreement than we are fine to upload e-d-s and g-s-d after ff due to other transitions blocking us [13:26] k [13:26] would that be slangasek? [13:26] -queuebot:#ubuntu-release- New binary: libkf5incidenceeditor [armhf] (bionic-proposed/universe) [17.12.2-0ubuntu1] (kubuntu) [13:27] let's see if somebody else steps up or if Steve has an opinion when he wakes up [13:27] apw, thanks [13:27] slangasek or infinity sort of thing yeah, someone who has dones actaul release side of it [13:27] and can explain to me why this would be a bad plan should it be such a thing [13:27] seb128, did you see that pastebin, that has the actual error i saw indeed for bolt [13:28] apw, yeah, I did thanks, that has a weird umockdev warning, wonder if that has been an issue with it ... but I'm going to keep an eye to see if I can reproduce [13:29] it seems like an odd one for sure [13:29] -queuebot:#ubuntu-release- New: accepted libkf5incidenceeditor [amd64] (bionic-proposed) [17.12.2-0ubuntu1] [13:29] -queuebot:#ubuntu-release- New: accepted libkf5incidenceeditor [armhf] (bionic-proposed) [17.12.2-0ubuntu1] [13:29] -queuebot:#ubuntu-release- New: accepted libkf5incidenceeditor [arm64] (bionic-proposed) [17.12.2-0ubuntu1] [13:29] -queuebot:#ubuntu-release- New: accepted libkf5incidenceeditor [i386] (bionic-proposed) [17.12.2-0ubuntu1] [13:30] jbicha, thanks. I have done it now: bug 1752579 [13:30] bug 1752579 in brlaser (Ubuntu) "Needs sponsoring: Upload brlaser 4" [High,New] https://launchpad.net/bugs/1752579 [13:31] it sounds like this kde mess is close to done [13:32] jbicha, I hope the subscription to ubuntu-sponsors worked, as the subscribers to the bug do not get displayed currently. [13:32] apw: I think you're speaking a bit soon :( [13:32] apw: can you approve bolt/amd64 NEW too now? [13:32] jbicha, only repeating what i was told, accuracy may be lacking :) [13:34] -queuebot:#ubuntu-release- New: accepted bolt [amd64] (bionic-proposed) [0.1-0ubuntu1] [13:39] tkamppeter: to see you upload rights: bzr branch lp:ubuntu-archive-tools [13:39] cd ubuntu-archive-tools [13:39] ./edit-acl query -p till-kamppeter [13:41] sigh, there are several python modules that need a MIR to unblock pyasn1 et al [13:41] how much paperwork do these need? [13:43] at least pycryptodome and pysmi needed to unblock pysnmp4, which then should unblock pyasn1 et al [13:43] oh, already filed weeks ago [13:49] jbicha, thanks. [13:55] Laney: do you know if we're having general networking problems in adt on ppc64el? [13:56] -queuebot:#ubuntu-release- Unapproved: accepted friendly-recovery [source] (artful-proposed) [0.2.36ubuntu1] [13:56] -queuebot:#ubuntu-release- Builds: Ubuntu MATE Desktop amd64 [Xenial 16.04.4] has been marked as ready [13:56] -queuebot:#ubuntu-release- Builds: Ubuntu MATE Desktop i386 [Xenial 16.04.4] has been marked as ready [13:57] sil2100: ^ [13:57] -queuebot:#ubuntu-release- Unapproved: accepted friendly-recovery [source] (xenial-proposed) [0.2.31ubuntu1] [13:59] flexiondotorg: \o/ Thanks! [14:04] sil2100: Lubuntu got done last night ;) [14:11] stgraber: Hi! I proposed an MP to lp:queuebot a few weeks back, can I please get a review? [14:13] are we going to release 16.04.4 today? [14:14] tseliot: Hopefully, yes. [14:15] tsimonq2: ok, thanks [14:57] sil2100, netboot s390x is good, both GA & hwe kernels. [14:58] sil2100, i still want to test-boot LPAR and a VM, but these should be good, given that d-i booted. [14:59] xnox: \o. [14:59] Excellent [15:24] xnox: have you had the chance to test that issue with logind not letting go of fds? [15:38] valorie: hey! Could you give me a ping once you're up? [15:38] valorie: could you test .3 i386 kubuntu on your hardware? [15:44] tseliot, nope, sorry, been busy =/ [15:44] xnox: ok, np [15:50] sil2100, frank tested lpar/kvm. so s390x can be marked as ready, both d-i/netboot & server.iso [15:50] xnox: excellent, thanks o/ [15:55] seb128, apw: yes, I won't penalize for the transitions starting late due to -proposed being a jumble. help getting the current transitions through is of course always appreciated [15:55] slangasek, right, we have been doing that until the kde/qt piled up today [15:55] slangasek, but thanks for confirming [15:56] slangasek, apw, anyway bug #1752578 which is sort of ffe request if you want to +1 it already [15:56] bug 1752578 in evolution-ews (Ubuntu) "Evolution 3.27/3.28 transition" [Wishlist,New] https://launchpad.net/bugs/1752578 [15:59] seb128: what is the kde/qt pile-up today? [16:00] sil2100: if there is time today it looks like the cloud-init SRU only got half uploaded. rmadison sees 17.2.35 in artful-updates but not in xenial-updates [16:00] bug in question https://bugs.launchpad.net/ubuntu/+source/cloud-init/+bug/1747059 [16:00] Ubuntu bug 1747059 in cloud-init (Ubuntu Xenial) "sru cloud-init (17.1.46-g7acc9e86-0ubuntu1) update to 17.2-35-gf576b2a2-0ubuntu1~16.04.1" [Medium,Fix committed] [16:01] blackboxsw: we're generally not publishing things to xenial-updates now because of the point-release [16:01] But I can see if we can actually publish cloud-init or not [16:02] so mesa depends on https://launchpad.net/ubuntu/+source/mir/0.30.0.1-0ubuntu1 on s390x :( [16:02] ohhh interesting, ok i wasn't aware of that being an issue. post a point release does that open up again? [16:02] slangasek, https://launchpad.net/ubuntu/+source/libkf5pimcommon/4:17.12.2-0ubuntu1 [16:03] hmm [16:03] slangasek, that hit mesa and others [16:03] should we remove mir from proposed and then rebuild mesa and the 2 qt libraries that depend on mir too? [16:03] alan_g, ^ what's the deal with mir 0.30 failing to build and sitting in bionic-proposed? [16:03] slangasek: PIM stack taking much longer to build than was anticipated [16:06] acheronuk: next time, consider finishing the qt transition before starting a kde transition [16:07] acheronuk: Qt is blocked by lots of red autopkgtests and y'all said you needed Qt updated to let mesa through :( [16:07] jbicha, i love it -> built nowhere, but on s390x =) [16:08] acheronuk: qt 5.9.4 was released over a month ago, so it's not clear why it wasn't done earlier [16:11] ^^ tsimonq2 [16:12] Qt or mesa would not migrate even if all the tests passed [16:13] seb128, i've put some words in there ... [16:15] apw, thx [16:21] jbicha: Things got busy, plus Meltdown/Spectre. Why, is there a problem? [16:28] seb128, I knew nothing