[00:30] alecu: 184.82.108.14 port 8888 [01:53] lovely, thanks. === JackyAlcine_ is now known as jalcine === jalcine_ is now known as JackyAlcine === JackyAlcine is now known as jalcine === jalcine is now known as JackyAlcine === JackyAlcine is now known as jalcine === jalcine is now known as JackyAlcine === JackyAlcine is now known as jalcine === jalcine is now known as webjadmin_ === webjadmin_ is now known as jalcine [09:36] Good morning all! Happy Friday! :D [09:37] morning all! [10:21] Aloha [11:20] good morning! [11:21] dobey: when you're about, if you get a chance could you take a look at http://paste.ubuntu.com/886159/ which shows (line 1060) that rb is scanning a folder I dont want it to. This is resulting in double copies of my music in rb ☹ [11:23] gatox, morning! [11:23] so, is anyone here willing to test u1 with proxies? [11:23] gatox, popey ? pleaaaaaase :P [11:27] mandel, ok..... but i didn't save the ralsina ip.... so you will need to wait until i ask him that again :P [11:27] gatox, which irc client do you use? [11:28] gatox, 'cause if it is xchat you can find the longs under ~/.xchat/xchatlogs/ and then the channel :) [11:28] mandel, xchat..... but i can't find it in the logs [11:28] gatox, really? bummer :( [11:28] mandel, let me check in the file..... but at least is not in the history that the client load [11:29] gatox, oh, but in the files you have A LOT more :) [11:29] mandel, it's .xchat2/scrollback for the logs [11:29] mandel, no, is ~/.xchat2/xchatlogs/ [11:30] mandel: oooh, yeah [11:31] mandel, i always had the logs in ~/.xchat2/scrollback/... [11:31] mandel, and npo.... is not there either..... maybe i have the logs configured to don't save much [11:31] popey, sweet! do you have a proxy server of any type? maybe a squid or something running on P [11:31] i do [11:32] gatox, oh :( I have a stupid about of logs [11:32] mandel: the worry I have though is that the machine can alsso get to the web without going via the proxy [11:33] so i need to do some iptables fu to block port 80 to anything but the proxy IMO [11:33] popey, great, lets get our hands dirty then, you need to get the following branches: lp:~alecu/ubuntuone-client/proxy-tunnel-cookies lp:~mandel/ubuntu-sso-client/ lp:~mandel/ubuntu-sso-client/use-qt4-webclient-backend and lp:ubuntuone-control-panel [11:34] popey, don't worry, we can attempt first without the iptables foo, in theory it does not work, but in my machines does.. so I need a 'cleaner' env to test :) [11:34] popey, does it use auth? the proxy I mean [11:34] no [11:35] just squid on an ubuntu server [11:36] popey, ok, lets first start with nonauth, we can quickly later set the auth in a diff port with not too much work :) [11:36] * popey needs to fix his proxy first [11:36] popey, let me know when you are ready, should be easy to get you with a non-auth proxy [11:44] gatox, I'm noticed that running control panel from trunk the loading text is HUGE does this happen to you? [11:46] mandel, haven't tested that yet..... maybe the branch that was modifying the font sizes landed.... i'm going to check that now [11:46] gatox, please so, in my is crazy.. I know I'm old but I can still read, you guys did not need to do that ;) [11:48] mandel, oh yes..... is big.... that for the font thing that dobey was working i assume [11:48] gatox, he made it bigger? [11:49] mandel, let me check the qss [11:49] bzr: ERROR: Invalid url supplied to transport: "lp:~mandel/ubuntu-sso-client/": ~mandel/ubuntu-sso-client is too short to be a branch name. Try '~/+junk/', '~// or '~////'. [11:50] popey, ups, sorry i miss type the full branch name, it is lp:~mandel/ubuntu-sso-client/use-qt4-webclient-backend [11:50] popey, I guess I had a paste fail, sorry [11:51] mandel, i'll have to check with nessita if that was intended or what [11:53] gatox, I don't think it is, right? it does not even fit in the control panel window.. [11:53] mandel, maybe you are seeing something different than i.... i see the font big.... but fit in that area [11:57] mandel, i'm seeing that the part that specifies the font size of the loading label is missing in the u1-cp qss [11:58] gatox, hm.. weird [12:11] hello everyone [12:12] nessita, hi [12:12] hola gatox [12:12] nessita, mandel was telling me that the text from the loading overlay is too big, and i saw that the font size for that in cp was removed fromm the qss.... was that intended? [12:13] nessita, morning! [12:13] hola mandel! [12:13] gatox: yes, we need to use the font from the system [12:13] gatox: you can propose a branch changing xx-large for x-large if you want [12:13] I'm oj with it [12:13] ok* [12:14] nessita, ok [12:14] gatox: since now we should never ever use hard coded fonts, just relative fonts (small, large, etc) [12:18] gatox: what bug are you working on atm? [12:19] nessita, i was looking at this ones: https://bugs.launchpad.net/ubuntuone-control-panel/+bug/887284 - https://bugs.launchpad.net/ubuntuone-client/+bug/907479 [12:19] Ubuntu bug 887284 in Ubuntu One Control Panel "UI: improve handling of "'s" when appending to names ending in 's' in the Cloud Folders panel" [Medium,Triaged] [12:19] Ubuntu bug 907479 in Ubuntu One Client "u1sdtool -q should not start syncdaemon when syncdaemon is not running" [Medium,Triaged] [12:19] nessita, but wanted to do the font thing first [12:19] nessita, do you have anything better in mind? [12:20] gatox: if it's ok with you, I will ask you to look a couple of bugs regarding translations [12:20] gatox: https://bugs.launchpad.net/ubuntu/+source/ubuntu-sso-client/+bug/951376 [12:20] Ubuntu bug 951376 in Ubuntu Translations "Several strings untranslatable in ubuntone-client registration" [High,Triaged] [12:20] gatox: https://bugs.launchpad.net/ubuntu/+source/ubuntu-sso-client/+bug/951371 [12:20] Ubuntu bug 951371 in Ubuntu Translations "Translated strings shown wrong, character set not UTF-8" [High,Triaged] [12:20] gatox: I will mark them as High and assign to you if that's ok [12:20] nessita, yep....... shoot [12:21] i'll take those [12:22] gatox: thanks. Assigned 2 in ussoc, but I might assing a couple more for the controlpanel (regarding translations as well) [12:22] nessita, assign assign..... i don't mind [12:22] :-D [12:23] nessita, I would literally love you if you could look at line 89 of http://paste.ubuntu.com/886222/ [12:23] gatox: 184.82.108.14 [12:23] mandel: looking! [12:23] and good morning1 [12:23] ralsina, i will write that down [12:23] gatox: I suggest un tatuaje tumbero in your belly [12:23] ralsina, jeejeje [12:24] nessita, is the one with the executiong of login only [12:24] ralsina, morning! [12:24] gatox, on the back of your head, like a hit man! [12:24] gatox: hope you remember the high security user/password! [12:24] ralsina, yes [12:24] :P [12:24] mandel: line 89 seen. DO you also have the -gui logs? [12:25] nessita, let me ask, where can those be found? [12:25] popey, ^ [12:25] mandel, in the back of my head i have: 優柔不断な人 ñandú [12:25] jejeje [12:25] mandel: same dir as sso logs, file is called sso-login-gui.log [12:25] nessita, ok [12:26] mandel, hey! i actually have a chinese symbol in my back :P.... my body supports unicode :D [12:26] mandel: ? [12:26] popey, can we get the logs from the proxy test you did? [12:26] Failure: ubuntuone.platform.credentials.CredentialsError: dbus.Dictionary({dbus.String(u'errtype'): dbus.String(u'CredentialsError')}, signature=dbus.Signature('ss')) [12:26] thats it [12:26] hello, all! [12:26] alecu, hi [12:27] gatox: do you unicodedecodeerror sometimes? [12:27] mandel: not sure what other logs you want/need? [12:27] popey, the logs are here: ~/.cache/sso/sso-login-gui.log [12:27] hola alecu (we're not derailing, no....) [12:27] oh [12:27] popey, sorry, I'm assuming too much from you :) [12:27] alecu, morning! [12:27] yeah, assume I'm an idiot and you wont go far wrong ☺ [12:28] alecu, apparently you have to kick me in the balls, yet I don't know why.. [12:28] nessita, that would be painful :P [12:28] sso-client-gui.log sso-client.log [12:28] not sso-login-gui.log ☺ [12:28] mandel, I've spent the whole night practicing kicking myself between the eyes, too, if that helps. [12:28] mandel, but it seems I need shorter legs... [12:29] popey, can we get those please, pastebin in canonical just so that is not too public [12:29] alecu: or more