UbuntuIRC / 2012 /01 /13 /#ubuntu-devel.txt
niansa
Initial commit
4aa5fce
[00:13] <lamont> ScottK: I'll work on that this weekend
[00:14] * nonix4 ponders whether debug symbols for precise grub2 exist somewhere; debugging grub-install segfault without such is PITA
=== dendro-afk is now known as dendrobates
[00:24] <cjwatson> barry: Would you like me to take the python-debian work item? Not if you've already started or anything, but I wouldn't mind getting a bit of porting practice here
=== bladernr_ is now known as bladernr_afk
[00:58] <psusi> so unity is a shell and it is normally run with compiz as the wm... but gnome-shell is both a shell, and a clutter based wm in one... do I have that right?
=== bkerensa_ is now known as bkerensa
=== funkyHat_ is now known as funkyHat
[01:48] <psusi> why the hell is there no bzr branch for mdadm?
=== Laney is now known as Guest22075
[02:28] <ScottK> lamont: Great. Thanks.
[03:57] <maxolasersquad> I am working on packaging a game an have three remaining lintian errors being thrown.
[03:57] <maxolasersquad> http://paste.ubuntu.com/802571/
[03:57] <maxolasersquad> This is not an app I wrote, I'm packaging it for ARB.
[03:58] <maxolasersquad> Specifically I'm trying to resovle the invalid category in the .desktop file.
[03:59] <maxolasersquad> The file is being created automatically when I run debuild, but I'm not exactly sure where it is getting Qt;Game;Memory;
[04:00] <maxolasersquad> The code is at http://bazaar.launchpad.net/~maxolasersquad/+junk/qml-memory-game/files
[07:02] <vibhav> While compiling wxGTK I get this error : http://paste.ubuntu.com/802665/
=== zyga-afk is now known as zyga
=== bdmurray_ is now known as bdmurray
[08:20] <slangasek> barry: bug #915843
[08:20] <ubottu> Launchpad bug 915843 in linux (Ubuntu) "ThinkPad X201: intermittently powers off instead of suspending when closing the lid" [High,Incomplete] https://launchpad.net/bugs/915843
[08:33] <geser> Sweetshark: are you already aware of bug #915271? it breaks upgrades on precise
[08:33] <ubottu> Launchpad bug 915271 in libreoffice (Ubuntu) "package libreoffice-core 1:3.4.4-0ubuntu2 failed to install/upgrade: rmdir: failed to remove `usr/lib/libreoffice/basis3.4/program/': Directory not empty" [Critical,Triaged] https://launchpad.net/bugs/915271
[08:38] <mvo> slangasek: I am on the apt issue again this morning and would love to get the error message you have. I created a testcase for what I get (Internal Error, Could not early remove crda) http://paste.ubuntu.com/802717/ and would love to hear what kind of error you see
[08:48] <pitti> infinity: you can get your netbook back, I can reproduce on my machine by explicitly calling it
[08:55] <pitti> ogra_, infinity: FYI, it's bug 913085 if you want to follow
[08:55] <ubottu> Launchpad bug 913085 in pkgbinarymangler (Ubuntu) "Translation domain detection from config.h is broken" [Critical,Triaged] https://launchpad.net/bugs/913085
[08:57] <stgraber> slangasek: bug 24061
[08:57] <ubottu> Launchpad bug 24061 in apt (Ubuntu Precise) "GPG error with apt-get/aptitude/update-manager behind proxy (BADSIG 40976EAF437D05B5)" [High,In progress] https://launchpad.net/bugs/24061
[08:57] <infinity> pitti: Err. That's special.
[08:57] <slangasek> mvo: hi; so at this point I'm manually dist-upgraded to the new wireless-crda+crda and no longer see the issue myself :/
[08:57] <slangasek> stgraber: yep - thanks :)
[09:00] <sagaci> pitti: regarding ubuntu-defaults-image, is there any configuration to specify a mirror?
[09:01] <pitti> sagaci: not right now
[09:07] <mvo> thanks slangasek - no worries, the testcase definitely catches one more instance of the bug, I was mostly wondering if there is yet another corner case for it
[09:07] <infinity> pitti: Thanks for the pointer; removing the double-quotes fixed it locally.