about it [16:29] Where's the failed build? [16:30] alan_g: https://launchpad.net/ubuntu/+source/mir/0.30.0.1-0ubuntu1 [16:34] jbicha: why would mesa need a rebuild for mir? there's nothing there requiring 0.30 [16:36] tjaalton: on s390x, libegl-mesa0 depends on libmirclient9 (>= 0.30.0.1) [16:36] it's mir's fault, not mesa's [16:36] ah [16:37] oh right, built on s390x [16:37] It doesn't help but libmirclient9 ABI is unchanged in 0.30, so (>= 0.30.0.1) isn't actually needed [16:38] I'd prefer it if we could fix mir rather than going round on mesa again [16:38] :) [16:38] yes [16:39] The failures I've checked look spurious (mostly tests that ran so slowly that timeouts kicked in before they finished). But there are so many this time around. [16:39] alan_g: you run dh_makeshlibs -V in mir, that causes this kind of dependency [16:39] If -V is specified with no dependency [16:39] information, the current upstream version of the package is plugged into a dependency that looks [16:39] like "packagename (>= packageversion)". [16:40] Laney, it's all "magic" to me. I just repeat the incantation and it works. [16:40] just saying, you're getting what you ask for :-) [16:40] Ack. I know that [16:42] Can we try rebuilding? I see no pattern to the failures and this has all built fine in our PPAs and COPRs [16:43] sure, let's give rebuilding a try… [16:44] (amd64 had already been retried once last night) [16:44] I just looked and it had been tried 6 hours ago ... [16:45] :( [16:45] So even if it works this time it's very flaky [16:45] maybe we get lucky though [16:45] got to get a train, back in a bit o/ [16:46] acheronuk: "Qt or mesa would not migrate even if all the tests passed" - that is not a reason to pile on more transitions, quite the contrary [16:48] * alan_g had hoped the build infrastructure was unusually heavily loaded at the time. [17:02] slangasek: I take that point, just blaming me for all ills when when other people's ships are not in order either, seems a bit unfair [17:03] not you I hasten to add [17:03] and I'm still sitting here trying to sort this [17:05] tjaalton: hmmm freeipa-client has a hard-coded dep on libcurl3, and libcurl4 doesn't show up at all in shlibs:Depends. I'll switch this from libcurl3 to libcurl4 but maybe it should be dropped altogether? [17:06] damn is auto-import not disabled? [17:06] slangasek: from proposed? I'll check [17:06] freeipa still needs libnsspem to pass tests (server install..) [17:06] :/ [17:06] I did sync something based on the fact that freeze was today [17:07] I'm thinking of filing a tech-ctte bug about nss [17:07] -queuebot:#ubuntu-release- New binary: libplacebo [s390x] (bionic-proposed/universe) [0.4.0-2] (kubuntu) [17:07] LocutusOfBorg: freeze is usually about 9pm UTC is it not? [17:08] -queuebot:#ubuntu-release- New binary: libplacebo [amd64] (bionic-proposed/universe) [0.4.0-2] (kubuntu) [17:08] -queuebot:#ubuntu-release- New binary: libplacebo [i386] (bionic-proposed/universe) [0.4.0-2] (kubuntu) [17:08] tjaalton: yes, understood; but I'll fix the bug in front of me first [17:08] slangasek: ok [17:08] anyhow, they weren't new packages, so I didn't cause any extra work for AA folks [17:09] FYI ^^ libplacebo is needed for vlc 3.0.1, I plan to merge it as soon as MoM is happy [17:09] slangasek: and I had a look at memcached which still fails tests on armhf even after a new merge, so I'll just override running tests on armhf so it can migrate [17:10] -queuebot:#ubuntu-release- New binary: libplacebo [ppc64el] (bionic-proposed/universe) [0.4.0-2] (kubuntu) [17:12] -queuebot:#ubuntu-release- New binary: dumb-jump-el [amd64] (bionic-proposed/universe) [0.5.2-1] (no packageset) [17:14] -queuebot:#ubuntu-release- New binary: golang-github-muesli-smartcrop [amd64] (bionic-proposed/none) [0.2.0+git20180228.f6ebaa7+dfsg1-1] (no packageset) [17:15] -queuebot:#ubuntu-release- New binary: libplacebo [arm64] (bionic-proposed/universe) [0.4.0-2] (kubuntu) [17:15] -queuebot:#ubuntu-release- New binary: libplacebo [armhf] (bionic-proposed/universe) [0.4.0-2] (kubuntu) [17:15] slangasek, somehow it seems to me that python is busted on arm64 & s390x and that vim is possibly busted on arm64. Can we remove subversion/arm64 subversion/s390x vim/arm64 from bionic release? [17:15] nobody uses vim on arm64, right dannf ? [17:15] alan_g: retry didn't really help https://launchpad.net/ubuntu/+source/mir/0.30.0.1-0ubuntu1 [17:16] xnox: you're going to have to provide some context, and then the answer is probably still no wrt removing vim [17:16] xnox: heh... well, i don't :) [17:16] xnox: but i'd expect a lot of screams if it went away [17:17] slangasek, no change rebuilds for ruby2.5-only failing =( http://people.canonical.com/~ubuntu-archive/transitions/html/html/ruby2.5-only.html [17:21] xnox: are you avoiding scheduling those for other things currently in -proposed? [17:21] slangasek, there are no more rebuilds left to do. [17:21] just fixing failures and autopkgtest regressions [17:22] k [17:32] make: execvp: /bin/sh: Argument list too long [17:32] fun autopkgtest failure [17:33] tyhicks: libseccomp autopkgtests fail with glibc 2.27; looks like a real test regression, can you investigate? [17:38] slangasek, please process http://people.canonical.com/~ubuntu-archive/priority-mismatches.html [17:39] slangasek, but do lint them [17:40] valorie: cyphermox just tested the i386 kubuntu install on his machine and it went fine [17:42] slangasek: hi - how quickly does it need to be fixed? [17:44] sil2100: should I not be releasing xenial SRUs? [17:44] tyhicks: glibc 2.27 should get into bionic release pocket this week. If you can tell me it's ok to have this test regress because it's a test incompatibility and not a runtime incompatibility, that would unblock [17:44] xnox: wow, yes, linting them which I think is more than someone did when promoting them the other direction :) [17:45] slangasek: ack - I'm tied up for the next couple hours but will be able to make that determination after [17:45] tyhicks: or if there's someone else on the security team you can delegate to/ :) [17:45] slangasek, then publisher run, then rebuilding ubuntu-base should become better again. [17:46] slangasek: yeah but they're in the same situation :) [17:46] I'll try to take a quick look now [17:47] tyhicks: pff ;) [17:47] xnox: yep, thanks :) [17:48] bdmurray: we're almost