[05:08] rye: figued you are not on know buut wanted to leave you this for in the mtning. Was looking at bugs that OPs have responded to. bug 769176 [05:08] Launchpad bug 769176 in ubuntuone-client "Ubuntu One Reporting Incorrect File Storage Statistics (affects: 1) (heat: 31)" [Undecided,Incomplete] https://launchpad.net/bugs/769176 [08:35] ralsina: ping [09:03] morning [09:26] be back in 20 [10:02] back [10:02] Happy Friday, everyone! :D [10:03] JamesTait: dont make me sing some rebecca black [10:04] Ok I think im after being sent back to the 90s La Noire comes with 3 disks [10:06] * fagan is going to distract himself for an hour and then get back to twisted [10:11] hi guys, got a couple of questions regarding #ubuntuone. [10:11] is this the right place to ask? [10:11] yep [10:11] ask away and someone will answer when they get the chance [10:14] great, will it work with my old android phone? (2.1) [10:14] I want to move all my music to ubuntuone [10:14] (about 22gb) [10:19] crodas: well I dont know myself but you can check in the marketplace on the phone [10:19] just search for ubuntu one and it should give you it [10:19] I have it installed [10:19] already [10:19] but just want to be sure before buy the service [10:19] also, does it have any sort of limits? [10:20] I want to copy the 22gb of musics [10:20] then yep it will work but the 22gb should take a long while to sync [10:20] fagan, I know :P [10:21] I will use it as a backup, my laptop was stolen last week here in Spain. [10:21] and I never backup my mp3 files [10:22] crodas: nope doesnt have any limits stream all you want [10:22] fagan awesome! [10:22] will try tonight [10:23] and I will write here if it works [10:25] cool [10:41] crodas, are you using ubuntu 11.04? [10:46] rye, yes, I bought a new laptop and I installed the last one [10:46] I believe it's 11.04 [10:46] crodas, then it will be syncing faster :) [10:47] rye, cool! [10:47] does it provide any sort of api? [10:47] crodas, https://one.ubuntu.com/developers/ [10:47] It'll be cool to implement another client, only for fun [10:47] crodas, sorry, https://one.ubuntu.com/developer/ [10:47] crodas, file sync api based on rest interface which is FAST! [10:48] rye, what is it using? nodejs or python? [10:48] I've got a friend in canonical [10:48] so I bet python :P [10:48] crodas, the server-side is in python [10:48] rye, cool [10:48] :) [10:49] ‎is it a new service? I'm new in the ubuntu world [10:49] I was using suse in my old-stolen notebook [10:49] it's pretty cool === teknico is now known as teknico_away === Chipaca` is now known as Chipaca === teknico_away is now known as teknico [12:14] hello everybody! [12:14] mandel, ralsina: ping? === Chipaca_ is now known as Chipaca [12:22] is ralsina here? cool! hi from Spain! [12:55] ralsina: ping [13:03] nessita, there's no ralsina nor mandel around [13:04] alecu: oh [13:05] alecu: hi BTW, how are you? [13:05] im around if you need any stuff with pyqt or anything alecu [13:06] nessita, hi hi! and hi fagan! [13:06] fagan: thanks! A couple of questions: [13:06] cool cool [13:07] alecu: can you, by any chance, review the branch I linked yesterday? https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/move-logic-to-backend/+merge/61642 [13:08] fagan: questions about designer. In ralsina's pyqt tutorial, he provides a window.ui that I haven't been able to match. How you make the treeview resize with the window that contains it? [13:09] nessita: you have to set a layout [13:09] fagan: how? [13:10] nessita: scroll to the top of the box of stuff on the left [13:10] nessita, ok, I'll review it. [13:10] nessita: then drag it to the window [13:10] fagan: right, I tried that, but. BUT: if you see the window.ui that ralsina provides, there is no layout in the window! :-) [13:10] so that confuses me [13:10] nessita: then you can right click the window and set the layout [13:11] nessita: yeah thats a weird thing that confused me too [13:11] nessita: you have to put the layout there before it lets you set the window layout [13:12] fagan: me no understand :-) [13:12] nessita: hah well you can try just right click and there is a layout thing there but its greyed out if you dont have a layout on the window already [13:13] fagan: let me try it [13:14] nessita: if that doesnt work just put the layout on the window and right click and set it there [13:14] fagan: ok, next question. After adding a QTreeView, how can I set the columns (name, type, etc)? [13:15] nessita: give me a sec I forget this one [13:15] fagan: sure [13:17] nessita: well setheader works for a column thats made already [13:17] right, but so far I have no columns :-D [13:19] nessita: oooh you mean in the designer /me just blanked and went looking for code :D [13:19] nessita: you should be able to right click for that [13:19] or maybe double click it [13:20] duanedesign, I like your writeup about the U1 application developer stuff :) [13:20] duanedesign: I liked it too :) [13:21] nessita: ok