[02:34] bdmurray: yes, it will [03:36] barry: okay, I did that then [04:59] pitti: Howdy. [05:07] Good morning [05:07] hey Unit193, how are you? [05:08] I think I'm alive, but I'd have to check. :) [05:18] doko: no, those other Qt packages are not from qt.io upstream, so not part of Qt as such. but our stack has a lot of other Qt/QML modules too. [06:17] I'm stunned, what am I doing wrong? [06:18] -rw-r--r-- 1 root root 289 Jul 16 08:11 /etc/init/autopkgtest-worker.conf [06:18] $ sudo initctl status autopkgtest-worker NAME=1 [06:18] initctl: Unknown job: autopkgtest-worker [06:18] I even tried initctl reload-configuration and rebooting [06:18] (that's on trusty) [06:18] s/status/start/ of course (it fails with any command) [06:19] pitti: tru initctl --system [06:19] same [06:20] pitti: iirc trusty introduced the user session stuff, and automatically uses --user if the right environment variable is available [06:20] sarnold: right, but no $UPSTART_SESSION and no user upstart instance running [06:20] either way, --system doesn't help, and I'm calling this through sudo [06:22] "initctl status ssh" works fine, it just doesn't seem to like my custom job [06:23] this is an "instance" job, but trusty has other jobs with "instance" [06:26] Jul 16 08:25:04 autopkgtest kernel: [ 768.186277] init: /etc/init/autopkgtest-worker.conf:12: Unknown stanza [06:26] aah! [06:26] here's to useful error messages [06:28] aha :) [07:00] hi folks, how do you feel about syncing virtualbox from experimental? [07:00] I uploaded yesterday 5.0 [07:04] I don't see a way to sync from experimental with requestsync [07:05] LocutusOfBorg1, -d experimental? [07:06] oops bad monday :) thanks [07:06] that kind of monday that persists until saturday I mean [07:06] lol, I see [07:07] usually with backportpackage -d means "destination" [07:07] so I got confused ;) [07:07] bugs filed, thanks [07:08] yw! [07:13] good morning [07:18] hey dholbach [07:18] hey seb128 === marcusto_ is now known as marcustomlinson === timchen1` is now known as timchen119 === psivaa is now known as psivaa-afk [11:00] Laney, seb128, Mirv: poppler ftbfs with the current qt5, see http://paste.ubuntu.com/11887104/ [11:00] triggered by [11:00] # enable all the hardening options [11:00] export DEB_BUILD_MAINT_OPTIONS = hardening=+all [11:01] that's what adds the -fPIE [11:03] right, Qt 5.4.2 bans now -fPIE usage [11:03] more info at http://code.qt.io/cgit/qt/qtbase.git/commit/?id=36d6eb721e7d5997ade75e289d4088dc48678d0d + http://code.qt.io/cgit/qt/qtbase.git/commit/?id=3eca75de67b3fd2c890715b30c7899cebc096fe9 [11:04] I'm now just uploading a poppler with that commented out to the silo 16 [11:04] with GCC 5 [11:09] doko: not "-pie"? [11:10] no, -fPIE triggers the macro [11:17] FYI, all newly-created amd64/i386 LP builds will now be dispatched to scalingstack (lcy01-*/lgw01-*), even if they're in non-virtualised archives such as the primary Ubuntu archive or ci-train silos [11:18] Retries of builds created prior to this change will still go to the bare-metal build farm for now, until we're confident enough in things to do the more invasive database surgery === _salem is now known as salem_ === MacSlow is now known as MacSlow|lunch [11:41] doko: Sorry, was checking - I mean hardening=+all,-pie === zequence_ is now known as zequence === MacSlow|lunch is now known as MacSlow [13:16] Laney, sure, but's the default in Ubuntu anyway [13:16] well we're not using that here, so needs fixing === gammax90 is now known as gammax [13:36] Laney: does that look correct? http://paste.ubuntu.com/11887745/ [13:36] kde4libs is blocking too much, and apparently there's no interest in fixing it [13:37] Laney: so far I modded results.history, but this should last a bit longer [13:38] pitti: looks right to me - do you have a hint entry in the britney config file? [13:38] Laney: no, I'm going to commit that next [13:38] right [13:38] Laney: http://paste.ubuntu.com/11887766/ [13:39] nod [13:41] Laney: both committed; I'll check excuses [13:42] Hello! I'm anxiously awaiting the arrival of AMDGPU from the 4.2 kernel. If I were to compile and test it myself (for my R9 285 GPU), would I also need to compile and install the AMDGPU branch of Mesa as well? [13:59] #ubuntu-devel [13:59] hello i am packing a Qt webkit based application am confused what should i put as dependencies my app executable ldd output is here in pastebin http://pastebin.com/mfgdrEtz [13:59] popey, hello i am packing a Qt webkit based application am confused what should i put as dependencies my app executable ldd output is here in pastebin http://pastebin.com/mfgdrEtz [14:03] try ##ubuntu-app-devel ? [14:03] oops one # [14:03] not two [14:04] ikonia, thanks i joined it [14:04] yeah i got it [14:45] barry: python-apt has an autopkgtest regression, did something else need uploading first? [14:46] bdmurray: it's unclear. it builds right? i wasn't able to do any dep-8 tests in the ppa, so my first order of business will be looking at clearing python3-defaults promotion issues [14:47] barry: yes, it builds. here is the failure - https://jenkins.qa.ubuntu.com/job/wily-adt-python-apt/lastBuild/ARCH=amd64,label=adt/artifact/results/log/*view*/ [14:47] bdmurray: ah, PendingDeprecationWarning printed to stderr [14:47] shouldn't be hard to fix [14:48] bdmurray: i'll take that one first [14:48] barry: okay, thanks! [14:48] infinity, Laney, slangasek: FYI: "run-autopkgtest --help" on snakefruit if you want to (re-)run an autopkgtest; e. g. "run-autopkgtest -s wily -a i386 libpng" (if you skip -a it'll run on all arches) [14:48] infinity, Laney, slangasek: ^ in the cloud, that is [14:49] infinity, Laney, slangasek: it's very simple-minded right now, and we'll probably extend it to do stuff like "re-run all failed/tmpfail tests" or something such, we'll learn the use cases over time [14:49] but that should be good enough for next week while I'm away [14:56] infinity: yay, the ddeb-retriever run from hell finally finished [15:00] pitti: cool, thanks - I guess the eventual goal is to provide something for uploaders in general to retry stuff? [15:02] Laney: if we can somehow hook u1 auth to that, sure [15:03] sso, I mean [15:04] Laney: then again, we've never had this anyway, so it's not a regression [15:06] no, but it's been an annoying wart of the current setup that only (~canonical ∩ (~ubuntu-core-dev ∪ ~ubuntu-release)) or so can retry things [15:06] Laney: well, I'm not sure whether we want so many people hit retry buttons all the time, given our limited resources [15:07] but in practice we actually have the opposite problem, not enough people watching excuses.html [15:07] but then again, cleaning up requires lots of overrides, which is ~u-release [15:07] I spent an hour yesterday to flush out tons of stuff that got stuck [15:10] not pretending that retrying can fix everything, does help sometimes though [15:10] yes, especially with uninstallables [15:11] Laney: anyway, it's a question of "how to get SSO there", not "I want to keep my toys for myself" [15:11] * Laney nods [15:12] for me it's "I don't want to have to do button clicking for people" and "I don't like requiring membership of ~canonical for Ubuntu dev tasks". :) [15:12] WTF, 16:12 and I haven't had lunch [15:12] * Laney brb [15:14] mterry, hey - thanks for all the MIR reviews - much appreciated [15:14] jamespage, yw, good break from other work for me :) [15:15] and if there are any archive-admins around, we have a hole load of main promotions that can be done for wily to unblock dep-waits for numerous openstack bits and pieces [15:29] pitti: will there be a way for developers to rerun a test from the web ui, like exists today for jenkins? [15:32] slangasek: lots of work and doesn't seem like a blocker, so not at the top of my list right now [15:32] if we can get it behind SSO, sure [15:35] pitti: right, no need for it to be at the top of the list, I just don't like the idea of retries having to be managed by a small set of people with ssh access to the right server [15:35] slangasek: extending this to "everyone in canonical" is reasonably easy, but as it requires VPN extending it beyond that is lots more work [15:36] err, not actually VPN, we just need to open the firewall; but again, need to find a way to put SSO in front of rabbitmq [15:36] right, so having that in the web interface would be nice :) [15:37] ac [15:37] k [16:07] * Laney likes that the same discussion just happened twice :P [16:08] mvo: regarding apt-clone if o.origin is empty and archive is not now, is it fair to say the package was installed from a .deb file? [16:18] bdmurray: most likely in this case the Release file has nothing/not much set in its headers. there is no distinction right now for "from-some-archive" or "from-dpkg-directly" [16:20] mvo: got it === psivaa-afk is now known as psivaa [18:59] where is codespeak-lib 1.4.30-1? this failure makes no sense: https://launchpadlibrarian.net/211833349/upload_7660667_log.txt [19:07] barry: Hm... If you search for the binaries, you can find that they do indeed exist http://packages.ubuntu.com/wily/python3-py, but seem to have come from a completely different package https://launchpad.net/ubuntu/+source/python-py [19:08] hjd: ah from that changelog it looks like maybe codespeak-lib source package should be removed from the archive? [19:09] Looks like it. Just found https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=792109 [19:09] Debian bug 792109 in ftp.debian.org "RM: codespeak-lib -- ROM; renamed to python-py" [Normal,Open] [19:10] hjd: looks like i should remove-package -s it from ubuntu too [19:11] * barry something else since apparently -s doesn't mean "source" any more [19:19] slangasek, infinity perhaps has permission for ^^ ? [20:09] pitti: Do you have any plans for an apport upload before your holiday? I'm thinking about bug 1470572. [20:09] bug 1470572 in apport (Ubuntu Vivid) "native-origins.d information causes apport to strip origin information from Package" [High,Triaged] https://launchpad.net/bugs/1470572 === WReTcH is now known as Guest84644 [22:44] slangasek: in MIR bug 1315313 oxide-developers ended up being subscribed to the package which isn't a team in the package to team mapping. Do you think the MIR process should include a list of team names? [22:44] bug 1315313 in ninja-build (Ubuntu) "[MIR] ninja-build" [Undecided,Fix committed] https://launchpad.net/bugs/1315313 === salem_ is now known as _salem === _salem is now known as salem_ [23:05] bdmurray: as long as there's a process for adding other teams to the list also === DrKranz is now known as DktrKranz