bendy ankles [12:29] http://paste.ubuntu.com/886237/ [12:29] oops [12:29] or more ankles [12:30] ralsina, mandel: do you guys have 5 mins for a quick mumble? [12:30] alecu, certainly! [12:30] alecu, although, give me 5 mins that other person at the office is screaming.. [12:31] popey, that is the sso-client.log, right? and the sso-client-gui.log? [12:31] -rw-rw-r-- 1 alan alan 0 Mar 6 13:15 /home/alan/.cache/sso/sso-client-gui.log [12:31] empty [12:32] nessita, ^ [12:33] mandel: is there a -d debug option on ubuntuone-control-panel-qt ? [12:33] alecu: sure [12:33] alecu: let me start it [12:33] popey: no, but you can set U1_DEBUG=True [12:33] popey, that is the U1_DEBUG=True we did [12:35] popey: can you please pastebin the output of: apt-cache policy ubuntu-sso-client* [12:36] ralsina: I'm reviewing your 'tweaks' branch, and I have a question [12:37] mandel: nessita i didnt set U1_DEBUG for the control panel line [12:37] $ PYTHONPATH=~/u1/use-qt4-webclient-backend/:~/u1/proxy-tunnel-cookies/:. ./bin/ubuntuone-control-panel-qt [12:37] popey: you're running a custom branch? :-/ [12:38] mandel: are you trying to debug a custom branch with popey? [12:38] nessita, yes [12:38] mandel: uh [12:38] mandel: and does it work for you? [12:38] :-) [12:38] popey, "iptables fu to block port 80" -> "port 443" (since that's the port u1 uses) [12:38] mandel: apparently the UI is not starting, from the 0-sized logs file [12:39] nessita, yes, it works for me, and the same error 1 was given to alecu, right alecu? [12:39] ralsina: having this: [12:39] /* Compensate for border so text doesn't move */ [12:39] padding-left: 8px; [12:39] padding-right: 23px; [12:39] isn't that absolutely dependent of the font size? (ie how many pixels are needed so text do not move) [12:39] popey, try with: U1_DEBUG=True PYTHONPATH=~/u1/use-qt4-webclient-backend/:~/u1/proxy-tunnel-cookies/:. ./bin/ubuntuone-control-panel-qt [12:39] mandel: 1 means the app crashed before it finished starting [12:39] alecu: good point :) [12:40] mandel, right. The "1" is the exit code of the ui process, as far as I debugged yesterday. [12:40] mandel: are you making popey also run the sso service from ~/u1/use-qt4-webclient-backend ? [12:40] nessita, yes [12:40] to run the UI in isolation, you all could: [12:41] cd ~/u1/use-qt4-webclient-backend/ [12:41] U1_DEBUG=True PYTHONPATH=. ./bin/ubuntu-sso-login-qt --app_name=foo [12:41] http://paste.ubuntu.com/886250/ [12:41] gatox, guess who is the winner of a ticket to review two spectacular branches? [12:41] /home/alan/.cache/sso/sso-client-gui.log still empty [12:41] and that will show how the app is crashing at startup [12:41] * gatox is offline [12:41] alecu, jejeje [12:41] alecu, shoot [12:42] popey: can you try what I mentioned earlier? please always run the sso service first in another terminal [12:42] to run the sso service, you should: [12:42] cd ~/u1/use-qt4-webclient-backend/ [12:42] U1_DEBUG=True PYTHONPATH=. ./bin/ubuntu-sso-login [12:43] yup, I'm running that in another window [12:43] --app_name=foo <- what goes there? [12:45] popey: anything, just trying to see the crash [12:45] ralsina: ping? [12:47] http://paste.ubuntu.com/886256/ nessita [12:48] popey: je, the error is simple, just run inside that folder: [12:48] ./setup.py clean build [12:48] mandel: ^ [12:48] nessita, why? [12:48] mandel: he just branched and ui files are not built [12:48] no? [12:49] nessita, he did a python setup.py build, right popey ? [12:49] he did? [12:49] alecu, ?? the branches? [12:49] gatox, sorry, I was on mumble. [12:49] alecu, ah ok [12:52] i did [12:53] http://paste.ubuntu.com/886262/ [12:53] popey: you're likely missing a package [12:54] popey: let me give you the list of packages you should have [12:54] popey: can you please run: [12:54] apt-get build-dep ubuntu-sso-client ubuntuone-control-panel [12:54] 0 upgraded, 0 newly installed, 0 to remove and 44 not upgraded. [12:54] hum, we have have a missin dep, let me list the qt4 deps :-) [12:55] alecu, http://paste.ubuntu.com/886256/ [12:56] mandel: what's the needed package to build the rc files? [12:56] ralsina, i just bought "the dig" for the kindle..... it looks pretty cool [12:57] nessita, you need the uic from PyQt4 [12:57] mandel: yes, and what's the deb package name? [12:57] nessita, let me look [12:57] pyqt4-dev-tools? [12:57] nessita, you can't just do from PyQt4 import uic ? [12:58] gatox: not me, another user [12:58] nessita, if noy... install the dev-tools [12:58] ah ok [12:58] pyqt4-dev-tools is already the newest version. [12:59] nessita, popey yes, it should be pyqt4-dev-tools [12:59] gatox, https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-auth/+merge/97763 [12:59] gatox, and after that, https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-cookies/+merge/97791 [12:59] popey: can yout try running the same command as before? (the one with app_name=foo) [13:00] hey all! [13:00] nessita: nothing has changed since I last ran it though [13:00] any progress yet on the ubuntuone login issues? [13:01] nessita, perhaps not all ENV vars are being passed from the first process to the second? [13:01] ooh, more debug stuff [13:01] nessita, my bet is on the DISPLAY env var not being passed around. [13:01] popey, can you ls /usr/lib/python2.7/dist-packages/PyQt4/bin [13:01] nessita, ^ should be where the rc compiler is [13:01] nessita, what API is spawning the process? [13:02] alecu, ack [13:02] oh, a missing rc file. [13:02] alecu: did you see the import error? [13:02] nessita, I've seen it now. [13:02] http://paste.ubuntu.com/886273/ [13:02] popey: is working!!! [13:02] mandel: ^ [13:03] popey: you got an UI, didn't you? [13:03] yes [13:03] nessita, popey what changed? [13:03] not quite sure [13:03] you sure thats using the proxy? [13:04] popey: what we did so far was to debug why the UI was not being opened befoe [13:04] popey: so now, you and mandel can go back to the way you were debugging before [13:04] nessita, alecu look at line 111 of setup.py in ubuntu-sso-client, /usr/lib/python2.7/dist-packages/PyQt4/bin is wrong, in my system is /usr/lib/python2.7/dist-packages/PyQt4/uic/ [13:04] ralsina: so, how do we begin? [13:04] nessita, popey so, what changed for the ui to work again? [13:05] popey, can you try doing what we did last time with proxies? [13:05] hang on [13:05] the UI for the panel _did_ open before, just not the sign up screen [13:06] mandel: the PATH thing does not matter on Linux, the script is in the PATH already [13:06] nessita@dali:~$ which pyrcc4 [13:06] /usr/bin/pyrcc4 [13:06] mandel: the PATH may be relevant to windows, though [13:06] ralsina: heeeeellllloooo :-) [13:07] nessita, probably.. we should check that, maybe is outdated if you install a new version of PyQt on windows.. [13:07] mandel: right [13:07] popey, yes, you got stuck in the login one.. which is when everything went bananas [13:09] ah [13:13] nessita, what did you do to get it working, because alecu was having the same CredentialsError 1 issue [13:13] mandel: I just saw the trace and aske popey to run ./setup.py clean build [13:14] nessita, funny, 'cause I know he ran ./setup.py build [13:14] mandel: didn't you install some dep in between, perhaps? [13:15] nessita, maybe [13:15] alecu, can you test that you can launch the ui for login? [13:15] sure [13:16] alecu, if it does work, can you test lp:~mandel/ubuntu-sso-client/use-qt4-webclient-backend [13:17] mandel, I'm testing it on that branch. [13:17] popey: yes the dup songs from u1 purchases is a known bug. are you using nightlies, or plain ubuntu packages for u1? [13:17] alecu, make sure that you are on revno 917 [13:18] dobey: plain ubuntu packages [13:18] popey: also, thanks for killing my computer [13:18] haha [13:18] log less [13:18] :p [13:19] popey: fix should show up in ubuntu on tuesday or wednesday then. [13:19] yay! [13:21] mandel, I'm on revno 916. But it worked! [13:21] alecu, wait, with proxy and all? [13:21] with proxy and all [13:22] mandel: what is the spanish translation for the "Getting information..." string in the loading overlay? (if you're using spanish locale) [13:23] dobey, sorry I don't use spanish [13:23] spanglish? :) [13:23] alecu, ok, I'll run the tests and will propose.. [13:23] dobey, mandelish :) [13:23] haha [13:24] dobey, I don't use spanish either, but guess it can't