[09:21] morning [12:06] hello crowd! [12:07] nessita: your here [12:07] nice [12:07] fagan: yes :-) [12:09] how is it going? [12:09] good good [12:09] how was the trip back [12:11] it was good, and we were really lucky, since our flights operated normally, but the airport closed a couple hours after we landed [12:11] nessita: ralsina is still stuck in brazil [12:12] Chipaca: yeah, I figured from his tweets :-( [12:12] nessita: yeah. I thought you weren't working today? [12:12] Chipaca: I am, I wasn't working yesterday [12:12] nessita: kewl [12:13] so, I just checked lan.com and the buenos aires-córdoba flight right next to mine was cancelled on Sunday [12:13] I feel special :-P [12:13] nessita: cue mandel singing 'i feel pretty' [12:13] nessita, you're special :) [12:13] (good morning) [12:13] Chipaca: lol [12:14] nessita: i need to go over some dates/planning stuff wiv you [12:14] nessita: when would be a good time? [12:14] nessita: I have been looking at trunk of control panel and u10client and all tests seem to pass, did they get fixed? [12:15] or am I doing something wrong? [12:15] mandel, pign [12:15] Chipaca: depending on your lunch, it can be as soon as I finish catching up with stuff, like in 30 minutes? [12:15] *ping [12:15] facundobatista: pong [12:15] mandel: I would say your machine is either too slow or too fast and the timing issues are not presenting there [12:15] nessita: sure. ping me. [12:15] mandel: as far as I know, everything is still "broken" re tests [12:16] which is :-( [12:16] mandel: unless dobey fixed that yesterday? though I think he wasn't working yesterday [12:16] nessita: really? dammed I ran them several times in my vm and worked.. that is annoying [12:17] mandel: do you have u1-devtools up to date? because that is were the issue is (we fixed the base test case) [12:18] nessita: good point, I think I reverted to r33 at some point to find out when all this started [12:18] mandel: right [12:18] * mandel looks at the r version [12:18] mandel, just a reminder of that branch for you to run the tests in windows :) [12:18] facundobatista: oh, I forgot, can I have the url? [12:18] facundobatista: I'll do it right now [12:20] mandel, this one https://code.launchpad.net/~facundo/ubuntuone-client/waiting-send-id/+merge/63979 thanks! [12:20] np [12:23] mandel: how is it going the SyncDaemonTool for windows? (trying to build a mental map of our pending assignments re wcp) [12:24] nessita: goo progress but I'm waiting for the test to fix and land the new code that has been blocked for a while, I'll like to have something ready for this evening tom morning [12:24] facundobatista: since running al tes tests on windows atm is a pain, which test are the ones you wrote for that branch you gave me? [12:25] mandel: ack, I'll work on the tests as soon as I finish with what Chipacaa needs [12:26] mandel, actually, didn't write a new test for the windows part: changed what windows/ipc.py receives in some part, and fixed the code so the *current* tests go ok [12:26] mandel, see line 90/91 of the LP diff [12:28] facundobatista: ok, then I know which to run :) [12:30] facundobatista: looks like the tests pass if I merge it with https://code.launchpad.net/~mandel/ubuntuone-client/provide_credentials_management which has been blocked due to the tests situation.. [12:30] major PITA... [12:34] morning! [12:34] hi ralsina [12:34] ralsina: how/where are you? :-) [12:35] hi nessita, greetings from foggy sao paulo [12:35] Or rather Atibaia, wherever that is [12:36] ralsina: and your family, where are they? [12:36] nessita: in another hotel in Sao Paulo proper [12:37] ralsina: have you get together? [12:37] since the airlines are paying we can't change hotels, and I am 1 hour by car away from them [12:37] oh [12:37] and neither can leave the hotels since the airlines want us near when the plane gets scheduled [12:37] so, maor suckage [12:37] s/maor/major/ [12:38] ralsina: are you getting back to ARG in the same flight? [12:38] nessita: nope [12:38] mayor oh [12:38] nessita: not even the same airline :-( [12:38] major* [12:40] I wonder how I describe this on canonicaladmin (yes, I had swap scheduled for today but I am stuck in another country) [12:42] ralsina: you should ask your boss ;-) [12:42] I'm so down I am about to go on a free-shop chocolate binge