[05:36] Good morning [06:29] good morning [06:36] bonjour didrocks, comment vas-tu? [06:36] pitti: ça va bien, et toi? comment s'est passé le voyage de retour? [06:37] didrocks: tout etait bien, merci [06:38] c'était une voyage calme [06:38] et j'ai faire beaucoup de hacking :) [06:40] oui, j'ai lu ton post sur g+ ;) [08:25] hey larsu, uneventful flying back home? :) [08:36] didrocks, yup, all 40 minutes of it ;) And for you? [08:37] larsu: all good for me, no surprise ;) [08:56] I think I left my heart in San Francisco... and my body clock in Copenhagen :P [08:58] duflu: take it easy until you can get used to your timezone again :) [09:00] hey desktopers [09:02] ;p [09:06] heya [09:06] hey seb128, sil2100, mlankhorst! [09:07] Morning everyone! [09:07] my body clock is slowly shifting towards amsterdam again, waking up later than before :p [09:07] good morning all [09:07] hey didrocks sil2100 mlankhorst BigWhale [09:07] waking up at 7 was fun while it lasted [09:07] hey seb [09:08] So, are there any guidelines about unit testing GTK? [09:08] hey [09:08] hey Laney [09:14] hey desktop team [09:14] * Laney saw Skyfall yesterday... some creative computing in there [09:14] * Laney strips off the packet headers to decypher the GPS trace [09:14] hey chrisccoulson, Laney! [09:17] hey chrisccoulson, Laney, how are you? [09:17] hi didrocks, Laney, seb128 [09:18] seb128, yeah, not too bad thanks. did you get back ok? [09:18] good thank you [09:18] going through a forced mobile phone detox ;-) [09:19] chrisccoulson, going back was fine, I was home at 11am on saturday, got some ubuflu though :-( [09:19] Laney, you broke your phone? [09:19] yeah I did it after the UDS party [09:19] ouch [09:19] Laney: it wasn't exactly christmas for creative computing =) but I liked that it was mostly in London & Scotland. [09:20] waiting for the nexus 4 to come out [09:20] xnox: yeah, the locations were good [09:21] seb128, ah, that sucks :( [09:21] i've not been too bad since i got back. and i biked for more than 100km since friday too :) [09:21] you crazy man ;-) [09:22] training for the tour de france next year? ;-) [09:22] heh [09:23] hmmmm, all of my daily builds failed last week because my PPA's all ran out of space :/ [09:23] https://launchpad.net/~ubuntu-mozilla-daily/+archive/ppa/+packages [09:23] you are fighting with Sweetshark on disk space? ;-) [09:23] not sure where more than 20GB is going there :/ [09:24] and it seems a coincidence that they've all ran out of space at the same time [09:24] all 3 PPA's :/ [09:25] launchpad bug? ;) [09:26] oh yes, it's full of old builds: http://ppa.launchpad.net/ubuntu-mozilla-daily/thunderbird-aurora/ubuntu/pool/main/t/thunderbird/ [09:28] chrisccoulson: old builds... you mean like 5 days old? =) [09:29] xnox, daily builds from 24th october ;) [09:29] hey seb128, how are you? [09:29] hey Laney, morning chrisccoulson [09:29] hi pitti, how are you? [09:29] pitti, hey, I'm ok, got some ubuflu though ... you? [09:30] chrisccoulson: quite fine, thanks! [09:31] seb128: je suis desolé à entendre que [09:32] xnox, it's bug 1071562 (from #launchpad) [09:32] Launchpad bug 1071562 in Launchpad itself "UEFI signing failures cause binaries to be republished continuously" [Critical,Triaged] https://launchpad.net/bugs/1071562 [09:44] on my phone connection as it seems my ISP provider decided to stop working… [10:09] desrt: hey Ryan, how are you? got a minute today? [10:19] hmm, were do I ask for a fresh canonicaladmin password again? [10:19] I think #hr internal [10:20] Laney: thx === ChipAway is now known as Chipaca === MacSlow is now known as MacSlow|lunch [12:29] Hi all. [12:29] Just need to know how to integrates pidgin in unity notification area : any idea a? [12:33] Anyone there ? [12:36] Alpic: I think the broken integration is bug 1040259 [12:36] Launchpad bug 1040259 in skype-wrapper "FFE: libmessaging-menu transitions for quantal" [High,In progress] https://launchpad.net/bugs/1040259 [12:37] thanks for the tip [13:03] seb128, we lost the pkexec patch from gnome-system-log again? :/ [13:05] pitti: hi [13:05] jbicha: pong [13:05] hey desrt, how are you? had a safe trip? [13:05] mhr3, seems