[07:43] rye: hi rye, are you there? [08:29] psypher246, hi, yes [08:33] hey rye, how are you? [08:34] psypher246, well, it is still morning over here but doing fine, thanks, you? were you able to get the backup of the files that you deleted from Ubuntu One folder? [08:42] rye: where in the world are you? No my files have still stopped after 5000, nothign since then [08:54] psypher246, and did you receive my message from yesterday? The one that stated that if you put the files back into Ubuntu One folder then if they are identical, they will not be reuploaded but will appear immediately after MakeFile processing/ [08:54] ? [08:56] rye: no i didn't get that one [09:01] I would have to restore my data from another cloud provider anyway [09:10] rye: ok i have started my restore from crashplan [09:11] rye: really good speed they have, is there any plans on improving U1 download speed? [09:30] Good morning all! [10:06] morning ! [10:23] Hola mandel [10:35] facundobatista, buenas! === Guest18047 is now known as zumbi [11:09] good morning! === eu is now known as Guest8379 === Guest8379 is now known as adorilson [11:32] gatox, morning! [11:32] mandel, hi [11:32] gatox, can you help me with something [11:33] mandel, of course (if i can :P) [11:33] gatox, should be very small.. I have problems doing the point 2 point tests of the creds dialog [11:33] gatox, in the setup.py of the sso project we copy the bins to /usr/lib/ubuntu-sso-client [11:34] gatox, and therefore the bin is not present in my path, is that correct? [11:34] let me check [11:37] mandel, mmmm i'm not following.... i did a setup.py install from trunk to my machine the other day, and the result went to that folder, yes [11:37] gatox, ok, are the bin files in your path [11:37] nop [11:37] that is can you execute ubuntu-sso-login-qt [11:38] nop [11:38] gatox, meh.. so how is spawn_process suppose to find them? [11:38] is not in the path [11:38] mandel, hmmmm good question [11:38] gatox, I suppose I can find the full path [11:39] mandel, yes..... you are always going to know (or suppose to know) where they are in linux and windows [11:40] gatox, but that seems ugly.. [11:40] mandel, agree [11:40] mandel, but what choices do you have??.... i assume in windows was going to be even more difficult [11:41] to add it to the path [11:42] gatox, well, adding to the path is easy in the setup.py [11:42] but I'll work using the get_data_file call [11:44] gatox, should get_data_file return the correct path? as in get_data_path('my-creds-bin') [11:45] I get an error.. [11:45] mandel, which error? [11:47] gatox, File "/usr/lib/python2.7/posixpath.py", line 68, in join [11:47] elif path == '' or path.endswith('/'): [11:47] exceptions.AttributeError: 'NoneType' object has no attribute 'endswith' [11:49] mandel, maybe i'm wrong, but it seems the method is trying to do some folder operation with the string, and doesn't have the separator char [11:49] maybe not with your string, but with something inside [11:51] i've seen errors that like before in other project [11:51] gatox, hm.. I'll try to find what is going on.. [11:51] gatox, to set debugging for control panel is U1DEBUG, right? [11:51] mandel, yes [11:52] mandel, i'm pretty sure that any of the past that this function is checking is trying to determine at somme point if that string is a folder or something, and the python lib that does that using the string try to check to some info in the string for example ending with '/' and can't find that char that was expected [11:54] gatox, pdb for the rescue :) [11:54] gatox, although with the stupid qtreactor is a huge pain in the ass [11:57] ah osrry..... i didn't see the nonetype :P [11:57] gatox, sys.modules.get('ubuntu_sso.constants') [11:57] that returns none [11:57] hola! [11:57] the idea is the same..... python lib try to check if the string has folder-shape.... but in this case is none [11:57] alecu, hi [11:57] good morning, #ubuntuone! [11:57] morning gatox, mandel. [11:58] alecu, morning! [11:58] gatox, so, we are using sys.modules that just checks if the module has been loaded in this instance of the