=== dduffey is now known as dduffey_afk [05:03] What is the suggested way to preserve the execution bit on synced scripts? === jono is now known as Guest29470 [07:33] morning [07:56] Good morning :) === pedronis_ is now known as pedronis [10:00] hi guys [10:00] is http://pastebin.com/XHu7YAJc normal when started 'ubuntuone-installer' in terminal? [10:01] pl_PL.UTF8 locale [10:03] Mornin' all! [10:07] hello [10:17] hrw, this is not normal, do you have ubuntuone-control-panel-gtk installed? [10:18] I just installed ubuntu-client-gnome [10:18] on precise === zyga_ is now known as zyga-xchat [10:22] hrw, ubuntuone-client-gnome is only for the nautilus plugin - the control panel is now a separate package [10:22] ok [10:23] did not had it installed [10:23] with it installed I got ui [10:23] so missing deps or missing check? [10:25] i have noticed users having to install that seperately [10:25] hrw, it looks like the code in installer does not handle encoding properly, installing Polish locale to test [10:26] hrw, was this a clean install or an upgrade to precise? [10:26] upgrade [10:27] duanedesign, ^ the control panel might have been uninstalled [10:27] rye: it was uninstalled for sure === yofel_ is now known as yofel === Guest50763 is now known as p === p is now known as Guest62719 [10:47] hmm... amount of free storage depends where you check [10:47] https://one.ubuntu.com/account/ says "0% in use (5.1 MB of 45.0 GB)" [10:48] ops, sorry - wrong check in one place [11:03] karni: http://marcin.juszkiewicz.com.pl/~hrw/video_lut_02_2012_0.mp4 shows crazy notifications which I mentioned [11:04] hrw: sweet, checking it out [11:08] hrw: this is clearly a problem of clearing the notification when not all transfers have been completed. acutally, I'm currently working on new queues, so this should be resolved soon. sorry, and huge thanks for the mp4 ! [11:14] karni: no need to be sorry [11:14] karni: I have good mood for helping this week ;) [11:23] hrw: :) [11:27] karni: and too many times lacked a 'simple' way to share files phone<>machines<>machines [11:27] karni: looks like 5GB of dropbox will work just fine so far ;D [11:32] karni: added screenshot to bug 853975 so you have it available [11:32] Launchpad bug 853975 in ubuntuone-android-files "Notification text color should use system default (affects: 1) (heat: 6)" [Low,New] https://launchpad.net/bugs/853975 [11:43] is there a way to browse my U1 folders from the Ubuntu desktop? [11:50] jml, what do you mean? === Guest62719 is now known as mandel [11:50] jml, sorry needed to chante the nickname :P [11:50] jml, if the machine is in sync you have your files in the system, so you just browse them like normal files [11:52] jml: FWIW, you can also browse U1 from any Internet-enabled device with an Internet browser ;) http://one.ubuntu.com [11:52] or better, http://one.ubuntu.com/files === eu is now known as Guest67608 [12:11] mandel, ping [12:11] facundobatista, pong [12:11] facundobatista, buenos dias! [12:14] @ping [12:14] pong [12:15] facundobatista, hello? [12:16] mandel, hola :) [12:24] good morning! [12:25] ralsina, buenas [12:40] ralsina, got time for a super quick mumble? I have some design questions [12:40] mandel: sure, starting mumble [12:40] ralsina, and having a qt guy to help would be nice :) [12:41] let me find my qt guy hat... [12:42] mandel: hey there! [12:42] mandel: have you started with the qt dialog? [12:42] mandel, i can help you too if you want [12:51] mandel: ping [12:55] nessita, mumble witn ralsina [12:55] nessita, give me some mins [12:55] gatox, is not qt specific [13:04] nessita, ok, done with mumble [13:04] nessita: ping, can we have a quick mumble around 12? [13:05] nessita, I have talk with lissete regarding the design and with ralsina about some concerns I have with the spawning of the dialog regarding modlaity of it and other ui integration [13:05] mandel: the dialgo will not be modal, but the controlpanel will block putting the loading overlay [13:05] ralsina, nessita: hola, do you know u1sdtool is kind of broken in precise? [13:06] mandel: like it already does when querying the web, and