[00:00] DiegoSarmentero, I tried it like this [00:00] stop: 0 #e44e19,stop: 1.0 #fecfc2); [00:00] alecu, i tried [00:00] it [00:00] alecu, looks nice [00:00] alecu, do you want me to submit that? [00:00] yup, it's not perfect but will do till they choose a better combination. [00:00] DiegoSarmentero, sure [00:00] alecu, definitly looks better [00:00] :P [00:03] alecu, ready! [00:05] DiegoSarmentero, cool, I've approved it. thanks! [00:05] alecu, :D [00:07] karni: ok :) [00:07] duanedesign: Josh took care of it [00:10] aha, i was just about to ask :P [00:11] karni: i updated the wiki to reflect the change to mobile contacts...it will be nice when the site reflects it as well [00:12] duanedesign: Josh takes care of these kind of support messages (concerning contacts) [00:13] /s/messages/questions [08:27] morning all, ill probably not be too attentive on IRC today since im working on 1 monitor because im waiting for a package downstairs but im working on testing the bundle maker all day anyway [08:37] Good morning all! [08:38] happy friday JamesTait [09:46] * fagan break [10:19] Wow windows takes bloody ages to install [10:19] I gave it 2 processors and all and its still taking an hour [10:20] ~, ]; [10:21] * fagan just made an emoticon by dropping his keyboard [10:34] ooh its working on the clean install of XP [10:35] Ill wait till ralsina_ gets in to +1 the branch just in case [10:47] fagan: here I am :-) [10:53] ralsina_: working perfect [10:54] yay === ralsina_ is now known as ralsina [10:54] ralsina_: so I can +1 the branch if thats all thats needed [10:56] ralsina: so is it ok to +1 the branch then since its working or do I need to code review too [10:57] Do a +1, I will get a code review from nessita in a bit [10:57] kk cool [10:59] ralsina: so whats the task now? [11:00] I was thinking if there isnt any task I can do up some examples on using the APIs or what ever if you dont have any task in mind [11:00] fagan: could you uninstall, and re-try with the latest version of that branch? It has a few changes [11:00] ralsina: what do you mean by uninstall? [11:01] did you update the branch more? [11:01] oh,right, you tried the bundles [11:01] yep did [11:01] ok, now try the installer :-) [11:01] kk [11:01] Get bitrock installbuilder qt edition and follow the README [11:01] ralsina: will do [11:01] create an installer, try it n the clean VM [11:25] morning all [11:27] morning duanedesign [11:29] * duanedesign needs to install the Windows Client today === teknico is now known as teknico_away [11:30] duanedesign: I can give you the .exes for it if you want [11:30] you just need to unzip it and double click to test [11:30] ill pm you a link [11:31] fagan: cool, thank you [11:35] * fagan break [11:57] good morning everyone! [12:02] good morning nessita [12:02] nessita, we have (another) big problem, and sorry to start the day this way :-( [12:02] basically, logging is broken in the way we use it on windows [12:03] bug #811006 [12:03] Launchpad bug 811006 in ubuntuone-client "Problem with logfile rollover on windows (affects: 1) (heat: 6)" [High,In progress] https://launchpad.net/bugs/811006 [12:03] We are using a RotatingFileHandler and we are spawning subprocesses. That mix is a no-no. [12:04] Hola nessita [12:05] ralsina, spawning subprocesses? [12:05] facundobatista: yes, to activate ussoc from syncdaemon. [12:06] facundobatista: we don't have dbus to do it for us :-( [12:07] ralsina, so, SD spawn a process for the sso part, why it complicates the logging? [12:08] facundobatista: basically that doesn't work on windows, it's complicated and I understand half of it: http://bugs.python.org/issue4749#msg101822 [12:09] ralsina, ah, because the other process also wants to log in the same file? [12:09] facundobatista: not even needs to do that,it seems [12:09] facundobatista: it seems the file handle is