like somebody re-enabled that yes :-( [13:06] pitti: yup. nice and uneventful. :) [13:06] seb128: greetings [13:06] desrt, hey, how are you? [13:06] everyone else back home safe or still some lingerers behind? [13:06] desrt: the best kind :) [13:06] seb128, how come debian is fine with pkexec? did they patch it or glib or something? [13:06] desrt: I wondered if you could have a look at https://bugzilla.gnome.org/show_bug.cgi?id=686149 and check it's what you meant [13:06] Gnome bug 686149 in general "fix class initialisation" [Normal,Assigned] [13:06] seb128: very happy! i'm still riding the happy jetlag that lets me wake up at 8am and feel like it's past noon :) [13:07] lol [13:07] it _is_ past noon! :-) === attente_zzz is now known as attente [13:07] attente: hey. what's up? back home? [13:07] mhr3, they are not, they "fixed" that recently: http://packages.qa.debian.org/g/gnome-system-log/news/20121012T153227Z.html [13:07] desrt: yes :) [13:08] attente: ĉhu venos vi al esperanto vespere? [13:08] desrt: i didn't practice any [13:09] desrt: better i just skip [13:09] nenioj levionoj? [13:09] *lec [13:09] hey attente, had a good flight back? [13:09] looks like i am out of practice as well :) [13:10] seb128: i'm alive, that's enough for me :) [13:11] hehe [13:12] seb128, hmm, and the reporter lies about StartupNotify, it doesn't fix it :P [13:15] didrocks: I noticed that on the desktop-r-2d-test-environments I got a workitem "fix ubiquity-frontend-gtk to not pull in metacity" [13:16] didrocks: while in the "foundations-r-arm-ubiquity" I got a workitem "fix ubiquity-frontend-gtk to optionally use compiz/unity instead of metacity" [13:17] didrocks: Is that the same thing or do you want ubiquity to work with lxce window manager? === MacSlow|lunch is now known as MacSlow [13:26] xnox: it's the same one I would say :) [13:27] xnox: so pick the blueprint your prefer [13:27] didrocks: hmm.... I'm picking your blueprint at it has higher priority and I don't want this work item to sink. [13:27] xnox: thanks! [13:37] desrt: hey, I think we need to patch yelp to convert help:gnome-help URI prefixes to help:ubuntu-help when running Unity [13:38] I expect that somebody good at string manipulation could do that in a few minutes but it would take me a whole lot longer to figure it out [13:38] instead of having patches all over various desktop components like nautilus? [13:38] http://bazaar.launchpad.net/~ubuntu-branches/ubuntu/raring/yelp/raring/view/head:/src/yelp-application.c#L379 [13:38] desrt: yes :) [13:38] jbicha: a worthy cause :) [13:38] jbicha: that's a pretty nasty place to be doing the tweak... [13:41] oh it seemed to be the right place but you're welcome to do it somewhere else [13:48] jbicha: it depends on what you want to do [13:48] do you want to redirect *all* uri requests from one place to the other? [13:49] or only the ones given on the commandline directly [13:49] what if, for example, the help of some application made a link to the gnome docs? [13:50] seb128, thanks for updating the descriptions of those indicator-messages SRU bugs [13:51] larsu, yw, thanks for providing the testcases ;-) [13:51] desrt, hey! had a good flight back? [13:51] seb128, np :) [13:51] desrt: wouldn't that be the same place? doesn't yelp interpret those the same way as commandline args? [13:53] I just want the help links to work; I'm fine with the command line working w/o being mangled but I didn't think it worked that way [13:55] larsu: yup. you? [13:55] jbicha: the commandline arg thing is only for commandline args... [13:56] i doubt that yelp reinvokes itself on the commandline when clicking a link in a document [13:56] desrt, yeah. Had a bit of an ubuflu this weekend, though [13:56] desrt: so the main point of https://bugzilla.gnome.org/show_bug.cgi?id=686149 was to not add any interfaces after g_type_class_ref(), right? the patch does that now, I believe; or was there more to it? [13:56] Gnome bug 686149 in general "fix class initialisation" [Normal,Assigned] [13:56] hey larsu [13:56] larsu: urgh, ubuflu again? gute Besserung! [13:56] pitti: sorry. under a pile of bugs/patches/reviews at the moment :) [13:56] hi