[05:56] good morning! [06:14] Good morning [06:37] good morning desktopers :D [06:38] hey hikiko [06:38] my isp just gave me 10GB mobile internet for free until they fix the problem so that I can use my mobile as a hotspot until they fix the issue :) [06:38] \m/ [06:38] hi didrocks :) [06:42] good morning desktopers [06:42] hey hikiko, re didrocks [06:43] re seb128 [06:45] hi seb128 [07:21] bonjour seb128 [07:21] * pitti -> appointment, bbl [07:21] pitti, salut pitti, ça va ? [08:02] good morning! [08:03] hey larsu! [08:05] bonjour didrocks! [08:14] Buon giorno! [08:16] hey Trevinho [08:16] morning Trevinho [08:17] hey larsu, wie gehts? [08:17] hey Trevinho [08:17] morning seb128! good thanks - and you? [08:19] Hi larsu, didrocks, seb128 ;-) [08:27] hey all [08:35] hey darkxst [08:40] * Trevinho hates when can't connect to routers... [08:53] interesting those frameworks which keeps changing their icon path :p [08:54] (great that tests enable us to discover this) [09:00] morning [09:01] willcooke: bonjour [09:02] morning willcooke [09:03] hey willcooke Laney [09:03] hey ;-) [09:03] a pre hello hey from seb128! [09:03] :-) [09:03] i think you meant larsu [09:04] no, but it was within the minute you usually say hello [09:04] morning Laney :) [09:04] so I though it was good timing ;-) [09:04] seb128 is clever [09:05] a cunning young man [09:05] some would say too cunning [09:05] hi larsu! [09:05] haha [09:05] morning Laney [09:08] what's going on with autopkgtests, why is there so much backlog? [09:09] hey didrocks [09:09] * seb128 looks at pitti [09:10] one of the cloud regions is broken [09:10] so there are 10 less workers than normal [09:10] I see [09:11] need to to feed more hamsters [09:12] still doesn't seem to have gotten to most of the gtk tests though [09:12] that's been like 1.5 days now [09:13] which is a bit suspicious [09:13] yeah, some are in progress for ages [09:13] and the queue only shrunk by like 60 overnight on xenial/amd64 [09:14] I think one problem is that kde tests all do a rebuild from source [09:14] don't know if that explains all this [09:14] there were quite a lot on trusty too because of some kernel uploads [09:14] bleh [09:15] did you see my request yesterday for some promotions? [09:15] yeah, I was still around but not in full work mode [09:15] and I'm unsure I feel like promoting things without a MIR [09:15] is that "perl sources can be freely promoted" written somewhere? [09:16] it feels weird [09:16] like why do we need to bother for things like cmake rules or small python wrappers, but perl is ok? [09:17] ok, never mind, I'll ask someone else [09:17] thanks! [09:17] I'm tempted to just do it for the sake of avoiding work though :p [09:17] sorry, ping me again if you don't manage to get a reply for e.g infinity [09:17] I'm a bit reluctant but it's not going to take a lot to convince me ;-) [09:17] what is pulling those in now btw? [09:17] I missed the context [09:18] lintian [09:19] Hi all [09:19] Writing planning documents feels like going back 10 years in software development... [09:20] hey robert_ancell [09:20] hey robert_ancell! [09:20] what a nice surprise! [09:20] how are you? (out of the planning documents writting ;-) [09:20] hey robert_ancell! [09:21] what's the planning document for? [09:21] seb128, sore. I spent the evening putting up insulation in my house renovation. Wild times for me! [09:21] Laney, GNOME Software [09:22] ohai robert_ancell ! [09:22] seb128: +1 on being annoyed by those CI downtime without warnings (spent an hour this morning because all machines were on weird state) [09:26] * Laney tries didrocks appstream commit [09:27] no fear! :-) [09:28] is using a virtualenv best practice? [09:29] Laney: if you want to control exactly what deps are in, yeah [09:29] I got all the deps from the archive anyway [09:29] so it made me wonder a bit [09:29] (and if you use --site-packages, it's just a portable way to have your binaries installed in a non root dir) [09:29] so I would say it's still the best to do [09:29] because with setting PYTHONPATH it works directly from the checkout [09:29] better than /usr/local/… [09:29] but it's not hard to install it [09:29] yeah, I would still suggest to install it [09:30] in this virtualenv [09:30] unless there is weird path stuff [09:30] Laney: btw, I guess the service isn't scalable with just served by one server, isn't it? [09:30] what's the plan for that, giving