/srv/irclogs.ubuntu.com/2012/07/31/#ubuntuone.txt

=== beuno_ is now known as beuno
mandelmorning all!07:50
JamesTaitGood morning all! :)08:58
mandellunch time!12:26
* mandel lunch12:26
ryealecu: bug #1021098 - how can i notice that the proxy needs authentication if user is not aware of that12:27
ubot5Launchpad bug 1021098 in Ubuntu One Windows Installer "Installer 3.0.2 can't use an authenticated proxy" [High,Triaged] https://launchpad.net/bugs/102109812:27
alecurye: authenticated proxies are not working right in the PyQt version used in our latest windows release.12:29
alecurye: if the user tries to use an authenticated proxy and proxy credentials are missing or wrong, then our dialog to enter credentials should pop up.12:30
alecurye: it currently does not, but our next release should.12:31
ryealecu: is there any way to find that out from the logs12:33
rye?12:33
* alecu looks12:34
alecurye: "how can i notice that the proxy needs authentication" -> when trying to log into the u1 account, the control panel will request the proxy credentials with a popup. In 3.0.2 on windows, due to the PyQt bug, both good or bad credentials throw our program into a loop, and it never connects.13:17
alecurye: in the log in C:\Users\%USER%\AppData\Local\xdg\cache\sso\sso-client.log a line similar to this shows up:13:19
alecu2012-07-31 09:49:51,118:118.999958038 - ubuntu_sso.utils.webclient.qtnetwork - INFO - auth_required None, PyQt4.QtCore.QString(u'192.168.1.10')13:19
* mandel back from lunch13:30
mandelalecu, can i have a review for https://code.launchpad.net/~mandel/ubuntuone-client/correct-namespace/+merge/11626413:38
mmcchi folks13:42
ralsinaLooks like I forgot to say hi. Hi!13:51
diogobaederHi! :-)13:54
mandelhello o/13:56
ralsinamandel: I may be near your town in november13:56
mandelralsina, how come?13:56
ralsinaas in "same continent" at least. So let's have a beer ;-)13:56
mandelralsina, hahaha eu is small but not that small13:57
ralsinamandel: I am fulfilling my mom's lifelong dream of going to Paris.13:57
mandelralsina, if you take here to eurodisnay you might fulfill tatos too hehe13:57
ralsinamandel: we are taking tato there too ;-)13:58
ralsinamandel: but he has bee to real disney, so gabacho disney may be a disappointment ;-)13:58
mandelralsina, hahahaha13:58
mmccralsina: which disney is real disney?14:04
ralsinammcc: well, I would say disneyland, but disneyworld qualifies because walt was there :-)14:04
mmccheh. Yeah, I think disneyworld counts too14:06
mmcclots of Californians disagree :)14:06
ralsinammcc: heh, interesting... "our smaller, older and more boring disney park is more *authentic*" does sound californian :-)14:09
* dobey prefers to just stay away from the disneyverse14:11
ralsinadobey: it's more aimed at us breeders and our litters.14:12
ralsinaI wouldn't go there by myself. Unless they opened a Phineas & Ferb ride. Then I would totally go there.14:12
dobeyisn't that not disney?14:13
ralsinadobey: It's Disney.14:13
ralsinaLike Pixar, Marvel, and everything else ;-)14:14
mmccI'd go to Club 33 without the kid…14:14
ralsinammcc: couldn't go to the grownup parts because, well, the kid. :-)14:15
dobeypixar didn't used to be disney14:17
=== mmcc_ is now known as mmcc
joshuahooverralsina: any ideas what might cause this error on precise? CredentialsError DBusException(dbus.String(u'Process /usr/lib/ubuntu-sso-client/ubuntu-sso-login exited with status 1'),)14:34
dobeyjoshuahoover: have to look at sso logs14:36
ralsinajoshuahoover: logs14:36
joshuahooverdobey: ok14:36
=== mmcc_ is now known as mmcc
briancurtinme15:01
alecume15:02
dobeymeh15:03
mmccme15:03
mandelme15:06
alecuralsina, thisfred, standup!15:06
thisfredme15:06
ralsiname (nonotes)15:07