pitti! Did you enjoy Copenhagen? [13:56] pitti: yours is next on the list :) [13:56] desrt: ack, not that urgent [13:57] larsu: yes, I did! http://www.piware.de/fotos/UDS-Kopenhagen-Oct2012/ [13:57] desrt: oh, so yelp_application_new_window is what we want? [13:57] jbicha: probably even deeper... [13:57] jbicha: i'd talk to shaunm [13:57] it may even be that you need some kind of webkit hook. i don't know how it's implemented. [13:58] ok, I'll ask him, I think my main problem is that I didn't know how to properly sed a string [13:58] i can help with that part [13:58] so we have like gnome-help:foo and we want ubuntu-help:foo, right? [13:58] pitti, you saw the mermaid! Awesome, isn't it ;) [13:59] * larsu likes the picture of seb128 best [13:59] larsu: well, I was a bit underwhelmed by that one :) [13:59] and let's say we have the string in an input string 'instr' and want to produce a new string 'outstr' [13:59] so first, the check: [13:59] if (g_str_has_prefix (instr, "gnome-help:")) [13:59] larsu: https://plus.google.com/107564545827215425270/posts/CHKYc34MGt7 :) [13:59] ya that one [13:59] in that case we want to drop the first 11 characters ("gnome-help:") and concatenate the remainder with our own prefix, so: [14:00] outstr = g_strconcat ("ubuntu-help:", instr + 11, NULL); [14:00] yeah I kind of figured out the first line, it's the next that I wasn't sure about [14:00] else [14:00] outstr = g_strdup (instr); [14:00] ^^ straight copy [14:00] so either way we end up with a freshly-allocated 'outstr' that contains the new value [14:00] desrt: ok cool [14:00] it's a bit of a point of style in glib-ish circles not to do conditional memory allocation [14:01] ie: even though it's a waste to allocate the exact copy, we do it anyway for consistency with the other case [14:01] because it helps avoid leaks caused by weird logic bugs [14:01] it's easier to just always free.... [14:06] mterry: I tried to start on the System Settings Appearance panel but I didn't know what I was doing so it doesn't really work [14:08] jbicha, OK. You mean to make it a separate panel from Background? [14:08] yeah; maybe it doesn't actually need to a separate panel [14:09] hey mterry, had a good flight back? [14:10] seb128, yeah :) [14:10] jbicha, why not? [14:10] * kenvandine waves [14:10] kenvandine, hi! [14:10] kenvandine, hey, how are you? [14:10] tired :) [14:10] and you? [14:10] good, though ubuflued slightly [14:10] ugh [14:11] hopefully that passes quickly :) [14:11] it had to happen, wouldn't be UDS without that :p [14:11] indeed [14:11] mterry: just because I couldn't get it to work and maybe it would be simpler to just have a separate .ui and .desktop and add some lines of code to the background panel [14:12] seb128: is it common to catch ubuflu? [14:12] attente: if you didn't catch it, you're lucky :) [14:12] attente, yeah, it happens to a bunch of people at every UDS [14:13] where bunch is a non trivial % of the team usually [14:13] jbicha, eh, they are so different I think that if we can get separate panels to work, it would make sense. [14:13] heh [14:14] seb128, did you have an opinion on where the Ubuntu Appearances panel should live once it's separated? Still in g-c-c? [14:14] * desrt managed to avoid it this time [14:15] mterry, no strong opinion but I suppose it would be easier to maintain as a standalone component, would it only be for having its own vcs commit logs etc [14:15] seb128, sure [14:17] desrt: it's all because you had an awesome roomate :) [14:18] :D [14:19] haha [14:19] didrocks, did you escape the ubuflu too? [14:20] well I think there was not a lot of flu around [14:20] i think i did... i just had alergies while i was there [14:20] kenvandine: yeah :) I think that's because I had a 4 week flu just a month before UDS. So strong body :) [14:20] which are gon enow [14:20] but pitti came with the flu and I was rooming with him so... [14:20] well, classic UDS story ;-) [14:20] * pitti hugs seb128 [14:20] je suis desolé [14:20] * seb128 hugs pitti back [14:21] seb128: pitti: stop hugging! that's how it started :) [14:21] hehe [14:21] pitti, il n'y a pas de problème, maintenant ou plus tard ... comme ça c'est fait pour la saison [14:21] IRC hugs are BzgA approved! [14:21] héhé :) [14:22] we definitely need to quarantine dholbach during flu season === m_conley_away is now known as m_conley [14:41] My secret to avoiding Summit-disease was to be near the path of an enormous hurricane, get allergies from all the stirred-up airborne pollen and fungal spores, and be so sniffly and paranoid about getting other people sick at UDS that I washed my hands every hour. So, now allergies are gone, and no "flu". [14:41] I'll plan a hurricane every time, if I can. [14:44] qengho, :) [14:44] qengho, i got on the plane in denmark sniffling... like i had been for nearly two weeks [14:44] and got off the plane clear [14:44] no more sniffles :) [14:45] You may be alergic to umlauts, kenvandine. Consult a doctor. [14:45] hehe [15:27] hello, where does one file bugs regarding plymouth? LP is telling me that there really isn't an LP project for it [15:30] pitti: can you explain this unknown signal thing to me? [15:31] pitti: i guess it's a _good_ thing that this warning is now emitted? [15:31] the testcase appears to be engaging in some intentionally-incorrect behaviour [15:31] hmmm, my pandaboard feels like it's going to catch fire! [15:32] achiang: https://launchpad.net/ubuntu/+source/plymouth/+filebug ? [15:32] achiang: upstream isn't using Launchpad so you won't find a launchpad.net/plymouth project [15:32] chrisccoulson: firefox build? :) [15:32] desrt, heh, how did you guess? [15:33] chrisccoulson: because i know that you're not insane enough to try libreoffice? [15:33] stgraber: thanks [15:39] Ooo, sexy pandaboard. Some madman suggested I try to set up this Nexus 7 as a build machine. [15:39] qengho: nexus 7 makes a nice build machine [15:39] There he is! [15:40] beats the pants off of a pandaboard in terms of performance [15:41] Is cross-compliation such a mess that this even makes sense? [15:42] cross-compilation sucks [15:42] More than slow hardware does? [15:42] no. [15:43] there are all kinds of problems with build systems that try to compile and then use tools [15:43] and even larger problems with things like gobject-introspection which need to open the just-built library in order to poke around its insides [15:43] doesn't work so well when the library that you just built won't run on the machine that you're on [15:43] hi [15:44] any idea what's wrong on a package that complains about missing m4 dir when running aclocal in build stage? [15:44] creating the m4 directory makes it build ok, but I guess that's not the correct fix :) [15:44] rodrigo_: 'mkdir m4' ? [15:44] rodrigo_: i have 'mkdir -p m4' in some of my autogen scripts... [15:44] yeah, but I don't want to add that to debian/rules [15:44] *shrug* [15:45] building from git or a tarball? [15:45] tarball [15:45] a package [15:45] weird. [15:45] and it didn't dist any m4 macros? [15:45] locally creating the m4 dir makes it build ok [15:45] but launchpad then fails with that 'missing m4' error [15:45] i guess the answer is that you should no attempt to run aclocal on a tarball.... [15:45] desrt, I guess that's it, checking [15:46] ok, so it's autogen.sh that creates the empty m4 dir, so I guess debian/rules should be running autogen, not configure [15:47] configure is not the problem.... [15:47] aclocal is [15:48] there is one thing supported after you extact a tarball: ./configure [15:48] re [15:48] rodrigo_: are you running autoreconf, or just aclocal? [15:48] if that's not working, file a bug upstream because their package is broken [15:48] dobey, dh_autoreconf [15:48] desrt: I'm not sure why the warning wasn't shown before [15:48] dh_autoreconf on a tarball is somewhat wrong-headed.... [15:48] desrt, well, it might be my fault, I am packaging a new upstream release [15:48] desrt: I stepped through with it in gdb, and it's running through the exact same code [15:48] the previous package should work, it's in quantal [15:49] pitti: i suppose you consider it to be a good thing that we now get the warning, though [15:49] desrt: you always get the python exception, but an additional warning doesn't hurt [15:50] ah. i see. [15:50] okay [15:50] pitti: what's the easiest way for me to exercise this code? [15:50] run the pygobject testsuite? [15:50] desrt: git am, and make check, yes [15:50] k. [15:50] I hope it still applies, we have made quite some changes today [15:50] * pitti checks [15:50] i'm adding my warning to glib to make sure that it breaks [15:51] then i'm applying your patch and making sure it unbreaks :) [15:51] ah, nice [15:51] yep, still applies [15:52] hm. it's not breaking. [15:52] the tests fail on glib warnings (except those where we explicitly expect and thus ignore them, but there's only a few) [15:52] is it possible that you lack a testcase that implements interfaces? [15:52] desrt: you mean it doesn't trigger your warning? [15:53] explicitly in tests/test_interface.py, and I'd think implicitly all over the place [15:53] desrt: when I removed either of the pyg_type_add_interfaces() calls, it failed all over the place [15:53] fair enough [15:54] either i'm loading the wrong libgobject or my warning is not working [15:54] you are building/trying in jhbuid? [15:54] ish [15:55] strace says it's getting the right one [15:55] if you build pygobject on quantal, you'll get some failures [15:55] lemme poke around a bit more [15:59] ahah. it's actually getting the wrong one. [15:59] hrmph. [15:59] * desrt wonders if library paths get hardcoded in typelibs or something [15:59] hm, I don't think so [15:59] I remember hacking on glib while running sth. like LD_LIBRARY_PATH=~/upstream/glib/.libs python3 -c '...' [16:00] ya. my LD_LIBRARY_PATH is fine... [16:00] but "make check" sets its own library path [16:00] seems to be favouring a different flavour of the library, though [16:00] tests/Makefile.am: LD_LIBRARY_PATH=$(builddir)/.libs:$$LD_LIBRARY_PATH \ [16:00] this is meant to preserve an already existing one [16:00] i get the impression that this is libtool's fault [16:01] yup! [16:01] * desrt has such intense hatred for rpath... [16:01] urgh, rpath? where? [16:01] everywhere, if you're using libtool [16:02] OMFGreally? [16:02] as soon as you do --prefix=/something/outside/of/usr it sets the rpath on everything it installs [16:02] so you get libfoo depending on a rpath'd version of libgobject in a particular place [16:02] overriding your LD_LIBRARY_PATH [16:02] utter utter braindamage [16:02] $ chrpath -l ./gi/_gi.so [16:02] ./gi/_gi.so: RPATH=/home/martin/upstream/pygobject/gi/_glib/.libs:/home/martin-scratch/gnome/lib64 [16:02] eek [16:03] so presumably when I did above tests, I just did ./configure, not a jhbuild build [16:03] i desperately hate .la files [16:03] like, very very badly [16:03] but it's nowhere near the level of my hatred for the fact that libtool sets rpath [16:06] pitti: your patch seems to do the trick [16:06] \o/ [16:07] let me upload my gobject patch... [16:07] and see if you can come to the same conclusions [16:07] yep, will do [16:08] desrt: to https://bugzilla.gnome.org/show_bug.cgi?id=687659 ? [16:08] Gnome bug 687659 in gobject "drop support for adding interfaces after class_init" [Normal,New] [16:08] yup [16:08] attaching a patch there in just a moment [16:09] okay. patch is attached. [16:10] * pitti runs jhbuild buildone glib [16:10] pitti: before your pygobject patch I see: [16:10] (process:22791): GLib-GObject-WARNING **: attempting to add an interface (TestInterface) to class (test_interface+MyObject) after class_init [16:10] Trace/breakpoint trap [16:10] after your patch, it works again [16:11] confirmed, it's unhappy with the glib patch applied [16:12] Ran 700 tests in 9.854s [16:12] OK (expected failures=1) [16:12] and that's with my patch [16:12] * pitti ^5s desrt [16:12] sweet :) [16:12] i've commented about the positive test results on the bug [16:12] it's up to you if you want to go ahead with pushing your patch [16:12] in its own right it's a nice patch [16:13] it will take a little more discussion on the glib side before i can push my patch.... [16:13] the glib testsuite itself is in the foul, for example [16:13] desrt: I pushed it [16:13] awesome [16:13] I like it a lot more that way [16:13] ya. me too. [16:13] while I'm in the mood of simplifying code :) [16:13] even without the glib