get as uglier than "Getting information..." [13:24] alecu, dobey it can: Tio, recibiendo la info de las pelotas... [13:24] :P [13:24] mandel, that's funnier, not "ugly" [13:25] heh [13:25] i'm guessing that text isn't wrapping [13:25] and is extraneously long anyway [13:26] We should show a pretty animation and a spinning CD that says "Loading…" next to it, like on XBox games [13:28] mandel, now that it works: some issues. [13:28] alecu, shoot! [13:29] mandel, if I click "save and connect" without entering proxy credentials, crap would be saved on the keyring, and it will get blocked. [13:29] all the font size discussion reminds me of this: http://www.youtube.com/watch?v=tY6dIxQTaf8 [13:29] mandel, the same if I just enter the username, without any password. [13:30] alecu, let me test that.. but there are tests for that.. [13:35] dobey: "Obteniendo información, por favor espere..." should be [13:36] yeah, it's too long/formal i think [13:36] dobey: ah, unrelated, you saw https://bugs.launchpad.net/ubuntu/+source/ubuntuone-client-gnome/+bug/953119 ? [13:36] Ubuntu bug 953119 in ubuntuone-client-gnome (Ubuntu) "valgrind invalid read error" [Undecided,New] [13:36] but eh, no more string changes this cycle [13:36] nessita: yes [13:36] ack [13:40] alecu, sorry to drop out but had to reboot, windows had to reboot, then configure updates, then reboot :-/ [13:41] ralsina, no prob. [13:41] ralsina, got a minute for mumble now? [13:41] mandel, too ^ [13:41] alecu: sure [13:41] alecu, sure, lets do it! [13:42] alecu, I'm already there [13:46] gatox: spawn-evil has tests fixed [13:46] alecu, i have a failure running the tests.... let me know if that is related to your u1-client branch [13:46] ralsina, ack [13:48] ralsina, mandel: my mumble seems broken [13:48] alecu: you dropped [13:48] ralsina, mine is the same [13:49] mandel: you still appear there [13:49] ralsina, alecu I cannot hear a thing every few mins.. [13:49] ralsina, mandel: anyway, I think we were closing the meeting. [13:49] mandel: that's the brain damage from rugby [13:49] ralsina, I think is mumble :P [13:49] ralsina, alecu so in summary, I fix the use-qt branch and then I disable the ssl stuff, right? [13:50] ralsina, alecu shall we really disable it, or pass a param to allow pinned certs or something like that? [13:50] alecu: yes, actions: 1) if it works at the level of chromium (ssl-to-proxy) I say it's ok to leave it. If it works any worse, drop it. 2) Don't pin certs from u1 [13:51] ralsina, ssl-to-proxy is not supported by qt. (nor by firefox for that matter) [13:51] alecu: then let's just pretend it doesn't exist ;-) [13:52] alecu, ralsina so I remove it and we are done with it? [13:53] mandel: yep [13:53] ack [13:53] gatox, failure running the test? [13:53] gatox, can you paste it? [13:53] ralsina, alecu I'm off yo have lunch, I should be done with the qt issue in a few mins [13:53] alecu, it's in the mp [13:53] hmm [13:53] alecu, I'll file a bug about the issue when the save creds dialog gets nonesense [13:53] mandel: cool [13:53] mandel, can I review it already? [13:54] mandel, great. There are a few small bugs we should consider next week. [13:54] alecu, no yet, I need to fix one test :) [13:54] alecu, after lunch I'll fix it and will ping you [13:54] mandel, for instance, socks support is not enabled on ussoc, but is on the sd tunnel. [13:54] gatox, thanks. [13:55] alecu, lets 'virtually sit down' and make a list of those small ones, I don't think they are terribly hard ones to fix (socks means changing the proxyfactory) [13:55] mandel, right! [13:55] mandel, we can do it after the qt branch lands. [13:56] alecu, exactly, that is the important one [13:56] ok, I'm off to get energy from dead animals [13:56] * mandel lunch [13:57] * rye starts understanding why go does not use exceptions [13:57] * rye fixes the validation script to deal with different types of errors, it's been a while... [13:58] ralsina, +1 [13:58] nessita: I am putting the tweaks branch on hold until monday because I need to kick windows into shape today [13:58] * gatox needs to relogin..... brb [13:58] ralsina: ack [13:58] nessita: thanks for catching that, too! [13:58] gatox: thanks [13:58] ralsina: bad news is that means we're not releasing to ubuntu that... :-/ [13:59] nessita: if that's the case, I will try to jump into linux 15 minutes to fix it today [13:59] gatox, Looks like the ussoc you are using is not trunk. [13:59] Try updating nightlies or configuring this branch using "--with-sso=/path/to/sso/trunk" [13:59] alecu, ok [13:59] ralsina: let me know, otherwise we can explore options [13:59] nessita: sure [14:08] nessita, gatox: super trivial review https://code.launchpad.net/~ralsina/ubuntuone-windows-installer/no-hack/+merge/97882 [14:09] ralsina, mandel: On a sunnier note, I've just IRL tested the very latest nightlies, and for a nonauth proxy, SD and SSO are working great. [14:10] alecu: \o/ [14:10] ralsina: apparently "u1sdtool -s" doesn't know about the SD it starts. i stepped into the twisted select code and select.select never returns that anything is ready - always ([], [], []) [14:10] briancurtin: interesting [14:10] briancurtin: blocks on IPC would explain the u1cp locking up too [14:11] briancurtin: there are test failures in SSO for IPC, want to take a look at them? Maybe that's the underlyig cause [14:11] ralsina: will do [14:12] alecu, ok, then I wont jump trough the window ;) [14:13] mandel, ralsina: http://www.youtube.com/watch?v=-TUt5n3sU14 [14:13] alecu, mandel: if you have a little time later today, I may need you to put onyour windows/twisted/IPC hat (remember where you put it?) [14:14] ralsina, ok, no problem :) [14:14] ralsina, I think I sat on it. [14:14] luckily it was not my pointy wizard hat. [14:14] alecu: at least there is no h in sat [14:15] loh [14:18] Hi, all. Sorry I'm late. [14:20] hi laser-vision dude! [14:20] nessita: if trunk is lp:blablah then what's "stable 3"? lp:blabla/stable-3 ? [14:21] ralsina: yes [14:21] urbanape: how's the eyes? [14:21] ralsina: no, stable-3-0 [14:21] urbanape: if you can read this, of course [14:21] So, I don't know what they use in metric countries, but I'm guessing I'm 7/5 [14:21] dobey: right [14:21] ralsina: yes, sorry, what dobey said [14:21] I see at 7m what other see at 5m [14:21] so, better than perfect [14:21] dobey: ok, and the current one we are releasing is stable-3-0 ? [14:21] we use feet here, so I'm 20/15 [14:21] ralsina: yes [14:21] urbanape: yay, so if I have 5 apples, you see 7? Awesome! [14:22] little bit of haziness around bright lights, but that should go away in a week or so. [14:22] urbanape: that's the pupil being too relaxed [14:22] urbanape: I have that twice a year when they check for glaucoma (no luck yet!) [14:23] dobey, nessita:thanks so much [14:24] joshuahoover: tomboy sync validation script can now fix xml errors too [14:27] oh i need to fix the conflict [14:28] urbanape: while you were gone I hijacked briancurtin for windows work [14:28] urbanape: and I can't release him today, so you will have to hack alone on mac today [14:30] will do, though it'd be really keen to get a pair for a few minutes to get me unblocked on a few things. I've been nibbling around the edges on some things, but need a little help getting stuff all tied together. Probably doesn't need to be brian... [14:32] gatox, did the newer sso help? [14:33] alecu, i have 2 failures with credentials now [14:34] alecu, do you want the paste of those failures or i put them in the mp? [14:35] gatox, I think a paste is better, because otherwise the MP gets huge [14:36] alecu, ok [14:37] ralsina: you left a print here: https://code.launchpad.net/~ralsina/ubuntuone-windows-installer/no-hack/+merge/97882 [14:37] alecu, http://paste.ubuntu.com/886399/ [14:37] nessita: argh,sorry [14:38] * ralsina is rushig things too much.Time to cool down and meditate. [14:39] gatox, spanish? my code did not touch any of those tests, so would you mind running the tests again with an english locale? [14:39] alecu, ahhhhhhhh sorry.... i'm with an spanish locale now because i'm testing another bug [14:40] alecu, gatox: we may have random breakage of tests in spanish. I tried to fix all of them, but surely missed some, specially when translations change [14:40] so please run tests with LANG=en_EN.UTF-8 but if you find