[09:08] <ogra_> same here
[09:08] <pitti> right, I tested that here
[09:08] <pitti> funny bug
[09:08] <infinity> Certainly an unusual symptom for the problem.
[09:09] <slangasek> mvo: ah, ok
=== cjwatson_ is now known as cjwatson
[09:34] <hallyn> poolie: http://package-import.ubuntu.com/status/libvirt.html#2011-05-26%2020:07:23.558315 'bzr import-dsc' by hand works fine, package importer fails.
=== enrico_ is now known as enrico
[09:42] <mvo> slangasek: fwiw, I proposed removing debian/ from trunk for command-not-found and have it only in a packaging branch to avoid confusion again (confusion caused by absent minded people like me)
[09:43] <pitti> cjwatson: I think we should remove the current libo 3.5 binaries to make upgrades a little less broken; does that sound ok to you
[09:43] <pitti> ?
[09:44] <cjwatson> pitti: The NBS ones?
[09:44] <pitti> cjwatson: no, all binaries
[09:44] <pitti> bdmurray says the upgrade bug alrady has 130-something dupes
[09:44] <pitti> cjwatson: but building the fix will take at least 10 hours on the buildds
[09:44] <cjwatson> So effectively no LibO in precise for a while?
[09:45] <cjwatson> That will break image builds
[09:45] <infinity> This sounds like a good long-term solution.
[09:45] <cjwatson> infinity: :-P
[09:45] <pitti> cjwatson: they are already broken
[09:45] <pitti> it doesn't make it any less urgent to unbreak LibO
[09:45] <cjwatson> Well, if I can blame the desktop team for it :-P
[09:45] <pitti> but it might ease upgrades a little bit
[09:45] <pitti> cjwatson: well, you can blame us either way :)
[09:46] <pitti> ah, bug 915271
[09:46] <ubottu> Launchpad bug 915271 in libreoffice (Ubuntu) "package libreoffice-core 1:3.4.4-0ubuntu2 failed to install/upgrade: rmdir: failed to remove `usr/lib/libreoffice/basis3.4/program/': Directory not empty" [Critical,Triaged] https://launchpad.net/bugs/915271
[09:46] <cjwatson> If we remove it, people who upgrade to precise in the relevant interval will be prompted by update-manager to remove LibO, I think
[09:47] <cjwatson> Maybe we can tolerate that but I thought it should be mentioned
[09:47] <pitti> ah, so it's not -l10n, but due to the broken postinst
[09:48] <cjwatson> pitti: Well, I'm a bit nervous of the removal, but if you think it's a good idea, I can't actually think of solid arguments against
[09:49] <pitti> it's only a recommends of u-desktop, so if u-m removes it, nothing would pull it back in
[09:49] <pitti> cjwatson: bdmurray, Sweetshark and I are currently discussing reuploading 1:3.5.0~beta2+really3.4.4-0ubuntu1
[09:49] <cjwatson> That sounds intrusive
[09:49] <pitti> it "only" takes ~ 2.5 hours to build instead of 10
[09:50] <cjwatson> Despite the low build time, I think I like that even less
=== smoser` is now known as smoser
[09:51] <cjwatson> Why the massive build time spike in 3.5?
[09:51] <pitti> we still have the corresponding -l10n source in the archive
[09:51] <pitti> cjwatson: unknown yet; on Sweetshark's machine it builds in 2
[09:51] <cjwatson> It might be different on different builders?
[09:51] <pitti> possibly
=== masACC is now known as maswan
[09:52] <pitti> cjwatson: beta3 is already released, so we wouldn't have to carry the +really version number for long
[09:53] <pitti> so it seems both binary removal and the status quo breaks upgrades quite badly
[09:55] <pitti> cjwatson: OOI, why do you think a reversion is worse?
[09:56] <pitti> hm, it'd need a libo-l10n upload as well, as libreoffice-l10n-XX Depends: libreoffice-common (<< 1:3.5~)
[09:58] <geser> what about fixing the broken preinst?