neither of them are right, its in the properties thing [13:22] fagan: one second, installing designer in this machine.... (yesterday I worked from another machine) [13:22] nessita: ok not that either [13:22] fagan: stop guessing! :-D [13:23] yeah im in there now [13:24] nessita: well ill keep looking but my best guess is you have to do it in code [13:24] nessita: my guessing was from the way text boxes and all work [13:24] fagan: no way, open the window.ui from ralsina, it has the columns in it [13:24] without need of coding nothing [13:25] nessita, approved. [13:25] nessita: ok then im just blind and cant find it [13:26] alecu: thanks! [13:29] fagan: you lied to us! you said you knew qt! :-P [13:30] nessita: hah I dont remember treeviews. Everything else is good [13:31] fagan: excuses! :-P [13:31] I did it for 2 weeks and didnt do 1 more :) [13:32] nessita: it should be on the properties thing but default size and all that is pixels not rows or anything [13:32] nessita: do you have any more questions? [13:32] morning all! [13:32] fagan: not at the moment, thanks [13:33] fagan: no worries, I'll chase ralsina when he comes back. By the way, have you "seen" him today? [13:33] hola thisfred! [13:33] Wow. It took a while, but my notmuch mail setup is close to perfect now. Inbox 0 achieved, and I think I can maintain it! [13:33] hola nessita! [13:33] thisfred: I'm making some mate, you're just in time! [13:33] :) [13:33] nessita: I saw mandel looking for him at 8am my time but other than that no [13:33] I'm still on coffee [13:34] fagan: ok, thanks [13:34] nessita: ahh I didnt do treeviews I did a listview [13:34] thats why I got confused [13:34] more excuses!!!! :-D [13:34] nessita: can you not do a listview [13:34] * fagan rocked that one [13:35] fagan: I'm following http://lateral.netmanagers.com.ar/stories/BBS47.html, so I need a treeview [13:36] nessita: yeah actually thats the first result on google for pyqt treeview example [13:36] :D [13:41] nessita: oh and if you want to know how to do css ping me im awesome at that now [13:41] (in pyqt) [13:41] fagan: you may do it yourself! :-) [13:44] nessita: hah actually I got the designers to do it themselves [13:44] nessita: I made a thing called css magic tester that does that [13:56] are we going to do a standup or does it matter? [13:57] fagan: of course we are [13:57] fagan: why wouldn't we? [13:58] nessita: missing mandel and ralsina [13:58] fagan: well, they should appear, right? as far as I know they don't have any license setup for today [13:59] nessita: maybe couldnt be sure [14:00] mandel: ping [14:00] me [14:01] alecu, dobey, thisfred, mandel, ralsina, fagan [14:01] me [14:01] me too [14:01] me [14:02] wow my computer is slow at the moment [14:02] * fagan is running updates [14:03] DONE: propsoed and landed branch to move some logic from the GTK layer to the backend, and to move all the "stringfication" logic (needed by DBus) to the dbus layer. Started with the pyqt-by-example tutorial, I've got so many questions... ralsina, help! [14:03] TODO: plan control panel for windows development, learn QT? something else? [14:03] BLOCKED: a little, I need a quick qt4-designer guide [14:03] NEXT: thisfred [14:03] DONE: caught up with email TODO: fix some bugs in ubuntuone-couch reported by Michael Terry | blueprinting BLOCKED: no NEXT: alecu [14:04] alecu: ? [14:05] nessita: other than the treeview bit I should be able to answer a lot of the questions [14:05] alecu: you here? [14:05] isnt it thisfred first [14:05] DONE: mumbled with mandel, proposed a branch for qt + twisted in u1cp. [14:05] TODO: fixes for that branch, work more on qt designer [14:05] BLOCKED: no [14:05] NEXT: fagan [14:05] DONE [14:05] * Got a good poke at the twisted code in u1 to figure out how all the pieces fit together [14:05] TODO [14:05] * nope [14:05] * do some twisted [14:05] Blocked [14:05] thisfred: go [14:06] fagan: thisfred already gave his [14:06] (since we skipped him) [14:06] ohhhh [14:06] we didnot [14:06] dobey, mandel, ralsina? [14:06] any closing comments? [14:06] nope other than if you guys have any twisted words of advice share away [14:07] LIKED: notmuch+emacs+orgmode, but I won't bore you with details unless you want me to. (Also I won't start doing standups in orgmode format again, I promise ;) [14:07] I havent wrote a line of code yet but I think I understand it now [14:07] ok, I think this is eom [14:08] ill be writing some code for the rest of the day [14:10] * fagan break [14:12] ok, ralsina is taking the day off [14:13] ahhh ok that explains it [14:32] λ DONE: fixed some 11.10 nightlies, #785327 [14:32] λ TODO: more 11.10 nightlies fixes [14:32] λ BLCK: Still recovering from cold. [14:32] sorry, i overslept. damn cold [14:33] dobey: we forgive you (?) [14:34] lol [14:34] nessita: you made me spit out some tea with that (?) [14:35] * nessita is like that [14:54] heh [15:46] hmm [15:47] alecu: is your branch moving forwards? [15:48] is mandel not working