[00:02] no patches that I see, i'm trying to fix it with extra symlinks Ia32 -> IA32 [00:02] because case insensitive filesystems cannot usually have symlinks ;-) [00:05] slangasek: that makes it fails... later. [00:13] slangasek: i think nobody uses gcc to build Ia32... [00:18] xnox: yes, djgpp is more suited to that platform [00:19] slangasek: ah, you should be at Intel. This is some classy trolling. Alan Cox would love it ;-) [00:19] hah [00:23] slangasek: i'll see evan tomorrow. any tips on best thanksgiving trolling? [00:23] slangasek: my plan is to pretend it's christmas [00:23] hum [00:24] I don't know how you troll Thanksgiving [00:24] oh wait yes I do [00:24] https://www.youtube.com/watch?v=dZlwKkturgQ [00:32] turkey through chimney baked potatoes in the stockings, gotcha. === mark- is now known as mark === mark is now known as Guest74905 === robert-ancell is now known as robert_ancell === Punna is now known as Pwnna === BenC- is now known as BenC [05:53] mwhudson: bug 1396070, did you want to push the packages for trusty and utopic? (I'm out until dec 10, so the whole shebang could be sru'd before i get back :) [05:53] bug 1396070 in libvirt (Ubuntu) "Libvirt patches for launching VMs with 'ppc64le' architecture." [High,Confirmed] https://launchpad.net/bugs/1396070 [06:09] Good morning === Guest11916 is now known as happyaron [06:44] hallyn: that would be good, i'm not sure about the process though [06:44] hallyn: i guess last week would have been a good time to talk about it, doh [06:55] er, i appear to have the result of expanding ${shlibs:Depends} in the package not be the same as the value from debian/$pkg.substvars? [06:57] oh, i'm looking at the wrong version of the package [08:01] good morning [08:15] Mirv: hi, I am looking at disabled tests in qtsensors. They fail because make check does not honor QT_PLUGIN_PATH we path to it. [08:15] I.e. we do `xvfb-run -a dh_auto_test -- QT_PLUGIN_PATH=/build/buildd/qtsensors-opensource-src-5.3.2+dfsg/plugins' [08:16] But the actual test command is `QT_PLUGIN_PATH=/usr/lib/x86_64-linux-gnu/qt5/plugins ./tst_qsensorgesturepluginstest' [08:16] Tried passing it as env variable (not make variable), but that doesn't work as well. Do you know how to fix that? [08:29] mitya57: hmm, no, I don't know. I guess iahmad didn't find a way either when he enabled those tests a year ago. I wonder if it would need hacking the .pro files directory or where it comes from. [08:30] Ok, will try to dig that. [08:33] mitya57: I just checked that it seems to be pretty much the same everywhere, also in those packages like qtdeclarative that call make check more directly [08:33] dh_auto_test -- foo is the same as make check foo [08:34] so that is not a problem === MacSlow is now known as MacSlow|lunch === _salem is now known as salem_ [12:21] * zyga is unsure who to talk to about this bug [12:21] systemd provides hostnamectl with set-hostname command [12:21] that works ok on fedora, on Ubuntu/Debian though it breaks [12:21] as there are three operations that need to happen [12:22] 1) change the current hostname (works) 2) update /etc/hostname (works) update /etc/hosts (doesn't work, custom debian/ubuntu patch?) [12:22] this breaks stuff like sudo [12:22] because hostnames cannot be resolved and it's all downhill from there [12:23] I'd like to think how 3) can be solved [12:23] or if it's really needed (perhaps we could remove the $hostname entry from that file?) [12:24] https://bugs.launchpad.net/ubuntu/+source/gnome-control-center/+bug/1162475 [12:24] Launchpad bug 1162475 in systemd (Ubuntu) "[hostnamed] Changing hostname doesn't update /etc/hosts" [Low,Triaged] [12:24] * zyga looks [12:24] also https://bugs.launchpad.net/ubuntu/+source/libnss-myhostname/+bug/1162478/comments/6 [12:24] Launchpad bug 1162478 in libnss-myhostname (Ubuntu) "[FFe] [MIR] libnss-myhostname" [Undecided,Confirmed] [12:25] ah, nice, so with libnss-myhostname that entry will be obsolete [12:25] and it will all work [12:25] nice! [12:25] mgedmin: thanks [12:26] oh [12:26] except slangasek doesn't like that solution [12:28] slangasek: would your opinion on this topic change after a re-evaluation done today? [12:28] slangasek: I don't quite agree that parsing /etc/hosts to fix that is a sound approach (it's not reliable) [12:29] slangasek: but perhaps the better question is why is that needed in the first place (that entry?), if