while watching friends reruns :-( [12:44] * mandel walking dog [12:54] * fagan break === teknico is now known as teknico_away [13:16] mandel: let me know when you get back [13:38] store bought scones are not as good as the ones from london :( [13:42] nessita: i had yesterday and today as swap, so no i didn't work on the tests. :-/ [13:48] dobey: ack. Is tarmac running? [13:48] yes [13:48] but u1client is broken; and i don't want my workstation crashing because the tests ate all my RAM [13:54] hello #ubuntuone! [13:55] hi alecu [13:55] hi nessita. I thought you were on swap today. [13:55] alecu: not, only yesterday [13:55] :-) [13:58] * fagan needs to head off for a few hours for errands [13:59] stand up in 2' [13:59] fagan: stand up first! [13:59] nessita: will do [14:00] me [14:00] alecu, mandel, fagan, thisfred? [14:00] me [14:01] me [14:01] me [14:02] dont think mandel is back yet [14:02] DONE: london sprint [14:02] TODO: windows control panel [14:02] BLOCKED: tests are not passing on u1client [14:02] NEWS: we're going back to the "work only on bugs". IE, do not submit a branch without a bug attach to it, idelly created before the branch was started :-) [14:02] NEXT: fagan [14:02] DONE [14:02] * played about more with that branch [14:02] TODO [14:02] * Get the branch merged in and the tests finished [14:02] Blocked [14:03] * by stupidity [14:03] next alecu [14:03] DONE: team sprint [14:03] TODO: catch up with windows port [14:03] BLOCKED: no [14:03] NEXT: thisfred [14:03] * fagan should have added the team sprint to the done item :D [14:04] mandel, alecu: we're having a mumble meeting re the windows port right after the stand up (well, a couple of mins later so we can grab some tea) [14:04] DONE: worked on ubuntuone-couch release TODO: get ubuntuone-couch uploaded, help anyone with anything BLOCKED: no [14:05] nessita, ack. Do you know if anyone besides ralsina is on leave? [14:06] alecu: ralsina and dobey are on leave, though dobey may do some work today to try to fix the tests [14:08] everyone: please say "ack" if you understood what I meant regarding "work only on bugs" [14:10] mandel, alecu, fagan, thisfred: ^? [14:10] ack [14:11] mandel, alecu: meeting in mumble in 20 minutes [14:11] ack [14:11] ack [14:11] thanks! [14:11] * thisfred files bug: Work On Stuff [14:11] :P [14:11] lol [14:13] guess i should maybe make a thing for tarmac to require a linked bug with --fixes to land stuff [14:22] mandel: you back? [14:29] does anyone has the cell phone number from mandel? if so, please tell me in private === m_conley_away is now known as m_conley [14:34] I repeat: does anyone has the cell phone number from mandel? if so, please tell me in private (the one in the directory is old from Belgium) [14:34] ralsina: ping [14:34] nessita: directory? [14:34] oh [14:34] :-) [14:39] rodrigo_: ping [14:52] alecu: until we have the meeting, what are you planning on working on? [14:57] nessita, right now I'm catching up with warthogs, and will do my pending expenses and will start looking into the pending gnome keyring 3 bug. [14:57] nessita, what about you? [14:58] alecu: I just filed a couple of bugs and I'm trying to remember where I left the cloud-folders tab, to create the merge proposal [14:58] alecu: the keyring 3 bug will have to be postponed, we have some hard dates on the windows port [14:58] alecu: that was the meeting about, I'm "desperately" looking for mandel to have the meeting and sync up [15:00] nessita, I know that the keyring 3 bug can wait, but I was planning on taking a look at it while we are idle on the windows side... [15:00] alecu: let's un-idle you in the windows side! :-D [15:00] alecu: do you have any pending branch? [15:02] alecu: I'll review now https://code.edge.launchpad.net/~alecu/ubuntuone-control-panel/more-devices-tab/+merge/63292, what about you starting with the preferences/settings tab? [15:02] nessita, https://code.launchpad.net/~alecu/ubuntuone-control-panel/more-devices-tab [15:02] that one [15:02] on it [15:03] nessita, I don't like that we keep building and have not tried this on windows yet. [15:04] alecu: I agree. But until we have a proper planning meeting with mandel, who for now we depend on, there is not clear path for testing this on windows [15:04] nessita: sorry I'm back had an emergency to sort out... [15:04] and