syncdaemon [13:06] verterok: broken how? [13:06] verterok: Idon't know that it is broken in any specific way [13:06] verterok: mainloop not returning when quitting is a bug in glib, FYI [13:06] ralsina, nessita: after executing it just sits there and never exit [13:07] verterok: yeap, not broken [13:07] oh [13:07] * verterok really hates glib [13:07] not by us, at least :-) [13:07] nessita, well, we need to talk about this, later we can mumble about my fears :) [13:07] verterok: you can see that GTK apps are not exiting either (try executing ubuntuone-control-panel-gtk in the terminal and closing that up) [13:07] nessita, ralsina: give alecu some free time to work on pydbus :p [13:07] mandel: what fears? [13:08] nessita: that works [13:09] verterok: hum... perhaps is the new GLib mainloop that is not working [13:09] verterok: and controlpanel is still old glib [13:09] so, it was broken by us after all? :) [13:10] verterok: I'll debug after feature freeze. Can you please open a bug? [13:10] nessita, interaction between the dialog and the control panel, if webclient spawns the dialog in a diff process the dialog shows and then the user closes control panel, what happens there? [13:11] nessita: u1sdtool is using the glib2reactor [13:11] mandel: the spawned process will quit, since the parent process will die [13:11] verterok: then we should debug this further (after feature freeze) [13:11] verterok: anyways, we did not change anything in that, so I'm not sure what that is [13:11] nessita, also, the dialog will not be shown as the top window, how are we going to do that? [13:13] nessita, the implementation of the details is what I want to get right, the dialog is already done with qtdesigner and will get it running this afternoon with storing the creds etc.. ui is not a problem :) [13:13] mandel: so, I would advice not worry about that now, we need the dialog built before the next friday so we can try to make it before the freezes and avoid asking freeze exceptions [13:14] nessita, as I said, the ui is not a problem at all, it will be there on monday, I already have arranged a meeting with lisettte to look at the implementation [13:14] mandel: when the dialog is landed, we can worry about the other stuff, anyways we'll need to sort that out for sso as well (if that turns out to be an issue) [13:14] mandel: we can make the popup alter the launcher, for example [13:14] mandel: or be raised on top of other windows [13:15] nessita, certainly, we need to think a little about these interactions but we will be there on time for ui freeze that is not to worry [13:15] mandel: there will be relationship between the callers and the proxy settings, but there is not much we can do [13:15] nessita: this fixes it: http://paste.ubuntu.com/826339/ [13:16] mandel: I understand your concern. Anyways, those interactions are not top priority in my head, since we still don't have proxy support working. If later we have a bug report about this, we can fix. [13:16] nessita: I think it might be defer.inlineCallbacks related issue [13:16] verterok: that's ugly! and also, we're having the same issue with plain mainloops in sso, where we have no reactor at all [13:16] nessita, sure, I just wanted to make sure we are doing the right thing from the beginning, or that we at least have the possible problems in the schedule [13:16] verterok: so there should be some other thing stuck in the mainloop, and the callLater may be giving an extra spin to the mainloop, no? [13:17] nessita: no idea, just tried it and it works :) [13:17] mandel: right. We don't have much other options considering that the proxy dialog will be in sso [13:18] mandel: there are ways to set a window transient for other window, and if needed we'll try to make that work. But again, I would like to see proxy actually working before worrying about these things (if proxy does not work, then a pretty shinny dialog is useless) :-) [13:18] ralsina: not sure how you feel about that ^ [13:19] nessita, true, pretty and not working is like nothing at all [13:19] nessita: the main problem is that because of focus prevention, usually