it charmed to IS? [09:31] or handling it in canonistack for some amount of time [09:31] like this https://github.com/ximion/appstream-dep11/blob/master/dep11/generator.py#L386 [09:31] and handling replications and such [09:31] what do you mean? [09:31] launchpad will just pull the data into the archive [09:31] re [09:31] seb128: 'sup? [09:31] users don't hit it directly really [09:31] seb128: oh, the backlog -- well, look at excuses.html :/ [09:31] pitti, yeah, that's what I'm looking at [09:31] Laney: ah, so one consumer (being launchpad), but still, who will host it? [09:32] pitti, things seem to be on "in progress" for days [09:32] $cloud [09:32] Laney: I have some suggestions if it's on canonistack to ensure it's always up :) [09:32] seb128: and bos01 died over night, so it didn't make as much progress as it could have [09:32] (even one instance) [09:32] it'll be on prodstack or whatever I guess [09:32] ok [09:32] so not us [09:32] debci-xenial-amd64 472 [09:32] canonistack is no good for a production service [09:32] debci-xenial-armhf 30 [09:32] debci-xenial-i386 256 [09:32] debci-xenial-ppc64el 8 [09:32] it's been randomly shut down many times already [09:32] Laney: the path stuff doesn't really shock me TBH [09:32] and only been up for like 1 week [09:32] Laney: it's either using prod data or local checkout one, which is common [09:32] seb128: yeah, I know -- we are still heavily underpowered and Monday's KDE landing causes an awful backlog as the tests take very long [09:33] Laney: yeah, that's why if you wanted to use something like kubernetes meanwhile… [09:33] (providing one or multiple pods and keeping them up) [09:33] didrocks: this breaks if it's in the virtualenv though no? [09:33] because ".." won't be what it wants it to be [09:34] Laney: yeah, wierd that it worked, though, right? [09:34] I don't know that it did [09:34] because of the bug you fixed it never ran [09:34] so would have found this later on [09:34] pitti, I see, would be nice to have some smart logic to not let big jobs DoS the infra and take all the builders, blocking the small/easy ones [09:34] oh well [09:34] no hurry, let's wait [09:34] Laney: I told you I tried it in a virtualenv on the machine, no? [09:34] did you look at the output? [09:35] seb128: believe me, I'm harrassing IS regularly about getting moar powah :/ I know it sucks [09:35] yep, it generated some new content, but I didn't look in detail (there was some warning, but similar to the one you had with your global run IIRC) [09:35] I guess the system fallback would have worked [09:36] * Laney removes the package ;-) [09:36] yeah [09:36] if you installed it :p [09:36] the best practice is first to factorize that in a get_template_dir() [09:36] and have the logic poking here [09:36] seb128: so KDE and lots of other xenial packages plus new kernels in all stables (which also trigger several hundred tests each), plus little capacity -> I don't think reordering gains us that much there [09:36] how do you get the root directory of the virtualenv? [09:37] seb128: is there a package that's particularly pressing? we can hint it [09:37] Laney: $VIRTUAL_ENV [09:38] neat [09:38] isn't it? :p [09:38] Laney: want some help or you are handling the patch? [09:39] as you wish [09:39] pitti, no, it's ok, mostly mir and gtk+ but they are not that urgent [09:39] there are some other cases e.g. dep11-hints.yml [09:40] Laney: can handle a small refactoring of this if you wish [09:41] pitti, btw, other topic ... could we have a cron job for xenial langpacks updates? [09:41] seb128: oh, do we have exports now? sure! [09:41] * seb128 wants his new gedit translated :p [09:41] (I need to build the first one by hand, but that's simple) [09:41] https://translations.launchpad.net/ubuntu/xenial/+language-packs :( [09:41] pitti, unsure, but translations are open [09:41] -> #u-devel [09:41] k [09:50] seb128, did you see that Yorba went under? :( [09:51] willcooke, no, I didn't see, but I assumed they did since none of the projects are active and we didn't get a reply for the shotwell key issue since june [09:51] where did you read that? [09:51] willcooke: aren't they the guys that do shotwell [09:51] seb128, G+ [09:51] davmor2, exactly [09:52] so worth having a quick discussion about Shotwell? I mean, nothing has changed for ages, so we could keep it and not expect any serious breakage. But might be worth thinking about a change