it wasn't there is something going to stop working? [12:36] resistance is futile! =) [12:36] [12:40] cjwatson, bug 1330410 is driving me a bit crazy :) The packages i build with pbuilder and in LP PPA are not quite the same in terms of the result of DEBIAN/config [12:40] bug 1330410 in Ubuntu Kylin "All keyboard types in "Keyboard layout" are not localized during installation" [High,Triaged] https://launchpad.net/bugs/1330410 [12:40] hope someone can explain to me the build environment used in LP [12:40] sbuild [12:41] I'd suggest starting with https://wiki.ubuntu.com/SimpleSbuild === MacSlow|lunch is now known as MacSlow [12:42] okay, is that similar to what LP use? [12:42] Yes [12:42] from the wiki, it should, thanks! [12:43] LP uses a very old fork for various reasons (hopefully we'll be able to fix that next year) but the differences almost never matter [12:44] ypwong: cjwatson: note however that src:ubiquity has it's own copy of console-setup which it builds during it's build and uses it at runtime.... [12:45] ypwong: thus the build of src:console-setup doesn't matter as much as the build of console-setup at src:ubiquity build time. [12:46] ypwong: reproducible build-environment is ~= $ mk-sbuild vivid; sbuild -A -d vivid *.dsc [12:46] ypwong: there is also sbuild-launchpad-chroot package which uses identical chroots downloaded from launchpad to do identical local builds as in launchpad. === pete-woods is now known as pete-woods|away [14:23] any chance of getting packagekit1 in vivid? [14:26] sitter: needs https://code.launchpad.net/~cjwatson/click/native-dbus or similar first or else it'll break the phone stack - I think mvo was working on continuing that? [14:28] wow [14:28] pbuilder-dist sid login [14:28] $reboot [14:28] "Running in chroot, ignoring request." [14:29] if I run reboot on pbuilder-dist vivid login [14:29] It reboots my machine [14:29] sid's will be systemd [14:29] mmm so? [14:31] LocutusOfBorg1: chroot is not a different machine.... e.g. try $ cat /proc/1/cwd/etc/hostname [14:32] LocutusOfBorg1: it's trivial to break out of a chroot if ones goal is to cause mischief and one is root in the chroot. [14:33] xnox, I understand that [14:33] the question is: why vivid behaves differently from sid? [14:34] reboot is provided by the init system [14:34] which is why my comment about systemd was pertinent [14:34] it's a completely different implementation, and it so happens that systemd's version of reboot includes that test [14:34] so they are both on systemd but the debian one has a check? [14:34] vivid is not on systemd [14:35] not yet, right? [14:35] correct [14:35] when it is, it will include that test [14:35] so it is just a matter of time/default [14:35] correct [14:35] wonderful, I'll try to reproduce on wheezy then :) [14:36] thanks, wonderful to know that systemd adds some little security there ;) [14:39] well. I think I'd call that a safety catch rather than security, but yes [14:41] yes of course [14:41] :) [14:42] so trying, my scientist mind forces me to try even if I know I'll reboot :) [14:45] lol --> wheezy WARNING: could not determine runlevel - doing soft reboot (it's better to use shutdown instead of reboot from the command line) shutdown: /run/initctl: No such file or directory init: /run/initctl: No such file or directory [14:46] trying squeeze [14:46] nope [14:52] pitti, hi === pete-woods|away is now known as pete-woods [14:58] mvo, hi === quadrispro_hds is now known as quadrispro [15:40] doko, hi === roadmr is now known as roadmr_afk === roadmr_afk is now known as roadmr === roadmr is now known as roadmr_afk === benonsoftware is now known as \b === \b is now known as benonsoftware === salem_ is now known as _salem [20:48] Can someone please force migration for gasic? It has been changed in debian to only build on amd64 as it depends on bowtie which is now only on amd64 === roadmr_afk is now known as roadmr [21:36] Noskcaj: Force? No. Carefully usher? Sure. [21:38] infinity, ok :) [21:38] Noskcaj: Fixed. [21:38] ty [21:38] hopefully someone will sort out bowtie properly soon [21:38] Noskcaj: I know people like the word "force" cause it's, well, forceful. But 99% of the time, when you're asking that, you're effectively asking to make the archive broken. [21:39] ok [21:39] Noskcaj: Examining what needs fixing (in this case, removing the NBS !amd64 binaries from -proposed) is the saner thing.