[09:59] <pitti> that needs to happen, of course; currently discussing with Bjoern
[10:00] <pitti> if it's faster to fix 3. with -l10n (that already is confirmed to work) and the human theme (that still causes problems), or revert to 3.4
[10:00] <pitti> "fix 3.5"
[10:02] <slangasek> cjwatson: per bdmurray's suggestion, I'm going to mark ubuntu-foundations-team as restricted rather than moderated so I don't have to tell people "no" anymore... we'll just have to invite people instead when we want them in :)
[10:04] <cjwatson> slangasek: good idea!
[10:04] <cjwatson> pitti: I guess it's not so terrible if there's no risk of us releasing with it; LibO is sufficiently high-profile that I do think people might actually vaguely care about the version number
[10:05] <pitti> in the worst case we'll release alpha-2 with it, but this should land next week
[10:05] <pitti> but at this point it gets dangerously close to the weekend for my taste
[10:06] <cjwatson> I agree we need to fix this before the weekend ...
[10:06] <cjwatson> barry: did you see my offer last night to take the python-debian port?
[10:06] <geser> and on monday you can't open that bug anymore as LP times out due to the amount of duplicates
[10:08] <pitti> this is an apport generated bug
[10:08] <pitti> I'll create a bug pattern
[10:11] <pitti> committed
[10:11] <cyphermox> jodh: stgraber: re the upstart jobs for bluez I was talking about earlier -- bluetooth-daemon.conf: http://paste.ubuntu.com/802774/ -- bluetooth-device.conf: http://paste.ubuntu.com/802775/
[10:11] <cyphermox> if you could take a look that would be great.
[10:13] <jodh> cyphermox: will do, thanks.
[10:13] <cyphermox> thanks to you -- the initial code was great help
[10:14] <cyphermox> fwiw: if you want to test you'll need to obviously remove /etc/init.d/bluetooth, but less obviously get /lib/udev/rules.d/97-bluetooth.rules out of the way
[10:14] <pitti> so, 3.4 took 6:40 hours to build, seems it's really buildd dependent
[10:14] <pitti> i386 takes 2.5 hours
[10:14] <pitti> but most of the dupes are in fact on amd64
[10:15] <pitti> (not surprisingly, given the target audience of current precise)
[10:17] <bdmurray> pitti: the pattern doesn't match may of the duplicates
[10:18] <pitti> ah, e. g. bug 915455
[10:18] <ubottu> Launchpad bug 915271 in libreoffice (Ubuntu Precise) "duplicate for #915455 package libreoffice-core 1:3.4.4-0ubuntu2 failed to install/upgrade: rmdir: failed to remove `usr/lib/libreoffice/basis3.4/program/': Directory not empty" [Critical,Triaged] https://launchpad.net/bugs/915271
[10:19] <pitti> bdmurray: I'll check the logs instead
[10:20] <barry> cjwatson: i missed that, but that would be awesome
[10:20] <barry> cjwatson: btw, i'm going to go hang out with the bzr guys now. do you have a package w/quilt that you'd like us to try to merge?
[10:21] <cjwatson> barry: not off the top of my head that I'm not already handling adequately, sorry :)
[10:22] <cjwatson> pitti: if you can find a builder that does a decent job, we can stick the farm on manual just before the uploa
[10:22] <cjwatson> d
[10:22] <barry> cjwatson: okay ;) we'll hunt around for a fun one to hammer on
[10:22] <pitti> cjwatson: *nod*
=== lan3y is now known as Laney
[10:25] <pitti> bdmurray: added another pattern, matches a lot more now
[10:32] <bdmurray> pitti: great, thanks
[10:43] <hrw> eglibc failed to crosscompile for me: http://paste.ubuntu.com/802791/
[10:54] <slangasek> hrw: is that the current precise version of the package?