ready, but just to be safe for now release only unseeded thingies [17:48] i've checked everything, and then i'm like, there is no trace as to why it is still minimal after keyring fix.... unless priority-mismatches... [17:50] xnox: yes, minbase is precisely based on priorities [17:50] -queuebot:#ubuntu-release- Builds: Mythbuntu Desktop amd64 [Xenial 16.04.4] has been marked as ready [17:50] -queuebot:#ubuntu-release- Builds: Mythbuntu Desktop i386 [Xenial 16.04.4] has been marked as ready [17:50] -queuebot:#ubuntu-release- Builds: Ubuntu Base amd64 [Xenial 16.04.4] has been marked as ready [17:50] -queuebot:#ubuntu-release- Builds: Ubuntu Base arm64 [Xenial 16.04.4] has been marked as ready [17:50] -queuebot:#ubuntu-release- Builds: Ubuntu Base armhf [Xenial 16.04.4] has been marked as ready [17:50] -queuebot:#ubuntu-release- Builds: Ubuntu Base i386 [Xenial 16.04.4] has been marked as ready [17:50] -queuebot:#ubuntu-release- Builds: Ubuntu Base powerpc [Xenial 16.04.4] has been marked as ready [17:50] -queuebot:#ubuntu-release- Builds: Ubuntu Base ppc64el [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Ubuntu Base s390x [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Ubuntu Server amd64 [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Ubuntu Server arm64 [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Ubuntu Server armhf+raspi2 [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Ubuntu Server i386 [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Ubuntu Server powerpc [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Ubuntu Server ppc64el [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Ubuntu Server s390x [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Ubuntu Studio DVD amd64 [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Ubuntu Studio DVD i386 [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Netboot arm64 [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Netboot armhf [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Netboot powerpc [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Netboot ppc64el [Xenial 16.04.4] has been marked as ready [17:51] -queuebot:#ubuntu-release- Builds: Netboot s390x [Xenial 16.04.4] has been marked as ready [17:57] xnox: I don't see why priority-mismatches wants to promote libharfbuzz0b to important [17:58] -queuebot:#ubuntu-release- New binary: gst-plugins-base1.0 [s390x] (bionic-proposed/main) [1.13.1-1] (desktop-core, ubuntu-desktop) [17:59] * tsimonq2 wonders how 16.04.4 is coming along... [18:00] xnox: libxml2 -> libicu60 -> libharfbuzz0b -> {libgraphite2-3, libfreetype6}. Looks like icu needs fixing [18:00] * acheronuk cries as new VLC breaks KDE tests for slower arches [18:00] slangasek, it's trying to promote libicu-le-hb0 and... that [18:01] xnox: do you agree with fixing icu to not pull these deps in? [18:01] slangasek, if i were you, i would only demote things =) [18:01] * xnox looks [18:03] slangasek, that's not libicu60 fault there. [18:03] slangasek, libicu-le-hb0 is from icu-le-hb [18:03] ok [18:03] slangasek, maybe libxml2 should use libicu60? [18:03] it does [18:03] and libicu60 depends libicu-le-hb0 [18:03] -queuebot:#ubuntu-release- New binary: gst-plugins-base1.0 [amd64] (bionic-proposed/main) [1.13.1-1] (desktop-core, ubuntu-desktop) [18:03] sorry, I missed a link in the chain [18:03] sigh [18:03] -queuebot:#ubuntu-release- New binary: gst-plugins-base1.0 [ppc64el] (bionic-proposed/main) [1.13.1-1] (desktop-core, ubuntu-desktop) [18:03] yes it does [18:04] wtf [18:04] + cc -o tools/generate_nlmsg tools/generate_nlmsg.c -I ../include/ ../lib/libnetlink.a ../lib/libutil.a /usr/lib/x86_64-linux-gnu/libmnl.a [18:05] cc: error: /usr/lib/x86_64-linux-gnu/libmnl.a: No such file or directory [18:05] ok seriously, now. in an arm64 autopkgtest? [18:05] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=879820 [18:05] Debian bug 879820 in src:icu "configure icu with --enable-layoutex" [Important,Fixed] [18:05] 32 * Build with ICU Layout Engine API (closes: #879820): [18:05] 33 - add libicu-le-hb-dev build dependency, [18:05] 34 - add libicu-le-hb-dev dependency to libicu-dev package, [18:05] 35 - update layout extension tests, [18:05] 36 - add libiculx to shlibs. [18:05] * xnox ponders if that can be a separate package [18:06] oh, it's not icu-little-endian, ok [18:06] i think /usr/lib/x86_64-linux-gnu/libiculx.so.60 needs to go into a separate package [18:06] -queuebot:#ubuntu-release- New binary: gst-plugins-base1.0 [i386] (bionic-proposed/main) [1.13.1-1] (desktop-core, ubuntu-desktop) [18:07] xnox: +1. will you make that change? should I? [18:09] mwhudson: I've come back around to where fpc and glibc/s390x itself are the last test failures to look at [18:09] slangasek, it needs thought. because spliting that into separate package is easy, but we need to declare breaks on everything that depends on the layout engine by now, and rebuild things such that layout dependant things declare new dependency..... [18:10] xnox: if anything at all actually does depend on the layout engine [18:10] xnox: yes it needs thought - but do you have time to do that, or should I put it on my stack? :) [18:10] > OpenTTD that is failing due to the layoutex component. [18:10] yes, I didn't find any others. [18:10] possibly just the one. [18:11] slangasek, i am end of day, i can look at this tomorrow. [18:11] xnox: ok. before you go, have you seen the glibc/s390x autopkgtest regression? Does this look ignorable? === alan_g is now known as alan_g|EOD [18:12] -queuebot:#ubuntu-release- Unapproved: accepted python-cliff [source] (artful-proposed) [2.8.0-0ubuntu1.1] [18:13] slangasek, if the log.gz will ever load [18:13] i might see it [18:15] xnox: FAIL: malloc/tst-malloc-tcache-leak [18:15] original exit status 1 [18:15] INFO: 624 (bytes) are in use before starting threads. [18:15] error: xpthread_check_return.c:32: pthread_create: Resource temporarily unavailable [18:15] error: 1 test failures [18:16] found it. [18:16] xnox: with a different set of tests failing in the preceding run [18:16] our canonistack overcommitted with memory? [18:16] this is scalingstack, not canonistack [18:16] i meant scalingstack [18:16] infinity, you did ask me for s390x before for glibc rebuild; did you rebuild glibc, and did all the tests pass? [18:16] and s390x is almost never near capacity on units [18:17] slangasek, well..... [18:17] xnox: a memory overcommit problem would be showing up all over the place, not just in a glibc testsuite [18:17] slangasek, it's the one mainframe, with a single pool, of overcommitted ram, on HMC level... [18:17] slangasek, i do not believe we segregated RAM for scalingstack. [18:18] 19/* The point of this test is to start and exit a large number of [18:18] 20 threads, while at the same time looking to see if the used [18:18] 21 memory grows with each round of threads run. If the memory [18:18] 22 grows above some linear bound we declare the test failed and [18:18] 23 that the malloc implementation is leaking memory with each [18:18] 24 thread. This is a good indicator that the thread local cache [18:19] 25 is leaking chunks. */ [18:19] sounds like it is racy on purpose, and shouldn't be reliable in kvm [18:19] xnox: my biggest concern is that the run with that failure finished in half the time of the previous 2 runs which had a different set of (matching) failures [18:19] -queuebot:#ubuntu-release- New binary: gst-plugins-base1.0 [arm64] (bionic-proposed/main) [1.13.1-1] (desktop-core, ubuntu-desktop) [18:20] slangasek, well, right now launchpad queue is empty & autopkgtest queue is empty [18:21] slangasek, maybe it is a good time, to rerun it again, without much load. [18:21] slangasek, and infinity did suspeciously ask for an s390x on the weekend on a weird channel. [18:21] suspiciously [18:25] slangasek, the error "Resource temporarily unavailable" in normal conditions, is a retryable error, no? one should not fail on unavailable? [18:27] mdeslaur: fwiw xmltooling got Complicated™; I have to disable xmlsec to get it to build at all, and if I do that there's actually no longer a library conflict.... otoh it causes the testsuite to fail because no one ever builds without xmlsec, and that's probably a serious regression in functionality [18:28] xnox: yes, I have already requeued [18:29] slangasek: :( [18:29] slangasek: it was an issue with one of the libseccomp autopkgtests and I've just uploaded a fix [18:30] LocutusOfBorg: Could you explain this version number to me? 5.1.34-dfsg-0ubuntu1.17.10.1~17.10.1 [18:30] slangasek, the test code looks wrong to me, at least upstart testsuite for memory allocations like that did handle EAGAIN [18:30] slangasek: however, you are safe to proceed without waiting on the new libseccomp upload [18:31] slangasek, i am asserting, that the test code does not handle pthread_create returning EAGAIN correctly; and thus the result of the test case should not be FAIL, but SKIPPED couldn't test. [18:32] -queuebot:#ubuntu-release- New binary: kopanocore [s390x] (bionic-proposed/universe) [8.5.2-1ubuntu1] (no packageset) [18:33] tyhicks: cheers [18:36] -queuebot:#ubuntu-release- New binary: kopanocore [ppc64el] (bionic-proposed/universe) [8.5.2-1ubuntu1] (no packageset) [18:38] -queuebot:#ubuntu-release- Unapproved: accepted apt-xapian-index [source] (xenial-proposed) [0.47ubuntu8.4] [18:39] -queuebot:#ubuntu-release- New binary: gst-plugins-base1.0 [armhf] (bionic-proposed/main) [1.13.1-1] (desktop-core, ubuntu-desktop) [18:42] -queuebot:#ubuntu-release- New binary: kopanocore [i386] (bionic-proposed/universe) [8.5.2-1ubuntu1] (no packageset) [18:43] -queuebot:#ubuntu-release- New binary: kopanocore [amd64] (bionic-proposed/universe) [8.5.2-1ubuntu1] (no packageset) [18:47] bdmurray, sure, it is mostly in sync with the previous SRU we used [18:47] -queuebot:#ubuntu-release- New binary: kopanocore [armhf] (bionic-proposed/universe) [8.5.2-1ubuntu1] (no packageset) [18:47] was this meant to autosync? https://launchpad.net/ubuntu/+source/blogilo/4:17.08.3-1 [18:47] reason was: have a version that has *always* been lower than Ubuntu+1 and for sure higher than Ubuntu-1 [18:48] considering I had the ubuntu version removed earlier today [18:48] and doing this, like a "backport of a future release that has been probably in Debian and Ubuntu" was a good versioning [18:48] I did mostly the same in Debian, and I keep the same packaging to avoid more headaches [18:48] -queuebot:#ubuntu-release- New binary: kopanocore [arm64] (bionic-proposed/universe) [8.5.2-1ubuntu1] (no packageset) [18:49] we discussed the versioning probably in the previous xenial SRU, in the bug template (I did use a wrong one, and somebody from Release Team suggested that one IIRC) [18:49] can an AA accept libplacebo from new queue? this is an internal vlc transition, so I don't have to upload a new vlc after it gets accepted [18:50] the previous one has .16.04.2 no ~release.number [18:54] ok, so you can probably drop the ~foo at the end [18:55] it was because I did a *lot* of uploads in my ppa, and I wanted to have people upgrading with the Ubuntu official one after the upload [18:55] so, I used that versioning, and forgot to drop it :( [18:55] it affect virtualbox/xenial virtualbox/artful virtualbox-hwe/xenial [18:56] -queuebot:#ubuntu-release- New binary: nvidia-graphics-drivers-390 [amd64] (bionic-proposed/restricted) [390.25-0ubuntu2] (no packageset) [19:00] slangasek: fyi, the kopnaocore uploads (now in new) should resolve the missing dependencies in bionic-proposed [19:01] LocutusOfBorg: I think you own the work of dropping ~foo from the uploads. [19:03] already doing it :) [19:05] -queuebot:#ubuntu-release- Unapproved: virtualbox (xenial-proposed/multiverse) [5.0.40-dfsg-0ubuntu1.16.04.2 => 5.1.34-dfsg-0ubuntu1.16.04.1] (ubuntu-cloud) [19:06] -queuebot:#ubuntu-release- Unapproved: virtualbox (artful-proposed/multiverse) [5.1.30-dfsg-1 => 5.1.34-dfsg-0ubuntu1.17.10.1] (ubuntu-cloud) [19:06] -queuebot:#ubuntu-release- Unapproved: virtualbox-hwe (xenial-proposed/multiverse) [5.0.40-dfsg-0ubuntu1.16.04.1~16.04.4 => 5.1.34-dfsg-0ubuntu1.16.04.1] (no packageset) [19:08] we should be good now, wrt versionig :) [19:08] sorry again! [19:10] acheronuk: is it an accident that libkf5libkdepim again builds on ppc64el+s390x? [19:11] and who broke the yaml that retry-autopkgtest-regressions needs to parse >_< [19:16] -queuebot:#ubuntu-release- New: accepted debos [amd64] (bionic-proposed) [1.0.0+git20180112.6e577d4-1] [19:16] -queuebot:#ubuntu-release- New: accepted debos [i386] (bionic-proposed) [1.0.0+git20180112.6e577d4-1] [19:16] -queuebot:#ubuntu-release- New: accepted fonts-monoid [amd64] (bionic-proposed) [0.61-1] [19:16] -queuebot:#ubuntu-release- New: accepted gst-plugins-base1.0 [amd64] (bionic-proposed) [1.13.1-1] [19:16] -queuebot:#ubuntu-release- New: accepted gst-plugins-base1.0 [armhf] (bionic-proposed) [1.13.1-1] [19:16] -queuebot:#ubuntu-release- New: accepted gst-plugins-base1.0 [ppc64el] (bionic-proposed) [1.13.1-1] [19:16] -queuebot:#ubuntu-release- New: accepted libplacebo [amd64] (bionic-proposed) [0.4.0-2] [19:16] -queuebot:#ubuntu-release- New: accepted libplacebo [armhf] (bionic-proposed) [0.4.0-2] [19:16] -queuebot:#ubuntu-release- New: accepted libplacebo [ppc64el] (bionic-proposed) [0.4.0-2] [19:16] -queuebot:#ubuntu-release- New: accepted nvidia-cuda-toolkit [amd64] (bionic-proposed) [9.1.85-1] [19:16] -queuebot:#ubuntu-release- New: accepted debos [arm64] (bionic-proposed) [1.0.0+git20180112.6e577d4-1] [19:16] -queuebot:#ubuntu-release- New: accepted golang-github-muesli-smartcrop [amd64] (bionic-proposed) [0.2.0+git20180228.f6ebaa7+dfsg1-1] [19:16] -queuebot:#ubuntu-release- New: accepted gst-plugins-base1.0 [i386] (bionic-proposed) [1.13.1-1] [19:16] -queuebot:#ubuntu-release- New: accepted libplacebo [arm64] (bionic-proposed) [0.4.0-2] [19:16] -queuebot:#ubuntu-release- New: accepted libplacebo [s390x] (bionic-proposed) [0.4.0-2] [19:16] -queuebot:#ubuntu-release- New: accepted puredata [amd64] (bionic-proposed) [0.48.1-3] [19:16] -queuebot:#ubuntu-release- New: accepted qgis [amd64] (bionic-proposed) [2.18.17+dfsg-1] [19:16] -queuebot:#ubuntu-release- New: accepted qgis [armhf] (bionic-proposed) [2.18.17+dfsg-1] [19:16] -queuebot:#ubuntu-release- New: accepted qgis [ppc64el] (bionic-proposed) [2.18.17+dfsg-1] [19:17] -queuebot:#ubuntu-release- New: accepted dumb-jump-el [amd64] (bionic-proposed) [0.5.2-1] [19:17] -queuebot:#ubuntu-release- New: accepted gst-plugins-base1.0 [s390x] (bionic-proposed) [1.13.1-1] [19:17] -queuebot:#ubuntu-release- New: accepted nvidia-cuda-toolkit [ppc64el] (bionic-proposed) [9.1.85-1] [19:17] -queuebot:#ubuntu-release- New: accepted qgis [arm64] (bionic-proposed) [2.18.17+dfsg-1] [19:17] -queuebot:#ubuntu-release- New: accepted qgis [s390x] (bionic-proposed) [2.18.17+dfsg-1] [19:17] -queuebot:#ubuntu-release- New: accepted gst-plugins-base1.0 [arm64] (bionic-proposed) [1.13.1-1] [19:17] -queuebot:#ubuntu-release- New: accepted python-funcsigs [amd64] (bionic-proposed) [1.0.2-4] [19:17] -queuebot:#ubuntu-release- New: accepted libplacebo [i386] (bionic-proposed) [0.4.0-2] [19:17] -queuebot:#ubuntu-release- New: accepted qgis [i386] (bionic-proposed) [2.18.17+dfsg-1] [19:17] LocutusOfBorg: libplacebo accepted, but I don't see how this doesn't still require a vlc reupload; vlc-plugin-video-output depends on libplacebo3 [19:18] slangasek, https://release.debian.org/transitions/html/auto-libplacebo.html [19:18] don't know, it is listed here [19:19] LocutusOfBorg: I don't see how a Debian transition tracker is supposed to tell you anything about whether Ubuntu binaries need rebuilt [19:19] LocutusOfBorg: all that seems to tell *me* is that the vlc binaries in Debian were built against the new libplacebo [19:19] I use britney for that, I didn't check Ubuntu, I don't upload if not needed :) [19:20] slangasek: not an accident. but just a consequence of it's build-dep akonadi-contacts removing it's requirement to build itself against qtwebengine [19:20] LocutusOfBorg: ok; that's fine, it's just that you earlier asserted that a reupload wasn't needed [19:20] acheronuk: ok. I've stabbed some more autopkgtest retries [19:20] sadly same can't be sais for the rest [19:20] *said [19:21] slangasek, I have to leave for ~3h, I wanted to avoid a "possible" and useless rebuild, sorry if in my words it seemed to be "necessary" :) [19:21] xnox: glibc s390x retest completed, and we're back to the original 4 failures instead of the racy one. So I guess we need some analysis of those [19:21] anyhow, tonight I'll learn something new in case :P [19:31] turned off the Debian importer [19:43] ...point release publishing [19:43] re-queued --all-proposed for ruby-related failures [19:44] * sil2100 types in random publish-release commands and hides [19:49] slangasek: hi === fossfreedom_ is now known as fossfreedom [19:50] mwhudson: good morning! [19:50] slangasek: so fpc did you say? [19:50] mwhudson: ayup [19:51] huh it actually has revdeps so there goes idea #1 /s [19:51] :) [19:52] slangasek: so it seems to me that it is slightly incompatible in that glibc 2.27 now expects the compiler's crt0.o (or whatever it's called) to define __dso_handle and fpc's doesn't [19:53] slangasek: but it's obviously not very incompatible because the only effect this has is to transform one test that currently segfaults in release to one that fails to compile in proposed [19:55] slangasek: so uh, shrug? [19:56] mwhudson: would it be worth sanity-checking that conclusion by doing a rebuild test of the things using fpc in the archive, against latest glibc? [19:57] mwhudson: or does it even need a rebuild, vs. testing the existing binaries? [19:57] slangasek: hm yes that probably would make sense [19:57] slangasek: rebuilding i think, it's a link against libc_nonshared.a that fails... [19:57] i don't understand at all how the fpc test suite works i think [20:08] what are the chances that a packge name matching golang.*ratecounter.* is going to have flaky tests? [20:24] slangasek: so i've copied-with-binaries glibc to ppa:mwhudson/devirt [20:24] slangasek: i'll copy-without-binaries rev-build-deps of fpc when that's published [20:25] mwhudson: you could also just configure the ppa to build against -proposed, without needing to copy glibc [20:25] faster than waiting for a publication run [20:25] slangasek: is http://qa.ubuntuwire.org/ftbfs/rebuilds/test-rebuild-20171220-bionic.html new enough for comparison? i guess i hope so [20:25] slangasek: true but less hygenic i guess? as in that will get other bits of -proposed [20:26] mind you for a rebuild test that's ok i guess given builds build against proposed anyway [20:26] yeah [20:27] oh heh this ppa was configured against proposed anyway :) [20:30] bdmurray: hey! Can you update http://changelogs.ubuntu.com/meta-release xenial to 16.04.4? [20:30] sil2100: sure [20:31] bdmurray: thanks [20:35] you'd think my now I'd sort out the server name change [20:35] s/my/by/ [20:38] sil2100: all done [20:42] -queuebot:#ubuntu-release- Packageset: 397 entries have been added or