breakage in other languages, file and assign to me [14:41] nessita: removed the print and pushed. Again, sorry, I am a bit too accelerated today, will slow down a bit and doublecheck things now [14:41] ralsina: thanks! :-) [14:41] en_US [14:41] what country is EN? :) [14:42] ok, en_UK ;-) [14:42] proper english! ;-) [14:42] what is UK? [14:42] you mean GB? [14:42] en_EN actually defaults to en_US, just like en_XY does ;-) [14:42] hahaha [14:42] alecu, +1, those were the only tests failing... and not related to this.... and the code seems fine [14:42] en_UK might be interesting though [14:43] ralsina: everything that fails defaults to en_US because en_US is POSIX C [14:43] \o/ [14:43] "blimey mastah! we 'av'a'issu 'ere!" [14:43] dobey: therefore en_EN works, neener neener ;-) [14:44] ralsina: well, it works sort of. it falls back to C, not en_US.UTF-8 :) [14:44] also, the fact that we have en_US.po files for things, is frightening [14:44] gatox, don't go too far away! here's the other one for you :-) https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-cookies/+merge/97791 [14:45] alecu, yesssssss i was starting with that [14:48] thisfred, are you around? [14:48] thisfred, can I ask you to review that branch too? ^^^ [14:49] gatox, mind that it has the previous branch as a prerequisite. [14:49] sigh [14:50] briancurtin: while you check those test failures, I am cleaning up the windows build starting from your branch (removing the windows-installer stuff and other bits) [14:50] fix conflicts; commit; update trunk; merge new one in; oh look, more conflicts :( [14:50] dobey: what branch? [14:50] alecu: I am, though elbow deep in segfaults [14:51] ralsina: sounds good. the only things im seeing on SSO are "Reactor was unclean" test errors fyi [14:51] alecu, ack [14:51] alecu: will review in a sec [14:51] briancurtin: yes, those are the ones I saw [14:51] ralsina: sso system-font [14:52] dobey: oh [14:52] resolving conflicts in qtdesigner ui files is not fun. [14:52] but i think i got them fixed for now [14:52] just NOGBODY BETTER BREAK IT AGAIN [14:53] NOG WILL NOT BREAK BODY [14:53] dobey: ack! [14:53] * mandel back [14:54] nessita, u'test-me-more-♥' => lol [14:55] mandel: we needed complex unicode everywhere :-D [15:00] me [15:00] me [15:01] me [15:01] me [15:01] meh [15:01] me [15:01] nessita: slacker [15:02] me [15:03] mandel: go [15:03] DONE: Fixed the qt issue where the wrong webclieng was used. Fixed bug 933081 and bug 956185. [15:03] TODO: Find a good way to test is qt running. Remove ssl support. 1-1 ralsina, after this. [15:03] BLOCKED: no [15:03] gatox, please [15:03] Launchpad bug 933081 in ubuntu-sso-client (Ubuntu Precise) "GTK UI: URL for terms and conditions never finishes loading" [Undecided,In progress] https://launchpad.net/bugs/933081 [15:03] Launchpad bug 956185 in Ubuntu Single Sign On Client "When viewing the term conditions twice an error occurs" [Medium,Fix committed] https://launchpad.net/bugs/956185 [15:03] DONE: [15:03] Lots of reviews, all my branches landed, working on some translations issues. [15:03] TODO: [15:03] Finish with the translation issues and keep killing bugs from my queue unless someone needs helps with anything else. [15:03] BLOCKED: [15:03] No [15:03] alecu, go [15:03] 𝕯𝖔𝖓𝖊: branch to use creds from keyring: landed, branch to protect tunnel with a cookie: being reviewed. Found SSL issue. IRL tested SD and SSO thru a nonauthenticated proxy. Always bug #929207. [15:03] 𝕿𝖔𝖉𝖔: working on branch to pause syncdaemon on credential errors. [15:03] 𝕭𝖑𝖔𝖈𝖐𝖊𝖉: no [15:03] 𝕹𝖊𝖝𝖙: ralsina [15:03] Launchpad bug 929207 in ubuntuone-client (Ubuntu Precise) "[FFE] Proxy "tunnel" for syncdaemon" [Undecided,Confirmed] https://launchpad.net/bugs/929207 [15:03] DONE: team call, nessita 1:1 dobey 1:1, chipaca 1:1, windows testing, tweaks, reviews, bunch'o'things TODO: kick windows butt, take windows name, chew gum. BLOCKED: windows NEXT dobey [15:03] λ DONE: team meeting, ping about rb-scope, reviews, bug #951425, fix conflicts in sso system-font branch [15:03] Launchpad bug 951425 in rhythmbox-ubuntuone (Ubuntu Precise) "Ubuntu One plugin has problems when being enable/disabled" [High,Triaged] https://launchpad.net/bugs/951425 [15:03] λ TODO: bug #934206, bug #932103, bug #953119 [15:03] λ BLCK: none. [15:03] Launchpad bug 934206 in rhythmbox-ubuntuone (Ubuntu Precise) "track duplication from RB-U1" [High,Confirmed] https://launchpad.net/bugs/934206 [15:03] Launchpad bug 932103 in rhythmbox-ubuntuone (Ubuntu Precise) "Cannot install MP3 playback support; silently fails" [High,Confirmed] https://launchpad.net/bugs/932103 [15:03] Launchpad bug 953119 in ubuntuone-client-gnome (Ubuntu) "valgrind invalid read error" [Undecided,New] https://launchpad.net/bugs/953119 [15:03] nessita [15:04] DONE: weekly call, some other random mumbles, lot of reviews and re-reviews, more on more-pages on controlpanel [15:04] TODO: finish the above [15:04] BLOCKED: nopes [15:04] NEXT: ? [15:04] DONE: tried to get windows install working so it can be released [15:04] TODO: make it work... (working on SSO right now) [15:04] BLOCKED: none [15:04] NEXT: nessita [15:04] sorry briancurtin, seems like I skipped my turn [15:04] alecu, writes fanzy standups [15:04] also, urbanape, thisfred: standup! [15:05] gatox, :-) [15:05] oops that shows as one hour ago in my calendar too [15:06] me [15:06] me [15:06] uh.. dst, fun! [15:06] alecu: didn't we cancel standups, also? [15:07] thisfred, you didn't get the memo? [15:08] DONE: investigate segfaults in u1db | reviews TODO: more of the same BLOCKED: not sure what is causing the segfaults [15:09] urbanape: c'est a vous! [15:09] DONE: GOT LASER VISION PEW PEW PEW! [15:09] TODO: Get brain dump from Jason Foreman (threeve) whose last day is today. [15:09] BLOCK: My setup is not letting me run tests. I'm missing something stupidly simple, I'm sure. I'd appreciate a pair for a few minutes sometime today. [15:09] eom? [15:10] ralsina, 1-1? [15:10] mandel: sure! [15:10] ralsina, mumble! [15:11] alecu, +1 for the other branch too [15:12] ralsina, im there [15:12] mandel: you can't hear me? [15:12] ralsina, no... stupid mumble, let me restart it.. [15:12] mandel: sube el volumen! [15:12] mandel: ack [15:16] nessita, trivial: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/translation-problem/+merge/97901 [15:17] gatox: ack [15:18] gatox: doesn' this bug also fixes the other ussoc report about missing translations? [15:19] nessita, this one is about missing translations... the other one is about encoding problems [15:19] gatox: ah, ok, thanks [15:20] nessita, i didn't wanted to put them together, so we can land this one and start with the translation while i fight with the other one [15:20] gatox: yes, +1 === yofel_ is now known as yofel [15:34] * gatox lunch! [15:34] is it just me or are "reactor was unclean" issues like a needle in a haystack (maybe i just dont get it?) [15:36] briancurtin, they usually mean that something was not properly closed or disposed of. [15:36] yeah thats what im trying to hunt down [15:36] briancurtin, do you have a paste for that error? [15:37] alecu: lot of them in sso - https://pastebin.canonical.com/62479/ [15:44] briancurtin: those are usually caused by a slow or loaded system [15:45] briancurtin: can you please try either (or both) increasing the suite timeout and/or closing everything in windows and rebooting, to free mem? [15:45] briancurtin: I bet they pass after that. Those tests require some (real) IPC interaction, and in windows those are *slow* (which is a bummer, yes, but we prefer that and have the whole ipc stack tested) [15:46] nessita: i do need to reboot anyway since this update dialog keeps popping up. i'll try it and see if the timeout needs to be changed [15:46] nessita, briancurtin: I'm getting those errors too in my windows VM, and I did not usually get them. [15:46] briancurtin: thanks!!! [15:46] nessita, so my guess is that something else changed. [15:46] alecu: reboot... windows mem management sucks (IMHO) [15:47] alecu: the mem in windows seems to run out just because the OS is running [15:47] nessita, I've *just now* started windows on this machine. [15:47] alecu: is your host machine loaded? [15:47] nessita, and I think it was not even paused. [15:47] nessita, no. [15:47] alecu: let me try then [15:47] nessita, it's got 8 cores, and it's mostly idle. [15:47] alecu: ack [15:48] testing in my VM now [15:48] it seems like the same tests consistently fail, so i think its just that something isnt cleaned up properly, but i do need to reboot anyway so brb [15:48] * alecu needs to pick amelia from the new kinder, see you in a few minutes. [15:49] gatox: approved [15:50] anyone could review this trivial (in code, not in semantics) branch? https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/translation-problem/+merge/97901 [15:52] alright, needs lunch. bbiab [16:01] briancurtin: I think you are right, the lock is because syncdaemon is not taking connections [16:02] briancurtin: so, u1cp is trying to connect to it and failing miserable [16:02] yeah CP just sits there. if you let it go long enough it'll eventually get some popup message (that i cant remember), but it takes like 10 minutes [16:03] the good news is, syncdaemon seems to work [16:05] ralsina: so should i keep looking at that SSO unclean reactor stuff, or dive into SD? [16:05] briancurtin: doesn't even seem to be opening its server port, AFAICS [16:05] briancurtin: let's try SD [16:05] ralsina: i seem to remember this to be true that in the past you could only open one SD at a time, no? [16:06] briancurtin: yep [16:06] because i noticed now if you just keep opening u1sdtools it'll open more SDs [16:06] briancurtin: that's not supposed to happen. But if it's failig to open its server port, it's understandable [16:06] So I guess tcp activation is broken [16:06] yeah now it makes sense [16:06] or something similar [16:09] briancurtin: process monitor doesn't show syncdaemon with any LISTENING port [16:09] hmm [16:11] since i don't know where off the top of my head yet, i'm going to step through SD startup and try to find it [16:19] briancurtin: it seems it should log "IPC Initialized" when it happens. And I don't have that on my logs === JackyAlcine_ is now known as jalcine [16:23] ralsina: yeah, i have a breakpoint in the __init__ where that happens. IPCInterface never gets created [16:26] briancurtin: that should be called from interaction_interfaces.py line 1167 [16:26] i was just opening that :) [16:27] briancurtin: ok, you follow it, drop breadcrumbs here :-) [16:27] sounds good [16:28] mandel, ping [16:28] alecu, pong [16:28] mandel, I'm about to go to lunch. Any progress on the qt branch? [16:28] mandel, can I review it already? :-) [16:28] * ralsina goes get a mate working [16:29] alecu, got problems updating the proxy when there is an error because of the caching of the network access manager.. [16:29] briancurtin: I am going to drop off windows to work on a linux branch, let me know if you need me [16:29] alecu, it was 'easy' to fix listening to the proxy signal, not that easy in the errback [16:29] ralsina: sounds good. i think we're going in the right direction now [16:29] nessita, do i need another revision for this? https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/translation-problem/+merge/97901 [16:30] gatox: I'd say merge [16:30] nessita, ok [16:30] mandel, ok. I'll be back in 30 mins aprox, twit me if you need me sooner. [16:30] alecu, ok [16:50] ralsina: any idea why controlpanel tests, the qt4 suite, is segfaulting on widnows? (using trunk) [16:59] mandel, back [16:59] well, in fact I never left :P [16:59] alecu, ack [17:00] alecu, I hate QNetworkAccessManager and its creds cache.. [17:00] I just had some fruits, now having some soup while risking the keyboard [17:06] thisfred, ping [17:07] alecu: pong [17:07] sry, haven't gotten to the branch yet [17:07] thisfred, ack [17:07] will do it now [17:08] thisfred, let me know as soon as you find any issue, because I'd like for that branch to land today so it makes it into beta 2 [17:08] ok, lunchtime here [17:08] right, switching system to proxy now [17:09] nessita: was not segfaulting for me yesterday IIRC [17:09] nessita: it was locking up though [17:09] ralsina: ubuntuone/syncdaemon/main.py:113 - creating that ActionQueue causes something to go wrong (trying to figure it out in debugger, it's in some huge twisted loop) [17:10] briancurtin: great [17:10] i'm going to grab lunch before i get sucked into this issue and forget to eat [17:13] alecu do I need a special version of sso for this or just trunk? [17:14] thisfred, trunk should be fine. Also the sso in nightlies is fine too. [17:14] (I think) [17:14] kk [17:16] need to restart [17:17] no need to restart anymore :P [17:18] alecu, also, since this is the client and not the control panel, what exactly do I run? Just shut down u1sdtool, and then run that from t [17:18] he branch? [17:21] shoot, don't have builddeps and they won't install through the proxy... :S [17:22] thisfred, I usually run it with: reset;U1_DEBUG=True PYTHONPATH=. bin/ubuntuone-syncdaemon --debug [17:22] kthx [17:24] ok, *now* going for lunch (must. stop. debugging.) [17:25] yay has tickets [17:26] tickets for? [17:28] uds [17:29] thisfred, ready for a review? [17:29] thisfred, https://code.launchpad.net/~mandel/ubuntu-sso-client/use-qt4-webclient-backend/+merge/97934 [17:30] mandel: in a bit, reviewing alecu's branch right now [17:30] alecu, puto! puto! puto! ;-) [17:30] alecu, the told blode is mine! [17:30] tall... u brainfuck [17:31] mandel, what's a blode? "A bodyless head with spiky blond hair and black feet" ? [17:31] alecu, my brain is stuck with the bloody QNetworkAccessManager.. [17:31] stupid qt! [17:32] alecu: so I get 'Nonetype' object has no attribute 'host'. with your branch as well, but may [17:32] be I'm not testing the right thing [17:32] alecu: blode is a one-char typoe of bloke :P [17:33] thisfred, that is from libsoup.. [17:33] nessita, ralsina unicode fix: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/unicode-args/+merge/97936 [17:33] thisfred, where are you getting that error? on the log? [17:33] Also: why can't I select and copy text from the cp? [17:33] thisfred: what text? [17:33] alecu no the control panel [17:33] gatox: looking [17:33] dobey: the error [17:33] i guess we aren't setting that label to be selectable [17:34] alecu: I started sd, and it didn't connect so I figured I needed to use the cp to sign in [17:34] thisfred, oh, ok. [17:34] thisfred, try "u1sdtool -c" [17:34] kk [17:35] alecu does the device need to be registered or not? [17:35] ralsina, briancurtin: reading a little backlog, it rings the bell that facundo added code in u1client to have 2 queues of commands, where the whole queue (which may be xtra long) is in disk, and a subset of it in memory [17:35] nessita: yes, that is breaking somewhere [17:35] ralsina, briancurtin: so since you mentioned 'AQ seems stucked', perhaps that's it? [17:35] ralsina, briancurtin: also, another not minor change on AQ was the tunnel adding [17:36] ralsina, briancurtin: not sure if you talked with alecu about it [17:36] about what? [17:36] nessita: not yet, maybe after briancurtin cmes back from lunch [17:36] thisfred, if it's not registered then you'll be testing sso too. [17:36] alecu same error happens in sso when I'm not registered [17:37] thisfred, is the proxy authenticated or not? [17:37] alecu: syncdaemon is not starting in windows, apparently is getting stucked creating the ActionQueue, so I listed the 2 mayor changes we had in that class: the tunnel adding and the split of two queues (disk and in mem) [17:37] ok proxy off, reregistering [17:37] alecu: it is ralsina's proxy so yes [17:38] if anyone needs unauthed proxy I would need his IP [17:38] thisfred, ok. Try installing a default squid3 with apt-get it will listen on localhost:3128 and ask for no passwords. [17:40] mandel, reason = 'QNetworkAccessManager is buggy.' !!!!!! [17:40] mandel, why skipping *so many* tests? [17:40] alecu, QNetworkAccessManager is buggy. [17:41] alecu, the stupid network access manager does not like when you change the settings of a proxy while requests are in the way.. [17:41] nessita, the remove title for reset page: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/remove-title/+merge/97940 [17:41] gatox: thanks! [17:42] alecu, then, we (you and me) know that the proxyAuthenticationRequired really does not work, specially with the yield, so the situation is, we get a crazy warning from the auth cache manager and everything goes a la mierda [17:42] alecu, I've