as far as we foresee there should not be (tm) differences for that code (assuming we have the SyncDaemonTool) [15:04] mandel: mumble please? [15:05] nessita: sure on it [15:08] alecu: mumble please? [15:08] one minute === teknico_away is now known as teknico [15:08] sure! thanks [15:12] nessita, pong [15:12] rodrigo_: is ok, no need now :-) [15:14] nessita, ok :) [15:14] thanks! [16:13] alecu, mandel: first bug tagged with u1-zomg-windows! bug #797199 :-P [16:13] Launchpad bug 797199 in ubuntuone-control-panel "Implement cloud folders tab in the QT version (affects: 1) (heat: 6)" [High,In progress] https://launchpad.net/bugs/797199 [16:13] and the tag is now official [16:14] is also official for u1client [16:14] is there anyone in here that works on the windows client? [16:15] Uber_Geek: yes :-) [16:15] mandel: ping [16:15] nessita: pong [16:15] Sweet, i would like to get involved, and help make the windows version work better. [16:16] mandel: what's with the https://code.launchpad.net/~mandel/ubuntu-sso-client/use_txnamedpipes/+merge/61935? wasn't ralsina doing reviews for it? [16:16] nessita: no one was I'm going to go through by ranches and take a look, when done I'll send an email to you guys [16:17] I have not been bossy enough with the reviews [16:17] Uber_Geek: that's great! right now we're really hard on having a complete new version, using the same daemon as the one we use in linux. So I would advice waiting until we release this code, and then help us test it and improve it [16:17] mandel: PLEASE! :-) [16:17] mandel: also, if not present, add test instructions to the merge proposal so we can actually test the stuff (even if you already EOD'd) [16:17] nessita: anything I can do to help in the meantime? [16:17] ok [16:18] Uber_Geek: well, are you familiar with QT and css styling for QT? [16:19] no :( [16:19] Uber_Geek: is ok :-) [16:19] Uber_Geek: then there is not much you can help with at the immediate moment. Could you please re-ping us after June, 24th? [16:19] (if you're still available) [16:20] mandel: can I mark as approved this one https://code.launchpad.net/~mandel/ubuntuone-client/provide_credentials_management/+merge/62627 ? [16:20] I plan on following this for the forseable future. [16:20] nice [16:21] nessita: let me take a look [16:22] nessita: go ahead, it looks correct [16:22] actually I can do it, face palm [16:24] mandel: well, I trust you will update all your branches, so I'll now stop digging into them :-P [16:24] * nessita goes back to her stuff [16:24] hehehe [16:40] alecu: ping [16:44] nessita: ping [16:44] mandel: pong [16:44] nessita: can you review the following: https://code.launchpad.net/~mandel/ubuntuone-control-panel/use_correct_reactor/+merge/62961 [16:44] mandel: sure! [16:45] nessita: allows to use the correct reactor according to the platform. Later we can add a bug that states that the reactor is no needed on linux [16:45] that way we try to fix the dbus issue [16:45] right [17:06] nessita: ping [17:07] mandel: pong [17:07] mandel: Text conflict in ubuntuone/controlpanel/gui/qt/gui.py [17:07] 1 conflicts encountered. [17:08] nessita: hm, something must have changed, will fix right now before I forget [17:08] nessita: this is a hard one, but can you take a look at: https://code.launchpad.net/~mandel/txnamedpipes/add_qt_integration/+merge/61923 [17:08] nessita: if you feel you are not sure about it tell me and I'll get ralsina + alecu do it [17:08] mandel: I will certainly try [17:08] to do what? [17:09] oh, to review it. [17:09] alecu: review https://code.launchpad.net/~mandel/txnamedpipes/add_qt_integration/+merge/61923 [17:09] :) [17:09] ack [17:09] mandel: also https://code.launchpad.net/~mandel/ubuntuone-control-panel/use_correct_reactor/+merge/62961 for alecu, right? [17:10] yes [17:10] nessita: but let me fix the stupid conflict :) [17:11] mandel: why calling QObject.