/srv/irclogs.ubuntu.com/2012/04/20/#ubuntuone.txt

* rye_ logs out for reboot08:09
JamesTaitHappy Friday, folks! :)08:28
rye_sweet, somebody stole my IRC nickname08:49
=== rye_ is now known as rye
ryelooks like i got it back08:54
mandelrye, are you sure it was not you :)08:57
* mandel mandel08:58
ryemandel: http://ubuntuone.com/5TVaiFtH9sddWDWTnLwnRT08:59
ryemandel: running autoupdate binary directly08:59
mandelrye, ok, looking09:00
mandelrye, uhhh did we deploy that!09:00
mandel???09:00
ryemandel: yep, 3.0.0 official09:00
mandelrye, la puta!!! ok, we have to make a new installer, is probable that brian had the trial install when he built it09:01
ryemaaan, my windows vms became so slow it is painful09:01
mandelrye, is just a small mistake, but makes us look kind of bad :)09:01
mandelrye, do you have a bug number for that?09:01
ryemandel: well, it definitely raised a wtf look on my face09:01
ryemandel: filing one09:01
mandelrye, please add briancurtin as the person to take care of it, I'll talk with him and ralsina to fix it asap09:02
ryemandel: bug #98607809:08
ubot5Launchpad bug 986078 in Ubuntu One Windows Installer "[3.0.0] AutoUpdate was created with an evaluation version of BitRock InstallBuilder" [Undecided,New] https://launchpad.net/bugs/98607809:08
mandelrye, thx!09:08
gatoxgood morning!11:09
mandelgatox, morning, me needs you :)11:11
gatoxmandel, what happend? :P11:12
mandelgatox, can you please approve the following: https://code.launchpad.net/~mandel/ubuntuone-dev-tools/fix-runtests/+merge/10282811:12
* gatox looking....11:12
mandelgatox, also, cna you confirm that with the new changes that we have for lp:~urbanape/ubuntu-sso-client/initial-darwin-port we have a segmentation fault?11:12
gatoxmandel, yes..... i confirm that yesterday..... and also a lot of tests fail in windows because an api change..... do you want to see the trace.... i already have them in a paste11:13
mandelgatox, I think is because glib and qt are being imported at the same time which makes the code import versions 2 and 3 of glib/gtk which brakes things :)11:13
mandelgatox, no need, as long as it happens to both, I'm ok :)11:13
gatoxmandel, i'm getting this for linux: http://paste.ubuntu.com/937410/ this for windows: http://paste.ubuntu.com/937423/11:14
gatoxmandel, ahhhh :P ok11:14
mandelgatox, if you approve the ubuntuone-dev-tools we will have that in jenkins too, I recommend you set you irc client to listen for u1-jenkins in u1-internal11:14
gatoxmandel, roger that11:15
mandelgatox, in xchat is under prefs->alets->nick names to always highlight :)11:15
gatoxgood11:15
gatoxmandel, done11:16
mandelgatox, the approval or the highlight?11:16
gatoxmandel, highlight11:18
gatoxmandel, and now the approval11:18
mandelgatox, superb!11:19
mandelgatox, and since is your review day, can I have a review for https://code.launchpad.net/~mandel/ubuntuone-dev-tools/squid-errors/+merge/102825 and for https://code.launchpad.net/~brian.curtin/ubuntuone-client/run-tests-buildout-cleanup/+merge/10239411:21
gatoxmandel, jeje i can't say no :P11:21
mandelgatox, I know, and also the good thing is that you are the first one usually to arrive :)11:22
mandelgatox, cna you in the mean time look at the one I gave you from brian?11:33
gatoxmandel, i'm reviewing the both branches you gave me11:34
mandelgatox, sweet!11:34
gatoxmandel, +1 to your branch11:39
mandelgatox, great :)11:40
mandelgatox, when that lands I'll be able why we cannot build on L u1-dev-tools11:40
mandelgatox, this one https://code.launchpad.net/~mandel/ubuntuone-dev-tools/params-runtests/+merge/102836 passes the parameters to runt-tests and skipslint :)11:43
mandelgatox, if you do me a quick review I would be very very happy11:43
gatoxmandel, ok, on it...... i was looking at brian's branch making sure that doesn't brake already installed system to run-tests...... but i should install a new vm to fully test that :S11:44
gatoxmandel, ohhhhh the lovely GOTO jeje..... +1 on the last branch11:47
mandelgatox, yeah... batch, beautiful!11:48
gatoxyeyyyyy..... i already have a cleanvm..... i can fully test brian's branch11:49
mandelgatox, I'm off to have lunch12:34
* mandel lunch12:34
gatoxmandel, ok12:34
ralsinarye, mandel: we have a license for it, and yes, that means we have to do a new build and sign it. Ok, let's take it as an opportunity to sign the autoupdater while we are at it.12:44
ralsinaand good morning12:44
ralsinathisfred: you up already? I want to give you some Qt lessons :-)12:44
gatoxralsina, good morning12:45
ryeralsina: from PyQt4.QtLesson import QtSome12:47
ralsinarye: hehe12:47
thisfredralsina sorry yes am up :)12:48
ralsinathisfred: ok, here's the deal in 6 lines12:48
ralsina1) Do the .ui using designer (and add it to the repo)12:48
thisfredI did the first half :)12:48
ralsina2) compile the .ui using pyuic4 from setup.py and don't add the .py to the repo.12:49
ralsina3) In your own code, import the generated .py and use setupUi to give you UI12:49
thisfredI did the first half of that as well :)12:49
ralsina4) Then connect stuff using signals and slots12:49
ralsina5) ...12:49
ralsina6) Profit!12:49
ralsinathisfred: I have a tutorial on how to do most of this stuf here: http://lateral.netmanagers.com.ar/stories/BBS47.html12:50