today? [15:48] nessita, moving all right. [15:48] alecu: shall I re-review? [15:48] nessita, not yet. [15:49] ack [15:49] dobey, he was having some personal problems this week, so he's probably out. [15:50] alecu: I'll run a quick errand while you finish [15:50] * nessita brbs [15:50] ack [15:53] alecu: did you get any answers about qtreactor? [15:54] dobey, no, I've got no response from mandel nor ralsina yet. [15:54] i would answer but I know nothing about it :) [15:54] dobey, I'll take it out of my branch, and have devs install it for now. [15:54] fagan, don't worry, no need. [15:55] nessita: btw: you know about conky? I think that would solve your system monitor issues, so you could run unity. [15:56] conky, gkrellms [15:56] err [15:56] s/s$// [16:00] alecu, ralsina is traveling afaik [16:00] no [16:01] alecu, last time I talk with him he was on London or Estambul [16:01] or at least he just said "hi" so i doubt it [16:01] he is on-line though, generally speaking [16:03] crodas, he is in Estambul, I know. But I heard he took the day off today. [16:04] alecu, nice. we're tweeting yesterday night, so probably he will be back tonight or so. [16:04] crodas, or "Istambul" if we want to sound less "spanish-speaking" :-) [16:05] alecu, sorry, I'm from South America, but I'm Europe now :P [16:05] alecu, paraguay [16:07] crodas, de que parte de Paraguay sos? Yo sólo conozco Ciudad del Este :-) [16:07] alecu, Asuncion, but I'm in Barcelona now for a couple of meetings :) [16:07] alecu, and you? Argentina? [16:07] crodas, si, soy de Buenos Aires [16:08] nice, I'll be there next month to see Les Luthiers. We should gather if you have time :) [16:08] alecu, I'm more PHP/C though, I assume you're a pythonist as ralsina, isn't it? [16:09] crodas, right :-) [16:09] alecu, that's fine with me though, I'm not a zealot. [16:09] heh, pythonist [16:09] crodas, and definitely we should gather when you come. [16:10] it does seem to be a very religious thing, so an apt name for it [16:10] alecu, also I'm nosql guy :P [16:10] alecu, I'm crodas in freenode, twitter and @php.net [16:10] I'll ping to ralsina when I arrive [16:10] crodas, cool. [16:11] it's nice to meet people from Canonical :) [16:13] hi crodas! [16:13] ralsina, hi! [16:13] this irc channel looks cool! [16:13] thanks for the recommendation [16:13] I didn't know you were so popular here :P [16:14] well i am here, so it must be cool. [16:14] dobey, :P [16:15] ralsina: hi there. Do you know if mandel is working today? [16:16] alright, am off to get some lunch, and errands. bbiab [16:16] no, personal issues [16:17] he will be back on monday [16:17] crodas: BTW, mandel is in Barcelona, to [16:18] s/to/too [16:18] ralsina, really? where? [16:18] do yo have a phone? [16:18] alecu is a big Les Luthiers fan, BTW. Can quote every act ;-) [16:18] I'm in front of the Plaza Universitat [16:18] crodas: I do, but I have no idea where he is physically [16:19] and he is probably not at home until monday [16:20] ralsina, I'm not a "big Luthiers fan" at all! I can hardly remember the lyrics for my favourite LL song. :P [16:20] hahaha [16:20] "la mojó en el arrolluelo... y cantando la lavó" [16:20] "la colgó de un abedul!" [16:24] still looking for my phone number [16:33] +34649275387 [16:35] ralsina, that was for you [16:36] crodas: well, I am pretty far away from Barcelona ;-) [16:43] ralsina, I misread it, you said mandel [16:43] k [16:43] crodas: if you want to drop by, it's a 4 hour flight ;-) [16:44] ralsina, that's not an issue [16:44] money is an issue [16:44] one of my credit card and my laptop has been stolen last weekend [16:44] crodas: ouch [17:05] I'm starving. bbiab [17:18] * nessita too, brb [17:21] hey nessita :) [17:23] I haven't received a response from Matt, have you heard or received any response about that (u1store)? [17:26] kamusin: no, let me ping them [17:27] thanks nessita [17:27] :-) [17:31] about what? [17:38] dobey: community members need/want a way to test the music purchase [17:38] oh [17:39] kamusin: mattgriffin will get back to you soon, he has been travelling [17:39] i think that's really only possible for people in the US, where there are some free songs available (unless they want to make actual purches with credit card/paypal) [17:39] ok, lunchtime! [17:39] this is (udd) documentation gold, I can't believe I hadn't found this before http://people.canonical.com/~dholbach/packaging-guide [17:39] the free songs don't tend to hit all the corner cases where there might be problems (like the one utf-8 bug) [17:44] you're saying we need more free metal? [17:46] no [17:46] i'm saying testing without spending money is hard, because we have no control over what is actually available in each country [18:06] alecu: can I re-review your branch? if not, can I help somehow with whatever is blocking it? [18:09] nessita, there were a few tiny things blocking it, most resolved; but I'm currently trying to find out what is the best way to have run-test run twice, once for