[10:58] <hrw> slangasek: yes and fail is during cross toolchain bootstrap
[10:58] <hrw> slangasek: I am doing test now with older version to check when it started to fail
[11:00] <slangasek> hrw: ok, thanks
[11:00] <slangasek> barry: hem, have to reject this python-dbus upload - E: python3-dbus-dbg: non-standard-toplevel-dir debian/
[11:01] <slangasek> barry: there are several other lintian errors to be fixed at the same time
[11:09] <tjaalton> let's try here; any DD's in budapest willing to sign my gpg key?
[11:09] <tjaalton> one is enough :)
[11:16] <infinity> tjaalton: Depends on how nicely you ask.
[11:17] <tjaalton> infinity: pleeease.. ? :)
[11:18] <tjaalton> infinity: no rush with the actual signing, but I can drop you a paper with the fingerprint and stuff..
[11:18] <infinity> tjaalton: Bring extravagant gifts to the ARM room.
[11:18] <tjaalton> salmiakki that is then :)
[11:19] <hrw> salmiakki aka 'what is wrong with Finnish people' :D
[11:22] <nigelb> lol.
[11:22] <nigelb> "extravagant gits to ARM room."
[11:23] <nigelb> I can almost imagine some kind of primitive initiation ceremony happening near infinity :P
[11:25] <tjaalton> hrw: the same :)
=== _salem is now known as salem_
[11:28] <cjwatson> Riddell: In kdeplasma-addons 4:4.7.90-0ubuntu1, did you mean to also drop plasma-widget-kimpanel's dependency on plasma-widget-kimpanel-backend-ibus while dropping the latter package?
=== axp2_ is now known as axp2
[11:30] <Riddell> cjwatson: hmm there's a whole spec to sorting that out which involves finding and evaluating an update to kimpanel. I'll take a quick look at kdeplasma-addons today to answer your question
[11:31] <cjwatson> Riddell: OK, I just noticed because it shows up in NBS. Thanks
[11:34] <slangasek> tjaalton: one is never enough
[11:35] <slangasek> tjaalton: I'm sure you want to give me salmiakki in exchange for signatures as well
[11:36] <tjaalton> slangasek: hehe, well I got Mirv to sign it earlier, but another one wouldn't hurt I guess
[11:36] <tjaalton> let me "print" the key first
[11:36] <tjaalton> er, fingerprint
[11:37] <slangasek> oh, I forgot that Mirv is also a DD; clearly both of us failed at having our fingerprints displayed prominently on the outside of our jackets
[11:39] <pitti> cjwatson: just FYI, we discussed reversion, and it seems it's not any less risky or less work than fixing 3.5 (it's building for an hour already)
[11:39] <pitti> we have the preinst fixed, and -l10n* are generated properly
[11:40] <tjaalton> slangasek: ok, where are you atm?
[11:40] <slangasek> tjaalton: foundations room (across from ARM)
[11:40] <tjaalton> ok
[11:42] <infinity> tjaalton: Sent to pgpkeys.mit.edu
[11:44] <tjaalton> infinity: excellent, thanks
[11:44] <pitti> infinity: which of the amd64 builders is the fastest we have?
[11:44] <cjwatson> pitti: *nod* thanks
[11:46] <slangasek> infinity: without validating the email addresses! <gasp>
[11:49] <doko> pitti: allspice
[11:49] <pitti> doko: danke
[11:51] <bdmurray> pitti: /msg pitti do you have any plans to upload apport in the near future?
[11:52] <pitti> bdmurray: there are a few fixes in trunk, but nothing urgent; but I can do an upload on Monday or so if you want to get in the kerneloops stuff
[11:52] <Mirv> tjaalton: :P
[11:53] <bdmurray> pitti: that and the ubiquity package hook change would be nice
=== lool- is now known as lool
=== yofel_ is now known as yofel
[12:10] <barry> cjwatson: you might want to give lp:bzr and lp:bzr-builddeb a try. i got a demo of `bzr merge` in the presence of quilts. *really* nice
[12:12] <cjwatson> cool, ok, will have a look next time I have occasion
[12:12] <barry> cjwatson: i think it will land in precise in a few days
=== akgraner` is now known as akgraner
=== jbernard` is now known as jbernard
[13:24] <infinity> slangasek: I already knew his IDs were his. All I would have been validating is the lack of an MITM.