thisfredso what I did (wrong) was throw away the .ui and add the generated code to the repository12:50
thisfredneat! thx12:50
ralsinathisfred: you can re-generate the compiled .py manually after you touch the .ui and add it to the repo, too12:52
ralsinathisfred: but that's bad practice because mybe future versions of the ui compiler are better12:52
thisfredralsina, right12:52
ralsinaQuoting myself from that tutorial: "Let me put that in these terms: IF YOU EDIT THE UI FILE (WITHOUT USING DESIGNER) OR THE GENERATED PYTHON FILE YOU ARE DOING IT WRONG! YOU FAIL! EPIC FAIL!. I hope that got across, because there is at least one tutorial that tells you to do it. DON'T DO IT!!!!"12:52
thisfredso the makefile should take care of generating the code?12:52
ralsinathisfred: right12:52
ralsinathisfred: you could do a Makefile or add something in setup.py like we have in u1cp12:53
urbanapemandel, gatox: yeah, I'll get to the windows API changes this morning.12:53
thisfredthe reason I did not do that was, it lives in u1db for now, and I don't want to pollute that with qt builddeps12:53
urbanapehi, all12:53
thisfredbut I still want the example code to be runnable12:53
gatoxurbanape, ok12:53
ralsinathisfred: right12:53
ralsinathisfred: then maybe adding the compiled code for the time being is not horrible12:54
ralsinathisfred: alternative: load the .ui file dynamically12:54
thisfredok, sounds like  a decent compromise12:54
ralsinathisfred: and sidestep the whole issue. It's not hard12:54
ralsinathisfred: it does make it slightly harder to package for windows though12:54
thisfredok, that sounds even better, is that in the tutorial?12:54
ralsinathisfred: I can find you an example12:55
thisfredthat would be awesome12:55
ralsinathisfred: https://pastebin.canonical.com/64666/12:56
ralsinathisfred: sorry about the comments in spanish but it's from a book I was writing ;-)12:56
ralsinathisfred: it assumes the .ui file is in the same folder as the .py that's loading it. Adjust accordingly12:57
ralsinathisfred: after you do that, all the ui components  are in self.ui.object_name and such12:57
ralsinathisfred: and that's it.12:59
thisfredthanks!13:00
thisfredGood for my technical Spanish :)13:02
thisfredcargamos los iconos DEL DIABLO!13:02
thisfredI find putting that at the end of every sentence spices things up a bit.13:03
dobeyhmm13:06
ralsinathisfred: there's a lot more where that came from: http://nomuerde.netmanagers.com.ar/5.html13:16
ralsinathisfred: it's a media player, so you will like it ;-)13:17
thisfredit's not quodlibet, so it has one thing going against it :P13:18
thisfredheh, the gui is the easy part. I only partly agree with that :)13:18
ralsinathisfred: hehe13:19
ralsinathisfred: it's the most minimalistic media player humanly possible. It has one icon and you can only listen to one radio station by default :-)13:19
=== IAmNotThatGuy is now known as Sidewinder1
=== Sidewinder1 is now known as IAmNotThatGuy
* dobey wonders when u1db server will be widely usable, and when android u1db will be ready13:21
dobeyeww13:22
dobeyla GUI es la parte no es tan facil13:23
ralsinadobey: it depens on what GUI and what not-GUI it is13:27
ralsinadobey: but it's easiest to do a functioning GUI than a functioning backend for almost everything13:27
ralsinadobey: although a good frontend is harder than most good backends13:27
ralsinadobey: so, frontend has much more variance13:27
dobeyit's very easy to make a very bad GUI, yes13:28
dobeybut it's also very easy to make a very bad back-end13:28
dobeymaking a great GUI is very hard. and harder if you're doing it in English and you're not a native speaker13:28
* ralsina is guessing thisfred's UI is going to be easier than u1db ;-)13:28
ralsinadobey: that's why I only use icons in sign language13:29
dobeyyeah, demo app doesn't need great UI13:29
thisfredyeah, especially since the ui is not the point of the example code at all13:29
dobeyit needs great comments13:29
thisfredIt does :)13:29
ralsinadobey, thisfred: it needs doctests13:29
ralsinaso it can have truthul comments13:30
ralsinatruthful13:30
thisfredralsina: meh, I think doctests are good for API, I find them annoying in docstrings13:30
thisfredbut I'll keep an open mind, maybe they'll come in handy13:31
dobeydoctests are good for stabbing yourself in all your limbs13:31
ralsinahaha13:33
ralsinadoctests that are not exhaustive but are good examples, are awesome13:33
dobeywhat you need, is a lintian tool which verifies that everything is properly documented13:37
dobeyof course, python also makes that extremely hard to do 100%13:38
ralsinadobey: there is no tool in the universe that can verify anything is properly documented. But I understand what you mean.13:41
* mandel back13:41
dobeywell, not necessarily understandably documented, but all inputs/outputs/exceptions are in the documentation for a function. and the spelling/grammar is reasonably correct13:42
mandelralsina, briancurtin ping13:42
briancurtinmandel pong13:42
ralsinamandel: pong13:42
dobeymandel: what is all this runtests-params nonsense?13:43
mandelralsina, briancurtin, got a very cure bug for you, is bug bug #98607813:43
ubot5Launchpad bug 986078 in Ubuntu One Windows Installer "[3.0.0] AutoUpdate was created with an evaluation version of BitRock InstallBuilder" [Undecided,New] https://launchpad.net/bugs/98607813:43
