[02:35] robert_ancell: do you know why/how a packaging branch gets out of date? I thought it was all automatic [02:35] mfisch: Bugs in the package importer. [02:36] for those packages do we fallback to the old style of just attaching a debdiff to the bug? [02:39] Yup. [02:42] RAOF: one more question, when we hold patches for a package and then they're no longer needed because debian or upstream takes them, do we just sync from debian and lose the changelog entries? [02:42] Yup. [02:42] If we no longer need the divergence from Debian, we sync and lose any Ubuntu changelog. [02:43] okay cool [02:43] theres certainly no reason to hold a divergence just for a changelog [02:44] findutils, for example, took all our changes in debian/experimental [02:49] Sync away! [02:50] I thought we sync'd from unstable [02:51] We can sync from any Debian series; the autosync syncs from unstable (or, sometimes, testing). [02:52] okay let me write up a sync request later [02:52] You're aware of “requestsync”? [02:54] yep [02:54] I'm in the middle of another merge right now, but I will do it later === VD is now known as Guest67615 === zoktar_ is now known as zoktar [03:44] so, i have an issue with online-accounts. They won't hook in with the lenses [03:47] swiss: most of the team is probably in bed at this hour unfortunatel [03:47] mfisch: well, i've switched over to gnome3 for now, so it's not urgent [03:47] and i don't part, i'm hooked in through my server [03:47] swiss: did you ever file a bug? [03:48] i've seen enough similar bugs that i was wondering if i had done something stupid first [03:48] and I don't like submitting incomplete data [03:49] strace data will be better than 99% of bugs filed [03:52] yeah, but i can give better data if needed. I want someone to be able to look at the bug, and not have to ask any more questions [03:54] understand [03:54] more people will be around tomorrow during uds hours [03:54] kk [05:05] Good morning [05:07] mornin [05:16] pitti, hi [05:17] apport is not working for me atm, when trying to process crash reports [05:19] it doesnt seem to be crashing as stuff, just silently exiting when it should be uploading [05:31] darkxst: it uploads to errors.u.c. [05:31] we haven't enabled LP crash reports yet [05:31] eventually we only want it to report to errors.u.c. [05:31] pitti, oh I meant via ubuntu-bug [05:32] bug reports should go to LP; crash reports not === Ursinha is now known as Ursinha-afk [05:36] ok === Ursinha-afk is now known as Ursinha [06:09] pitti, good morning :) [06:13] robru: you are making my life hard btw! :) [06:14] * didrocks wonders if he should put "udaily" [06:23] didrocks, how can you say that?? this problem was caused by security team uploading direct to distro! [06:23] robru: heh, I know, just kidding and trying to make you feel guilty :p [06:24] robru: I didn't want to support those 42ubuntu1 TBH ;) [06:24] robru: so as I can do -> 42daily [06:24] because d < u [06:24] I think I'll be forced to do: [06:24] 42udaily [06:24] until we bump to 43 [06:24] for instance [06:25] robru: making sense? (yeah for ugly version :/) [06:25] didrocks, it makes sense, but I think we're actually ok right now because the offending versions are 2.2ubuntu1 and we're already at 2.4.x [06:26] robru: yeah, I added the support in case we don't bump though [06:26] we never know how ugly things like that can happen apparently :p [06:26] didrocks, yeah. maybe "zdaily" or something ;-) [06:27] robru: heh, like "you can get higher than I": zzzzzzzzzdaily :p [06:28] oh sure, force us to use Zdaily to overwrite your packages.. :-P [06:28] ;) [06:28] sarnold, if you are making daily packages that are more frequent than ours.... [06:31] * didrocks just needs to create a test for the download part now [06:32] robru: haha :) [06:38] didrocks: morn [06:39] good morning Mirv! [06:43] robru: http://10.97.0.1:8080/view/cu2d/view/Head/view/WebApps/job/cu2d-webapp-head-1.1prepare-unity-webapps-facebookapps/ [06:43] robru: so in prod and working :) [06:44] great [06:44] didrocks, did you run the whole stack? [06:45] robru: I let the nightly daily happening, so only rerun those 2 (the whole stack ran a couple of hours ago) [06:45] robru: http://10.97.0.1:8080/view/cu2d/view/Head/view/WebApps/job/cu2d-webapp-head-2.1build/92/console [06:45] so if everything builds fine, we'll get it :) [06:46] didrocks, where can I find the build log? (as opposed to the *source* package build log, which is what I'm seeing in jenkins). [06:46] robru: in the ppa [06:46] robru: https://launchpad.net/~ubuntu-unity/+archive/daily-build [06:46] like https://launchpad.net/~ubuntu-unity/+archive/daily-build/+build/4582836 ;) [06:47] didrocks, ok, great. that's where we have to look to see what the lint script complains about. [06:48] robru: exactly. Basically, the build really happens in the ppa (to have the same configuration than distro) [06:48] I guess at this point we should review the lint script and decide which problems should be fatal (currently it only warns and never fails the build regardless of what problems it finds) [06:48] robru: what daily release is doing is preparing the source package, checking everything, monitoring the build, the integration tests and then eventually publishing [06:48] yeah [06:49] robru: right, the lint script should abort the package build IMHO if we get "this is fatal" [06:49] didrocks, yeah, I agree, but I haven't had time to figure out what is really fatal and what isn't ;-) [06:49] so it only warns right now [06:49] oh, I see a problem though [06:49] make[1]: /usr/share/unity-webapps/tools/po2json.py: Command not found [06:50] that's not so good... that's provided by unity-webapps-dev [06:50] good morning [06:50] didrocks, and without it, the built package actually is broken in chromium... [06:50] hi jibel [06:51] salut jibel [06:51] Hey robru, didrocks [06:51] robru: is this in https://launchpadlibrarian.net/139996744/buildlog_ubuntu-saucy-i386.unity-webapps-facebookapps_2.4daily13.05.16-0ubuntu1_UPLOADING.txt.gz? I grep for po2json and find nothing [06:52] didrocks, [06:52] $ dpkg -S /usr/share/unity-webapps/tools/po2json.py [06:52] unity-webapps-dev: /usr/share/unity-webapps/tools/po2json.py [06:53] robru: I mean, I don't see your make error on the above log [06:53] didrocks, oh, right, that one is good actually [06:54] robru: can it be that you invoke directly the script and you don't have python installed in the other one? [06:55] didrocks, you should grep for PLACEHOLDER. if you see sed replacing the PLACEHOLDER with emptiness, then that's very, very bad. in that log you linked, the PLACEHOLDER gets replaced with '%7B%7D', that's fine (just means no translation data is available for that app). some apps with translations will have a huge blob there instead. [06:55] didrocks, nope, can't be, because they all have identical debian/rules (provided by -dev package itself), and that is where po2json.py is called from [06:55] robru: do you have a link to the log having that failing? [06:56] didrocks, https://launchpadlibrarian.net/139992170/buildlog_ubuntu-saucy-i386.unity-webapps-facebookmessenger_2.4.12daily13.05.16-0ubuntu1_UPLOADING.txt.gz here for one [06:57] robru: python is not installed in this chroot [06:57] robru: the -dev package should dep on python I guess [06:58] didrocks, it does [06:58] so that's why it can't execute it, with vm, it doesn't work as good as it can be :-) [06:58] didrocks, oh, there was a bug in debian/control, missing comma on previous line made it not pick up the python deb. fixed in trunk. I guess rerun the stack and it should be fine. [06:59] robru: this is in which package? [06:59] the -dev one? [06:59] well, webapps-applications [06:59] didrocks, yes [07:00] didrocks, yes, there is a bugfix in lp;webapps-applications that fixes the deps for unity-webapps-dev to properly pull in python that should fix these package builds. [07:00] robru: ok, it seems it's part of this daily release [07:00] robru: however, we need to rebuild what didn't work on the packages depending on it [07:01] yeah [07:01] robru: fortunately, you have those copyright issues still to fix, right? :) [07:01] didrocks, just checking which packages failed right now [07:01] robru: great! :) [07:01] didrocks, what? no, I fixed that all 12 hours ago :-P [07:01] robru: you are too fast, ok, so a dummy commit I guess will be needed (like a changelog build to force the rebuild) :p [07:02] didrocks, actually most of them are