skipped those so I don't block you and I fix the double dialog issue [17:43] alecu doesn't seem to want to connect when I do t [17:43] hat [17:44] thisfred, can you paste the sd log? [17:47] alecu: it appears I cannot: pastebinit breaks on the proxy too, so I guess network settings is screwed up again [17:47] logging out and ing [17:48] mandel, I'm wary of approving a branch that skips so many tests, at this point. [17:49] mandel, so let's do something: I'll try IRLing it [17:50] alecu, I was going to propose the same :) [17:51] alecu: http://paste.ubuntu.com/886635/ [17:52] alecu: that's ~/.cache/ubuntuone/log/syncdaemon-debug.log [17:52] is that the right log? [17:53] gotta go pick up the kid, will be back soonish [17:53] (pastebinit now works through squid with no password, but the client still does not, it seems [17:55] alecu, in my IRL if you give the correct creds for the proxy everything goes like a charm [17:55] alecu, those tests are just when you screw the creds [17:56] thisfred, can you check for a proxy.log ? [17:56] mandel, ack. with creds it works. [17:56] mandel, I'm testing no creds now [17:56] alecu, hombre de poca fe.. [17:56] :) [17:57] nessita: actually yeah it's in the tunnel stuff, a problem with a spawnProcess. i closed it to go to lunch but i'll re-run and get the exact spot [17:57] alecu: http://paste.ubuntu.com/886643/ [17:57] briancurtin: awesome that you have that isolated [17:57] alecu, if you give the wrong creds, it will get stuck [17:57] briancurtin: please ask alecu for insight about it if you can't debug any further [17:58] thisfred, thanks, I'll check it out in a few minutes after reviewing mandel's branch. [17:58] ok guess I'll switch to that too then [17:58] alecu, and in the sso, when that happes if will get a QNetworkAccessCache telling that you are not allowed to update settings in a proxy.. [17:59] alecu, which is due to: https://bugreports.qt-project.org/browse/QTBUG-14850?page=com.googlecode.jira-suite-utilities:transitions-summary-tabpanel [18:00] but with a factory it is not as easy as that bugs description states.. [18:01] ugh, jira [18:01] mandel, have you tried entering wrong u1 credentials? It hangs up indefinitely (both using a nonauth proxy and an authenticated one.) [18:02] alecu, let me test that [18:02] mandel: so how do I test this: run the trunk of CP with this in the PYTHONPATH? [18:03] thisfred, yes [18:04] mandel, entering wrong u1 creds also fails with no proxy configured. [18:05] mandel, luckily you can close the window and retry connecting. [18:05] alecu, wtf? let me see that [18:06] mandel: logging in through non-auth squid works [18:07] mandel, this seems to happen in trunk too, when using the qt backend [18:07] alecu, you got me testing on trunk too, and I can confirm that! [18:08] dobey: to come back to earlier: why in Zod's name would anyone ever make text non-selectable. [18:08] that that option even exists is a bug [18:08] thisfred, a professor at a uni so that you cannot finish your paper [18:08] thisfred: because having a cursor drawn somewhere that isn't a text entry, is fugly [18:09] why would there be a cursor? [18:09] it's not editable [18:09] sorry, we'll rewrite everything in HTML5 for you [18:09] yay! ;) [18:09] because i didn't write the widget library, so i didn't get to make the decision about what it should draw, and where [18:09] dobey: needs fixing in https://code.launchpad.net/~dobey/ubuntu-sso-client/system-font/+merge/97264 for a couple of font sizes [18:10] gatox, briancurtin: may I please have a review? https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/default-folders/+merge/97949 [18:10] nessita, yep [18:10] nessita: will look in a few mins [18:11] all I'm saying is that whoever did write it does not get a cookie [18:11] briancurtin, gatox: thanks! [18:12] mandel anything more you want me to test? [18:12] thisfred, the qt branch, but I'm going to talk first with alecu.. [18:12] alecu, quick mumble? [18:13] kk [18:16] mandel, the "qtnetwork" hanging up on wrong u1 credentials seems to come from a long time ago. [18:16] mandel, thing is, we never found out because it was not being used. [18:16] mandel, I'm on mumble. [18:17] alecu, on my way.. [18:17] alecu, my mumble is failing, one sec [18:19] alecu, sorry, mumble crashed again.. let me update the system [18:19] nessita: is there some way to make that balloon thing expand and shrink properly for the text? [18:19] mandel, ack, let me know. [18:19] thisfred, I'm puzzled by this line on your proxy.log: 2012-03-16 13:50:39,551 - ubuntuone.proxy - ERROR - connection error: Proxy server not found (17) [18:20] dobey: I'm not sure. But even if there is (which would be great for translations), the font there needs to be smaller than what currently is in your branch... [18:20] alecu maybe squid's running on a different port after all? [18:20] alecu: but everything else seems to work is the thing [18:20] alecu, in theory, I'm there.. [18:21] nessita: yes, but with on my branch instead of normal size, half that balloon is empty and just dead space [18:21] nessita: on my computer anyway :) [18:21] nessita: and if someone has set their fonts to a larger size, we still have the same issue [18:21] thisfred, try: telnet localhost 3128GET http://slashdot.org HTTP/1.0 [18:21] dobey: i guess without the you see it completly? [18:22] nessita: yeah, without the small it looks fine for me; but i'm using tiny fonts [18:22] alecu: works [18:23] dobey: right. So, since I now you have more in your plate, I would say: use to match what we currently have, and let's report the bug about the ballon needing to shrink and expand for larger fonts and/or larger texts [18:27] dobey, nessita: +1 those are two separate bugs [18:27] QWizardPage::addField: Duplicate field 'email_address' [18:27] what is that all about? [18:27] dobey: that has been there forever, nevermind [18:28] dobey: basically, on a QWizard you are not supposed to have two fields with the same name [18:28] dobey: it's harmless [18:28] you shouldn't be able to have 2 fields with the same name on anything [18:29] harmless or not, it's ugly, and makes the tests output ugly, because it gets printed like a billion times [18:29] dobey: ok, file a bug I promise I will fix it after the release [18:29] dobey: I am not renaming things today ;-) [18:29] hrmm, the sso tests also seem to be pretty slow [18:29] ralsina: after the release == after windows .exe building, or after tuesday? [18:30] dobey: after tuesday [18:30] dobey: since that is crossplatform code [18:31] heh [18:31] i /am/ changing the font handling, you know ;) [18:31] dobey: sure, what's that have to do with that message? [18:32] ralsina: i'm pretty sure it's cross platform code [18:32] dobey: I have a list of things I expect to get done by today. Changing the fonts? In it. Renaming widgets? Not in it ;-) [18:33] ralsina: does that include a linux.qss/windows.qss fix for sso as well? [18:33] dobey: yes, branch is proposed with yours as prereq [18:34] nessita: fixed/pushed [18:34] dobey: ack [18:34] ralsina: ah ok. didn't see it yet, i'll look now [18:34] dobey: you have already approved it :-) [18:34] Oh, wait, that's u1cp [18:34] right, that's cp not sso :) [18:35] dobey: I can do one for sso in 10 minutes after yours lands [18:35] alecu, ping? [18:35] ok [18:35] ralsina, alecu: so tunnel_runner.py tries to start "ubuntuone-proxy-tunnel" - reactor.spawnProcess chokes on this (i'll figure the exact details out). i modified TunnelRunner.start_process to start "ubuntuone-proxy-tunnel.PY" (note extension) by way of python.exe - it works [18:36] so the SD startup was just hanging around inside of reactor.spawnProcess forever [18:36] briancurtin: yay, then this should not be terribly hard to fix! Good job! [18:36] briancurtin, cool. But please don't land that on trunk yet as it might break the linux release! [18:36] alecu: the windows release is supposed to be done from the same tarballs [18:37] alecu: so we better make it work for both :-) [18:37] ralsina, cool, then let's fix it in a sane way for both [18:37] alecu: right [18:37] alecu: in fact, it's trying to spawn the tunnel even without proxy configured! [18:37] ralsina, that why I said "not yet" :-) [18:38] i'll work up a patch to build the right command lines and arguments [18:38] ralsina, that's fine. Qt knows