mandelbriancurtin, ralsina, and that was deployed to the users :)13:43
briancurtinuh13:43
* ralsina looks at the mail sending brian the license key. Asks for that to be added to the build docs13:43
gatoxbriancurtin, hi! i'm trying to test this https://code.launchpad.net/~brian.curtin/ubuntuone-client/run-tests-buildout-cleanup/+merge/102394   i've installed the buildout in a clean vm, and then merge your branch with the one of u1-client inside parts..... how i should run the tests to test this branch?13:43
dobeyheh13:44
briancurtinmandel, ralsina: i definitely applied that license, so i'm not sure if it went away or what the deal is...looking now13:44
ralsinabriancurtin: ack, please take  a look. But don't worry too much.13:44
dobeybitrock's "we should respond within 12 hours" thing doesn't hold very well does it :(13:44
ralsinadobey: about the version logic?13:44
mandeldobey, so, the run-tests of ubuntuone-dev-tools was not working (not finding python) and was not forwarding the parsm to u1trial, which mean we could not pass the reporter and get nice reports when running the tests on jenkins13:44
ralsinadobey: they have been responsive for us in the past13:44
dobeyralsina: about support requests, but that's what i asked for support on yeah :)13:44
ralsinadobey: ack13:45
mandeldobey, so now you can do run-tests /skip-lint --reporter=subunit13:45
mandeldobey, and we have the tests of ubuntuone-dev-tools in jenkins like the other projects /cc ralsina13:45
dobeymandel: oh, that seems rather nasty13:45
mandeldobey, is fixed, it was a very simple fix in the .bat13:45
mandelwe can now trust that jenkins will tell us when we broke u1-dev-tools on windows13:46
dobeymandel: yes i saw the branch. but i mean that fix seems nasty13:46
mandeldobey, oh, yes, is batch, with gotos and all that..13:46
mandelralsina, briancurtin, I would not be to worried, but is something to consider for whenever the .exe comes out of jenkins :)13:48
dobeywe probably need some better magic for that13:48
dobeyoh well13:48
briancurtingatox: so if you already have the lp:~brian.curtin/ubuntuone-windows-installer/buildout-env branch merged, make sure you run "env.bat" from the devsetup folder. that sets the path and some other things up. then in u1client you just run "run-tests.bat" and it should work13:48
gatoxbriancurtin, great..... i'll test that13:48
mandeldobey, I wonder how.. but is the runtests.bat, so I expect some ugliness in there as long as it is just in there13:50
mandelralsina, 1-1 whenever you want :)13:50
mandelbriancurtin, FYI we have u1-dev-tools in jenkins, I did some hacks to get run-tests.bat there working, would be nice if at some point you can dome some of your magic in that one too :)13:52
briancurtinmandel: what's the issue there?13:53
mandelbriancurtin, well, there I had to fix the way it finds python and pass parameters to the u1-trial, maybe doing the same approach you have done in the buildout branches is better13:54
briancurtinmandel: can i see what changes you had to make?13:55
gatoxbriancurtin, wow! everything working... nice branch!13:55
gatoxbriancurtin, +1 approve13:55
briancurtingatox: cool, thanks!13:55
mandelbriancurtin, for the u1-dev-tools, yes here is a diff: https://code.launchpad.net/~mandel/ubuntuone-dev-tools/params-runtests/+merge/10283613:56
gatoxdobey, (are you in your tests day?) can you review this one? https://code.launchpad.net/~brian.curtin/ubuntuone-client/run-tests-buildout-cleanup/+merge/102394   i already run this in a working env... and in a clean env installing everything using the buildout13:57
mandelbriancurtin, approved: https://code.launchpad.net/~brian.curtin/ubuntuone-client/run-tests-buildout-cleanup/+merge/10239413:57
mandelgatox, I did that one too, so dobey  no need13:57
gatoxdobey, no need..... mandel already approve it13:58
gatoxthat! :P13:58
thisfredralsina: pushed the changes so the ui is loaded  dynamically. Now 578% nicer!13:58
thisfredthanks!13:58
ralsinathisfred: just doing my job! :-)13:59
mandelgatox, dobey, would be nice if you take a look at https://code.launchpad.net/~brian.curtin/ubuntu-sso-client/run-tests-buildout-cleanup and https://code.launchpad.net/~brian.curtin/ubuntuone-control-panel/run-tests-buildout-cleanup13:59
gatoxon it!13:59
* dobey needs to take a vacation14:00
dobeyralsina: https://launchpadlibrarian.net/102670998/Captura%20de%202012-04-20%2008%3A00%3A00.png <- waht does that say?14:02
ralsinadobey: it's going to merge a UDF with an existing flder14:03
ralsinadobey: and what language is that???14:03
* ralsina can kinda read it14:04
dobeyralsina: it's not spanish?14:05
ralsinadobey: nope14:05
gatoxdobey, nop14:05
ralsinadobey: I would guess romanian14:05
mandeldobey, is catalan :)14:05
ralsinadobey: or esperanto14:06
dobeyor catalan14:06
ralsinamandel: same thing ;-)14:06
ralsinamandel: made up languages all of them14:06
mandelralsina, as old as spanish: http://en.wikipedia.org/wiki/Catalan_language#Middle_Ages:_origin14:06
ralsinamandel: which is also a made up language14:07
dobeyjust like all language14:07
ralsinamandel: we should all be speaking in the original indostanic14:07
mandeldobey, also, dpm is from valencia, so you should have been able to guess it :)14:07