removed [20:45] \o/ [20:53] FYI, LXD is going to miss the feature freeze by a couple of hours, sorting out some autopkgtest failures locally before uploading. [20:56] anyone looked at why retry-autopkgtest-regressions scales so poorly? [20:59] slangasek: i have not, but i did notice it too (when php was particularly bad) [21:00] Feature Freeze /o/ \o\ /o/ \o\ /o/ [21:01] slangasek: since you have context, could you ack LP: #1752713 ? [21:01] Launchpad bug 1752713 in gosa (Ubuntu) "Please sync gosa 2.7.4+reloaded3-2 from Debian unstable" [Undecided,New] https://launchpad.net/bugs/1752713 [21:01] nacc: ack and this doesn't need an FFe [21:02] slangasek: oh ok, i wasn't sure -- since it's a 'feature' in the sense of a new package [21:02] nacc: looks like it's the yaml parsing itself that's slow [21:02] slangasek: i could see that [21:02] nacc: we don't auto-sync such packages anymore, but individual Ubuntu devs can exercise judgement on new packages that are "safe" to sync [21:02] slangasek: ok, thanks for clarifying and sorry for the noise [21:05] 16.04.4 officially released, sending out the announcement now [21:05] (sorry for the noise) [21:05] sil2100: thanks for doing the stuffs :) [21:05] sil2100: nice work! [21:06] sil2100: Thanks! [21:08] slangasek: Latin like last cycle's FF topic change? :D [21:08] slangasek: no new failures [21:09] nacc: wow ok, so it's actually not the yaml parse time, it's the download time of update_excuses.yaml <_> <_< >_< [21:09] slangasek: do you know who I should poke about ubuntu-announce/ubuntu-release moderation? [21:09] we need proper download caching in the ubuntu archive tools [21:09] slangasek: has it exploded again? or is it just 'slow'? [21:10] sil2100: ubuntu-announce moderation done [21:10] slangasek: thanks! :) [21:10] nacc: 52s here to stream the file to /dev/null. I don't *think* it's my network that's to blame [21:11] nacc: 39MB of yaml, tho [21:11] slangasek: I guess someone needs to approve the ubuntu-release@ one as well, I guess because I sent it out from an unsubscribed e-mail [21:12] Big thanks for your help everyone o/ [21:17] slangasek: took almost 2 minutes here [21:21] -queuebot:#ubuntu-release- Builds: Kubuntu Desktop i386 [Xenial 16.04.4] has been marked as ready [21:21] sil2100: yeah turns out I don't have mod bits on ubuntu-release :) [21:21] tsimonq2: which bit was (needs to be) latin? [21:25] sil2100: cyphermox, thanks! i've marked it as ready [21:29] slangasek: My terminology is off (I don't study specifics of any languages besides English and Spanish (which isn't my strong suit to say the least!)) but to jog your memory: https://irclogs.ubuntu.com/2017/08/26/%23ubuntu-release.html [21:29] tsimonq2: oh. the latin is still there, the other stuff is just IPA :) [21:30] slangasek: ahhhhh, that's the right terminology [21:39] i'm hoping someone can provide me some advice, as i'm pulling my hair our here trying to figure something out with lubuntu seeds (always a wonderful source of confusion). it looks like on our desktop image (not the alternate), xiterm+thai is showing up and ending up as x-terminal-emualtor. looking at the logs it looks like it's installed as a depend of apport-gtk (line 3859). perhaps this is because [21:39] lxterminal isn't installed first (doesn't happen until #3958) and apport-gtk requires some x-terminal-emulator? [21:40] however, i'd expect whichever terminal is installed last to be x-terminal-emulator, but maybe that's foolish of me. [21:41] slangasek, trying: libplacebo gives --> * amd64: browser-plugin-vlc, forensics-extra-gui, forensics-full, freeplayer, freetuxtv, hdhomerun-config-gui, kaffeine, libplacebo-dev, libplacebo4, phonon-backend-vlc, phonon4qt5-backend-vlc, vlc, vlc-plugin-video-output, vlc-plugin-vlsub [21:41] so, vlc needs rebuild? [21:41] LocutusOfBorg: yes it does [21:41] thanks! [21:42] wxl: funny, i think FurretUber in #ubuntu+1 reported that issue this AM :) [21:42] nacc: He popped into #lubuntu-devel and this is a result of that debugging :) [21:43] ah cool [21:43] tsimonq2: thanks for the context [21:43] nacc: yeah i thought it was crazy at first, but there it is :/ [21:43] it's only lubuntu, too, at least from a spot check of 3 other flavors (no surprise there) [21:44] wxl: i think the germinate output would confirm your suspicion [21:45] -queuebot:#ubuntu-release- New binary: gst-plugins-good1.0 [amd64] (bionic-proposed/main) [1.13.1-1ubuntu1] (desktop-core, ubuntu-server) [21:45] -queuebot:#ubuntu-release- New binary: gst-plugins-good1.0 [s390x] (bionic-proposed/main) [1.13.1-1ubuntu1] (desktop-core, ubuntu-server) [21:47] -queuebot:#ubuntu-release- New binary: gst-plugins-good1.0 [ppc64el] (bionic-proposed/main) [1.13.1-1ubuntu1] (desktop-core, ubuntu-server) [21:47] sil2100: per trying to fast-track cloud-init SRU into xenial let's hold. regression on GCE w/ user-data not being observed [21:48] -queuebot:#ubuntu-release- New binary: gst-plugins-good1.0 [i386] (bionic-proposed/main) [1.13.1-1ubuntu1] (desktop-core, ubuntu-server) [21:48] regression only affects artful/gce [21:49] nacc: to be clear, that's what exactly? [21:50] wxl: germinate is what ends up processing the seeds [21:50] wxl: it's at http://people.canonical.com/~ubuntu-archive/germinate-output/ [21:50] that's what i needed. thanks nacc [21:50] wxl: but neither package is preinstalled, so it's coming from the task at install time, right? [21:51] nacc: right. and i have NO idea why xiterm+thai is even getting int here at all. [21:51] wxl: yeah it's seeded in 'daily-live' `seeded-in-ubuntu` says [21:51] nacc: i explored the possibility of language-pack-th requiring it but that's not the case [21:52] wxl: whereas lxterminal is in daily and daily-live [21:52] nacc: which of these many files do you see that in? [21:53] wxl: i'm going off `seeded-in-ubuntu`'s output [21:53] oh! [21:53] wxl: https://wiki.ubuntu.com/SeedManagement may help too [21:53] (the debugging seed problems part) [21:55] wxl: i do see libthai0 and libthai-data in http://people.canonical.com/~ubuntu-archive/cd-build-logs/lubuntu/bionic/daily-live-20180301.log [21:56] wxl: so i guess i'd need to grab your seeds ot be sure [21:57] nacc: neither of those seem like they require xiterm+thai tho [21:57] wxl: agreed [21:57] wxl: i think your original assessment (x-terminal-emulator being needed) is accurate, the question is why xiterm+thai