the standard gtk reactor, and once for the qt reactor that u1trial can install. [18:10] alecu: what about doing that in another branch? so I can use the first branch to start doing some QT testing [18:10] nessita, currently I'm havint run-tests be run twice; once just run-tests, and then another with "run-tests -qt" [18:11] but I don't quite like having them run twice, because pylint and pep8 are run twice too. [18:11] anyway, I'll push it as it is. [18:12] alecu: thanks [18:14] kenvandine: haha, identi.ca puts your head on the courthouse for the locations map [18:16] lisette: ping [18:18] nessita, dobey: please re-review https://code.launchpad.net/~alecu/ubuntuone-control-panel/tx-qt-mainloop/+merge/61652 [18:18] alecu: on it [18:19] nessita, dobey: I've added detailed instructions on how to test. [18:19] alecu: thanks [18:22] alecu: you got the arguments to u1trial mixed up in run-tests [18:23] nessita: btw, what is control panel using twisted for? [18:23] dobey: nothing, why? [18:23] dobey, nessita: it will use it soon. [18:24] alecu: what for? [18:24] dobey: we use twisted deferreds, but not the twisted reactor [18:24] alecu: why? [18:24] dobey, nessita: we'll use twisted to talk to syncdaemon and with the webservices. [18:24] why? [18:24] dobey, don't understand why the args are "mixed up" [18:24] alecu: I thought we were using some QT stuff to talk to webservices... [18:25] nessita, we may, yes. [18:25] dobey, nessita: but we need twisted to talk to syncdaemon on windows. [18:25] alecu: you're passing the GTK tests when using the qt reactor in trial, and the QT tests when --qt is not passed to run-tests [18:25] alecu: oh, there's no way to do the named pipes in qt main loop? [18:26] dobey, that is the argument to --ignore-paths [18:26] alecu: and regarding syncdaemon, seems like I got that wrong, since I thought that all the work the contractor was doing was to provide a way to contact syncdaemon [18:26] hi. I have synced around 100gb of photos to my u1 account from my desktop. now i'm setting up my laptop, and i've copied all the images via an usb hdd cause u1 is too slow, thinking i'll save me some time, but now i see on the laoptop it's uploading thousands of files again? does it need to do that? does it overwrite the old ones, or? [18:26] dobey, so it's ignoring the GTK tests when passed -qt, and viceversa. [18:26] alecu: oh, right [18:26] Ignatus: it shouldn't upload all your files again, unless the hash of the files have changed [18:27] nessita, hm, so i'll try deleting them all again, and recopying them then, but they are the same files... [18:27] the hash is probably different if copied to usb drive then back to hard disk; they likely all have +x bit on now if usb drive was fat32, and they weren't +x before [18:28] thing is, i have a ext3 usb disk :) [18:28] Ignatus: do not delete them [18:28] can i check the hash somehow? [18:28] nessita, mattia is working on providing ipc on windows using iocp, as a twisted transport (instead of using tcp on localhost) [18:28] Ignatus: what ubuntu version are you running? [18:29] natty on both, fresh installs [18:29] nessita, that means we will need to use twisted on the control panel to talk to syncdaemon [18:29] Ignatus: please do this: disconnect file sync from the control panel (top right corner) [18:29] why do people think using twisted for this stuff is a good idea? [18:30] Ignatus: how do you know if files are being downloaded or uploaded? [18:30] nessita, i've stopped it from magiciada (can't use unity on laptop unfortunately, since beta1) [18:30] nessita, magiciada [18:30] hm, magicicada even :) [18:31] dobey, do you have a better idea on how to do ipc between a twisted app (syncdaemon) and a qt app (control panel) on windows? [18:31] nessita, so, is there a way to check the hash value of the files manually? i can compare them and see it it's changed? [18:31] alecu: well doing it directly in the qt event loop would be much better [18:32] Ignatus: ah. Your pictures, where are they located? inside Ubuntu One or any other custom folder? [18:32] in the pictures folder [18:32] Ignatus: you can't access the hash from the server, as far as I know. Let's do this: [18:32] nessita, sorry, i keep forgetting to put your name up front :) [18:32] alecu: using twisted to do that IPC doesn't give you any benefit, beyond introducting infinitely more complex code as a dependency [18:33] (no, that's not a benefit :) [18:33] Ignatus: start syncdaemon but do not connect it (or if it connects, disconnect it). Once that s done, open the control panel (is available on non-unity systems) [18:33] dobey, well, if you are proposing yourself to work on it, tell ralsina and do it. [18:33] nessita, started, and not connected [18:34] Ignatus: in the control panel, go to "cloud folders". Unchecked the Pictures folders. Once that's done (the UI is no longer unsensitive), re-copy all your pictures [18:34] alecu: why does my pointing out an obvious flaw in the design mean i have to