failing... [07:03] didrocks, so far I count 21 failures and I'm not done looking yet [07:03] didrocks, no way you can just force a rebuild? I really have to make 41 trivial trunk commits? [07:03] robru: it's the same way than for distro [07:03] robru: you need to bump the debian/changelog [07:03] didrocks, ok, I am just going to go through and version bump then. [07:03] robru: sounds good [07:04] robru: I'm waiting for this run to finish first [07:04] didrocks, actually but wait, the error message isn't that python isn't found, it's that the command itself is not found... [07:04] robru: yeah, IIRC, that's what happens if you run a python command directly and python isn't installed [07:04] robru: let me confirm in a terminal, one sec [07:05] $ ./foo.py [07:05] bash: ./foo.py: /usr/bin/python4: bad interpreter: No such file or directory [07:05] $ make build [07:05] ./foo.py [07:05] make: ./foo.py: Command not found [07:05] make: *** [build] Error 127 [07:06] so actually yeah, bad interpreter then [07:06] different error from make than from bash [07:06] ok, so it should work with the right deps, which are fixed in trunk. [07:06] robru: right, that's what I remember :) [07:06] robru: confirmed here as well [07:06] the message is really not clear [07:06] and I already lost some time for this :p [07:06] hence I remember now ;) [07:07] didrocks, do you know a way to make dch do a version bump non-interactively? like what options can I pass to it to force version 2.4.16 with message 'Version bump.' [07:08] robru: dch -v2.4.16-0ubuntu1 "Version bump for…" [07:08] robru: not sure if your setup is good though, do for one and check you have "UNRELEASED" [07:08] didrocks, ok, hang on [07:10] +unity-webapps-amazon (2.4.16-0ubuntu1) UNRELEASED; urgency=low [07:10] + [07:10] + * Version bump. [07:10] + [07:10] + -- Robert Bruce Park Thu, 16 May 2013 00:10:27 -0700 [07:10] + [07:10] looks good right? [07:10] robru: that looks gorgious! [07:11] robru: well maybe be a little bit more explcit [07:11] didrocks, you've been looking at the codes for too long ;-) [07:11] didrocks, time to get some fresh air today ;-) [07:11] robru: like Version bump to build against latest webapps-applications-dev [07:12] robru: ahah, it's raining though :/ [07:12] robru: now, I have to switch to otto :p [07:13] robru: btw: http://10.97.0.1:8080/view/cu2d/view/Head/view/WebApps/job/cu2d-webapp-head-3.0publish/ [07:13] manual publication mode (I guess because of debian/copyright changes) [07:13] I don't publish yet, right? Let's wait for the rebuild [07:13] didrocks, yeah, definitely wait [07:14] ah, as well, let me add cnn-news [07:14] (well deploy the branch you proposed) [07:14] deploying* [07:15] didrocks, ok, I just started a script to push all the branches, will be ready in ~10min [07:15] perfect :) [07:15] (bzr uploads are slow...) [07:15] robru: don't tell me, I'm doing the bzr config on the remote branch (remember this bugs so that lp-propose target the right branch? :p) [07:16] heh, ouch [07:17] didrocks, I can't count how many times this month I have types "for i in *; do cd $i; bzr commit -m "foobar" && bzr push; cd ..; done [07:17] robru: you typed? I thought it was a script and you didn't rely on Ctrl + R ;) [07:18] it's a script that I type into my shell prompt every time I need it ;-) [07:18] heh [07:18] but the cleaning you've done to the packaging is awesome [07:18] good work! [07:18] thank you! [07:18] also, nice to see we won't revert security fixes again :) [07:19] didrocks, well, probably the security team should be told to get their fixes accepted in trunk... [07:19] robru: well, in ubuntu, everyone has upload rights [07:19] I mean, I understand the urgency of getting things fixed in distro and pushed to users, but that one fix was just never committed to trunk, that's terrifying. [07:19] robru: so, you can't be sure people are committing to the vcs (if attached) [07:19] but yeah, this one should have been backported for sure [07:20] hey ricotz [07:20] bonjour didrocks [07:21] hello robru [07:21] didrocks, this is actually a really serious problem I think. There's this big mysterious "archive" that is somehow separate from the VCS... we should just drop the archive and declare "whatever