is vailable at all probably [21:57] wxl: if they are both avialable, i'm not sure how apt picks [21:58] nacc: well, right, there's the question of if there are more than one how does it choose, but the other factor is i don't know how xiterm got there at all! [21:59] wxl: yeah, double bogeys :) [21:59] wxl: reading, seeing if i can help [21:59] slangasek: is often my goto for this, but i think he's pretty busy right now [22:00] well, thanks for the help. i'll keep poking at it, but if you get any ideas, please ping me or tsimonq2 [22:02] wxl: will do [22:02] http://people.canonical.com/~ubuntu-archive/germinate-output/lubuntu.bionic/rdepends/ [22:02] wxl: --^ that is a useful directory, and is what i was looking for [22:02] you can see why packages get pulled in [22:02] great! thanks [22:04] nacc: except xiterm isn't there *cries* [22:04] wxl: yeah and i don't know why [22:04] wxl: i'm assuming you've grepped your verbatim seeds etc [22:05] nacc: yep [22:07] nacc: we do have (fonts-thai-tlwg) but that's not it [22:08] -queuebot:#ubuntu-release- New binary: gst-plugins-good1.0 [arm64] (bionic-proposed/main) [1.13.1-1ubuntu1] (desktop-core, ubuntu-server) [22:09] nothing in germinate_output, nothing in all+extra. ugh i'm going to be bald soon [22:12] wxl: just to be 100% you've d/l the lubuntu iso and you see the package on the iso, right? [22:13] nacc: i've grabbed both the alternate and desktop images from today and installed them. alternate doesn't have xiterm+thai. desktop does, and it's x-terminal-emulator. it's also in the manifest, fwiw. [22:13] wxl: ok [22:13] and just for grins [22:14] strings bionic-desktop-amd64.iso | grep xitermxiterm+thai 1.10-2 === sil2100 changed the topic of #ubuntu-release to: Released: Xenial 16.04.4, Artful 17.10 | Archive: open | Bionic Release Coordination | Please don't upload things during freezes where you shouldn't, or be prepared to apologise to the release team | We accept payment in cash, check or beer | melius malum quod cognoscis [22:14] ah well you get the idea :) [22:14] wxl: by any chance did you check any older isos? [22:14] wxl: i wonder if you might need to run germinate locally ... or you'd need someone with access to cdimage [22:14] nacc: i checked as far back as is available on cdimage for bionic, but there's no problem with artful and before [22:15] nacc: if anyone's tried to run germinate locally it's tsimonq2 so i'll leave that to him :) [22:15] wxl: hrm [22:15] python datetime tz handling is a special kind of hell [22:16] slangasek: yes it is [22:17] nacc: that said, somewhere betwen october and now there's a problem. i could bisect it, but it looks like that would be extremely time consuming because the images would need to be produced. it seems that the germinate-output isn't very telling.. [22:17] wxl: yeah i thought there was a 'better' output page, i'm sorry -- i know i've seen it [22:18] regression filed sil2100 https://bugs.launchpad.net/ubuntu/+source/cloud-init/+bug/1752729 [22:18] Ubuntu bug 1752729 in cloud-init (Ubuntu) "GCE: cloud-init 17.2.35 no longer processes user data" [Undecided,New] [22:20] nacc: what really gets me is the cd-build-logs don't show anything there either. [22:20] wxl: yeah i'm finding that quite confusing too [22:20] well they show a relative diff, but not the file contents [22:21] nacc: however there they are in the ubuntu-cdimage logs [22:21] wxl: where are those logs? [22:22] https://launchpad.net/~ubuntu-cdimage/+livefs/ubuntu/bionic/lubuntu [22:22] ^ nacc [22:23] wxl: very strange; i would assume that would only be a possible satisifier if they were already avilable on the seeds [22:24] nacc: yes but clearly logic here doesn't apply XD [22:24] wxl: heh [22:24] wxl: virtual packages aren't always handled that well, my initial thinking is that you need to change apport-gtk's Depends from simply "x-terminal-emulator" to something like [22:24] nacc: do you know how these two cd logs are related? [22:25] wxl: i have no clue :/ [22:25] "gnome-terminal | mate-terminal | tilix | lxterminal | x-terminal-emulator" or whatever [22:25] wxl: do you have a LP bug for this issue yet? [22:25] the ubuntu-cdimage one appears to start around 16:38:49 while the cd-build-logs one begins at 16:38:26 [22:25] so they seem to be covering the same info to some degree [22:26] my assumption would be that the ubuntu-cdimage logs are part of the cd-build-logs, but without the verbosity [22:27] jbicha: i think that is a 'fix', but it doesn't explain why this alternative package is on the ISO in the firs tplace [22:27] last time on ubuntu-cdimage is 16:56:49 but cd-build-logs 17:09:01 [22:27] wxl: you could also try a ! trick thing like I used at https://bazaar.launchpad.net/~ubuntu-gnome-dev/ubuntu-seeds/ubuntu-gnome.bionic/view/head:/live#L74 [22:27] jbicha: apport-gtk requires x-terminal-emulator, which includes lxterminal (and xiterm+thai) [22:28] jbicha: ah an explicit unseed? [22:28] nacc: my guess is that germinate or whatever just makes things up when it finds a virtual package like that [22:28] and i don't think that's a bad thing. i don't want to make it harder for thai folks to get what they need [22:28] also, again, this wasn't an issue before so thta's just.. weird [22:28] I had trouble getting xterm out of main because of weird virtual package issues [22:29] jbicha: ah i guess i could see that [22:29] i don't see any unseeds in that link fwiw :) [22:29] wxl: line 74 [22:29] oh hah missed it [22:29] wxl: it's hard to read :) [22:29] wxl: xiterm+thai has no rdepends and as long as you don't literally add "Conflicts: xiterm+libthai", you won't be hurting any one that wants to use that package [22:30] i guess that might be a possible solution, but it seems like an annoying band aid [22:32] wxl: the band aid might be a lot easier than going down the rabbit hole to try to teach these tools to handle virtual packages smarter [22:32] i hear you jbicha :( [22:32] here's the bug https://bugs.launchpad.net/ubuntu/+source/lubuntu-meta/+bug/1752733 [22:32] Ubuntu bug 1752733 in lubuntu-meta (Ubuntu) "Default terminal emulator is set to txiterm, which causes problems with certain characters" [Undecided,New] [22:33] -queuebot:#ubuntu-release- New sync: gosa (bionic-proposed/primary) [2.7.4+reloaded3-2] [22:34] I don't know how terminals work, but it feels to me like there are 2 problems here: 1. xiterm+thai shouldn't be installed. 