[13:24] <slangasek> pshaw
[13:24] <infinity> ;)
[13:25] <slangasek> you only knew that they belonged to some person calling himself tjaalton
[13:25] <slangasek> there are apparently 8 million of those in finland
[13:25] <infinity> No, I knew they were his, or belonging to someone who pretends to be him on a regular basis.
[13:25] <infinity> And quite convincingly.
[13:25] <infinity> If that's true, that guy can have the sig.
[13:25] <infinity> He's earned it.
[13:27] <tjaalton> :)
[13:28] <slangasek> hallyn: waah, why did you break virt-manager
[13:28] <hallyn> i did what?
[13:28] <slangasek> hallyn: Error starting domain: internal error no assigned pty for device charconsole0
[13:29] <slangasek> regression since Wednesday-ish
[13:30] <hallyn> slangasek: does 'virsh console <domain' work?
[13:30] <slangasek> hallyn: I don't know what I'm meant to specify as 'domain', this is why I use virt-manager ;)
[13:30] <hallyn> slangasek: zul says he's going to merge 0.9.8-2, hopefully that will pull in the *full* set of fixes.
[13:32] <slangasek> hallyn: "error: The domain is not running
[13:32] <slangasek> "?
[13:32] <slangasek> hallyn: 'virsh start' gives me the same 'Failed to start' error
[13:36] <hallyn> strgraber: the 'libvirt not showing running domains' - can you reproduce that at will right now?
[13:37] <stgraber> hallyn: no, that was yesterday, I then completely killed libvirt, manually killed anything that was still running and restarted it, worked fine after that.
[13:42] * om26er misses the patch pilots :/
[13:43] <ScottK> They were on holiday break, they'll be back.
=== Quintasan_ is now known as Quintasan
=== bladernr_afk is now known as bladernr_
[14:28] <hallyn> cyphermox: should netcf be depending on libnl3 (if src pkg depends on libnl-3-dev)?
=== rsalveti` is now known as rsalveti
[14:29] <cyphermox> hallyn: if you're using symbols from libnl3, you'd get a depends on libnl-3-200 or libnl-genl-3-200.
[14:29] <cyphermox> hallyn: depends mostly if it #includes netlink/netlink.h or netlink/genl.h (or any such thing)
=== kancerman__ is now known as kancerman
[14:32] <hallyn> cyphermox: guess i'm just confused as to why it wasn't in there to begin with. (or why it worked without it)
[14:32] <cyphermox> not sure, let me look
[14:34] <cyphermox> hallyn: right now it seems to me like it's there, in libnetcf1
[14:34] <hallyn> cyphermox: well i'm pretty sure i just messed up originally (and got away with it bc of spurious other pkgs)
[14:34] <hallyn> hm, i don't see it
[14:34] <cyphermox> Depends: libaugeas0 (>= 0.6.0), libc6 (>= 2.8), libnl-3-200 (>= 3.2.3), libnl-route-3-200, libxml2 (>= 2.7.4), libxslt1.1 (>= 1.1.25), augeas-lenses
[14:34] <cyphermox> that's from apt-cache show libnetcf1
[14:34] <hallyn> weird.
[14:35] <cyphermox> hallyn: coming
[14:53] * pitti shepherds a new libreoffice in and crosses fingers
=== elif is now known as elif_lunch
[15:47] <lifeless> why is 'for purchase' empty in software centre in precise ?
[15:48] <Laney> AFAIK stuff has to be manually rolled over and this has not happened yet. iamfuzz ^^^?
[15:49] <cjwatson> iamfuzz isn't responsible for it any more afaik, not sure who is
[15:49] <james_w> correct
[15:56] <pitti> bdmurray: found the bug, uploading fix now
[16:07] <iamfuzz> cjwatson, I think Scott Ritchie is still helping out with that part time
=== dendrobates is now known as dendro-afk
[16:22] <andrejz> hello! can anyone point me in the direction of checkbox developers ?
=== dendro-afk is now known as dendrobates
[16:34] <pitti> andrejz: cr3 ?