is in trunk VCS *IS* the archive", so you want something in the archive you get it in trunk [07:21] pitti, hi [07:21] salut pitti! [07:22] robru: yeah, the archive should always be authorative, that was the taken decision when we started to have our Vcs (for desktop team or anything else) [07:22] pitti, i was hoping you have a moment for libreoffice ppa builder tweaking ;), but i hope they work out already [07:22] robru: but I agree with you that Vcs should be the main source :) However, it means, we would need to fix perms everywhere [07:23] robru: as it blurries the line between distro maintainers and upstream devs [07:23] not an easy topic :) [07:23] ricotz: sure [07:24] pitti, so i might ping you later if they are failing on the current builders [07:24] didrocks, no no, not saying we merge upstream trunks with distro vcs, definitely there needs separation there. just that the archive itself needs to be in a public VCS that can be seen and have MP's proposed against in a meaningful way. not this "oh, here's an MP, please merge it, build it, upload it, sponsor it..." [07:25] robru: yeah, some people tried that 3 years ago, but it never reached to a point it was working top notch [07:26] robru: those are the lp:ubuntu/ branch [07:26] didrocks, yeah, backwards approach I think. lp:ubuntu/* tries to mirror what's in the archive. I think it needs to *BE* the archive. [07:27] quite complicated as 75% of the packages are coming from Debian [07:27] so from the archive :) [07:27] bah [07:27] didrocks, then debian needs to get itself hosted on launchpad ;-) [07:35] ahah :) [07:35] robru: want to start the discussion on debian ML? [07:36] didrocks, why do I have to do all the work? clearly this idea is brilliant and devs should be flocking to it ;-) [07:39] robru: a bit related, support for Debian package building in PPAs has been a very wished item for a long time, I wish someone would have time to implement that [07:39] robru: ahah, that's a manager's view :) [07:39] robru: well done :p [07:39] didrocks, great, time to promote me ;-) [07:39] robru: please, do a powerpoint with a plan :p [07:40] we don't have much perks for Debian devs to come to LP at the moment, including that plus lack of git support [07:40] didrocks, is there time to host a session about this tomorrow? ;-) [07:40] heh [07:40] * didrocks plans, set robru as leading the session [07:40] Mirv, yeah, I wouldn't mind being able to build debian debs in my PPAs, for sure [07:41] Mirv, what would "git support" look like? some kind of bridge to allow git tools to access bzr repos? or real git repos hosted on lp? [07:42] robru: either a very transparent bridge but probably real support so that one could easily merge from Debian packaging git:s to LP and vice versa [07:42] for example with Qt I'm now manually managing Debian git vs Ubuntu bzr [07:42] bah === tkamppeter_ is now known as tkamppeter [07:43] some teams like X team do have Ubuntu packaging in Debian git actually, but it wouldn't hurt that there wouldn't be that barrier between Debian and LP [07:44] Mirv, I had a real nasty issue recently just within LP, it wouldn't let me merge a +junk branch back into a real project branch even though it did share history. the +junk branch lost it's knowledge of it's heritage, and bzr couldn't merge them in spite of the fact that the commit histories were identical. had to manually copy files from one branch to another and re-commit. [07:45] frankly I prefer git, git would have seen the common commit history and merged happily ;-) [07:45] robru: I've gotten past such issues at times by pushing the same branch into another location. ie. first gotten "this branch cannot be merged into xxx", then after pushing into second location merge proposal has been possible [07:46] hmmm [07:48] robru: btw, can I rerun the whole stack? [07:49] didrocks, oh... well I didn't version bump libunity-webapps or anything, is that a problem? I did bump all of unity-webapps-* though, those definitely need to be rebuilt [07:49] robru: that's fine I guess, we already have the needed version in the ppa :) [07:49] didrocks, ok, so rerun it then ;-) [07:50] * didrocks click *the* button [07:50] clicks* [07:51] didrocks: sergiusens by the way asked