interpreter, which seems to be wrong [11:58] mandel, yes, returning None here too [11:58] mandel, I saw your branch had two reviews... but let me give it another anyway. [11:58] gatox, it should be using __import__, right? [11:59] alecu, yes please, I prefer to have 3 [11:59] gatox, alecu so, I see that get_bin_dir and get_data_dir won't work in ubuntu_sso.constants have not been loaded somewhere already [11:59] that seems to be a bug, since we won't find the paths [12:00] mandel, yep, it seems better (or at least working) with __import__ [12:11] brb! === gatox is now known as gatox_brb [12:14] nessita, may I pick your brain? [12:14] hello everyone! [12:14] mandel: sure [12:15] nessita, so, I've been doing a point to point test for the creds dialog and I'm noticed that the bin file is placed in /usr/lib/ubuntu-sso-client or /usr/local/lib/ubuntu-sso-client (if you use the setup.py and not the .deb) [12:15] yes, that is on purpose [12:15] nessita, that is not in the path, which means that spawn_process won't find it, right? [12:16] mandel: did you take a look to how we're using spwan_process in credentials.py? [12:16] nessita, so, I was using get_bin_dir form utils to pass the full path [12:16] nessita, no, I did not look at it, let me see [12:16] right, that's correct [12:16] credentials.py uses get_bin_dir [12:17] nessita, ok, the problem I was getting with get_bind_dir is that I don't import ubuntu_sso.constants anywhere, which means that sys.modules doesn't have it and there fore I get an attribute not found [12:17] nessita, all this installing sso via setup.py [12:19] mandel: that's fixed since yesterday, was a bug [12:19] nessita, oh, so I need to merge trunk :) [12:19] the problem is that tarmac may not be landing branches [12:19] let me check [12:19] nessita, ah.. ok [12:19] mandel: you can merge against this one https://code.launchpad.net/~nataliabidart/ubuntu-sso-client/find-me-bin-dir [12:20] ralsina: hola! you around? [12:21] nessita, ok, will do [12:21] nessita, hm.. then I'll block the merge 'til that is approve.. but will be good enough for my tests :) [12:21] mandel: yes, please depend your merge proposal on that branch [12:25] nessita, we are never going to support a local installation then? [12:26] mandel: what do you mean? [12:27] nessita: hello! [12:27] nessita: I finally managed to IRL the packages, they seem to work! [12:27] nessita: I finished the upgrade like 20 minutes ago [12:27] ralsina: oh uh [12:28] mandel: I'm asking since from my definition of local installation, we support it. But perhaps you mean other thing? [12:28] nessita, I mean in /user/local/lib [12:28] mandel: we support that [12:29] mandel: the BIND_DIR constant is built at installation time, you can check that constants.py.in file [12:29] nessita, hmm weird.. let me check [12:30] >>> from ubuntu_sso import constants [12:30] >>> constants.__file__ [12:30] '/usr/local/lib/python2.7/dist-packages/ubuntu-sso-client/ubuntu_sso/constants.pyc' [12:30] >>> constants.BIN_DIR [12:30] '/usr/lib/ubuntu-sso-client' [12:30] >>> [12:30] nessita, ^ [12:30] mandel: how did you install? [12:31] nessita, sudo python setup.py install [12:31] (in other order of things, I highly discourage having local installs of our projects, is misleading when testing and very easy to mix installations) [12:32] mandel: weird, at install time we replace lib_dir with what disutils tells us to [12:32] mandel: if you do: python setup.py install --prefix=/tmp/foo, it works [12:32] nessita, let me check [12:34] nessita, it does [12:35] yes, I tested that several times, but never tested what you did... perhaps is a bug in distutils [12:36] nessita, could be, I don't think is a problem at all with our code [12:47] back === gatox_brb is now known as gatox [12:47] I'm off to lunch, catch you in 30/45 min [12:53] nessita, ping [12:53] gatox: pong [12:54] nessita, are you able to run sso-qt in linux from trunk and get the captcha image?? it's failing here since yesterday (retrieving the captcha) [12:55] gatox: is working just fine here, used it heavily yesterday, and today [12:55] gatox: the GTk