the new window will popup below the other [13:19] mandel, nessita: I agree that we can fix it after we see the problem , though [13:20] nessita,mandel: since the solution is probably pretty minor coding, changing thing in or out of process [13:21] ralsina: I tested the spwanning of any UI process, and I got the new window on top of the currents one. Examples I did was spawning ksnapshot using a glib spawnner, and also using a qprocess [13:21] in both cases the ksnapsoht was raised on top of my terminal [13:21] (but yes, it was not related to it) [13:21] nessita: did you start it from a window that had input focus? [13:22] related as in "is child of" [13:22] ralsina: yes, I was executing the spwanner script in the terminal, so i hit enter in it right before the windows opened [13:23] nessita: the terminal i a special case, it's supposed to go to the background when you spawn things from it [13:23] nessita: but agreed about not worrying too much right now [13:23] ralsina: well, then I would suggest evaluate this in particular when the rest is working [13:23] nessita: yes. I will try tomake a toy example [13:24] ralsina: as long as we have the new dialog in trunk and released to ubuntu, we can not use it until after UI freeze. I would to have the dialog in trunk, and the projects working only with user:pass@foo.com [13:24] I would like* [13:25] so we can know that proxy work, at least for user:pass@foo.com settings [13:25] nessita: yes, at least that [13:26] ralsina: I'm worry that we can land features until next Friday 10th (and after that will require a feature freeze exception, which involves an exception per bug per branch) [13:26] worried* [13:26] nessita: ack [13:26] so, I think we should prioritize features and UIs on top of polish, that should come right after [13:28] yes, we can always get the polish in the 0-day SRU [13:28] thisfred: you will love http://www.youtube.com/watch?feature=player_embedded&v=x8bc_ZyORbM [13:29] me click in trepidation [13:29] ralsina: or between UI freeze (Feb, 23) and final freeze (March, something) [13:29] :-) [13:29] that timeframe will not require any SRU [13:29] nessita: I am trying to be pessimistic here! ;-) [13:29] dobey: hola! approved your client update branch [13:30] ralsina: nice [13:30] ! [13:30] hola nessita! thanks [13:30] mandel, nessita: toy experiment done, the child process spawns in front of the parent one :-) [13:30] ralsina: yey [13:31] mandel: so, worst case, we disable the whole UI until the child process exits [13:31] ralsina: hey! that's the good case :-D [13:31] nessita: hehe [13:31] ralsina: we put the Loading overlay on top of the parent, and remove it once the webclient deferred's returns [13:31] nessita: it's not truly modal, but it's close enough [13:31] yeap [13:32] thisfred: holland is such a cesspool of anarchy! [13:32] hahahaha [13:33] Next time someone asks where I'm from that's what I'll answer [13:33] haha [13:34] dobey: got 15 minutes for me? [13:37] nessita, approved: https://code.launchpad.net/~nataliabidart/ubuntu-sso-client/qt-in-linux/+merge/91164 [13:52] ralsina: would you have a chance to review https://code.launchpad.net/~nataliabidart/ubuntu-sso-client/qt-in-linux/+merge/91164 ? [13:52] nessita: of courses! [13:54] nessita: so i am not sure what's wrong with your branch in the precise tarmac. it's quite odd [13:55] dobey: I will try merging the 2 first test commands into one, since both use the gi reactor [13:55] dobey: but will do that once I get all the tarballs done [13:56] nessita: i don't think that's the problem [13:56] dobey: do you have some theory? [13:59] nessita: well i was trying to run the tests manually, in the precise tarmac instance, as the tarmac user, yesterday. the gtk+ tests were extremely slow after a certain point. individual tests were taking several hours, and the CPU was active the whole time. [13:59] dobey: huh? [13:59] that's what i said [14:01] dobey: heh [14:02] dobey: I'll debug it later today, no worries === dduffey_afk is now known as dduffey [14:53] * mandel back [14:59] ugh the time [15:00] dobey, it is indeed, what shall we do? [15:01] oh