inherited by the child process and then it's broken for the parent [12:11] my phone went deade for 10 minutes, if anyone spoke to me, please repeat :-) [12:15] nessita: bug #811006 [12:15] Launchpad bug 811006 in ubuntuone-client "Problem with logfile rollover on windows (affects: 1) (heat: 6)" [High,In progress] https://launchpad.net/bugs/811006 [12:15] ralsina: hey there! [12:15] hi! [12:15] I guess my hello was missing too :-) [12:16] ralsina: by any chance, did you cacth up with mandel? I Was wondering if he said something related to the email I sent last inght [12:16] didn't see him [12:16] is there a ubuntu one openid url I can use? [12:17] ralsina: he was online from 8-10:30 ish and then left [12:17] fagan: thx! [12:17] jo-erlend_: https://login.ubuntu.com [12:17] jo-erlend_: im pretty sure thats it [12:17] fagan, I mean in order to identify to other sites. [12:18] Then try login.ubuntu.com/~yourlaunchpadid [12:18] * fagan checks too [12:19] nope doesnt work [12:19] nessita, ralsina: http://bugs.python.org/issue4749#msg89218 [12:20] facundobatista: I can try that, see if it helps [12:20] facundobatista: does syncdaemon call subprocess? :-/ [12:20] nessita, no that I'm aware of [12:21] right [12:21] nessita, but ralsina tells me that we're spawning subprocesses [12:21] not within syncdaemon [12:21] ralsina: ^ [12:21] nessita: it calls Popen for the tcp activation of ussoc? [12:21] jo-erlend_: could you not do it with lp instead? [12:21] ralsina, where? [12:21] ralsina: but that's syncdaemon itself... [12:22] ralsina: I'm not sure that counts... does it? [12:22] nessita: well, since the bug happens it might [12:22] I am about to test it and tell you ;-) [12:23] No, doesn't help [12:23] ralsina: how it happened? I mean, it happened when closing syncdaemon? when running a new syncdaemon? or in the middle of a run? [12:23] nessita: I apparently left syncdaemon running [12:23] right [12:23] So I probably have a big log [12:23] And now when I start it, I get that [12:23] ralsina, I'm lost, where SD is spawning other process? [12:24] jo-erlend_: actually putting https://login.ubuntu.com works [12:24] ralsina: syncaemon always rotates logs at startup. YOu may have another syncdaemon running [12:24] nessita: I don't [12:24] ralsina: how do you kill the a running syncdaemon? [12:24] s/the a/a [12:24] nessita: with the process manager [12:24] But it doesnt have a fancy page like on lp when you click the link [12:25] ralsina: hum... [12:27] fagan, nice! Thanks :) [12:29] ralsina, I'm lost, where SD is spawning other process? [12:30] facundobatista: I don't know how it's done, but on windows, when sd needs ussoc, it's started. [12:30] facundobatista: I am trying to follow that code now [12:30] ok [12:32] facundobatista: when sd calls UbuntuSSOClient.connect() that checks if the sso port is unused, and starts sso using subprocess.Popen [12:32] nessita: ^ [12:32] unless I completely misunderstood the code, of course [12:36] And if I start ussoc manually before sd, it works perfectly. So I think that's it. [12:36] ralsina: but... [12:36] ralsina: ah! I understand [12:36] nessita, do you? [12:36] facundobatista: yes, let me explain quickly [12:36] facundobatista: on linux, we acess the SSO service by dbus activation, yes? [12:37] yes [12:37] facundobatista: on windows, there is no such thing, so we need to start the SSO service "by hand", that is done via subprocess.Popen [12:37] facundobatista: and sso may be inheriting syncdemon fds, which has no practical sense [12:38] yes? [12:38] bad news is, making it close_fds doesn't fix it [12:39] ralsina: you should kill sso now [12:39] ralsina: since sso already has access to the file [12:39] no? [12:39] nessita, yes [12:39] I killed it first.But I'll try again [12:40] ralsina, so, you tried close_fds on the Popen of tcpactivation.py, line 130 ? [12:40] facundobatista: right [12:41] good news, that does fix it! [12:41] ok, telecom my cut off my phone line any minute now [12:41] ralsina, -.