of app? [09:52] popey, ^^ [09:52] time to re-evaluate gphoto maybe [09:54] sad for them [09:54] but yeah, kind of expected [09:59] davmor2, gnome-photos I guess you mean? [09:59] we could, but changing in a LTS cycle... [09:59] and it's new GNOME style which doesn't fit well with our desktop [09:59] where shotwell is desktop style [10:00] seb128: yeah thought it was gphoto but yes the gnome photos app [10:00] old school desktop* [10:00] seb128, agreed. Sounds like a job for LTS+1 (to consider the alternatives) [10:00] also gnome-photos might depends on tracker or things we are not currently using (though I didn't look at it at all) [10:00] last I heard gnome photos didn't do importing from digital cameras yet [10:00] willcooke, yeah, also let's see if somebody picks shotwell up [10:00] seb128, yes, I suspect gnome-photos uses tracker [10:00] or any photo editing [10:01] willcooke: well by then we'll be unity8 based and have gallery right ;) [10:02] davmor2, it's a possibility [10:05] I doubt it [10:05] they were talking about deprecating gallery [10:06] the camera app is having most of the features [10:09] Hah! [10:09] Gallery - Written by Yorba [10:09] yes [10:10] didrocks: huh, I forgot to reply to you? [10:10] didrocks: feel free to handle it if you are willing :) [10:12] Laney: yeah, I was waiting for you to be back on that :) [10:12] Laney: ok, will do that this afternoon [10:12] willcooke, do you have an url to that g+ post? [10:12] seems ptyprocess needs a MIR [10:13] just curious, I'm probably not subscribed to the same people you are [10:13] didrocks: I at least typed something, must have deleted it :P [10:13] what is pytprocess? [10:13] seb128, sure, one sec.... [10:13] ptyprocess* [10:13] Laney: something used by pexpect to handle virtual shells [10:14] https://goo.gl/SsAhv [10:15] thanks [10:15] lol [10:15] https://plus.google.com/+NathanDyerdotMe/posts/67ts6ALTmTx [10:15] http://people.canonical.com/~ubuntu-archive/component-mismatches.svg [10:15] "AAAAAAAAAAAAAHHHHHHHHHHHHHHHHHHHHHH" [10:16] * Laney closes that tab forever [10:34] * robert_ancell -> sleep [10:34] see you robert_ancell [10:34] bye all === alan_g is now known as alan_g|test === alan_g|test is now known as alan_g === greyback__ is now known as greyback [10:56] seb128, Laney: ok, 45 mins with IS later I now have lcy01 and 8 extra workers back; let's see how long they'll hold :) [11:01] pitti: good luck [11:01] I noticed they re-enabled some builders there which all died again === mzanetti_ is now known as mzanetti [11:01] https://launchpad.net/builders <- lcy01-* Cleaning [11:03] Laney: yes, axino said he'd try to restart those and monitor [11:04] pitti: I mean that as a sign that it still wasn't stable at least earlier this morning [11:04] 11:54:58 axino | pitti: when they're all disabled, lcy01 tends to be stable [11:04] although maybe it is those that bring down the cloud and not the cloud itself being bad [11:04] "they" -> lcy01 buildds [11:05] what's the interesting difference between this and lgw01 that makes lgw01 reliable? [11:11] Laney: I have no idea [11:12] sort of rhetorical :) [11:13] * Laney is tail -f log/lcy*.log [11:13] Laney: as long as you don't see "ERROR"s in nova list, things should be okay [11:14] nod [11:14] man, one of these days I need to figure out why some tests use the ugly instance names [11:14] seems the --name option doesn't get passed along properly in one case [11:17] Laney, can you add gnome-builder to desktop-extra? [11:18] darkxst: if you mail devel-permissions [11:19] Laney, kk [11:19] but aren't you getting that fixed in debian? [11:19] i.e. autosync [11:22] Laney, yes, I am pushing the fixes through debian [11:23] but its kinda crap when I have to wait on debian to fix FTBS etc [11:24] (ones that only affect us) [11:24] get some sponsored uploads [11:24] then you can become DM [11:25] Laney, who is going to sign my key? [11:27] one of these kind people https://wiki.debian.org/Keysigning/Offers#AU [11:29] Laney, or I just come to the debian hackfest, if CC will sponsor it ;) [11:30] ^ insert pkg-gnome [11:31] heh [11:31] when is that? [11:32] no date set yet [11:34] which mailing list? [11:34] don't see one [11:34] just some irc the other day [11:35] Laney, only irc, but it will happen sooner or later [11:36] unless its like the git migration, that doesnt want to happen [11:37] that's one reason to have a sprint [11:38] yep [11:39] but then also must be a reason we don't have branches on alioth either ;) [11:41] the script is the hard part [11:41] ask tumbleweed [11:48] Laney, I was alluding to the point that it should in theory be really simple to create packaging branches for ubuntu GNOME/gnome3 ppa's, but it still hasnt happened [11:48] oh right [11:49] it's going to be cool to be able to git merge [11:49] * Laney does this for gstreamer already and it is g-r-e-a-t [11:49] I already git merge everything, but major hacky, since not merging real commits [11:50] mmm [11:53] works pretty well though, apart from the fact you have to resolve all conflicts before commiting [11:55] are you using rerere? [11:55] gbp and a custom script that pulls in bzr packaging branches if available [11:59] * darkxst really should sleep now, I meant to do that an hour ago [11:59] lxc -> xenial [11:59] * Laney looks at dbus merge now [12:23] * Laney loses utf8 sbuild on desktop too now :( [12:38] Laney, utf8 sbuild? [12:39] yeah it had nice output [12:39] now it's all ascii [12:40] :-( [12:42] hardly a big deal :P [12:42] yeah [12:43] oh, nice, gnome-desktop3 migrated, seems like autopkgtest backlog is getting smaller ;-) [12:43] gtk got some green too [12:46] nice [12:47] larsu, are you looking at the nautilus menubar thing or not yet? I'm trying to decide what I should do with the nautilus update, wait a bit/upload without menubar to start/... === hikiko is now known as hikiko|ln === davidcalle_ is now known as davidcalle [12:52] so, they cancel the light festival in Lyon… [12:53] they do?! [12:53] it's well in advance [12:53] well, it's in less than a month [12:54] like 2 weeks and a day [12:54] so not that much in advance [12:54] (on the contrary even, hotels are booked in general 6 months in advance) [12:54] seb128: does tomorrow work for you? [12:55] larsu, next week work as well, I just want an estimate if it's worth waiting or better to go without and do that later [12:55] seb128: ok. I'll do it tomorrow :) [12:56] larsu, danke [12:56] didrocks, right, I guess my comment was rather "it's probably not reacting to a particular plot they found but just being cautious/not wanting to risk it because it's difficult to secure a such event" [12:57] seb128: yeah, basically they fear that some noise of explosion due to anything create panicks [12:57] pitti is going to need to come to your place again next year then! [12:57] en effect ! [12:57] but the stollen! [12:58] double stollen! [12:58] one this year, one next :p [12:58] non, je vais toujour venir ! [12:58] tu as raison [13:01] ;) [13:01] * didrocks hugs pitti === alan_g is now known as alan_g|lunch [13:14] attente, hi, is thr [13:14] oop [13:15] attente, hi, is there a current patch-set of mir for gtk+ git master? [13:17] ricotz: for the ubuntu package? because it should be up-to-date as it hasn't had any activity in a while [13:17] attente, I mean gtk+ git master doesnt compile against mir in xenial [13:18] ricotz: i'll take a look [13:19] ./gdk/mir/gdkmir-debug.c:106:36: warning: implicit declaration of function 'mir_touch_event_get_cookie' [-Wimplicit-function-declaration] [13:21] ricotz: you'll probably have to build against mir trunk because that's new api [13:22] Laney, pitti, can you help me to understand http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html#ldb ? why aren't armhf/ppc64el trying for the current version? [13:22] attente, hmm, I see [13:22] Laney, seb128: bug 1512779 -- ATM these need to be retried manually, sorry [13:22] bug 1512779 in Auto Package Testing "Failing tests are not re-triggered when package with failures is re-uploaded" [Undecided,New] https://launchpad.net/bugs/1512779 [13:22] good morning, fellow desktoply -inclined people [13:23] hey desrt! [13:23] urgh, look at the wrapping in that bug [13:23] seb128: let me retry all armhf/ppc64 failures wholesale, the qeueus are emtpy [13:23] pitti, danke [13:23] ricotz: you can also just delete the line since it's just there for debugging purposes [13:23] well, not "empty", but "short" [13:23] note to self: never hard wrap when writing launchpad comments in the web UI [13:24] ricotz: but afaik, it won't behave any differently than what's in the archive for xenial [13:24] Laney, nag the launchpad team to have them to display a wrapping line in the ui so if you want to do it manually you know where ;-) [13:25] I probably used vimperator and vipgq it [13:25] attente, alright, will do that [13:25] * Laney does some rebuilds for liburcu [13:26] it'll be annoying if things get blocked on this transition after the tests run :P [13:27] pitti: "OSError: [Errno 12] Cannot allocate memory" - erm? [13:28] wut? [13:28] failure just now [13:28] oh, bos01-ppc64el-9 [13:28] Mem: 2050048 1912936 137112 1280 328800 624912 [13:28] doesn't seem terrible [13:29] erk, wrong machine [13:29] Mem: 2050048 1975464 74584 1228 7524 84096 [13:29] isn't that from the cloud worker? [13:29] still [13:29] Laney: yes, it is [13:30] * pitti shrugs, let's see if it happens on more workers [13:30] lcy01 is still up -- scary! [13:30] it really seems it's the buildds OR autopkgtest, there's no AND there [13:31] * Laney restarted it [13:32] cosmic rays === hikiko|ln is now known as hikiko [13:49] kenvandine, hey, we still get error reports for the unity-webapps-* migration scripts, didn't you put a landing up to drop those previous cycle? [13:50] seb128, i did, not sure if they landed [13:50] * kenvandine checks [13:52] seb128, yes, silo 38 [13:52] the branches are still awaiting review [13:52] kenvandine, can you nag someone? [13:52] dbarth? === alan_g|lunch is now known as alan_g [14:01] seb128: can you look at NEW for ltt-control please? [14:01] I fake synced it from git, will block liburcu, mir, etc [14:01] Laney, ok [14:01] thanks [14:01] yw [14:03] * Laney lunch [14:03] Laney, done [14:03] Laney, enjoy! === nudtrobert1 is now known as nudtrobert [14:22] google impresses me [14:22] they actually managed to make google+ _more_ difficult to use [14:23] oh true this looks different [14:23] morning desrt [14:23] g' morning [14:25] less easy to read :-/ [14:25] also less easy to do anything at all [14:25] ...and it was pretty bad to begin with [14:25] I'm not doing much with it out of browser through posts and reading sometimes comments [14:25] don't use it? [14:25] I gather their goal is to make everyone leave before they shut it down [14:26] that's sort of the impression i got [14:26] people will be less upset that way [14:27] unfortunately, google+/hangouts use is more or less mandatory for ingress operations :( [14:28] "unfortunately" :P [14:28] :D [14:28] i totally geeked out on the social aspects of this game [14:29] which unfortunately has dragged me into a world of G+ pain [14:29] isn't it just the account you need? [14:31] technically you don't need a G+ account at all -- just google login [14:31] but again, all the social aspects... [14:31] any kind of non-boring high-level play involves a fairly large amount of groupwork [14:33] a bit of it is kinda implicit "i see my team happens to be holding this territory, i'll help out" type stuff... but anything beyond that requires coordination [14:33] and since it's all hooked into the "G" ecosystem, this is the platform of choice for everything... G+ and hangouts [14:34] makes sense [14:34] everyone has a google account by definition [14:34] doesn't make sense to have another one [14:35] and it's just one simple click to "upgrade" to G+! [14:47] * Sweet5hark is back from berlin. [14:47] Twas awesome. [14:52] Sweet5hark: what was in Berlin? [14:54] qengho: 2. open it summit by the greens and OSBA. lots of members of parliament asking how they can help open source/open data/etc. [14:54] Nice. [14:59] Laney: lgw workers dying> looks like orphaned instances, I'm cleaning up and restart === dpm is now known as dpm-afk [15:35] didrocks, http://arstechnica.com/information-technology/2015/11/ubuntu-15-10-review-wily-werewolf-leaves-scary-experimentation-for-next-year/ [15:36] didrocks, "If you're looking for a quick way to get, for example, a basic Android development environment setup, you'd be hard pressed to beat Ubuntu Make's simple umake android command." [15:37] \o/ [15:39] didrocks, well done! [15:41] thx ;) [15:41] happy that it's started to get noticed by the press [15:41] (while I'm fighting pokfdpokfspodkfpods proxy on jenkins… not allowing me to access to localhost) [16:00] erm, maybe larsu or seb128 - will we get the Google Drive integration from GNOME apps in 16.04? [16:01] willcooke, yes, the current gvfs has a backend for it [16:02] though unsure if that needs online account integration/is going to work with uoa [16:02] also somebody needs to merge gvfs with Debian [16:02] I think it does [16:02] but xnox just put his name on that by doing a rebuild ;-) [16:02] lulz [16:03]