sso, no? [12:55] gatox: perhaps the Qt UI has a bug regarding captcha generation? [12:55] nessita, is working in gtk..... i don't understand why is failing in qt from trunk [12:56] gatox: BUG! :-) [12:56] nessita, but you said that is working for you [12:56] in Gtk [12:56] aHHHHHHH [12:56] (09:55:09 AM) nessita: gatox: the GTk sso, no? [12:56] * gatox debuggging mode ON [13:03] running oneric and i do not have the 'ubuntu one' option in the Synchronization tab under Presferences. normal? [13:11] ralsina: puedo setear una ventana en QT para que se abra centrada en la pantalla? [13:11] oops [13:11] spanish invasion [13:11] nessita: yes, but the WM may not let you [13:12] ralsina: is ok... how can I try? [13:12] nessita: you can use move() to put it anywhere, and there is a way to find the screen size, let me find it for you [13:13] nessita: QDesktopWidget has the info about the screen: http://developer.qt.nokia.com/doc/qt-4.8/qdesktopwidget.html [13:13] nessita: can't say I have *done* it though :-) [13:14] ralsina: ack, thanks [13:14] ralsina, i've done that [13:14] nessita, let me show you the code [13:15] I was hoping to use something like in GTk, where you say: window.set_position(Gtk.POS_CENTER) [13:15] nessita: not that I know of, sorry [13:16] nessita: not too hard: http://developer.qt.nokia.com/wiki/Center_a_Window_on_the_Screen [13:16] right [13:16] nessita: I would never had thought of that way to do it to be honest :-) [13:17] gatox: look at that solution ^ [13:18] ralsina, cuack.... i was used to the one calculating the coords :P [13:19] gatox: yes, me too! It's what alignedRect does anyway, but hey, it's neat :-) [13:21] mandel, ping [13:22] mandel, needsfixing on https://code.launchpad.net/~mandel/ubuntu-sso-client/webclient-use-dialog/+merge/93248 [13:22] mandel, please let me know if you need to discuss it, because I found some deep logic issues. [13:48] dobey: when you're around, let me know if you had any chance to review my controlpanel packaging branch [13:48] I will start uploading the ussoc packaging branch [13:57] nessita: the switch to plain dh looks fine [13:58] dobey: awesome, thanks! [14:06] ugh, hungry. :-/ === m_conley_away is now known as m_conley [14:16] briancurtin, nessita this branch needs yours re-review :P https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/931452/+merge/93003 [14:20] gatox: ack! [14:21] gatox: will do [14:27] thanks [14:33] alecu, ok [14:35] alecu, I understand the comments regarding libsoup, is not terribly hard, I simply did not know that we should pause the message [14:35] alecu, I do prefer to use is not None, I find it more explicit [14:38] nessita, ping [14:38] mandel: pong [14:39] nessita, I found an interesting thins with the runner package.. in sso the twisted.internet.reactor is in sys.modules yet is not running, so the if, elif finds that twisted is there and does not use the glib spawn_process [14:39] nessita, can you corroborate this? I might be wrong [14:40] mandel: hum, sorry, it may be the heat, but I could not parse you [14:40] mandel: when running the tests, only the twisted runner is used [14:41] mandel: when running IRL, on linux, either the glib or the qt will be used, never the twisted one [14:42] mandel: is that what you were asking? [14:43] ralsina: I'd need to mumble with you, nessita, diego and mandel, and lisette, in a few [14:43] nessita, well, I'm seeing when using runner in my branch that 'twisted.internet.reactor' is in sys.modules (maybe some package is pulling it [14:43] Chipaca, sure, when is a few since we have the weekly mumble in 16 mins more or less [14:43] Chipaca: we have the weekly call in 16 minutes... perhaps before or after that? [14:44] mandel: is that from tests? windows? [14:44] right now would probably be ideal [14:44] nessita, linux [14:44] lisettte: can you do now now? [14:44] and is diego around? [14:44] Chipaca: should be [14:44] Chipaca: yes [14:44] mandel: what branch? [14:44] nessita, let me push the latests version for you [14:45] nessita, I'd love to get somet help from you in this matter :) [14:46] Chipaca: happy to [14:47] Chipaca: mumble