briancurtinshall i?15:07
ralsinaso thisfred is last, go briancurtin15:07
briancurtinDONE: wrapped up and submitted py3-unicode-part-215:08
briancurtinTODO: fix test failures in py3-unicode-part-3, continue it15:08
briancurtinHELP: if anyone has time, i could use a second review on https://code.launchpad.net/~brian.curtin/ubuntu-sso-client/py3-unicode/+merge/112856 -- it's 470 lines, not very complicated, just needs a good look15:08
briancurtinNEXT: alecu15:08
alecuDONE: started with a branch series to make u1client run with python2 -315:08
alecuTODO: more py3k, do reviews for mandel15:08
alecuBLOCKED: no15:08
alecuNEXT: dobey15:08
dobeyDONE: releases15:09
dobeyTODO: finish releases/uploads, fix review issues in refactor branch, reorg to drop ubuntuone-installer for good, SRUs for lucid/natty/oneiric15:09
dobeyBLCK: None.15:09
dobeymmcc: go15:09
ralsina@ping15:11
mandelralsina, yes, you are here..15:11
briancurtinmmcc: you there?15:11
ralsinathanks mandelbot!15:11
mandelralsina, there is no @ping here AFAIK :P15:12
ralsinamandel: ha, ok15:12
briancurtinmandel is next since mmcc is gone15:12
mandelDONE: More refactoring for code reuse for the jenkins test daemon. Fixed a nmber of MPs15:12
mandelTODO: more work on the jenkins daemon.15:12
mandelBLOCKED: no15:12
mandelthisfred, please15:12
thisfredDONE: u1db example app, Bug #1030947 TODO: Bug #1030947, u1db example app, release 0.0.5 BLOCKED: no NEXT: ralsina15:12
ubot5Launchpad bug 1030947 in U1DB "documents should not be stored multiple times with the same value under the same index key" [High,In progress] https://launchpad.net/bugs/103094715:12
ralsinaDONE: drafted contractor contract, couple of reviews, got disconnected most of the afternoon, minor things. TODO: debug quantal bugs, objectives, other stuff BLOCKED: no NEXT noone15:12
mmcci'm here now, sorrry15:13
mmccDONE: test for reactor regressions, darwin dirspec, wrote mac cert_location15:13
mmccTODO: review day, test mac cert_location, first-run code15:13
mmccBLCK: none15:13
ralsinaso, NEXT mmcc15:13
ralsinaok, comments?15:14
mmccbriancurtin: I'll look at that py3-unicode branch15:14
briancurtinmmcc: cool, thanks!15:15
dobeyok, need to get lunch15:19
dobeybbiab15:20
=== salgado is now known as salgado-lunch
ralsina♫♪♬♩ Oh what a fun, it is to write, a ten page legal docume-ent! ♫♪♬♩16:00
mmccralsina: if you'd like a break, can you tell me if PyQt on python3 will return unicode when I access the value of a QLineEdit widget?16:08
ralsinammcc: it will return either a QString (which you can unicode() safely) or a unicode object, depending on the API level16:09
mmccnot sure if I'm asking that exactly right, but I'm looking at a test for the email verification page that checks essentially isinstance(self.ui.verification_code, str) , which might not be right for py3…?16:09
mmccso on py3 will it return a str?16:10
ralsinathat looks wrong16:10
ralsinacurrently the UI returns a QString, and we are converting it to "something", but I don't remember *what*.16:11
ralsinayou would have to find where we are setting verification_code16:11
mmccthe code is in ubuntu_sso/qt/tests/test_email_verification.py:5516:11
mmccin this test, it's being set on line 52 with self.ui.ui.verification_code_edit.setText('asd123')16:12
ralsinammcc: hmmm let me look16:12
ralsinammcc: because that doesn't set what you are reading :)16:12
mmcchrm16:13
ralsinaI mean, it surely sets it as a side effect16:13
ralsinabut there is probably some type casting in the middle16:13
=== beuno is now known as beuno-lunch
=== beuno-lunch is now known as beuno
mmccyeah, looks like it's getting set as str(self.ui.verification_code_edit.text()) in ubuntu_sso/qt/email_verification_page.py16:16
ralsinayep16:16