__init__(self) instead of super()? [17:12] mandel: and why is this for "type(data) == type(False)"? [17:13] alright, now where are we, so i can get back to my stuff :) [17:13] mandel: ah your branch failed for other reasons :) [17:14] does U1 have a mobile site for smartphone users? im not really interested in streaming music, just accessing my data. [17:14] dobey: ok going to look [17:14] dobey: what was it this time? [17:15] nessita: it's mucking up permissions on stuff under _trial_temp, so trying to remove some things fails [17:15] or maybe that also happens in trunk? [17:15] dobey: but is that mandel's changes? I don't think so... [17:16] dobey: I would guess your tarmac trunk has a dirty _trial_temp, can that be it? [17:16] no [17:16] dobey: any other idea? [17:16] or well, it's dirty now, becuase of mandel's branch [17:16] dobey: mandel's branch does not play with dir perms, so is odd [17:18] well also because if I dif play it would be for windows [17:18] well i don't know what it is. i am just reporting the facts. empty dir, merge mandel's branch, run tests, boom. [17:18] mandel: well, and linux doesn't have a "hidden" file permission :) [17:18] dobey: ok, I'm going to take a look locally :) [17:19] i'm trying trunk to see what happens there [17:19] dobey: what do you mean with hidden ? [17:20] mandel: on windows, files can be "hidden" via an attribute/permission thing. i was just trying to think of some actual file permission that exists on windows that's in the FS :) [17:20] dobey: oh, yes true, and you also have other strange ones… [17:20] let me first fix conflict I've got in a diff branch and then take a look at that [17:21] mandel: hooray for using utf-8! [17:21] dobey: you found the issue? [17:22] mandel: no, your sentence ended with "…" :) [17:22] haha [17:23] mandel: I asked you 2 questions before re your qt integration branch: why calling QObject.__init__(self) instead of super()? and why is this for "type(data) == type(False)"? [17:23] does U1 have a mobile site for smartphone users? im not really interested in streaming music, just accessing my data. [17:23] i wonder if we can parallelize tests [17:24] iheartubuntu: there is a files app coming for android if that's what you're looking for [17:24] aawww im still stuck with an iphone [17:25] nessita: QObject.__init__ in most cases you cannot use super in PyQt, but I've just tested it and it works, so no real reason for this exact class [17:26] nessita: the type is that I want to make sure it is a [17:27] mandel: then you should use type(True) == bool\ [17:27] type(something) == bool [17:27] nessita: what's the diff? [17:28] mandel: https://code.launchpad.net/~mandel/txnamedpipes/add_qt_integration/+merge/61923 line 89 [17:28] nessita: I have a question about the FakeSdtool class in control panel, what is it for? [17:29] mandel: before going to that, let's have a conversation at a time :-) Are you ficing the type() thingy? [17:29] fixing* [17:30] nessita, mandel: i am also seeing the file permissions things in trunk [17:30] dobey: you fixing it as well? :-D [17:30] if i knew what was wrong i'd fix it, but i don't [17:31] mandel: please answer me before I swicth my attention to something else :-) [17:31] problema por chicharros :) [17:31] dobey: can you please cry outloud for help to chicharreros? [17:31] nessita: let e think about the answer [17:33] nessita: I really do not see the difference between the code you propose and the one is present? is it a matter of style? [17:33] mandel: not only style but confusing people [17:33] mandel: someone can easily do: False = 'a; [17:33] False = 'a' [17:34] and then your code will be checking against the string type [17:34] nessita: hrmm, i wonder if some weird condition is being hit due to the broken tests setup/teardown/super usage all over the place [17:34] nessita: well, yes that is the same arg used when doing c, but if the assing Fals to a string, wel they are stupid… I can change it if you want [17:35] mandel: yes please [17:36] ok, I need to reboot due to font update breakage [17:37] alecu: nessita: I need to walk we dog will be back soon [17:44] mandel: I'm back [17:45] mandel: ping? [17:52] mandel: please ping back when you're here [17:55] nessita: am running tests on your testingit branch to see if it also has perms errors [17:56] dobey: let me know [18:02] nessita: not sure. your branch ate all my ram, and i got the inotify "too many open files" in it :( [18:02] it was 1.4G and climbing when i killed it just now :( [18:02] dobey: ouch [18:02] dobey: if you review the changes, you can check that they make sense... anyways, not funny to ran out of RAM [18:02] alecu: you're back! :-) I was feeling lonely [18:03] nessita: check what changes? [18:03] dobey: those from my