- [12:41] for a couple of minutes they say [12:41] bad news is, making it close_fds doesn't fix it [12:41] good news, that does fix it! [12:41] nessita, we'll miss you [12:41] facundobatista: would you prefer it to be the other way around? ;-) [12:42] ralsina, I don't understand what works and what doesn't [12:42] facundobatista: looks like I had a stale process somewhere [12:42] ralsina, ah, ok, ok [12:42] so adding close_fds makes the logging problem go away [12:42] so, close_fds works [12:42] makes sense [12:43] ralsina, what you will *not* be able to do is to log from that new process into the same file, but we don't want that, so we're happy [12:43] cool, this really scared me. The proposed "real" solution sounds like a lot of work. [12:44] ralsina, there's no "real" solution more than this [12:44] facundobatista: yes, do a socketlog server [12:44] ralsina, the other thing that is proposed in the python bug tracker [12:44] facundobatista: and use a sockethandler [12:44] ralsina, is for spawning different processess that all wants to log [12:44] in the same file [12:44] so, as there's no good way to synchronize them all to use the same file [12:45] facundobatista: or if you need access to stdin/out/err, which close_fds closes too [12:45] you need to log from a single process, and all the other ones to send the log lines to that one [12:49] standup in 10 [12:50] fagan: alecu and mandel won't be here, DiegoSarmentero finished his contract, nessita has obvious connectivity issues. Short standup! :-/ [12:51] ralsina: yeah sounds like it [12:52] And dob_ey is off [12:53] hello again!!! [12:53] hello nessita! It seems using close_fds works [12:53] facundobatista, ralsina: I missed everything from [12:53] (09:41:36 AM) facundobatista: bad news is, making it close_fds doesn't fix it [12:53] (09:41:41 AM) facundobatista: good news, that does fix it! [12:53] hey nessita standup with you, ralsina and I in 7 [12:54] nessita: *but* it means the spawned processes can't write to stdout/stderr. That should not be a problem, right? [12:54] .... [12:54] nessita, pasting in private [12:54] ralsina: well, nothing too serious I guess [12:55] nessita: at least we won't see it ;-) [12:56] nessita: also, while telecom lets you see it, is this one known,or I file it? https://pastebin.canonical.com/49806/ [12:57] ralsina: I just got a call from a Mr Telecom saying that I should not get more issues today (tm) [12:57] Nice of Mr Telecom :-) [12:57] ralsina: we had one like that that mandel said it was fixed [12:57] ralsina: so, please file a new one [12:57] nessita: ok, will file it [12:57] since it may be the same not fixed or may be a new onw [13:00] moi [13:00] nessita and ralsina standup [13:00] me === teknico_away is now known as teknico [13:01] fagan: go [13:01] DONE [13:01] * test the updated py2exe script IRL [13:01] TODO [13:01] * do the bitrock bit before that and test that bit [13:01] BLOCKED [13:01] * Nope [13:01] nessita: go [13:01] DONE: bug #810440, bug #810640, reviews, UDF testing + long report, meetings [13:01] TODO: interview windows candidate, UDF debugging by looking at logs [13:01] BLOCKED: nopes [13:01] NEXT: ralsina [13:01] Launchpad bug 810440 in ubuntuone-control-panel "QT UI: 'explore' button is too wide and looks ugly (affects: 1) (heat: 6)" [High,Fix committed] https://launchpad.net/bugs/810440 [13:01] Launchpad bug 810640 in ubuntuone-control-panel "QT UI: we need an asset for file sync error/alert (affects: 1) (heat: 6)" [High,Fix committed] https://launchpad.net/bugs/810640 [13:02] ralsina: iuhu? [13:02] @ping [13:02] pong [13:02] nessita: what does iuhu mean? [13:03] fagan: like "iuuuuuhhhuuuuuu" [13:03] nessita: hah [13:03] fagan: yoohoo, i guess you'd write [13:03] * fagan thought it was some spanish thingy [13:04] Chipaca: yep that would parse better [13:04] fagan: you parse in your lexer! that explains a lot [13:05] that's cheating! [13:05] :-P [13:06] hahah [13:06] * fagan is worried he might have to work downstairs on monday again if the package doesnt get here [13:06] :/ [13:09] ralsina: you too? :-/ [13:10] nessita: yes [13:10] ralsina: go! [13:10] nessita: but at least quassel saves my backlog :-) [13:10] sorry,1' [13:10] sure [13:10] DONE: * Windows call [13:10] * Fixed bug #810719 (url signing) [13:10] * Fixed bug #810674 (SSO timeout) [13:10] * Improved #809873 (bundle-and-installer-making script) [13:10] * Improved #810053 (credentials dont ping) [13:10] * Lots of reviews [13:10] Launchpad bug 810719 in ubuntuone-control-panel "Provide a function to oauth-sign URLs using the user's credentials (affects: 1) (heat: 6)" [Undecided,Fix committed] https://launchpad.net/bugs/810719 [13:10] Launchpad bug 810674 in ubuntu-sso-client "SSO-Client should live forever on windows (affects: 1) (heat: 6)" [Undecided,Fix committed] https://launchpad.net/bugs/810674 [13:11] Found a few bugs earlier [13:11] TODO: fix bug #811006 [13:11] Launchpad bug 811006 in ubuntuone-client "Problem with logfile rollover on windows (affects: 1) (heat: 6)" [High,In progress] https://launchpad.net/bugs/811006 [13:11] Mergemy peding branches, then finish the installer [13:11] and by "installer" y mean the wizard [13:11] BLOCKED: no [13:12] Oh, and I interviewed the candidate! [13:14] * nessita -> interview, brb [13:15] ralsina: that job listing has been up for a long while it seems [13:15] oh, yes [13:15] I suppose hr is a careful process [13:16] fagan: no, it's just hard to find the right person [13:16] ralsina: yeah I could only guess [13:16] * fagan has just been on the wrong side of the interview process [13:17] there is no right side of the interview process [13:17] it has only wrongsides [13:17] ralsina: haha [13:18] Im glad to see that a manager sees interviews as painful as the candidates [13:19] Ok, adding to my TODO: move the freaking process forward in Taleo before it all rots === zyga is now known as zyga-afk [13:59] nessita: do you know if the installer & client use images for buttons? [14:00] lisette: I don't but I can find out, one sec :-) [14:01] lisette: nopes, just CSS. Example: background: qlineargradient(x1: 0, y1: 0, x2: 0, y2: 1, stop: 0 #fedad1,stop: 1.0 #e47a55); [14:01] nessita: cool, thank you! === m_conley_away is now known as m_conley [14:10] facundobatista: just as a note, syncdaemon also spawns a subprocess on linux once: platform/linux/messaging.py [14:12] ralsina, that is aberrant :| [14:13] facundobatista: ? [14:16] ralsina, that there is code in the syncdaemon to start the ubuntuone-control-panel is very very ugly [14:16] oh, ok [14:25] hello! [14:27] Hola alecu! [14:28] hi alecu [14:28] hi all [14:29] ralsina, nessita: sorry again for the short notice. Luli told me about it yesterday noon, and I totally forgot. [14:29] alecu: no problem [14:30] Luli is a great name [14:30] :) [14:30] ralsina, so, is trunk running for you still this morning? [14:30] alecu: ran into bug #811006 [14:30] Launchpad bug 811006 in ubuntuone-client (and 1 other project) "Problem with logfile rollover on windows (affects: 1) (heat: 6)" [High,In progress] https://launchpad.net/bugs/811006 [14:30] fagan, her name is Lucila, but we call her Luli. [14:30] ralsina, ok. [14:30] nessita, is trunk still broken for you? [14:30] alecu: awh :) [14:31] how do I run one part of the test suite in ubuntuone-client? [14:33] ralsina, python c:\python27\scripts\u1trial --reactor=twisted tests\platform\windows\test_ipc.py [14:33] alecu: thanks [14:34] ralsina, I think u1trial also has a param to run a "match" of some tests... let me check. [14:34] alecu: let me confirm, I haven't