if the proxy is configured. SD should not use the qtreactor. [18:38] also has to consider when it's a Windows install so it'd be a compiled EXE [18:39] alecu: why is it ok to spawn a process it won't use? Lost you there [18:39] ralsina, the tunnel will start, and it will print a port number to stdout if proxy is configured. Or exit if no proxies are needed. [18:39] alecu, bug 957313 [18:39] alecu: ah ok [18:39] Launchpad bug 957313 in Ubuntu Single Sign On Client "If proxy settings are wrong the qt webclient cannot update the creds" [Medium,Triaged] https://launchpad.net/bugs/957313 [18:45] alecu, I'm leaving yet I have added bug 957317 [18:45] Launchpad bug 957317 in Ubuntu Single Sign On Client "When u1 credentials are wrong and the qt webclient is used the error is not propagated" [Critical,New] https://launchpad.net/bugs/957317 [18:45] alecu, and a possible hint of the reason, yet I need to know what is the sso backend backend (that is the server side not the sso backend) doing [18:45] mandel, why is test_anonymous_proxy_is_used also skipped? [18:46] alecu, proably I screwed up skipping that one, let me update it [18:47] mandel, I'm not sure it's an sso backend thing, since the libsoup webclient seems to work ok in this case. [18:47] and by "work ok" I mean, it returns the error expected when u1 credentials are wrong. [18:48] alecu, ok, I'll look closer on monday [18:48] ack [18:48] ralsina, do you have 5 mins for a quick status update? [18:48] alecu: sure, mumble? [18:48] ralsina, perfect [18:49] mandel, if you have to leave, go ahead. [18:49] alecu, is close to 8 and I'm been here since my 9 :P [18:49] alecu, and my head is about to explode [18:49] mandel: please go away then ;-) [18:49] mandel, yesterday my EOD was at my 1 AM. [18:50] alecu, you crazy! [18:50] alecu, push branch with the correct tests skipped [18:50] s/push/pushed [18:51] alecu, I'll look at the issue during the weekend.. we have waaaaaaaaay to many things interacting, looks like the windows ipc ;) [18:51] * mandel EOD [18:53] briancurtin: https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/default-folders/+merge/97949 fixed, great catch, thanks!!! [18:56] ralsina: i have a fix for running SD from Python, checking now that i fixed it for a py2exe'd build [18:56] briancurtin: awesome [18:56] nessita: looks good, approved :) [18:56] thanks! :-) [18:57] nessita: heh. i *just* posted a needs fixing on there too :) [18:57] let's see [18:58] dobey: which folders are not supported on XP? I specifically used the function with support on XP [18:58] nessita: not the function, but the constants. CSIDL_MYVIDEO seems to only exist for 6.0+ (or that documentation is confusing?) [18:58] dobey: let me re-check then [18:59] ah yeah i think he's right, those are the "libraries" that were added in vista (6.0) [18:59] ok, will try-except then [18:59] so they get listed on vista and above [18:59] right [19:00] dobey: nice catch with the .config dir [19:03] nessita: have 1' for a quick mumble? [19:03] ralsina: being 110% honest, no. But of course I can if you really need me [19:03] nessita: it's ok, ping me when you do [19:03] ralsina: I'm afraid I'm not gonna make it today with the more-pages branch [19:04] :-( [19:04] nessita: ok [19:04] nessita, +1 [19:04] will be closed, but not sure I can land that today [19:04] nessita: then we do need to talk a little, will bebrief [19:04] ralsina: ok, hooking up headset [19:05] oh mandel is gone [19:14] dobey, alecu, nessita, briancurtin, gatox, mandel: because thing have gone all explodey today, we are delaying today's trunk freeze to monday at 6PM UTC [19:14] ralsina: ack [19:14] then on tuesday we will do intensive QA and package the things on wednesday [19:15] ralsina, ack [19:15] ack [19:15] so, rest on the weekend and kick ass on monday, ok? [19:16] bah. test in production :) [19:28] dobey: that's going to be like 50% true next week ;-) [19:29] argh, the py2exe fix didn't work. close, though [19:29] heh [19:30] thisfred: what branch were you testing (or trying to test) earlier, involving proxies? [19:31] dobey: mandels use qt one and alecu's proxy tunnel one [19:32] ok [19:35] dobey: the encoding of the user-dirs file will be sys.getfilesystemencoding()? [19:35] or that just applies to file *names* (and not content)? [19:35] that only applies to names [19:35] dobey: then, have any suggestion? [19:36] encoding can not be guessed [19:36] nessita: fail if it's not utf-8? [19:36] I guess I can try utf-8 and ignore if not [19:36] right [19:36] i don't see any reason it shouldn't be utf-8, unless something broke pretty badly [19:51] ralsina, found the problem (and a solution!) for the bug with the u1 credentials with qtnetwork. [19:51] alecu: awesome! [19:51] ralsina, it seems that the wrong credentials were being retried *forever*. [19:51] alecu: ugh [19:51] a branch to fix it is on the way. [19:56] dobey: all issues addressed now on https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/default-folders/+merge/97949 [19:56] dobey: good catch, though I would have enjoyed fixing a lot more if I wasn't with so much rush [19:56] anyays, import codecs rule [19:59] anyone needs a hand, a review, anything? [19:59] cofee? cake? [19:59] cake would do [19:59] alecu: on it! [19:59] you'll start cooking it now? [20:00] alecu: http://bit.ly/FOa85J [20:00] until we get cake-over-ip that's the best I can do [20:00] * alecu is now officially hungry. [20:01] ha, a friend of mine started http://foodoverip.org/ [20:01] briancurtin: how's the spawner coming along? [20:02] ralsina: still need to figure out py2exe. i can start ubuntu-sso-login.exe and then start SD via Python, then ubuntuone-control-panel-qt.exe, and it seems to work "ok". im pasting a few errors i just saw a minute ago before i go figure them out [20:02] briancurtin: cool [20:03] the cake is a lie [20:04] ralsina: https://pastebin.canonical.com/62504/ and https://pastebin.canonical.com/62500/ [20:05] in any case, it's really close to working. need to add a bunch of logging or something to the py2exe build since it still can't find the right executable to start [20:05] briancurtin: that's the same exception seem from both sides of IPC right? [20:06] briancurtin: is that friend Benoît, or is he just a very enthousiastic user? :) [20:07] nessita: +1 [20:07] dobey: thanks! [20:08] thisfred: yeah its Benoît, the site was started at a sprint at Tarek Ziade's house where the focus became food and wine [20:08] heh [20:08] I know him a little from my couchd [20:08] b efforts [20:08] he's nice [20:09] i know tarek much more than benoit, but they're all a good group [20:09] EOD for me!! bye!! see you on monday [20:09] ralsina: i think it probably is the same exception, i'll see what changed there [20:09] see you latox alli [20:10] gatox [20:10] oi, it's 1600 already [20:11] thisfred, jejee abstract message :P [20:11] dobey: and nobody's brought me a mint julep yet! [20:13] trial is the awesome. Love it. [20:13] alecu: is that sarcasm? ;) [20:14] I've been looking at py.test again after I saw Holger's pycon talk... [20:14] It's nice [20:15] thisfred, no, seriously. I'm in love with it. I'll say it louder: [20:15] 𝕀 𝕝𝕠𝕧𝕖 𝕥𝕙𝕖 𝕥𝕨𝕚𝕤𝕥𝕖𝕕.