i don't know [15:01] ralsina: ? [15:01] let's skip it [15:01] I think I know what everyone is doing [15:01] doh! [15:01] and half the team is at sprint. [15:01] ok, I'll be there next time. Accounts is starting to wind down [15:02] and even if it weren't I need something besides that to keep my sanity [15:02] thisfred: I suggest knitting [15:03] In the interim, feel free to ask me for reviews, if they're not *too* windowsy [15:04] ralsina: I feel my time could be better used than to produce asymmetric itchy scarves. [15:04] thisfred: hehe [15:05] (Last time I tried knitting I was 10, and I sucked at it. I have no reason to believe that has improved) [15:05] thisfred: if it helps, I proposed you for the "bored out of his mind and has not gone on a killing rampage yet" award [15:05] haha [15:05] ralsina, so I ate fast to skip the stand up.. :*( [15:05] mandel: sorry [15:05] ralsina, nah, no worries hehe I like to complain [15:05] mandel: I have a few too many things in the air and forgot about the time [15:08] there is no time [15:08] also; ffs. i am not paying $3K for a bicycle. [15:08] unless it's made entirely out of nanobots [15:15] ralsina, FYI there is an info web that the web&mobile team has to do where we give information about proxy settings etc [15:15] ralsina, I know that lisettte already mentioned this to roberta [15:16] mandel: ack [15:18] ralsina, did you get lisettte mail? [15:19] mandel: only the one natalia responded [15:27] dobey: hola! did you u1client landed? [15:27] (did not check mail yet) [15:27] nessita: nope, missing package. trying again [15:30] ralsina, yep, that is the one, is the summary of the catch up I did with her after talking with alecu and nessita, trying to keep everyone in the same page [15:30] mandel: ok then [15:34] dobey: so, I can reproduce locally that the tests get stuck and trial is uging 05% of the CPU [15:34] 95% [15:34] dobey: perhaps is a bug in the gi reactor? does that make sense? [15:34] nessita: really? how so? [15:34] they run very fast for me locally [15:34] dobey: are you up to date re: packages? [15:35] dobey: mine is stuck in test_on_email_validated_is_not_called ... and top shows: [15:35] 3398 nessita 20 0 416m 215m 13m R 95 5.7 3:34.93 u1trial [15:35] sort of; i am holding some back because i actually want to use my computer and graphics drivers are a necessary part of that [15:35] dobey: X is for sissies! :_P [15:35] ;-P [15:35] :-P [15:35] (typing right is for sissies as well :-D) [15:36] heh [15:36] dobey: and the same test is also getting pseudo-stuck for alecu, in his precise install === m_conley_away is now known as m_conley [15:36] I've not updated since getting to the sprint [15:37] nessita: hrmm, maybe xvfb is the problem then [15:37] or zeitgeist [15:37] dobey: I tried without xvfb, same behaviour [15:37] dobey: ZG on sso? :-/ [15:38] oh right, no [15:38] what could it be [15:38] the kernel? [15:38] 64 bits? [15:39] i'm on 32 bit [15:39] I'm on 64 [15:39] is tarmac 64 bits? alecu and me are [15:39] tarmac appears to be 32bit in the vm at least [15:39] boo [15:39] :-) [15:45] dobey: so, once reproduced, alecu taught me how to strace a running process, and I've got this trace: http://pastebin.ubuntu.com/826507/ [15:45] dobey: "looks" like is stuck polling [15:46] nessita: right which means it's not really doing anything. what are those fds for exactly? [15:46] dobey: how can I find out? [15:46] look for the open() calls that result in those fd numbers (4/5) [15:47] dobey: http://pastebin.ubuntu.com/826510/ [15:47] there is no open() in that fragment (ie was already polling when I attach the strace to it) [15:48] oh, because you attached [15:48] yes [15:48] dobey: because I need to run the whole gui suite to reproduce, can't reproduce with only the testcase [15:48] kill it and do strace -o failing-test.log -Ff u1trial -r gi failing_tests.py [15:48] oh [15:49] ok [15:49] so [15:49] kill it and do strace -o failing-test.log -Ff xvfb-run u1trial -r gi --gui [15:49] dobey: reproduce agian and inspecting /proc [15:49] lr-x------ 1 nessita nessita 64 Feb 2 12:48 3 -> pipe:[127226] [15:49] lrwx------ 