about if all packages have been rebuilt in daily-build-next after the new Qt. it seems to me most Qt related packages, a couple haven't yet. [07:52] Mirv: they need a commit to be rebuilt, mind pushing something to them? (there is also something to clean up) [07:53] I found an awesome way how to crash unity/compiz! :) [07:54] didrocks: ok, I'll look at if some relevant package would use a push [07:54] BigWhale: :P [07:54] Have a program icon on the desktop. [07:55] Have a program icon on the desktop, drag it to Launcher, delete the icon on the desktop, move the icon in Launcher. [07:56] Trevinho: this is for you! ^ [07:56] BigWhale: a bug number (or looking for one, would be awesome!) [07:57] didrocks, I'll look into this now... :) [07:58] hey desktopers [07:58] salut seb128, ça va? [07:59] didrocks, lut, ouais, tjs gris et pluvieux mais ca va ... et toi, Lyon? [07:59] seb128: ça va bien! par contre, arrête de partager ta pluie, ça va, j'ai compris hein! :p [08:02] morning [08:03] didrocks, alright, you can get sun back tomorrow :p [08:03] Laney, good morning! [08:03] really? last time I checked, it wasn't really optimistic [08:04] * didrocks looks [08:04] well, far from stellar… [08:04] didrocks, it's a bit better tomorrow, really better starting next tuesday... [08:04] https://bugs.launchpad.net/ubuntu/+source/pidgin/+bug/1108056/comments/28 [08:04] Launchpad bug 1108056 in glib2.0 (Ubuntu) "fetching URLs freezes pidgin" [Undecided,Confirmed] [08:04] * Laney runs [08:04] Laney, isn't that the indicator-messages bug that we fixed yesterday? [08:05] the same as tb [08:05] likely :-) [08:05] wfm since at least [08:05] just thought it'd be worth checking [08:05] did pidgin break too for you? [08:05] nobody messaged me [08:05] let me give it a try [08:07] *yaaaaaaaaaaaaaaaaaaaaaaaaaaaaaawn*... 'night gents [08:07] robru, 'night [08:09] bonjour seb128 [08:10] pitti, salut, ça va ? [08:10] seb128: ça va bien, et toi? [08:11] pitti, ça va bien merci, toujours de la pluie ici par contre :/ [08:11] :( [08:11] ici aussi aujourd'hui [08:11] pitti, are you online on icq? [08:12] * seb128 is looking for somebody to im me to test that pidgin/libmessaging bug [08:12] seb128: no, I stopped using icq years ago, I got too much spam [08:12] but jabber doesn't behave [08:12] seb128: I can log in [08:12] pitti, it you could just send me a test message that would be nice [08:12] if* [08:12] hm, how do I add an ICQ account these days? [08:12] don't bother if you don't have it configured [08:12] I will wait a bit to have jabber back [08:13] ICQ still exists? [08:13] pitti, but if you need that at some point, you need to run empathy-accounts I think [08:13] ooh [08:13] that's indeed not the one that I get from the menu in empathy [08:13] you get the integrated online accounts by default [08:14] seb128, hello :) [08:14] it's just that icq doesn't have a telepathy connector [08:14] it's going through haze and libpidgin [08:14] people still use ICQ? [08:14] robert_ancell, hey, how are you? [08:14] robert_ancell, got my email? ;-) [08:14] seb128, hey, good. Thought I'd say hello since I'm around this time [08:14] robert_ancell, very few, I still have one friend using it regularly [08:14] yes... [08:15] robert_ancell, I'm glad IS figured out that versions was slow because you were DoSing the debian web viewvc UI :p [08:15] It didn't seem like that many requests... [08:15] they must have throttled us ages ago [08:15] the web ui sucks apparently [08:15] how did it only just come up now [08:15] seb128, did you see the MP from mfisch? It changed the Debian links [08:16] robert_ancell, no, I didn't, I guess I don't receive email for those [08:16] seb128, yeah, you might have to opt-in [08:16] I set all the ownerships to ~ubuntu-desktop but I think LP defaults to not notifying (we have the same problem in Mir) [08:16] seems like a feature [08:16] I wouldn't like to spam the whole team [08:17] seb128, I'd prefer opt-out because people miss things and it's not obvious how to opt-in [08:17] that's right [08:17] seb128: sorry, seems my ICQ account doesn't work any more, or I don't remember the password [08:18] pitti, don't worry, I will wait to have somebody showing up on jabber [08:18] seems like I've half