𝕥𝕣𝕚𝕒𝕝 𝕥𝕖𝕤𝕥 𝕣𝕦𝕟𝕟𝕖𝕣! [20:15] hehe [20:15] I don't dislike it [20:16] I like the repeat until failure option [20:16] great if you have randomly failing tests [20:17] what I like about py.test is that it makes for much less boiler plate in most tests [20:18] so they become a lot easier to read [20:18] and the failure messages are super too [20:18] thisfred, can I return a deferred from a py.test ? That's the killer feature for any test runner I use :-) [20:19] i don't think i'd want to a) convert devtools to py.test b) make py.test work with twisted [20:19] I bet you you can, but why would a test return anything? [20:19] thisfred: because it's twisted [20:19] dobey: I'm not proposing that we do [20:20] Just trying it out on a hobby project [20:20] i'm just saying i don't like twisted [20:20] I wasn't arguing on that point [20:20] also, alecu's branch name is making me hungry [20:23] this one? https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-cookies/+merge/97791 [20:24] yes [20:24] also, why the heck were we using a symlink for the music library in rhythmbox [20:25] also, where is sil so i can yell at him for making it a symlink [20:27] umm why does quickly depend on couchdb-bin? [20:27] ah, desktopcouch [20:27] why does quickly depend on desktopcouch then? [20:27] popping back to ubuntu, will be back in 10' or so [20:27] it used to have CouchGrid widget [20:28] rye: i think that widget has been removed now. maybe the package deps are not updated? [20:28] dobey: checking... [20:29] data/templates/ubuntu-pygame/project_root/python/hiscores.py:from desktopcouch.records.server import CouchDatabase [20:30] hmm [20:30] rye: talk to rickspencer3 about it then [20:50] dobey: system-font approved [20:54] yay [21:17] thisfred: I have a very small branch for review here https://code.launchpad.net/~ralsina/ubuntuone-control-panel/ubuntu-font-in-windows/+merge/97554 [21:17] 5 MLOC of XML later... [21:18] dobey: that's *2* lines of XML :-) [21:18] oops, I meant thisfred === m_conley_away is now known as m_conley [21:18] hehe, so we use the ubuntu font except when on Ubuntu? makes sense :P [21:18] thisfred: when the ubuntu font is the font in ubuntu, it will make more sense, agreed ;-) [21:19] it isn't? Maybe I set it by hand then [21:20] it is [21:20] the default is, anyway [21:20] ralsina: I'm proposing the branch with 2 of the 3 pages added (still pending the more complex and last one, the local to cloud) [21:21] dobey: I don't have it but may be because I updated from O? [21:21] dobey: as default, I mean [21:21] nessita: that's cool [21:21] ralsina: https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/cloud-to-computer-page/+merge/97990 [21:21] nessita: easier to test and everything [21:21] ralsina: well, kinda, I had to adapt the branch to have a page that will not be final to be final in this branch [21:21] now that I think about it, our embedded ubuntu font is probably not the latest [21:22] ralsina: did you set it to something else perhaps? [21:22] dobey: not that I recall [21:22] ralsina: so every branch results in a fully working wizard [21:22] nessita: good thinking [21:23] ralsina: +1 [21:23] thisfred: thanks! [21:26] thisfred: you dying to do another review, no? :-P [21:26] this one is fun, you will the premiere about the new wizard pages! [21:26] * nessita tries to be a sellwoman [21:26] nessita: sí! [21:27] YEAH [21:27] thisfred: will add IRL instructions and then share the link with you [21:27] kk [21:30] ralsina, thisfred: IRL instructions added to https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/cloud-to-computer-page/+merge/97990 [21:30] nessita: thanks! [21:30] * thisfred is on the case [21:30] alright all. am off for now. have a good weekend! [21:30] nessita: any suggestions about where to put a platform-specific constant on sso's gui? [21:30] later dobey [21:30] and I guess this is all for me today, will hang out for 15 minutes more but then I'm off to pilates and to have some nice night out [21:31] ralsina: which constant? [21:31] nessita: ":/windows.qss" on windows ":/linux.qss" on linux [21:31] nessita: to force the ubuntu font only on windows [21:31] ralsina: let me digg [21:32] nessita: I was thinking constant.py.in but I may have to make the setup.py a bitmore complex [21:32] ralsina: did you added the frozen thing to setup.\py? :-/ [21:32] nessita: no [21:32] ralsina: ah.... relief!!! [21:32] nessita: I added it on the runner [21:33] ralsina: hum... not to get_bin_dir? [21:33] nessita: oops, you are right, to get_bin_dir [21:33] ralsina: is that branch landed already? [21:33] nessita: yes [21:33] ok [21:33] nessita: it still needs some fixes by brian to add ".exe" at the end of binary names and such [21:34] ralsina: who approved? [21:34] nessita: hmmm don't recall outof memory [21:34] nessita: diego and dobey [21:35] I was told we never have to use this way: [21:35] 93 if hasattr(sys, "frozen"): [21:35] but instead: [21:35] 93 if getattr(sys, "frozen", False): [21:35] or [21:35] 93 if getattr(sys, "frozen", None) is not None: [21:35] by chipaca right? [21:35] yes [21:35] ralsina: hasattr is very evil [21:35] * briancurtin quietly changes the hasattr i just used for sys.frozen [21:36] nessita: when I check a folder to be synced locally, the screen flashes and then the checkmark is gone [21:36] briancurtin: change mine too please ;-) [21:36] nessita: briancurtin will undo that, sorry [21:36] thisfred: yes, the flash is the 'loading' screen, it should appear "the same" in the regular controlpanel folder screen [21:36] thisfred: and disapperas because is no longer "remote" (is local now that you subscribe to it) [21:37] briancurtin: thanks :-) [21:37] ralsina: is ok. SO, going back to the constant thingy [21:37] there does not seems to be a good place right now, I would advice: [21:38] ralsina: ubuntu_sso/qt/utils/{__init__.py,linux.py,windows.py, tests/test_linux.py,tests/test_windows.py} [21:38] nessita: I could make constants.py smarter by using Template (http://docs.python.org/library/string.html#template-strings) [21:38] nessita: I don't understand, but if that is known behavior then I'm fine with it. It looks to me like something didn't work. [21:38] thisfred: is known, will take notes to talk to design [21:39] ralsina: at this point, to be consistent, I would prefer the ubuntu_sso/qt/utils/{__init__.py,linux.py,windows.py, tests/test_linux.py,tests/test_windows.py} solution [21:39] nessita: ok [21:39] ralsina: any advantage in the template solution that I'm missing? [21:39] nessita: that the next time we need to add a constant that is not @prefix@ we will have to do it anyway [21:40] ralsina: woudl you be ok to wait till then? and then we may even apply the cleaner fix to controlpanel as well, since now we have the same @prefix@ trick [21:40] sure [21:41] nessita: also: http://ubuntuone.com/4MqS4Lr2pEscipn54Goup9 I don't understand what 'Sync locally?' without a checkbox means. Seems like I am being asked a question without a way to answer ;) [21:41] thisfred: those seem to be collapsed. What happens if you doble cick on "shared by cparrino"? [21:42] ralsina: 1. they should have an arrow then, and 2. nothin [21:42] g... [21:42] thisfred: ok [21:43] although I see log lines in the terminal appearing [21:43] thisfred: I assume you are already subscribed to everything you got from cparrino and that's why those are not there [21:44] ralsina: then neither should the question, right? :) [21:44] thisfred: exactly [21:44] nessita: want to fix that on this branch or should I approve this? [21:44] thisfred: looking [21:45] no crashes, the close dialog works [21:45] thisfred: oh nice bug... need to filter group headers with no folder in them :-) [21:45] awesome [21:45] thisfred: will fix in another branch, mainly because UIFe runs [21:45] ralsina: is that ok with you? ^ [21:45] kk, approving this then [21:46] thisfred: would you have the time to file me the bug as well? [21:46] sure [21:46] * nessita asks too much today [21:46] thisfred: thanks, really, I need to run to pilates ;) [21:47] nessita: I think you shouldnot remove the items when they are checked. That was not the behaviour it had before, and is weird. Will try to say something coherent about it before monday. [21:50] ralsina: ok... I may come up with something, but that may complicate the branch in a non trivial way... anyway, I haven't thought a lot about that [21:50] #957458 [21:50] nessita: ^ linked in mer [21:50] ge proposal and assigned to youh [21:50] thisfred: thanks!!! [21:50] nessita: it's ok, I will too. May have something for you to look at. [21:51] ok [21:51] now it's the dog's turn [21:51] later all [21:51] thisfred: just 1 sec [21:52] thisfred: nevermind [21:52] ok, I'm gone [21:52] bye all! [21:52] have a nice weekend [21:52] bye nessita! [22:04] so, anybody still up for a small review? https://code.launchpad.net/~alecu/ubuntu-sso-client/fix-wrong-credentials/+merge/97989 [22:06] anyway... EOW for me [22:06] bye all!!! [22:11] UnicodeWarning: Unicode equal comparison failed to convert both arguments to Unicode - interpreting them as being unequal success_filter=lambda info: info['path'] == path, [22:11] while trying to publish non-ascii file [22:12] what i dislike in twisted is that unless you specifically terminate it, it will keep the reactor running [22:12] even if there is a omg fatal error [22:14] im heading out of here. brain is fried. see everyone next week. === m_conley is now known as m_conley_away