branch [18:04] jajaja [18:07] nessita: i'm not saying they don't, but stuff is obviously broken :) [18:07] exceptions.AttributeError: 'FakeNetworkManager' object has no attribute 'busName' [18:07] like that [18:08] dobey: is that u1client trunk? :-( [18:08] dobey: yes things are broken. Not sure how to escalate this, can I suggest you try to define a plan of action with phil? [18:09] that was from your branch. some of the tests were failing with that, intermixed with the inotify issues [18:09] dobey: I would love to help, I like fixing this kind of messes. But for now, I can't [18:09] dobey: oh [18:09] weird [18:09] hum I think I removed that because as far as I saw it was not needed [18:09] well, that's fine, but we do need to fix the perms issue that's in trunk, and blocks branches now [18:09] yes [18:13] nessita: pong [18:13] tell me :) [18:13] * pfibiger just noticed he got pinged. [18:14] mandel: I added comments for both branches for minor fixing, the test request is not that minor and I would like to read your opinion on that one [18:14] i don't know if i'd call that a ping, so much as a dereferencing :) [18:14] hehe [18:15] nessita: which branch is the test request one? [18:15] mandel: in the txnamedpipes [18:16] nessita: ok, well that is a complicated thing to do… we could fake the integration, but I'm not happy about that. [18:16] nessita: other thing to do would be to test the logic of the diff test, which also seems a little off for me, but more doable [18:16] mandel: what about tetsing the class you're developing? not the integration itself. Can that be done? [18:17] nessita: yes, we could test the implementation of the event, eventFilter and postevent methods [18:18] mandel: would that be quickly doable? [18:18] nessita: not nice because is not very tdd, but we can do that… the problem is that that class does the interleave of the reator which I have no clue of how to test.. maybe if I look in twisted trunk .. [18:19] nessita: yes, is doable, and will brin some coverage to the code, it would stop from other people braking it, but I'm not a happy camper on how the test would look [18:19] nessita: I dont know how to explain it, but they will look artificial [18:19] mandel: not sure what that means at code level [18:19] I mean, when applied to code [18:19] nessita: fragile tests [18:20] mandel: can't we do plain unit tests where we don't depend on other components such as twisted? [18:20] nessita: yes, I'll write classic unittest that ensure the logic on the methods used for the inegration work as expected [18:21] should not be too much work [18:21] nessita: seems like facundo is only chicharros around today; and i don't see him on irc :( [18:21] mandel: thanks (do not spend too much time on that please) [18:21] dobey: facundo is on swap day today, now that you mention [18:21] nessita: wont [18:21] nessita: then nobody is around; lucio and verterok are on holiday, and the new guy is, well, new :) [18:21] nessita: I'll do that in an hour, I was going to go to rugby, tom I'll move to the argetina timetable [18:22] mandel: ok. Is there any chance you fix the conflict first? (in the other branch) [18:22] so alecu and I can do the review [18:22] nessita, dobey: before I go, in the test case of the sd in the tear down one of the methods ensures that the readonly files are read-write so that they can be deleted, maybe the issue is there [18:22] nessita: just pushing it right now [18:23] mandel: great [18:23] mandel: maybe; i suspect it's probably not getting called now, with the changes to devtools or something :( [18:24] ugh, there are lots of chmod calls [18:24] dobey: I think that the base test case they use for that does not inherit from the dev-tools test case, but I'm saying this from the top of my head [18:25] dobey: yes, lots of the, for each time you create the dirs used by the sd integration tests [18:25] dobey: I would guess the chmod-cleanup is being done in tearDown instead of inside a self.addCleanup call, so if there is a failure, tearDown is not called [18:25] s/the,/them [18:25] dobey: (but those cleanup functions registered with self.addCleanup will be called even in case of errors) [18:25] nessita: one question, the test_install_bookmarks_extension, does it always fai? [18:26] what I meant was: tearDown is not called if there is