dobeyralsina: hrmm, i guess that is a stock qt dialog we're creating via code?14:08
mandelbriancurtin, bad news: https://jenkins.errormessaging.com/view/Windows/job/ubuntuone-client-windows-test/36/console14:08
dobeyalso, what a horrible set of choices that is14:08
dobey"yes/no/cancel"14:08
ralsinadobey: not sure, would need to check the code.14:08
ralsinadobey: very not-verbs14:09
dobeythey're not woody at all. rather tinny in fact.14:09
mandelbriancurtin, looks like it cannot find sso, I'll see if I can fix it14:10
briancurtinmandel: is that running the new run-tests.bat? i think perhaps this grew a bit too quickly, because im not sure how the new run-tests.bat will run without an environment setup from the buildout14:12
briancurtinmandel: and jenkins already has its own deps and python setup underneath it14:12
briancurtinoh yeah it is running the new one14:13
briancurtinah i know why it fails14:13
gatoxbriancurtin, i tested your branches with a env without the buildout14:13
mandelbriancurtin, do you need me to get in the machine to fix it?14:13
gatoxand they work14:13
urbanapemandel: has your ubuntuone-dev-tools branch with the txwebserver landed?14:14
mandelurbanape, yes, is in trunk14:14
urbanapek14:14
briancurtinmandel: the run-tests.bat does SET PYTHONPATH=..\ubuntu-sso-client;. - so it overwrites the one jenkins setup14:14
mandelbriancurtin, ouch!14:15
briancurtinmandel: that makes it much easier to run the default case on a local machine, but on jenkins we don't want to touch it14:15
briancurtinmandel: perhaps we add a --jenkins flag or something, which skips setting of the path?14:15
gatoxmandel, are you sure your devtools is in trunk?? i've just run a buildout, etc..... and it keeps getting the txwebserver issue14:15
dobeygatox: it is in trunk. it is not in the tarball your buildout is pulling14:16
gatoxahhhhh it seems that the buildout is not downloading the proper dev-tools14:16
mandelgatox, http://bazaar.launchpad.net/~ubuntuone-control-tower/ubuntuone-dev-tools/trunk/files/head:/ubuntuone/devtools/testing/14:17
mandelgatox, is in trunk :)14:17
dobeygatox: dirspec and devtools should be moved to bzr checkouts under [sources] rather than tarball downloads14:17
gatoxmandel, yes yes..... sorry..... the buildout is downloading another thing14:17
* dobey said this the other day14:17
mandelgatox, I'm going to kill you for this!14:17
mandel:P14:17
dobeymandel: you have to destroy the resurrection ship first14:18
mandelbriancurtin, hm.. of a --jenkins or we can append the python path, can't we?14:18
gatoxjejejeee14:18
briancurtinmandel: i think we'd want to prepend it, to make sure jenkins' stuff is there first14:18
mandelbriancurtin, make sense, can you do a small branch + bug for this14:19
briancurtinmandel: yep, doing right now14:19
mandelbriancurtin, and do the same for the other two branches14:19
briancurtinyep14:19
mandelurbanape, let me know if you need any help with your branch, the segfault is probably due to using glib3 and qt at the same time14:21
ralsinagatox: could you add 986068 to your queue, set to medium?14:25
gatoxralsina, ack!14:25
gatoxbriancurtin, approve: https://code.launchpad.net/~brian.curtin/ubuntuone-control-panel/run-tests-buildout-cleanup/+merge/10239214:27
gatoxbriancurtin, and this one only a small need fixing: https://code.launchpad.net/~brian.curtin/ubuntu-sso-client/run-tests-buildout-cleanup/+merge/10239314:27
dobeymandel: there is no glib3 :P14:27
briancurtingatox: ah, merge error. i'll fix it14:27
gatoxbriancurtin, both tested in an env with and without the buildout, and everything is ok14:28
mandeldobey, well you know.. what ever is the new thingy14:28
mandeldobey, I'l terrible with those things..14:29
dobeymandel: well qt supports running the glib main loop. so it shouldn't be the cause of any segfaults. there's probably a deeper problem in the tests causing it14:29
mandeldobey, I'm seen that happening when you have a gi.repository import and import qt but I'm not 100% sure14:31
dobeymandel: it's possible, but i think that's not necessarily the cause, but merely what you see on top :)14:32
mandeldobey, yes.. we might need to look deeper. but for that branch makins sure that no gi and qt imports are done when they should not is ok14:33
briancurtingatox: fixed your copyright header comment14:37
gatoxbriancurtin, great.... approving it14:37
dobeymeh. next 10 minutes are the most agonizing14:54
ralsinabriancurtin, gatox, mandel, dobey, thisfred, urbanape: standup in 4'14:57
gatoxralsina, ack14:57
mandelack14:57
ralsinaalecu is on a conference today14:57
elopiohey, good morning.15:00
elopioI have a question.15:00
elopiothe control panel is the wizard and the thing where you connect/disconnect, remove devices, etc.15:01
gatoxme15:01
thisfredme15:01
elopiohow do you call that thing?15:01
briancurtinme15:01
urbanapeme15:01
ralsiname15:02
mandelme15:02
dobeyme15:03
gatoxDONE:15:03
gatoxWork on mac os (trying to figure it out how to run the tests easier, just using run-tests). Several reviews for mandel and brian. Start working on Bug #986068.15:03
gatoxTODO:15:03
gatoxFinish with Bug #986068, keep doing reviews (review day), try to improve the dev setup for mac.15:03
gatoxBLOCKED:15:03
gatoxNo15:03