ralsinaso that's a str. good thing our verification codes have no weird characters.16:16
ralsinabrb16:17
ralsinaback16:22
ralsinammcc: so, in short, that looks more correct in the py3 world than in the py2 world, where it mainly works by accident.16:22
mmccyeah, that's what I was thinking16:25
mmccso I'll leave that one alone16:25
ralsinamaybe a note saying "this can be safely converted to string because verification codes are ASCII"16:26
mandelok, EOD in spain, catch you all tom16:26
ralsina"in python 2"16:26
mandelplease take a look at my MPs :)16:26
ralsinagack, forget it, that note is awful16:26
ralsinabye mandel!16:26
mmccbye mandel, I've got your merges on my list16:34
mmccoh, I'm looking at stuff briancurtin hasn't gotten to yet, whoops16:37
briancurtinmmcc: yeah that branch you are looking at is 1/3, so if you stray from what's in the py3-unicode branch, it might be covered in part-2 (which is available) or part-3 (still working)16:38
mmccbriancurtin: are you converting tests in a later branch?16:38
briancurtinmmcc: some tests are covered in part-216:39
mmccbriancurtin: ok, thanks16:39
briancurtinmmcc: i tried to break them up by size so they're manageable to review, and they're somewhat grouped by package16:40
briancurtinmmcc: it may make it a bit hard to do because of that, but i didnt want to have one huge 5000 line change to cover it all16:40
mmccbriancurtin: ok, I'll check the other branches before complaining about anything :)16:43
mmccbriancurtin: +1 on part 116:48
briancurtinmmcc: awesome, thanks16:48
dobeyhmm16:49
mmccalecu, when testing the unicode branch, did you get stuck on displaying the captcha?17:19
alecummcc: yes. And it seems to get stuck while running trunk too, so it's not because of this branch.17:21
alecummcc: yestarday I thought it might be my setup, or something on the servers, but it seems to be something on trunk we should fix.17:22
mmccyes, it's the StringIO vs. BytesIO bug.17:22
mmccI found it a couple of weeks ago and I thought we fixed it. I guess we just talked about it17:22
mmccshould we fix it in this branch, since it is an encoding issue related to earlier py3 changes?17:22
mmccwell, a string type issue17:22
dobeyi'd say fix it in a separate, smaller branch17:23
alecu+117:24
mmccok17:24
mmccok then briancurtin2, +1 on part 217:25
dobeyhttps://code.launchpad.net/~dobey/ubuntuone-client/update-4-0/+merge/117497 needs a review or two17:26
briancurtin2mmcc: excellent, thank you very much!17:28
=== salgado-lunch is now known as salgado
* briancurtin2 lunch+taking GF to doctor17:34
mmcchere's the merge for the smaller branch fixing the captcha hang with stringio: https://code.launchpad.net/~mikemc/ubuntu-sso-client/fix-1031437-PIL-exception/+merge/11750417:57
alecummcc: great, I'll review it.18:12
mmccthanks alecu18:14
* mmcc lunches18:14
dobeythisfred: care to review https://code.launchpad.net/~dobey/ubuntuone-client/update-4-0/+merge/117497 ?18:19
joshuahooverralsina: will bug #1021098 be getting fixed in our next win update?18:25
ubot5Launchpad bug 1021098 in Ubuntu One Windows Installer "Installer 3.0.2 can't use an authenticated proxy" [High,Triaged] https://launchpad.net/bugs/102109818:25
alecujoshuahoover: yes. It gets fixed by replacing the pyqt version, so it should be fixed on our next windows update.18:36
joshuahooveralecu: very good, thanks!18:36
* alecu makes a note to IRL check that this bug is fixed on the next installer.18:37
alecuand I should also ping briancurtin2 about replacing pyqt too.18:37
ralsinaalecu: it should be fixed in the jenkins-generated bundles19:32
ralsinaalecu: or else, we need to update Qt there too19:32
dobeygotta run a few errands. bbiab19:40
briancurtin2back...19:44