changes it makes more sense [16:14] thanks a bunch [16:14] and another firefox tab gone :) [16:14] I had some 5 bz ones yesterday, and now it's down to 1 [16:14] ya. this morning is pretty productive for me too :) [16:17] didrocks, so did you want inline branches to be native or split? /me is not partial, but wants to be consistent [16:19] mterry: we discussed it a little bit earlier with ken, split seems to be the best IMHO [16:19] didrocks, k [16:19] thanks :) [16:19] * didrocks is fighting jenkins to have fork and join jobs workflow [16:28] didrocks, so this means no need to switch to format 3.0? [16:29] mterry: not really needed, indeed, this will make the SRU more readable to keep 1.0 if we need to cherry-pick at some point [16:40] GunnarHj: did you see my comment for bug 1035219? [16:40] Launchpad bug 1035219 in Baltix "In System Settings preference tool/keyboard layouts page automaticaly wrong language selectedGNOME" [High,In progress] https://launchpad.net/bugs/1035219 [16:42] while looking through all the gnome-control-center patches to get ready for 3.6 I stumbled across the 52 patch which needs work [16:44] jbicha, if you do g-c-c work please commit to the vcs so we don't dup work, I already reviewed the whole patch stack and took notes, I was going to look at continuing on that soon (I commited by g-s-d 3.6 work to the vcs) [16:46] seb128: my g-c-c work is already committed [16:47] jbicha, great ;-) [16:47] jbicha, do you know if anyone is looking at getting realmd uploaded btw? [16:48] oh do we need that? [16:50] I think we're pretty close on having g-c-c/g-s-d 3.6 ready, we need to figure out the Appearance panel & 2 g-c-c patches need updating [16:50] yes we need that, the user panel depends on it [16:50] we also need the new ibus [16:50] there's still the keyboard/ibus indicator rework to do but I don't think that should be a blocker at this point [16:51] I would like the ibus indicator to be updated if possible [16:51] the g-s-d one will be harder [16:56] seb128: how do you want to handle that? we could push what we have to the desktop ppa so that it can be worked on [16:56] jbicha, that works for me [16:57] or we could push to raring-proposed and as long as ibus has 1 nbs package, it won't migrate to raring [16:57] jbicha, somebody needs to have a proper look to ibus, I plan to do that in the next weeks if nobody beats me to it [16:57] jbicha, proposed is not meant for that, it's for things ready to be uploaded to the archive [16:57] let's use the ppa [16:57] seb128: Debian experimental has already started, the initial testing I did in an experimental PPA last month seems to work [16:58] ok, I'll push to the PPA later today [16:58] well, I don't think Debian ported our ibus indicator and tested if it still works under unity ;-) [16:58] thanks [17:05] jbicha, you had rb 2.98 ready for upload right? you should probably push it to raring [17:24] heh, 90.2degC! [17:24] will the panda die before the build finishes? [17:27] chrisccoulson: you should get the new samsung chromebook. dual A15s + 2GB of memory [17:27] only $250 USD [17:29] jbicha: Yes, I saw your bug 1035219 comment. Sounds as if you found a way to limit the importance of the bug for standard Ubuntu users, which is good, of course. [17:29] Launchpad bug 1035219 in Baltix "In System Settings preference tool/keyboard layouts page automaticaly wrong language selectedGNOME" [High,In progress] https://launchpad.net/bugs/1035219 [17:29] jbicha: As regards the g-c-c patch 52, please note that it's there for the User Accounts capplet only. Of course it will need work now when Region and Language is going to be used in Ubuntu. [17:37] pitti: ping [17:38] so after updating this morning unity is failing to start for me, compiz appeared to be running but has finally crashed [17:38] seb128: i guess you're watching on #g-h? [17:38] desrt, I'm subscribed to the distributor list, saw it there [17:39] cool [17:39] desrt, no surprise there [17:39] seb128: nope. [17:39] and good that they're discussing it so openly and making decisions early and firmly [17:39] though I didn't think they would drop the g-s-d "render the background" code [17:40]