ubot5Launchpad bug 986068 in Ubuntu One Control Panel trunk "Yes/No/Cancel messages appear untranslated" [Medium,Triaged] https://launchpad.net/bugs/98606815:03
gatoxthisfred, go15:03
thisfredDONE: u1db demo app UI iteration 1 TODO: add tags and done status to ui BLOCKED: no NEXT: briancurtin15:03
briancurtinDONE: got installer automation done, should be able to propose today along with a few other changes to make deployment easier (embedded CRT)15:03
briancurtinTODO: propose branches, keep an eye on how jenkins responds to changes15:03
briancurtinBLOCKED: none15:03
briancurtinNEXT: urbanape15:03
urbanapeDONE: Work on darwin branch. TODO: Fix some tests and shared code. BLOCK: None. NEXT: ralsina15:04
ralsinaDONE: kid's birthday, cake, reviews, alecu 1-1, administrivia, helped thisfred with qt, little coding, lotsa thinking (hey, it's my thinking month), reports TODO: reviews,  more thinking BLOCKED: tired brain, cake make belly hurt NEXT: mandel15:04
ralsinaOh, and weekly call15:04
mandelDONE: Fixed bug 985004 bug 986125  and bug 986144 which means that u1-dev-tools in on jenkins. Reviews, worked on a domain socket tcp-activation.15:04
mandelTODO: more tcp activation. 1-1 with ralsina.15:04
mandelBLOCKED: no15:04
mandeldobey, please15:04
ubot5Launchpad bug 985004 in ubuntuone-dev-tools trunk "Squid service startup hides errors" [High,In progress] https://launchpad.net/bugs/98500415:04
ubot5Launchpad bug 986125 in ubuntuone-dev-tools "runtests.bat cannot find python on windows" [High,Fix committed] https://launchpad.net/bugs/98612515:04
ubot5Launchpad bug 986144 in ubuntuone-dev-tools "runtests.bat does not take parameters" [Medium,Fix committed] https://launchpad.net/bugs/98614415:04
dobeyλ DONE: team call, bug #962860, review15:04
dobeyλ TODO: reviews, backports, SRUs, u1db packaging/buildsys15:04
dobeyλ BLCK: none.15:04
ubot5Launchpad bug 962860 in libubuntuone (Ubuntu Precise) "The prompt for installing MP3 playback support dialogue box does not fit in a 1024x600 display" [High,Triaged] https://launchpad.net/bugs/96286015:04
dobeyalso, the cake is a lie15:04
ralsinadobey: I beg to differ15:04
mandeland the spoon does not bend!15:05
* ralsina is eating delicious, delicious lies15:05
* thisfred wants lies15:05
ralsinathisfred: cake is good for you. There is a lie.15:05
ralsinaSo, comments?15:05
gatox_macdobey, jejeej the cake is a lie! great game! :D15:05
dobeyralsina: actually, the organic oreo-like cookies i have are good for you. they have 2g of whole grains per serving, and plenty of antioxidants! :)15:06
ralsina"Roughly translates to "your promised reward is merely a fictitious motivator"."15:06
dobeyyay green tea + sugar.15:06
mandelcomment, fuck twisted and disconnections! turns out that the stupid domains sockets disconnect in a diff way..15:06
ralsinadobey: antioxidants are actually bad for you, according to latest medical literature15:06
mandelsorry, frack twisted15:07
dobeyralsina: government conspiracies are like that.15:07
ralsinamandel: goes to show "domain sockets have good support in twisted" is a cake15:07
dobeyhaha15:08
ralsinaEOM it is then15:08
dobeyand now. lunch.15:08
dobeybbiab15:08
ralsinamandel, mumble?15:08
mandelralsina, there already :)15:08
ralsinamandel: 1 segundo15:10
* gatox quick lunch..... brb!15:38
=== gatox is now known as gatox_lunch
urbanapemandel: for the problematic names from ubuntu_sso/main/perspective_broker, should those just be explicitly skipped for pylint?16:20
urbanapeall the ones with studly caps are signal names, not instance methods16:20
urbanapeah, I see16:21
mandelurbanape, yes, my comment should have been, skip all those complaints that you consider stupid :)16:21
urbanapeI left that pylint exclusion in the windows module16:21
mandelurbanape, yes, you just have to move it and you should be fine16:21
urbanapeI think I got it all. pushing.16:22
mandelurbanape, did you run the tests on linux, I can do it for you if you done have a vm close to you16:23
urbanapeI haven't yet, but can16:23
mandelurbanape, ok16:24
dobeymandel: also please see my and sidnei's comments on https://code.launchpad.net/~mandel/ubuntuone-dev-tools/squid-errors/+merge/10282516:27
mandeldobey, ok16:27
mandeldobey, new version pushed, I added StringIO as the parent calss of Pipe since it implements a diff API, added \nstdout:\n%s and \nstderr:\n%s for the exception16:36
mandeldobey, I added the stdout and stderr to be more clear16:36
mandelurbanape, which revno?16:37
dobeymandel: cool16:37
urbanapejust pushed. 95916:37
dobeymandel: and what about the extreme indentation in the diff, now on line 80 of the diff in launchpad?16:39
mandeldobey, my style, I can put it to the same level of the ( I have no problems16:40
mandeldobey, I read it better, but I suppose it might be hated by people16:41
ralsinathisfred: needsinfo on your branch16:41
dobeymandel: oh i see you do that now on the msg = line. also you apparently typed /n instead of \n :)16:43
mandeldobey, oh my! that is stupid, fixing16:43
thisfredralsina looking16:44
ralsinathisfred: just suggested some style checking16:44
ralsinathisfred: other than that, it looks good16:44
mandeldobey, sorted and added more tranditional indentation16:44
dobeymandel: k, thanks16:44
mandeldobey, I guess this in the real purpose of rebase in git, hide how stupid one is at the end of the day :)16:45