1 nessita nessita 64 Feb 2 12:48 4 -> socket:[129100] [15:49] lrwx------ 1 nessita nessita 64 Feb 2 12:48 5 -> anon_inode:[eventfd] [15:49] or look in proc i guess [15:49] so it's a unix socket [15:50] dbus? [15:50] dobey: hum... will check, but I removed all dbus from this code (meaning that I now use a "multiplatform" getter for the backend, and I'm mocking that) === ralsina_ is now known as ralsina [15:52] oh. what would be using a socket then? [15:52] dobey: xvfb? the reactor? [15:53] the reactor shouldn't be. are those the FDs for xvfb or u1trial that you were looking at there? [15:53] dobey: those are from u1trial [15:54] dobey: want me to inspect also xvfb? [15:54] no. seems it's definitely in the tests [15:54] hrmm [15:54] dobey: iirc (I made several attempts), I think I reproduces this without even using xvfb [15:56] i just upgraded a bunch of stuff, trying again on my machine [15:56] dobey: ack [15:56] seems to be going fine here [15:56] what kernel version are you running? [15:57] dobey: the testcase that is hanging here is SignalsTestCase, and if we skip that put, the next testcase hangs [15:57] dobey: Linux ernst 3.2.0-12-generic #21-Ubuntu SMP Tue Jan 31 18:48:57 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux [15:57] nessita: do you still have 3.2.0-10 installed? [15:57] let me see [15:57] yes [15:57] want me to boot in that one? [15:58] can you reboot into the -10 kernel and see if it still fails please? [15:58] dobey: yes. FYI, alecu as -11 and he can reproduce [15:58] i am using -10 because of nvidia driver issues with the newer ones [15:58] I was able to reproduce, but three years ago when Amelia was born. [15:59] heh [15:59] dobey: fd's status from both u1trial and xvfb from proc: http://pastebin.ubuntu.com/826530/ [15:59] will reboot now to -10 [15:59] brb [16:00] ok, thanks [16:01] oh, fd 5 is inotify related i guess === eu is now known as Guest17386 [16:02] my branch is failing somewhere else now, but not sure where exactly. meh [16:04] dobey: your branch == u1client + gi branch? [16:05] yes [16:05] ok, will reboot to -10 kernel [16:05] bbrb [16:05] it's causing a unicode error in tarmac [16:10] ralsina: are we having another chat at 19:00Z still? or was that waht you asked about this morning? [16:11] dobey: it's what we talked about this morning [16:11] ok, cool [16:16] dobey: running the whole suite in -10 [16:16] ok [16:17] dobey: reproduced the bloackge [16:17] blockage [16:18] dobey: same data as before: strace shows polling on 4 and 5, and /proc shows: [16:19] lrwx------ 1 nessita nessita 64 Feb 2 13:18 4 -> socket:[15980] [16:19] lrwx------ 1 nessita nessita 64 Feb 2 13:18 5 -> anon_inode:[eventfd] [16:19] huh [16:20] any other ideas? :-/ [16:21] do you have a sacrificial dagger handy? [16:21] Ran 284 tests in 54.643s [16:21] :( [16:28] dobey: running the whole test_gui.py (so the stucking hapens) with strace [16:29] ok [16:30] will have lunch in the mean time [16:30] dobey: thanks for helping debugging this [16:32] sure. it's annoying me :) [16:32] and it's a matter of concern if we're supposed to switch the world over to gi :) [16:34] right [16:34] FYI, log file is 175M so far :-/ [16:35] yeah, stract logs a LOT of stuff [16:35] and i can't type it seems [16:35] dobey: is contagious [16:35] ok, I'm gone [16:35] ok [16:40] you finally realized that typing bad is the correct way to do it :P [16:47] i realize i need to make a decision on what i'm going to have for lunch [16:51] dobey, babies! [16:52] all, EOD for me, leaving a little early 'cause I'f not feeling that well. I think the flu got me.. bloody weather [16:52] see you all tom [16:53] mandel: of course, but it's which flavor of baby to consume, that is the question. [16:53] cow, shrimp, chicken. so many choices. [16:54] dobey, I will not comment due to fear of being fired due to braking the employee hand book in too many levels [16:54] * ralsina ====> 😋 (that is 'FACE SAVOURING DELICIOUS FOOD' (U+1F60B)' BTW) [16:54] lol === ralsina is now known as ralsina_lunch [17:40] alright, i need to get lunch [17:40] bbiab [17:54] dobey: I added a suggestion from