[16:34] <andrejz> ok, thanks pitti
[16:37] <cr3> andrejz: hi there, what's up?
[16:38] <andrejz> hello, i have some concern about the string to be translated. they require substantially more translator work then they could. I briefly explained it here - https://bugs.launchpad.net/checkbox/+bug/916096
[16:38] <ubottu> Ubuntu bug 916096 in checkbox "Checkbox requires a lot of translator work" [Undecided,New]
[16:38] <andrejz> do you think something can be done about it?
[16:40] <cr3> andrejz: I'm not sure, those strings are taken from the equivalent to rfc822 deb files
[16:40] <cr3> andrejz: however, you raise an extra good point that "PURPOSE" should not be translated but that's our fault, it's really not clear :(
[16:41] <cr3> andrejz: I'm really not happy with the pseudo-markup format which PURPOSE/STEPS/VERIFICATION are intended to be
[16:42] <andrejz> it's kind of annoying i need to translate some strings over and over again (PURPOSE, STEPS and VERIFICATION are most prominent, but there are others too)
[16:43] <cr3> andrejz: I'll raise this with the rest of the team for sure but I can't make any promises for Precise since we're approaching FF quickly. however, there are talks for a major refactoring in P+1
[16:43] <andrejz> ok, thanks
[16:43] <cr3> andrejz: there will be a blueprint for the refactoring listing some of the problems with the current implementation, so I'll make sure to raise your concern
[16:44] <andrejz> cool, i hope something nice comes out of it :)
[16:45] <cr3> andrejz: me too, I can appreciate the annoyance you have to deal with. thanks for taking the time to get in touch with me though, very nice!
[16:46] <pitti> cjwatson: (CC: RAOF) oneiric unapproved has some syncs, presumably from (failed?) experiments from you?
[16:46] <andrejz> well at first i started translating but then i got increasingly annoyed and then realised i am surely not the only one with this issue and decided to report it :)
[16:46] <pitti> cjwatson: we can't do anything but reject them, but not sure what the intent was?
[16:47] <pitti> cjwatson: qemu-kvm and compiz-plugins-main
[16:47] <cjwatson> hmm
[16:47] <cjwatson> oh! I wondered where those had gone, that explains a certain amount
[16:47] <cjwatson> yes, reject them
[16:48] <pitti> rejecting all five dupes :)
[16:48] <cjwatson> those were actually copies from oneiric-proposed - I did them a different way in the end after being unable to figure out what was going wrong
[16:48] <cjwatson> but if they landed in unapproved, that makes a kind of sense
[17:54] <hikenboot> anyone interested in seeing (besides me) Proxmox 2.0 kvm/openvz/ISCSI/ZFS working on the next version of Ubuntu? Here I have it working on DEBIAN see the last 4 articles posted at wanfuse.blogspot.com...should be easy enough to translate to ubuntu...problem is that most of it is touchy about kernel versions? I think with some help it would be a great inclusion in the next ubuntu release!
[17:54] <cjwatson> pitti: just so you know, you can now pass multiple -s arguments to sru-accept.py so that it only leaves one comment
[17:57] <hikenboot> or should i post that on #ubuntu-app-devel?
[18:00] <hikenboot> In order to get it working with ISCSI and able for the iscsi to run Microsoft clusters I needed to use IET ISCSTARGET instead of openiscsi target..
[18:01] <hikenboot> and had to get a patch for ZFS to work with ISCS since in its current form in the normal tree it does not
[18:30] <pitti> cjwatson: useful, thanks!
=== salem_ is now known as _salem
=== Alcine is now known as jalcine
=== bladernr_ is now known as bladernr_afk
=== dendrobates is now known as dendro-afk
=== dendro-afk is now known as dendrobates
=== dendrobates is now known as dendro-afk
=== dendro-afk is now known as dendrobates
=== dendrobates is now known as dendro-afk
=== dendro-afk is now known as dendrobates
=== rsalveti` is now known as rsalveti_
=== TraumDieb is now known as DreamThief
=== rsalveti_ is now known as rsalveti