thisfredralsina: I use flake8 in vim, which says the code is good...16:45
ralsinathisfred: there is at least a missing docstring around line 22416:45
ralsinathisfred: also, __init__ should have docstring if it has arguments16:46
thisfredralsina. Ah yes. Somehow flake8 strips that out. I wish I knew how to make it not do that16:46
ralsinathisfred: if pyflakes + pep8 gives different results, then let's not use lake816:46
ralsinaflake816:46
dobeyit's too flaky16:47
thisfredor send patches :)16:47
mandelurbanape, we have a segfault in the tests of that branch, so needs fixing again :(16:47
mandelurbanape, sorry16:47
dobeywhy didn't the flake8 people think of that? :P16:47
mandelok, EOD for me, all have a great weekend!16:48
dobeymandel: can you wait until we get your squid branch approved?16:48
mandeldobey, sure, go ahead :)16:49
=== gatox_lunch is now known as gatox
thisfredralsina:16:49
dobeymandel: or will you be checking it? would like to get it in and see what's happening with lucid :)16:49
thisfredhmm, my pep8 also does not show that warning  anymore?16:49
urbanapemandel: yeah, that's okay, I hadn't gotten to that bit yet. Need to follow that up.16:49
ralsinathisfred: pep8 doesn't complain of that. I expected pyflakes to complain.16:50
thisfredralsina: ah16:50
dobeythisfred: are you doing pep8 --repeat ?16:50
thisfredralsina: it does not16:50
thisfreddobey: yeah16:50
ralsinathisfred: carp16:50
dobeyhrmm16:50
thisfredpep8 should though.16:50
dobeypep8 is what complains about missing docstrings yes16:50
thisfredit's in pep816:51
ralsinareally? Ok then.16:51
thisfredis it an option?16:51
dobeyno16:51
dobeynot afaik anyway16:51
ralsinathisfred: silly question: are you sure you are checking that file? ;-)16:53
thisfreddobey: so pep8 should check for that, but it does not, nor does pyflakes16:53
thisfredpylint does16:53
thisfredralsina: ran it with verbose, and checked all the files16:53
dobeymandel: can you also fix the indentation on line 13 in the diff?16:54
ralsinathisfred: ok16:54
mandeldobey, that one is correct, is as close as possible to the upper opening (16:56
mandeldobey, where yould you like it?16:56
dobeymandel: it's not correct. it should be +4 of the parent line afaict. if i open emacs and go to that line and press tab, it moves it way over to the left to where i'd expect it to be16:57
dobeymandel: next line +4 more spaces16:57
dobeyis what i expect when i see a line end with an open paren16:58
mandeldobey, ok, is style and I'm not very worried about it :)16:59
mandeldobey, done17:01
dobeybriancurtin: https://code.launchpad.net/~brian.curtin/ubuntu-sso-client/run-tests-buildout-cleanup/+merge/102393/comments/22174017:03
dobeymandel: cool, thanks17:03
briancurtindobey: ah, screwed up merging it, i'll add it back in17:04
mandeldobey, let me know, I'm got a 'friend' waiting for me to get out of the office :P17:04
dobeymandel: ok, looks ok to me. i'll bug sidnei to review it again17:05
mandeldobey, sweet!17:05
mandeldobey, once that is in place we can debug the L building issue17:06
dobeyyep17:06
dobeycheers17:06
dobeyralsina: installbuilder support email is sad17:06
ralsinadobey: ?17:06
dobeyThe AutoUpdate uses incremental integers as versions to avoid issues with version formats (specially dealing with formats like 4.0.1-0beta2).17:07
dobeyso they only support that "version" being an integer :(17:08
ralsinadobey: so I was accidentally right all along17:09
mandeldobey, I really need to go now, but I'll follow the branch via phone17:10
dobeymandel: yeah, enjoy17:11
briancurtindobey: that batch file header is now corrected17:13
dobeyralsina: i suppose the question is what size integer it uses.17:13
dobeybriancurtin: cool. looking at it again17:13
ralsinadobey: if we limit ourselves to 3 version components, we could live even with 32-bit unsigned uint17:13
ralsinaunsigned int, of course17:14
ralsinaor just use date-based build numbers17:14
ralsinayyyymmddbb with bb being serials reset by day17:14
dobeywell i was thinking just use a time_t value17:15
dobeywhich will hold us until 2038 at least17:15
ralsinathat should have us covered until 4294967296 which means 4294 96 72 9617:15
ralsinaor a time_t yes17:15
elopiobriancurtin, ralsina, I'm late with windows testing this week. What's the most recent version?17:16
ralsinaelopio: 3.0.0 release which we just rolled back17:16
dobeybut i don't want to assume they didn't just use "int" in the code, so it's a signed 16 bit int or something17:16
dobeybecause then that would be a problem17:16
ralsinayeah17:16
ralsinawell, we could set a 2^17 release number and see if it's < 100017:17
ralsinaor something like tha17:17
dobeyheh, true17:17
ralsinat17:17
dobeywell testing is easy, yeah17:17
ralsinaok, my blood sugar is crashing, need to get some food in me17:17
dobeywas just thinking of how to automate generating the xml, and what to do incase we overflow it :)17:18
dobeybuen provecho ralsina17:18
ralsinaunix timestamp should work pending that test17:18
dobeyyeah17:18
ralsinaluckily python can generate unix timestamps on windows ;-)17:18
dobeywell it doesn't matter if we "release" ubuntuone-windows-installer from linux all the time17:19
dobeywhich is basically how it's been17:19
dobeyanyway17:19
dobeygo eat17:19