alecuralsina: it seems that the version built by jenkins uses the old pyqt.20:22
alecuralsina: also, it opens the console window, meaning that it's lacking the twisted fix for http://twistedmatrix.com/trac/ticket/572620:24
briancurtin2alecu: i haven't changed what pyqt is on the jenkins machine20:24
briancurtin2alecu: i can add that twisted patch, but it will still have the console windows for other things because it will have SHOW_CMD=True in the conf.py. we only set that to False when releasing20:25
briancurtin2alecu: is there a good PyQt version we can upgrade to?20:25
alecubriancurtin2: I think we should be using pyqt 4.9 (that includes Qt 4.8)20:26
alecubriancurtin2: jenkins is using pyqt 4.8 (that includes Qt 4.7) on the windows builds20:26
briancurtin2which is what i have on my machine and what we are releasing from20:27
dobeyback20:27
briancurtin2alecu: we're releasing with PyQt 4.8.4/Qt 4.7.2. if PyQt 4.9.4 contains good fixes i can upgrade everything to it20:28
alecubriancurtin2: I verified that pyqt 4.9.something fixed the issue with proxy credentials. But I also recall that 4.9.something had issues showing the errors in ssl verification failures, and since we were hitting a lot of those, that's why we moved back to pyqt 4.8.20:30
ralsinaalecu: yes, but we fixed those by including the certificate, right?20:30
alecuralsina: right20:31
alecuralsina: I mean, we fixed the ssl verification failures. I don't know if "showing the errors from ssl failures" is fixed in pyqt > 4.9.020:31
alecummcc: is there a way to add a unit test that shows the bug in this branch? https://code.launchpad.net/~mikemc/ubuntu-sso-client/fix-1031437-PIL-exception/+merge/11750420:39
mmccalecu, yes I'm working on that now20:39
mmccI can at least write a test that will show the exception we're seeing. Looking to see if there's a way to test more20:40
mmccI'd also like to just remove that PIL stuff and use Qt's PNG plugin, since that should work on OSX, but I'm not sure how to make it work on windows, where the original issue was. I suspect it's the same problem though20:43
dobeygrr.20:53
dobeytar: ubuntuone-client-3.99.2/ubuntuone/platform/filesystem_notifications/monitor/darwin/fsevents_daemon.py: file name is too long (max 99); not dumped20:53
dobeytar: ubuntuone-client-3.99.2/ubuntuone/platform/filesystem_notifications/monitor/darwin/fsevents_client.py: file name is too long (max 99); not dumped20:53
mmccmax 99? yow20:56
dobeyfilesystem_notifications is an insane name for a python module anyway. requires way too much typing for Python20:57
dobeyi'm surprised the branch that created those files, landed, though20:58
mmccdobey: it just squeaked by the lint line length limit: "from ubuntuone.platform.filesystem_notifications.pyinotify_agnostic import ("21:07
mmccaside from lint & style issues though, why are you surprised? is there other stuff like tar we need to worry about?21:09
mmccalecu: I've added a test to https://code.launchpad.net/~mikemc/ubuntu-sso-client/fix-1031437-PIL-exception/+merge/117504 that'll catch the regression, and should also be useful if we end up removing PIL, since it only tests the resulting QPixmap.21:19
dobeymmcc: i'm surprised it landed because we do make dist, and tar should have complained21:20
mmccdobey: ah.21:20
dobeyGLib (gthread-posix.c): Unexpected error from C library during 'malloc': Cannot allocate memory.  Aborting.21:21
dobeybut am also seeing this for some reason, and tests are getting aborted :(21:21
dobeysigh21:32
dobeyalright all, gotta go. have a good night21:51
briancurtin2see ya dobey21:51
mmccbye, dobey21:54
briancurtin2i'm out. time to go watch the cubs get beat by the pirates.22:39
mmccOK, I'm done now too. bye23:00

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