re tried [14:34] nessita, please try trunk of sso, sd, u1cp [14:34] alecu, ralsina, Chipaca: shall we mumble now? [14:34] nessita, or whatever was broken for you better. [14:34] nessita, mandel proposed to do it 3~4pm art [14:34] alecu, ralsina, Chipaca: I can catch up with mandel later [14:35] ok, then, let's [14:35] oks [14:35] nessita, mumble or skype? [14:35] mumble! :-D [14:35] alecu: argh, https://pastebin.canonical.com/49818/ [14:36] alecu: I'm at home, phone works *for now* [14:36] FINGER CROSSSSSSS [14:36] Chipaca: you available? [14:51] alecu: adding close_fds=True in ubuntuone-client/ubuntuone/platform/windows/tools.py line 443 breaks a test in a way I am not even close to understanding :-( [14:52] Whoa. [14:52] It doesn't even have fork(), right, so what does close_fds even mean? [14:52] ...assuming it's only run under Windows. [14:53] Er, exec(), I mean. [14:53] CardinalFang: yes, it's windows only [14:53] ralsina: trace of the failure? [14:54] ralsina: is probably a tgrace from MOck that are *no-understandable* [14:54] nessita: it's from mock alright [14:54] ralsina: show me! :-) [14:55] nessita: I am trying to pick it up in the 17 tests that fail there [14:55] lovely [14:55] (not) [14:55] ralsina, can you paste it? [14:56] alecu, nessita: https://pastebin.canonical.com/49822/ [14:56] I am changing start, so I assume it should make test_start fail ;-) [14:56] ralsina: the test is expecting... notepad to be pass to subprocess.Popen [14:56] => subprocess.Popen(['notepad']) - Performed fewer times than expected. [14:57] I am adding one argument to Popen. If I add that in the mock,I get 23 test fails [14:57] you need to tell mocker that you will call subprocess.Popen(['notepad'], close_fds=True) [14:57] ralsina, that .start() is not used by the tcp activation. [14:57] alecu: oh, even beter [14:57] ralsina, I think it could only be used by u1sdtool [14:57] (which we have not in windows yet) [14:58] alecu: then I am confused about how tcp activation works for sd [14:58] I know I fixed it in ussoc because it makes the error go away, I just assumed in sd it was done the same way [14:58] ralsina, it uses ubuntu_sso/utils/tcpactivation.py [14:59] Oh,cool, so it's only one fix [14:59] ralsina, very likely! [15:07] alecu, could you review https://code.launchpad.net/~ralsina/ubuntu-sso-client/close_fds_sso/+merge/68101 ? [15:07] ralsina, sure [15:07] It's trivial only in appearance [15:11] ralsina, I'm approving, but I have not tested it IRL. [15:12] alecu: the only possible bad side effect would be if sso or sd wrote to stdout/stderr, those will be closed [15:13] Ok, I'm off to see the taxman! [15:17] I'm not sure that this is the right place for this, but I seem to be having trouble downloading music I bought from the ubuntu one music store [15:18] I bought an album and one of the songs downloaded fine, but the rest have been stuck as "queued" for about 15 minutes now [15:18] similar to what was described in http://ubuntuforums.org/showthread.php?t=1496200 [15:19] ping [15:20] ralsina: shall the installer show the terms and conditions or that is not done yet? [15:22] vividviolin: could you send a message to our support and they can sort it out for you https://one.ubuntu.com/support/contact/ [15:22] okay, I'll do that. Thanks! [15:22] vividviolin: np [15:26] ralsina: bug #811099 [15:26] Launchpad bug 811099 in ubuntuone-windows-installer "Terms and conditions is not shown (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/811099 [15:29] * DiegoSarmentero is away: Not here... send me an e-mail! :P [15:32] ralsina: bug #811103 [15:32] Launchpad bug 811103 in ubuntuone-windows-installer ""Set up Account" can not be clicked and there is no indication of what's wrong (affects: 1) (heat: 6)" [Medium,Triaged] https://launchpad.net/bugs/811103 [15:33] @ping [15:33] pong [15:43] ralsina: bug #811109 [15:43] Launchpad bug 811109 in ubuntuone-windows-installer "The sign in form can be submitted even if no captcha solution was entered (affects: 1) (heat: 6)" [Medium,Triaged] https://launchpad.net/bugs/811109 [15:45] ralsina: bug #811111 [15:45] Launchpad bug 811111 in ubuntuone-windows-installer "There is no visual indication that the form is being processed (affects: 1) (heat: 6)" [Medium,Triaged] https://launchpad.net/bugs/811111 [15:49] ralsina: bug #811116 [15:49] Launchpad bug 811116 in ubuntuone-windows-installer "Last screen shows "are you sure" dialog when closing the wizard (affects: 1) (heat: 6)" [Medium,Triaged] https://launchpad.net/bugs/811116 [15:50] DiegoSarmentero: I finished the review, I added a couple of needs fixing, but very little ones [15:50] DiegoSarmentero: also attached some screenshots of some weird effects I see in the VM I tested that in [15:50] DiegoSarmentero: let me know if you have any question! === nessita1 is now known as nessita [15:53] DiegoSarmentero: did you see my messages? my internet went crazy for a couple of mins [15:54] ralsina: ok tested and working fine [15:55] the installer works nicely :) === nessita1 is now known as nessita [15:56] ralsina: I can give you the installer I built if you want === nessita1 is now known as nessita [16:02] ralsina: ill pm you the link anyway since you dont seem to be around [16:02] (and since im just about to EOW [16:02] ) [16:07] ok, I'm off for lunch [16:10] * fagan EOW === beuno is now known as beuno-lunch [16:34] * DiegoSarmentero is back (gone 01:04:22) [16:53] ralsina: ping [17:13] Chipaca: pong [17:16] nessita: I'll check the bugs you mentioned now [17:16] thanks fagan! [17:16] ralsina: no need to do it now, I just noticed all those when reviewing the installer branch from diego === beuno-lunch is now known as beuno [17:21] nessita: "Setup new account" should not be enabled until you accept the terms [17:22] ralsina: hum, that is not happening to me, unless the theme is hiding that? [17:23] nessita: the theme can not enable a disabled button. Here' how it looks for me: http://screencast.com/t/EchuDTVm [17:24] nessita: and here is the TOS being displayed [17:24] ralsina: can you please try with the branch from Diego? [17:24] nessita: sure [17:24] ralsina: about the TOS, I attached a screenshot as well, and my wins since it does not work :-P [17:24] mine* [17:26] ralsina: also, ping again about the bandwidth/file sync setting modification. Did you send the email yesterday? [17:26] no, sending it now [17:27] ack [17:28] nessita: TOS with diego's branch [17:28] http://screencast.com/t/rhF40o2Aec [17:28] ralsina: did you see my screenshot? [17:28] And "Setup Account is disabled" [17:29] nessita: dn't see links t screenshot in the backlog [17:29] ralsina: is in the bug report [17:29] ralsina: how can we debug my env? because that is not working here (in the VM) [17:30] nessita: that could be just that the page didn't load for some reason [17:31] DiegoSarmentero, nessita: a real problem is that the disabled button looks exactly like an enabled one ;-) [17:31] ralsina: then we need to handle that, otherwise the suer experience is ugly [17:31] user* [17:31] ralsina: also, I would expect some sort of spinner while the TOS loads [17:31] can we do that? [17:31] * beuno chears for nessita [17:31] ralsina, yes, we could specialized the buttons with an object name [17:32] nessita: of course we can [17:32] DiegoSarmentero: I think what ralsina says is that we need a specific style for disabled buttons... (I *think*) [17:32] nessita, ahhhh [17:32] yes [17:32] DiegoSarmentero: it's just the disabled state, it's not specific to a button [17:32] ralsina, yes... it's kind of confusing actually [17:32] beuno: thanks, you must know people is hating me already :-D [17:32] no [17:32] now [17:33] so I'm doing a great work! (?) [17:33]