do all the work? [18:35] nessita, ok, will do. i'll throw out a shout when it's done, ty [18:35] dobey: why would you say is "obvious"? I think that is the key [18:35] Ignatus: please do so, I'll be around [18:35] dobey, because yours is an opinion, not an obvious flaw on the design. [18:36] alecu: why are we ignoring "'.svn,CVS,.bzr,.hg,.git,*_ui.py'"? Do we have CVS stuff? [18:39] nessita, surely not, but it was the default for that option, and I added the *_ui.py [18:39] alecu: ah, ok [18:40] alecu: another question, why did you add the gettext code in utils.py? [18:41] sigh [18:41] there seems to be no need for that, as far as I can see [18:41] dobey, why did you abstain from reviewing the branch? [18:41] because i don't wnat to argue [18:42] dobey, please argue about that decision with mandel and ralsina; they worked out the details of the windows ipc. [18:42] dobey, but please review the branch having that in mind. [18:43] i understand why it should be done in twisted within syncdaemon. and that makes sense given it's already using twisted; though i think we should work on removing twisted from all our code instead, but alas [18:43] nessita, regarding the gettext: I took the code that compiles the .ui and the qt resources from sso; I guessed that since we are using gettext there we'll be using it in control panel. [18:44] alecu: I hope we don't need to, if we do we're doing something wrong [18:44] alecu: we may be having a bug/issue in sso then [18:44] dobey, again, that's an opinion. My opinion is that we should use twisted whenever it makes sense, and from a "time to finish" pov it makes a lot of sense to use it here. [18:44] alecu: i cannot consciously approve something i disagree with on such a basic level [18:47] alecu: and as per our meeting @ UDS, you need someone to test it on Windows anyway. [18:48] I'll do it, though not the windows review just yet, since I don't have a VM set up [18:49] dobey, thisfred: this is not working on windows yet. We'll start testing the control panel on windows probably on monday when ralsina and mandel are back. [18:49] right [18:49] (if they ever return! :-) ) [18:50] let's hope so, I don't want either of their jobs ;) [18:50] hahahaha, me neither! [18:50] ok, I'll get some lunch now, it's about 3pm here :P [18:50] * nessita neither [18:50] alecu: I have another q! [18:50] sure [18:50] alecu: at parrilla? [18:50] alecu: did you remove the gettext in utils then? [18:51] alecu: and also, what's with the DBUS constants inside ubuntuone/controlpanel/qt/__init__.py ? [18:51] nessita, I didn't remove it, since it's used in setup.py by the code that converts the .uis and the resources [18:52] alecu: why is used? I mean, it shouldn't be [18:52] alecu: well, now I don't know if you're talking about the gettext or the DBUS constants :-) [18:53] nessita, still about gettext [18:53] alecu: I don't see how is being used [18:53] nessita, it's used in _UICompiler in setup.py [18:53] nessita, o.write('from ubuntuone.controlpanel.utils import _') [18:54] alecu: ah, yes. That's ugly, why no importing dirctly gettext and setting the domain? [18:55] and even better, why are we generating python code programatically in that function? [18:55] nessita, the .ui are translated into .py [18:55] heh [18:55] alecu: right, but that is not related to the gettext thing, isn't it? [18:56] nessita, normally the .pys generated from .ui have no translation support [18:56] nessita, mandel has extended the translator to include another line before the rest of the .py [18:56] nessita, that line is "from ubuntuone.controlpanel.utils import _" [18:57] nessita, we need that line in every .py that's generated from a .ui [18:57] nessita, does it make sense now? [18:57] I understand now. Anyways, I don't see why we need to import gettext in utils and import that in the pys :-) [18:57] we should do the proper thing in each py [18:59] nessita, I don't understand why adding two lines in every .py is different than doing it in one place. [18:59] nessita, anyway, that's how it's done in SSO, and I try to follow the same idea here. [18:59] alecu: becasue if you see the getext code in utils, it makes no sense in having it, since nothing in that faile uses it, so it can be easily removed [18:59] I know I'll remove it, since _ is not being used in that module [19:00] s/faile/file [19:00] nessita, no, you won't remove it. You'll first search in our code to see who is importing that, and setup.py will. [19:00] alecu: in practice I might, but you get my point :-) [19:01] nessita, anyway: I can add a comment explaining just that in utils so you won't remove it tomorrow. [19:01] you import things where you need it, not adding a hop [19:01] alecu: I prefer we add the proper import gettext lines in the generated pys [19:01] is clearer, for anyone that reads the source code [19:02] nessita, that is bike shedding. :-) [19:02] alecu: not sure what that means :-) [19:02] alecu: http://pastebin.ubuntu.com/610723/ [19:02] fwiw: I agree with nessita [19:02] well the gettext package should befined as GETTEXT_PACKAGE = 'ubuntuone-control-panel' or something [19:02] and that should be imported from utils or something [19:03] dobey: we have that constant already defined [19:03] don't reimport things unless you're monkey patching them, and don't monkey patch things [19:03] nessita: ok, so then a) the bindtextdomain() call should use that define instead of a string literal === m_conley_away is now known as m_conley [19:03] dobey: I agree, it was my next point [19:03] thisfred, regarding the pastebin, I forgot to add "./setup.py build" to the testing instructions. Doing so now. [19:04] and b) what nessita already said [19:04] alecu aha, I knew I was missing something [19:04] alecu: you should make run-tests do ./setup.py build for the qt case [19:04] alecu: that would be good ^ [19:05] ok. [19:06] alecu: I'm getting ImportError: No module named qt [19:06] from dbus.mainloop.qt import DBusQtMainLoop [19:06] ImportError: No module named qt [19:06] nessita, package python-qt-dbus [19:06] nessita, or python-qt4-dbus [19:08] alecu: ok, from my end only 2 thing needed to get the approve: fix the gettext thingy, and remove the DBUS constants from the u1/__init__ [19:09] (the rest looks good!) [19:10] nessita, the TRANSLATION_DOMAIN in __init__ is the one that dobey mentions as GETTEXT_PACKAGE above? [19:10] sounds like it should be that [19:11] nessita, dobey: should I make the generated .pys import that constant? [19:12] alecu: yes to the first question, and to the second [19:12] you should only do the bindtextdomain() call where you start the main loop [19:12] dobey: and then only do the _ import? [19:13] if you need to translate things by other methods, such as dngettext() or something, then it will also be needed in those spots. but the general _() case doesn't take a domain argument [19:13] so it depends, and i doubt that only having the _() argument is sufficient for all translated strings, but maybe it is [19:15] alecu:eric@eric-Studio-XPS-9100:~/canonical/ubuntuone-control-panel/qt$ ./run-tests --qt [19:15] Running test suite for --qt [19:15] Usage: u1trial [options] path [19:15] u1trial: error: --qt-reactor option requires an argument [19:16] or do I not need to run it like that? [19:16] it's just -qt i think [19:17] +if [ "$1" == "-qt" ]; then [19:17] right just found that [19:17] but is weird that you got that error, since it does --qt-reactor=ui [19:17] isn't single dash just for single letter switches? [19:18] anyway with a single dash it breaks differently: [19:18] ideally yes [19:18] http://pastebin.ubuntu.com/610726/ [19:18] oh lovely [19:19] but that could be my unity bundle of joy [19:19] that is on 11.04? [19:19] yep [19:20] weird [19:20] does run-tests work in a clean cp branch for you? [19:20] or does it work without the -qt? [19:20] run-tests works [19:20] just not with the -qt option [19:20] do you have qtreactor installed? [19:21] not installed but added to the pythonpath as per the instructions [19:22] nessita, thisfred: revno 148 [19:23] thisfred: weird [19:23] ok, I need a bit of fresh air. And some food. [19:23] bbl [19:24] alecu: thanks, enjoy [19:31] hrmm [19:31] QLocalSocket [19:33] dobey: ? [19:34] qt class that uses named pipes on windows, and unix sockets on *nix [19:36] dobey: when you said bindtextdomain, did you mean "gettext.textdomain(TRANSLATION_DOMAIN)"? (that is what we're using right now on the control panel GTK) [19:36] sure [19:39] thanks [19:43] alecu: when you come back, I added a detailed comment on how the gettext code need to be tweaked (seems like we misunderstood here) [19:46] nessita: ah, i see what i did. mixed up bind_textdomain_codeset() and textdomain() calls [19:46] nessita: so yes, textdomain is what i meant there it seems [19:48] dobey: great, thanks for checking [19:53] thisfred: did you finish alecu's review? [19:53] nessita: not yet, but I think the errors I'm getting are not in the code [19:53] thisfred: what errors are you getting? the unity ones? [19:53] the main tests run, the window opens, so I think I'm ok [19:53] i'm not sure what your error is due to [19:53] nessita: well I think it may be unity, no idea [19:54] but it starts with an X error [19:54] it's not unity [19:54] what it looks like is the private dbus daemon not starting properly in the qtreactor case, which is weird [19:55] you know what would be awesome right now? [19:55] dobey: I'm afraid to ask [19:55] cake donuts with chocolate icing on top [19:56] ah, fiiiuu [19:56] also, working upgrade to O would be nice [19:56] grmbl [19:56] dobey: I'm getting this http://pastebin.ubuntu.com/610746/ [19:56] dobey: I concur. No good donuts near me though. Another reason to move to Portland [19:57] nessita: ah [19:57] thisfred: or atlanta [19:57] krispy kreme ftw [19:57] nessita: that's exactly what I got [19:57] also, portland sucks [19:57] dobey: KK is good, but not as good as http://voodoodoughnut.com/about.php [19:57] thisfred: but I run the gtk tests at least once, the qt one run :-/ [19:57] dobey: ^ [19:58] dobey: and I love portland, but then I have a high hipster tolerance. I could even live in Brooklyn I think [19:58] thisfred: i will cede that KK does not sell donuts with lots of crazy stuff on them, like bacon [19:58] thisfred: but i think they are different classes of donuts [19:59] thisfred: is like arguing about sicilian style vs flatbread style pizz [19:59] a [19:59] dobey: Voodoo have less exotic ones too, and they are still better :) [19:59] not really [19:59] but w/e [20:01] dobey: there seems to be a timing issue with the paste I shared [20:01] something it fails, sometimes it runs [20:01] any clues? [20:01] and any pizza is fine with me as long as it's not mattress style. Or chicago as some would have it [20:01] no idea [20:02] thisfred: i think you mean queso style [20:02] hehe [20:02] or, Buenos Aires as some would have it [20:02] that too [20:02] well the one we had anyway, I'm sure they have excellent pizza as well [20:05] anyway [20:07] i am hungry again now [20:08] Yeah me too. Thanks a bunch! [20:13] nessita, ok, i'm back. files recopied to laptop, haven't connected but i get around 9kfiles to Makefile in the metaqueue window, and 5k in content queue that wants to upload [20:14] nessita, so, quite fewer files, but still some that wants to reupload [20:15] Ignatus: somehow those have a different hash... [20:16] Ignatus: you can try now restarting syncdaemon, without connecting it, and see how many files are queued up [20:16] nessita, weird, as i just recopied them to hdd... [20:16] Ignatus: but not sure what else to advice. The expert on this is facundobatista, but I think he's not around right now [20:16] facundobatista: are you? [20:18] nessita, ok, thanks anyways mate :) 5k is abit less than all of them :) [20:18] Ignatus: :-) [20:19] volví [20:22] Ignatus: copying files, especially between different kinds of filesystems may change them. If you're able to, you could try using rsync -avz or somesuch, which will preserve all the bits [20:22] alecu: how was lunch/air? [20:22] well, it will try. fat32 throws information away I think [20:23] joshuahoover: ping [20:24] alecu: I added one more needs fixing comment to your branch, with detailed data of what's needed [20:25] nessita, I'm here, but not for long [20:25] nessita, I should have gone already, :| [20:25] facundobatista: manejalo ;-) [20:26] alecu: I did not find anything else, except an error in ./run-tests -qt (which probably should be --qt) which I don't know the cause of [20:26] alecu: so with nessita's proposed fixes done, you have my +1 [20:26] duanedesign: pong [20:27] joshuahoover: have a real quick minute for a PM? [20:27] Ignatus, we'd need logs in debug to check what hashes are being found [20:27] duanedesign: sure [20:31] facundobatista, nessita, status update, i quit the daemon again, and restarted, and this time no files were in the content queue. success :) [20:35] Ignatus: yey! [20:35] Ignatus: you resubscribed your Pictures folder, right? [20:35] otherwise is a false positive [20:36] nessita, yes i did [20:36] Ignatus: great news then :-) [20:36] nessita, indeed, thanks a bunch [20:38] you're welcome [20:41] nessita, revno 149 with the textdomain fix [20:41] alecu: gracias! [20:49] alecu: approved! [20:50] hmmm... lousy internet. [20:52] thisfred: when you have a moment, could you please check out logs in bug #781875? [20:52] Launchpad bug 781875 in ubuntuone-control-panel "ERROR - ReplicationSettingsChangeError: args ( looking [21:05] nessita: added my 2¢. Saw something strange in the dc logs, but have no way to explain it. [21:06] CardinalFang: if you have a second, do you perhaps have an idea what the tealeaves/coffeegrinds/goat entrails/logs in https://launchpad.net/bugs/781875 mean? [21:06] Launchpad bug 781875 in ubuntuone-control-panel "ERROR - ReplicationSettingsChangeError: args ( s/grinds/grounds/ [21:08] """ServerError: (500, ('shutdown', None))""" Wut?! [21:09] alecu: lint issues in tarmac for your branch [21:10] looking [21:10] CardinalFang: indeed [21:11] CardinalFang: I wonder if they were upgrading to natty at that time, and whether that uninstalled couch or something [21:11] nessita, seems like tarmac didn't run "./run-tests -qt" nor "./setup.py build" [21:11] thisfred, let's ask #couchdb [21:11] nessita, I'll have run-tests invoke "./setup.py build" all the times. [21:12] alecu: wait [21:12] alecu: tarmac will not run run-tests -qt unless you tell dobey to set it up to do so [21:13] nessita, right. that's why we need to change run-tests to run both the qt and the gtk tests. [21:13] that's not going to happen until qtreactor is packaged properly in our ppa at least, and ideally in ubuntu as well [21:14] alecu: let's run only run-tests (gtk only) tests for now on tarmac, until ralsina makes a decision aboue qtreactor [21:15] well we will have to package qtreactor [21:15] dobey, or put it in contrib :-) [21:16] no. [21:16] dobey, is pyqt installed in tarmac? [21:16] even if we didn't use qtreactor in control-panel-qt itself, we would still need it to run the tests [21:17] pyqt4 is yes, we already require it for sso lint to pass [21:18] alecu: but the tests fail on something gtk or am I misreading the failmail? [21:18] probably, nm [21:19] thisfred: tests passed [21:19] thisfred: lint failed [21:19] thisfred, when running the gtk tests, pylint is complaining about stuff in ubuntuone.controlpanel.qt.gui and ubuntuone.controlpanel.qt.controlpanel [21:19] nessita: and why is it not using u1lint?! [21:19] ah I see. Why are there weird control sequences in the lint errors? That's my mail client I guess [21:20] dobey: because you didn't propose that branch yet? :-D [21:20] thisfred, that seems to be ANSI color codes [21:20] urgh [21:20] dobey: bad dobey, bad :-P [21:20] i never said i would [21:20] dobey: you should! [21:21] thisfred: because nessita made it use pylint instead of u1lint, and has the pylintrc doing colored output [21:21] right [21:21] bueno bueno bueno [21:21] shall I make a branch before someone gets hurt? :) [21:21] I'd prefer that alecu's branch lands first [21:21] to avoid conflicts on run-tests [21:22] sure [21:22] "pylint --ignore ui ubuntuone/" [21:23] alecu: we won't be able to pass arguments to u1lint, I think. dobey, is that so? [21:24] hmmm [21:26] nessita, I've changed it to always build the *_ui.py files, even when running gtk tests. [21:26] nessita, that should fix this problem. [21:26] alecu: but not the qtreactor missing module, so you need a pylint disable there [21:27] nessita, right. [21:28] nessita: right [21:29] i guess i need to add an rcfile thingy so we can specify ignored paths [21:29] nessita, done. [21:29] and maybe add arguments for such things as ignored paths [21:29] alecu: groso, wanna re-approve it after the commit has be scanned by LP? [21:39] just did it [21:42] GRRRRRRRRRRRRRRRRRRR [21:42] on 11.10: gsd_ubuntuone_class_finalize used but not defined [21:42] on 11.04 with branch to fix that: gsd_ubuntuone_class_finalize defined but not used [21:43] sigh [21:51] ah, qtreactor [21:52] alecu: you need to pylint disable= that one, or package the reqactor :) [21:53] I'm not packaging the reaqtor [21:55] At some point it will start to look like the more attractive option :) [21:55] alecu: nessita dobey: In all seriousness, should I have a go at it? Shouldn't be too hard... [21:55] the packaging I mean [21:56] thisfred, sure! thanks! [21:56] Since I'm gonna go for Ubuntu Developer as one of my goals... ;) [21:56] kk, on it. alecu: is the git repo the officialest home? [21:57] anyway, the branch has already merged, but the qtreactor packaged sounds like something we may need. [21:57] thisfred: you can certainly try. I would advice some advice from dobey or rodrigo [21:57] thisfred, afaik, it is. [21:57] thisfred: it wouldn't be hard if it wasn't a totally broken repository [21:57] thisfred: it has no setup.py and such [21:57] dobey: wouldn't be fun if it was easy [21:57] ah, so it's not even a python package. Superfun! [21:58] exactly [21:58] thisfred, it is a python package, since it has __init__.py on it. [21:58] and i don't know if it's ever actually been released [21:58] alecu: fair enough. I meant the word that means the thing with a setup.py in ;) [21:58] thisfred, oh, you mean a python package :-) [21:59] package in the pypi sense [21:59] not the pypy sense [21:59] dobey: no setup means that it runs without building, which means writing a setup should be trivial. [22:00] I meant Cheese Shop :) [22:00] anyway, if we're going to use it, we should be active in upstream [22:00] cheese shop that doesn't sell cheese? [22:00] naming things, cache invalidation, character encoding, and off by one errors are the 3 hardest problems in CS [22:00] a cheese pi pie? [22:01] dobey: I'll kick the setup.py back to them [22:01] thisfred, very sensible to include "char encoding" [22:01] if there is a they [22:01] I'm not gonna be they [22:01] well there are commits in git by more than one person [22:01] but not sure what that means [22:01] sock pupperys [22:03] surprise, fear, and character encoding are the 3 hardest problems in CS [22:03] amongst the hardest problems in CS are surprise, fear, and an almost fanatical devotion to the pope [22:04] alright, i am outta here. have a good weekend all [22:04] later dobey === m_conley is now known as m_conley_away [22:20] * thisfred wonders how cross platform the qtreactor is, since it inherits from posixbase.PosixReactorBase [22:21] oh nm, it has a win32install so I suppose it's at least possible that it works [22:23] * nessita -> eods [22:23] bye all! [22:25] bye nessita [22:47] U1 on PC is broke lately :( I didn't even touch one folder, and got a self-conflict. My other webUI visible folders are not getting down to my PC (or they are, but very slowly, like, from time to time) [22:47] If I can provide some more logs/details, I'll be happy to.