alecu, and seems fixed in my computer now! I'm now destroying the window that is created on each test run === ralsina_lunch is now known as ralsina [18:26] nessita: heh, what was that? [18:26] the suggestion i mean [18:27] dobey: self.addCleanup(self.ui.destroy) :-) [18:27] dobey: so, apparently, we had tons of windows that we gathering tons of X events [18:27] heh [18:27] dobey: and with each test, we were having one more window to sent events to [18:27] right [18:28] dobey: branch landed now :-) [18:28] wonder why it wasn't slow for me though [18:28] dobey: yeah, it also worked for me in my desktop PC... so no idea [18:28] or why this issue only came up now [18:28] dobey: alecu and I are both running lenovo's... but his is amd and mine intel, so ??? [18:28] maybe gtk3 does something different, which made it an issue [18:28] so "???" <- no idea [18:29] dobey: right. Is fixed now, alecu reported the issue, and will fix controlpanel as well [18:29] (eventually) [18:29] I will? [18:29] no [18:29] :-) [18:29] sorry, *I* will [18:29] heh [18:30] dobey: I have not seen any email about the archive being unfrozen... shall I expect one? [18:31] nessita: i saw a mail asking to hold off on uploads as the distro rebuild is happening [18:32] dobey: ah yes, saw that too, from doko [18:32] yep [18:32] dobey: so, we'll get an email announcing the repo is "open" again? [18:33] nessita: i don't know. let's see what tomorrow brings :) [18:33] sure [18:34] dobey: i'm asking these things because I did not pay that much attention to these details before, I just asked for sponsorship and that happened [18:35] dobey, I'm getting this error when running u1trial with the gi reactor: [18:35] g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream returned 0 bytes on an async read (g-io-error-quark, 0). Exiting. [18:36] dobey, it happens after the "PASSED (successes=420)" message is shown. [18:36] nessita: yep, i understand why you're asking. i'm used to the hard freezes though where uploads are blocked until it's unfrozen again [18:36] alecu: hrmm [18:37] alecu: i don't see it, so not sure what it is exactly. you see that in sso i presume? [18:37] dobey, I see that when running tests [18:38] dobey, not only I see that, but tests get interrupted with nessita's latest branch [18:38] interrupted? [18:38] dobey, yes, since there are three invocations to u1trial in the same run-tests script in trunk now. [18:38] dobey: and the script has set -e [18:38] so it will abort on error [18:39] right, but there shouldn't be any errors [18:39] right [18:40] the 'foo' thing is really annoying though [18:40] dobey: yes, I agree [18:47] yay [18:47] that's much nicer without all the Gtk-WARNINGs [18:49] dobey: how did you fix it? [18:50] connected the signal [18:50] dobey: by hand? so builder is not doing it? [18:50] seems that way. pushing a branch now [18:51] dobey: ack, thanks [18:51] i hope nothing else is not getting connected up properly [18:51] dobey: shall we report the bug? [18:52] https://code.launchpad.net/~dobey/ubuntu-sso-client/avoid-foo-warning/+merge/91327 [18:53] nessita: not sure. it may just be something that's different in glade vs. gtkbuilder, should poke at the docs first to see what's correct [18:54] if the ui.glade file is correct as-is for gtkbuilder, then it's a bug. [18:56] the glade file has: [18:56] [18:56] [18:56] clicked will work [18:57] true, i guess clicked does work, since i managed to log in :) [18:57] so i suppose we should file a bug, yes [18:58] dobey: you do it or I do it? [18:59] you can :) [18:59] reminds me that i need to file a different bug though, for ubuntu [19:00] dobey: this bug is for glade, right? [19:01] nessita, review please https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/pages-migration/+merge/91331 (i executed the tests in all the platform like a gazillion times..... it should be fine! :P) [19:01] nessita: i'd say gtkbuilder. glade outputs the correct data there it seems [19:01] nessita, and this one too: https://code.launchpad.net/~diegosarmentero/ubuntuone-windows-installer/pages-migration/+merge/91333 [19:01] gatox: ack! [19:01]