dobeyalecu: should https://code.launchpad.net/~mandel/ubuntuone-control-panel/auto-update-looping-call/+merge/81015 be rejected at this point? :)17:22
gatoxralsina, can you repeat me please how to reproduce this bug: https://bugs.launchpad.net/ubuntuone-control-panel/+bug/986068 ?? i'm not being able to17:33
ubot5Ubuntu bug 986068 in Ubuntu One Control Panel trunk "Yes/No/Cancel messages appear untranslated" [Medium,Triaged]17:33
gatoxi'm already change the locale from the applicatoin and i get the dialogs in spanish when they are qt strings..... but i want to see that specific dialog17:34
gatoxralsina, i did it :P17:34
gatoxnevermind17:34
dobeyheh17:41
dobeygatox: i think i'd be more interested in getting the other issues fixed that i mentioned there17:43
dobeygrr mandel17:44
gatoxdobey, can you open a new bug for that..... because this bug report is actually wrong17:44
* briancurtin lunch17:45
gatoxdobey, did you open a new bug for that or do you want me to do it??..... from my point of view, is not directly related to this bug report, and i can fix it now, for the other issue i want some feedback from the user because something is missing in his machine or from qt translations18:08
dobeyi haven't filed it yet18:09
gatoxdobey, ok, no problem.... i'll do it18:09
dobeygatox: where are qt translations installed?18:10
gatoxdobey, /usr/share/qt4/translations18:11
ralsinagatox: what was it then?18:11
ralsinagatox: missing qt translations from the system?18:11
dobeyhmm18:12
gatoxralsina, i ask for some information to the user..... because in spanish the dialog is ok... so maybe he is missing some qt translations or qt translations are not complete, or something18:12
dobeywhat's the locale for catalan?18:12
dobeyit's ca_ES right?18:12
dobeyit seems there is no translation for catalan18:13
dobeyin qt, anyway18:13
ralsinadobey: no idea.18:13
gatoxi think that is probably the problem18:13
gatoxqt translation are not "complete"18:13
ralsinadobey: it's "ca"18:13
dobeyright18:13
dobeythere's no qt_ca.qm here18:13
dobeyno qt_sq.qm either18:14
ralsinadobey: ca, ca_ES, caES@valencia (WTF?)18:14
ralsinaSo, let's shift the bug to the catalan translation team or whatever18:14
dobeyralsina: well, you know, there are regional variations in spain :)18:14
ralsinadobey: I just had never seen the @ variation on l10n codes before18:15
dobeyit's not like they all speak es_IT like they do in argentina18:15
ralsinadobey: their loss! -)18:15
dobeyheh18:15
ralsinaca_ES@valencia|north18:15
gatoxralsina, here is my comment: https://bugs.launchpad.net/ubuntuone-control-panel/+bug/98606818:16
ubot5Ubuntu bug 986068 in Ubuntu One Control Panel trunk "Yes/No/Cancel messages appear untranslated" [Medium,Triaged]18:16
ralsinagatox: saw it18:16
ralsinagatox: looks like it's justthat Qt is not translated to catalan http://musescore-developer.685061.n2.nabble.com/Catalan-and-Qt-translation-td3840425.html18:16
ralsinagatox: and http://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg205958.html18:17
gatoxralsina, so.... which would be the proper answer? mark as invalid and leave a message saying this is a qt problem?18:17
ralsinafind the team that should handle and assign, I guess18:18
ralsinawe could do a workaround, but I refuse18:18
gatoxthat should be: "Ubuntu Translations Coordinators"?18:20
ralsinagatox: why not18:26
gatoxralsina, done18:26
ralsinadobey: yu know better? ^18:26
ralsinagatox: remember to put a nice comment saying "there seems to be a missing catalan translation for qt" or whatever18:27
dobeyit's already assigned18:27
gatoxralsina, yes, already did that too18:27
ralsinagatox: awesome18:27
dobeyeh18:27
dobeydon't assign the control-panel task to them18:27
dobeyassign the translations task to them18:28
dobeythey can't do anything in cp itself either :)18:28
gatoxdobey, ack18:28
dobeyand we can probably delete the cp task18:28
mandeldobey, good question, I don't think so18:35
mandeldobey, I'll remove the Mp18:35
dobeymandel: cool18:36
dobeymandel: also, can you review the 2 branches from briancurtin that you "claimed" ?18:36
mandeldobey, sure18:36
dobeythe pythonpath-jenkins ones18:36
mandeldobey, sorted, is an easy fix, we need to make sure that we don't step on the jenkins path18:38
dobeyyep is trivial, that's why i approved them :)18:39
dobeybut i didn't see you'd "claimed" them until after you left18:39
mandeldobey, I think I was proposed by someone else :)18:42
briancurtini think that might have been my doing. i think i set mandel as the reviewer when i proposed the merge...im not sure if that's a good thing to do though18:42
dobeyah18:42
briancurtini just figured since he was involved to set him in that manner18:42
mandeldobey, I don't like to claim them because it is blocking :)18:42
dobeymandel: good :)18:42
mandelbriancurtin, from my point of view, it is a flaw in the workflow of launchpad, is a good idea but should not block things from landing, I'm made the same mistake before18:43
dobeybriancurtin: it's usually better to just leave the default, except for special cases. trivial changes like this aren't special enough for it18:43
dobeymandel: it's not an LP problem. it's a user choice. by setting it to a specific person, you're saying you want to block on that person reviewing it :)18:44
mandeldobey, could be18:44