the people I usually have there [08:18] for some reason I can remember my ICQ UIN despite not having used it for about 10 years [08:18] I guess some server side issue on jabber.org [08:18] seb128: I am in jabber [08:33] ah, jabber is finally listing more people [08:34] I can see a didrocks [08:34] I think it's google+ :) [08:34] didrocks, can you reply? [08:35] didrocks, need to test that bug [08:35] hm, jabber doesn't seem to work for me ATM either [08:35] interesting, I don't even see that you pinged me on any tab (in fact, I don't have a g+ page open) [08:35] didrocks, it's maybe your phone or tablet? [08:35] seb128: ok, please try to ping me [08:35] seb128: yeah, probably my phone [08:36] hum no… [08:36] seb128: just try, I have a tab opened now [08:36] bah, got disconnected, jabber.org sucks [08:36] ah, back [08:36] didrocks, pinged you [08:36] ah, now [08:36] (Pidgin:4551): GLib-GIO-ERROR **: g_menu_item_set_detailed_action: Detailed action name 'im:prpl-jabber:seb128%40jabber.org%2F:pitti%40jabber.org%2F06baa216' has invalid format [08:36] Laney, ^ [08:37] so still there [08:37] seb128: nothing… [08:37] didrocks, pitti: can you ping me again? [08:37] Laney, no, that was the old libmessaging-menu0 to confirm the bug [08:37] done [08:37] works [08:37] seseI'm trying to look for you first, not sure with the new interface :p [08:37] ah [08:37] Laney, it's fixed [08:37] great [08:38] seb128: pinged [08:38] didrocks, doesn't work [08:39] shrug [08:40] new g+ doesn't work well for me [08:40] the hangout sidebar contain a "connect to talk to your friends" label with a "connect" button [08:41] which opens a new win, which tells me "you are logged in now, close that win and refresh the page you were on" [08:41] but that doesn't work [08:41] still get the same button [08:41] so I can't log in [08:41] which means I can't create an hangout [08:41] will be fun for vUDS later... [08:42] tried in chromium? [08:45] Laney, of course it works in chromium (though I'm using a personal account and not a canonical one there and the login process seems different since the login button sends me on an ubuntu.com url for auth) [08:46] ah, so could be some SSO interaction problem [08:46] is anyone using g+ canonical's account under firefox herE? [08:46] pitti, ^? [08:46] let me try [08:46] or did you all go chromium? [08:47] seb128: no, I'm firmly firefox [08:47] hangouts work again, too [08:47] no, I just never use my C google account really [08:47] pitti, hangouts bar is working for you? [08:47] nope, 'Unable to sign in" [08:47] :-( [08:47] for UDS I'll use my "public" account [08:47] (there it works) [08:47] bah [08:47] I don't want UDS videos to be posted on my personal account [08:47] I try to keep the work one for work [08:48] yeah, seems broken [08:48] #is probably want to know [08:48] *great* [08:48] yeah, I'm telling them [08:50] same issue in chromium [09:10] robru: webapps published [10:04] pitti, hey, https://launchpadlibrarian.net/134293864/xwiimote_0.3%2B20120630-2_0.3%2B20120630-2ubuntu1.diff.gz ... did you send that to Debian or upstream? (not sure if that diff is Ubuntu specific for some reason?) [10:07] seb128: ah, not yet, thanks for reminding [10:07] pitti, yw [10:23] seb128: who should be approver for https://blueprints.launchpad.net/ubuntu/+spec/client-s-xorg-general ? [10:24] mlankhorst, jason or me, pick one ;-) [10:24] I'll select you for being in the same timezone [10:24] it's not you jason, it's me! [10:25] playing with fire there [10:25] :-) [10:30] it is kind of hard to talk with people in different timezones on a regular sleep schedule, though :) [10:31] give up on that schedule then :P [10:33] but then I wouldn't wake up at 7 any more [10:33] depends which 7 you mean :) [10:34] 07:00 [10:35] it'll be close to 7am in some timezone :-) === MacSlow is now known as MacSlow|lunch [11:17] didrocks, jibel: ah, so by default there are indeed only 8 loop devices pre-created, one needs to request new ones from /dev/loop-control [11:17] (https://lkml.org/lkml/2011/7/26/148) [11:18] pitti: ah thanks, I had a script for that, but better to be able to put in a configuration file :) [11:28] jibel, didrocks: I confirm that "lxc-wait -s STOPPED" triggers on a reboot [11:28]