2. When it is installed, why is it default? [22:34] jbicha: like you said, #1 is a rabbit hole, but re: #2, i think it may have something to do with timing, like which is installed first/last. [22:35] jbicha: though logically i would expect the last one to be the default terminal and that's not the case here. lxterminal is last. [22:35] wxl: #2 that doesn't sound right to me. How do you set the default app for other apps? [22:36] jbicha: a lot of packages have post-inst scripts to make them default, or at least that's been my experience [22:36] installing chromium-browser shouldn't change the default browser from firefox for instance [22:40] wxl: btw, lxterminal 0.3.1-1 (this release cycle) lowered its x-terminal-emulator update-alternatives priority to be the same as xiterm+libthai [22:40] ah could that be why? [22:40] the postinst script in both sets priority to 20 [22:41] oh hah [22:41] gmta [22:41] what do you mean by "default terminal"? [22:41] what ultimately is symlinked as x-terminal-emulator [22:41] so here's my thinking: [22:42] 1. xiterm+thai is installed, and set to to 20 and since it's the first terminal, it's x-terminal-emulator by default [22:42] 2. lxterminal is installed, and set to 20 which is no greater than xiterm+thai, so it does not change the value of x-terminal-emulator [22:43] whereas before that, lxterminal would have taken precedence [22:43] that would answer 2) [22:43] that still does NOT explain why xiterm is on there tho XD [22:43] ok, I'll accept that explanation [22:43] still no answer for 1), afaict? :) [22:43] but i think the unseed is not a terrible idea [22:43] it would at least unbreak it for now [22:43] yepp [22:44] and you can always change it back if you figure out what's going on [22:44] who did you say would be the goto person for seed issues, nacc ? [22:44] slangasek is who i've asked in the past [22:44] i think any release team member or AA would know something about them, though :) [22:44] ok great. i'll bug him when he's awake again. thanks for the help to you and jbicha [22:44] wxl: also try specifically seeding lxterminal in 'live' ? [22:45] I'd point to any of cjwatson, infinity, or stgraber as deeper seed experts than myself [22:45] it may take several days to try out different seed fixes, good luck! :) [22:45] slangasek: thanks :) [22:46] nacc: meanwhile, I've just spent 1.5h hacking retry-autopkgtest-regressions, to save 1 minute off its running time :P [22:48] wxl: because of the nature of the bug, you may have to add a lof of ! rules [22:48] wxl: see apt-cache showpkg x-terminal-emulator for a list of all the packages that provide that virtual package [22:48] cyphermox: I've just landed some improved cache handling for update_excuses.yaml to retry-autopkgtest-regressions which I think you'll want to factor out and include in migration-assistant [22:49] i'll bug everyone in the morning when the usual parties are more awake then. thx slangasek [22:49] jbicha: i hope it doesn't come to that. [22:50] wxl: you could also email the ubuntu-release list [22:50] typically the way you work around "I'm getting the wrong implementation of $foo in my seeds" is to seed an alternative implementation "before" [22:50] slangasek: nicely done :) [22:50] the trouble is I never remember what "before" means in seedland [22:51] slangasek: i had that thought but i wouldn't expect that in lieu of no terminal emulator that xiterm+thai would be the goto XD [22:51] It'd be random [22:51] Essentially [22:51] oh [22:51] well maybe that's it then [22:51] ok, so now i just need to figure out the "before" thing [22:51] Should be sufficient to seed an alternative implementation in the seed that (indirectly) contains the dependency, or in an "inner seed" (i.e. one that that seed inherits from) [22:52] wxl: just try adding lxterminal to desktop-gtk then [22:52] germinate tries to follow roughly what apt would do if it were asked to install each seed in turn: that is, for each seed, it marks all the packages that are explicitly seeded for installation, and then it tries to unbreak any resulting broken dependencies [22:53] wxl: maybe sort of like https://git.launchpad.net/~lubuntu-dev/ubuntu-seeds/+git/lubuntu/commit/?id=6d424570 [22:53] how exactly do i determine the sequence of installation outside of going through all the seeds one at a time? [22:53] It prefers to pick packages that are seeded somewhere relevant (for some value of relevant ...) where it can, but failing that it'll be arbitrary [22:53] It's usually best to run germinate locally [22:54] i'm pretty sure tsimonq2 has germinate laying around on his machine [22:54] It's packaged, and its list of options is not impossibly long so it should be reasonably straightforward to duplicate what it does [22:55] Try to reproduce the bug first: come up with an invocation that shows the offending package being included in the output for some relevant seed [22:55] ... against a local but unmodified copy of the seeds [22:55] Then modify your local seeds until you get something that produces the desired output [22:56] Make sure to run germinate in an empty directory, as it'll spit out its many many output files all over it [22:57] And for a local run you probably want --no-rdepends or it'll take forever to emit all the gazilliojnn tiny little files [22:57] *gazillion [22:57] E: Failed to fetch http://ftpmaster.internal/ubuntu/pool/main/d/debconf/debconf-i18n_1.5.66_all.deb Could not connect to ftpmaster.internal:80 (91.189.89.99), connection timed out [22:57] aha, so the armhf autopkgtest problem isn't strictly a DNS problem, sometimes it randomly fails on other traffic too. :P [23:10] the one thing that's confusing is we haven't changed the location of apport-gtk or lxterminal in our seeds since 2016 [23:14] it's spectre's fault because we finally rebooted the machine germinate runs on! [23:14] (joking...) [23:21] hahahaha [23:30] -queuebot:#ubuntu-release- Unapproved: cloud-init (xenial-proposed/main) [17.2-35-gf576b2a2-0ubuntu1~16.04.1 => 17.2-35-gf576b2a2-0ubuntu1~16.04.2] (edubuntu, ubuntu-cloud, ubuntu-server) [23:34] -queuebot:#ubuntu-release- Unapproved: cloud-init (artful-proposed/main) [17.2-35-gf576b2a2-0ubuntu1~17.10.1 => 17.2-35-gf576b2a2-0ubuntu1~17.10.2] (edubuntu, ubuntu-cloud, ubuntu-server) [23:54] thx ^ [23:59] juliank: fyi, apt 1.6~beta1 fails its tests in bionic