mandelI wonder if there is a twitter plugin for jenkins, it would more useful than the irc one..18:45
ralsinasetting up a reviewer when proposing means "don't land until this guy sees it"18:45
ralsinatht's inherently blocking and 90% of the time a bad idea18:45
briancurtinespecially for my trivial one-liner. i'll save the assigned reviewer for big stuff that needs blocking18:45
mandelwell, since I'm not needed any more, I'm off to other japanese and some bad tv18:46
dobeymandel: there probably is, but it only twits using pictures of cats18:47
mandeldobey, lol18:47
mandelbriancurtin, ralsina, if you see that jenkins stays for ever running some tests it might mean that it cannot find qt4reactor (it has happened before) so just kill the build and look at what went wrong18:48
mandelbriancurtin, ralsina, the consoler does not give too much info18:48
ralsinamandel: ack18:48
briancurtinmandel: ack18:48
ralsinamandel: and how would we know it's stuck except going to the site and looking?18:49
mandelralsina, you have no way.. I wonder if we can set a timeout for that18:49
ralsinamandel: right18:49
mandelralsina, we should never be over 10 mins in tests18:49
ralsinamandel: I mentioned that. Let's give it a timeout.18:49
ralsinamandel: if jenkins doesn't have it, a python process that starts things and kills them in 15 minutes is trivial18:50
mandelralsina, from a quick look it does not have it as a default config, maybe as an add on, can you add a bug and I'll take a look on monday?18:51
ralsinamandel: sure. Do we have a place for jenkis bugs?18:51
ralsinamandel: and -> https://wiki.jenkins-ci.org/display/JENKINS/Build-timeout+Plugin18:52
mandelralsina, yep, plugin I can install it as soon as I start on monday18:52
ralsinamandel: awesome18:52
mandelralsina, I might also look at how to run the squid proxy tests there too18:54
ralsinamandel: awesomer18:54
ralsinamandel: I saw a one-file http proxy that may be handy18:54
ralsinamandel: called throxy, pure python18:55
elopioralsina: after uninstalling, shouldn't the C:\Program Files\ubuntuone folder be completely removed?18:55
mandelralsina, at some point we should start adding more tests for diff proxy settings so that we can test as many scenarios as possible18:55
mandelralsina, kind of integration tests18:56
ralsinaelopio: no, it keeps a rolling backup of something18:56
ralsinaelopio: never looked into it, really18:56
ralsinamandel: that's the future18:56
ralsinapresent is, it's 8PM in spain18:56
ralsinamandel: on a friday18:56
ralsinamandel: and you are still working18:56
mandelhaha18:56
ralsinamandel: go away18:56
ralsinamandel: have a nice weekend ;-)18:56
mandelralsina, and in 9  pm ;)18:56
mandelok, laters!18:57
ralsinabye mandel!18:57
elopio:) bye mandel18:57
elopioralsina: I'm just writing updating test case for uninstall, so the expected result should be to keep it or to delete it?18:57
ralsinaelopio: I need to think about it a bit and do some tests18:58
elopios/writing/(I need more chocolate)18:58
ralsinaelopio: file a bug asking it to be removed, assign to me18:58
elopioralsina: ok, thanks.18:58
ralsinaelopio: but hold on that test for a bit, in case we want to keep it18:58
urbanapemandel: shouldn't that test (test_qtwisted) be skipped on Linux?19:03
urbanapeit used to be test_windows19:03
gatoxneed to reboot...... brb19:04
urbanapegatox: I pushed a new rev (961) to my branch that should address the problems on Windows and the pylint stuff.19:06
urbanapealso, you should be able to skip test_qtwisted on linux, since it used to be test_windows.19:07
gatoxurbanape, awesome..... i'll run the tests now19:07
gatoxurbanape, the tests are ok in windows..... about linux, i'm still getting the segfault... if qtwisted need to be skipped, shouldn't we modify the run-tests?19:26
dobeyhrmm, i guess squid is just starting very slowly in the lucid buildd19:26
urbanapegatox: yes, probably.19:27
urbanapejust a sec19:27
ralsinagatox: yes, the skip is set on run-tests, and if the tests were renamed, that needs to be adjusted19:27
gatoxurbanape, thanks, let me know19:28
urbanapejust pushed 962 which excludes test_qtwisted.py and test_pykeyring.py19:28
* gatox testing....19:29
gatoxurbanape, i'm running trunk tests...... because i have 6 failures and i want to check with trunk19:38
urbanapek19:38
urbanapeI might have botched a bit of the windows stuff.19:38
gatoxurbanape, i'm getting this: http://paste.ubuntu.com/938784/19:38
gatoxin linux19:38
urbanapehmm, don't know about the proxy stuff.19:39
gatoxurbanape, i'll rebranch and check......19:40
gatoxgrrrrr.... my internet connection is working like crap19:41
gatoxurbanape, everything is ok...... it seems that the merge was wrong19:47
gatoxurbanape, approving the branch! :D19:49
gatoxurbanape, mmmm now i'm having some issues with tcpactivation on windows19:50
gatoxurbanape, http://paste.ubuntu.com/938806/19:50
urbanapeugh.19:52
gatoxurbanape, i'm in my EOD... but if you make changes to the branch, let me know and i'll run the tests to see if we can merge this today20:14
urbanapekk20:16
ralsinaEOW for me20:38
ralsinahave a nice weekend!20:38
gatoxralsina, you too!20:38
briancurtinenjoy, ralsina20:39
dobeyalright. have a good weekend all21:19

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!