an exception inside setUp [18:26] mandel: it never fails, it has nothing in it (I would guess) [18:26] * nessita checks [18:26] nessita: tearDown should be called still; afaik twisted adds tearDown with an addCleanup() [18:26] nessita: I get a timeout [18:26] mandel: where's the error? [18:27] dobey: I'm pretty sure tearDown is not added as a addCleanup [18:28] 768 d.addBoth(self.deferRunCleanups, result) [18:28] 769 d.addBoth(self.deferTearDown, result) [18:28] nessita: ubuntuone.controlpanel.integrationtests.test_dbus_service.OperationsTestCase.test_install_bookmarks_extension [18:28] dobey: is not added as a cleanup :-) [18:28] nessita: I just get tiral telling me that is still running after 3 [18:28] seonds [18:28] nessita: ok, whatever, doesn't matter, because we're not hitting the case you're talking about, since the tests are passing :) [18:28] mandel: are you up to date? and I mean every single package (u1devtools) [18:29] nessita: ie, the tests pass and tearDown itself is failing, it seems [18:29] or just not ever getting run [18:29] dobey: ok, then I better focus on this windows thingy [18:29] similar to the problem we are having with the InvalidDBusSession thing :( [18:29] nessita: yes.. but I;ve got strange issues with my testing vm, when I get back from rugby I'll test in a real up to date natty [18:30] mandel: ack [18:30] nessita: I'll ping you with the results :) [18:30] don't hate windows too much [18:30] mandel: please [18:30] too late! [18:30] :-P [18:37] hmm [18:49] need to buy more ram :( [19:06] nessita: hrmm, i've got part of it fixed it seems, at least :) [19:06] dobey: yeah? great. What was it? [19:06] nessita: what i last suspected; inlinecallbacks/setup/teardown fiasco :( [19:07] hu [19:07] uh [19:07] i don't know if i understand the dbus tests in syncdaemon well enough to fix it right though [19:08] dobey: if you have specific questions I can try to help you [19:09] http://pastebin.ubuntu.com/626733/ [19:10] those are the 2 failures i'm currently getting (some other weirdness as well, but this is the big blocker atm) [19:10] looking [19:11] i changed the tests to yield d and use inlineCallbacks, but didn't help [19:13] dobey: I would add several prints inside the if reply_handler: block to see what branch is entering in and what's the value of args_list [19:14] dobey: seems like what the dbus call is returning is not what the test expects. Can you please show me what changes you made re: inlinecallbacks+setup+teardown? [19:20] hmm [19:22] * dobey tries [19:26] nessita: seems those two tests block on the " config.get_user_config().set_udf_autosubscribe(not result) calls in the disabled reply handler [19:27] dobey: I'm lost, can you rephrase using two "? :-D [19:27] http://pastebin.ubuntu.com/626742/ [19:27] current diff [19:27] nessita: sorry, paste issue; closing " should have been after the (not result) [19:27] "config.get_user_config().set_udf_autosubscribe(not result)" [19:28] dobey: I know what the problem is [19:28] dobey: you removed the BaseTwistedTestCase as parent class, which takes care of restoring the user config [19:29] dobey: which is in turn used as a precondition (to be in a given state, the default) for the tests [19:29] hrmm, ok [19:29] dobey: the tests are failing because they assume they have an initial config that is not being there [19:30] (since that config is being messed up by other tests and no one is restoring it) [19:30] dobey: makes sense? [19:30] yes; but doesn't make sense that they were the only 2 that failed there :) [19:31] dobey: well, it depends on the order the tests executed, this time the previous tests left the config in a way ionly 2 are broken [19:31] only* [19:31] dobey: the easy fix is to restore the config at setUp, for example [19:31] well i've fixed that by fixing the inheritance [19:32] like BaseTwistedTestCase does in its setUp. So you can define the config default setting outside and call that from the 2 test cases, without inheriting from both [19:32] ok [19:32] now i have stuff like this all over the place: http://pastebin.ubuntu.com/626749/ [19:32] looking [19:32] DBusExceptions but tests still pass [19:34] dobey: ah, that is another bug that I think we need to report and ask chicharreros to fix. The test itself is not waiting for all the dbus communication to finish [19:34]