is rejecting my password [14:47] oh weather indicator [14:47] ralsina: quit and retry [14:47] ralsina: it lies [14:47] Chipaca, here! [14:48] gatox: hiya [14:48] gatox: mumblez? [14:48] how you consistently fail to be indicative [14:55] whoot. [14:55] hrmm, i guess i should test that the progress bar actually works now [14:59] meeting time! [14:59] * alecu fetches the laptop [14:59] dobey, I guess we are late for the meeting [14:59] alecu, ^ [14:59] bah [14:59] no [14:59] it's early still [15:00] *now* it's 10 [15:00] Chipaca, qt dialgos that look like gtk: http://ubuntuone.com/6Jy5HEW17OhGKB2WVPdI4O and http://ubuntuone.com/4QbSsMvydxLNt3zhRXgO1f [15:01] Chipaca, icon is grabbed from theme as well as font. link color is an issue due to the lack of propoer gtk3 theme support on qt [15:01] mandel: that looks a lot, lot better than what I get on my machine [15:02] Chipaca, those are the ones done for proxy with input from lisettte and mpt, no the ubuntu one account [16:20] quick lunch.... brb! [16:23] not quick lunch yet :P [16:29] noow quick lunch.... brb [16:32] briancurtin: free now if you are! [16:33] ralsina: yep, connecting in mumble now [16:46] alright; lunch time for me [16:46] bbiab [16:58] lunchtime here as well [17:06] end of lunch [17:13] and lunchtime around here. [17:28] alecu, do you have time for a mumble? [17:28] mandel, sure, lunch can wait. [17:28] mandel, let me open mumble [17:28] alecu, should be very very fast [17:35] mandel: so, you still need my help? [17:47] nessita, yes, super small quick question, give me a sec, I'm on mumble with alecu [17:47] nessita, ok, done [17:47] mandel: you wanted to look at a branch of yours? [17:48] nessita, yes, although, better to do a very quick mumble, ok? [17:48] nessita, is not a crazy amount of code [17:49] mandel: can't right now, having a parallel chat [17:49] nessita, ok, then we type :) [17:49] nessita, in ubuntu_sso/runner/__init__.py I'm getting a true for 'twisted.internet.reactor' in sys.modules when running the control panel [17:50] nessita, I think it will be nice, as a save guard, to check if the reactor is running [17:50] so, if 'twisted.internet.reactor' in sys.modules: import reactor; return reactor.running [17:50] nessita, I don't know what in control panel bring the reactor in, but is there :( [17:51] mandel: what has the control panel to do with the runner? [17:52] nessita, the control panel uses webclient, right? [17:52] nessita, so, when ti tries to get through the proxy it needs the creds [17:53] mandel: may be, I never tested that flow [17:53] mandel: but you're not calling the spwanner from the controlpanel, right? [17:54] nessita, so, how is webclient going to get the proxy creds? [17:54] mandel: from the sso service? perhaps I'm missing something? [17:55] nessita, hm.. adding ipc to call sso to get the proxy creds, is that what you are thinking? [17:55] mandel: no... having all the proxy creds isolated in sso's webclient. Not sure why you're spawnning from the controlpanel code [17:56] alecu: am I missing something there? ^ [17:56] nessita, lets say the following, sso webclient gets a 407 and needs to get the creds, how would you show the ui? [17:56] nessita, he just went to have lunch [17:56] mandel: what's a 407? [17:57] nessita, proxy-auth http error [17:57] mandel: my point is that the spawnning should be in sso webclient's code, not in controlpanel [17:57] nessita, it is in sso code, yet, is not sso how executes it, since it will be the control panel process [17:57] aahhh that's different [17:57] you scared me! [17:58] nessita, ;) [17:58] mandel: controlpanel does not use a reactor. So, if there is a reactor in sys.modules, some module added it to it [17:59] nessita, exactly, I don't know which, that is why I think adding that reactor.running test is a good idea [17:59] nessita, just to be super save [17:59] s/save/safe [17:59] mandel: but that just masquerade the problem, no? [17:59] why we're getting reactor in sys modules if no one installed it? [18:00] nessita, well, maybe there is an def function(): from twisted.internet import reactor somewhere [18:00] mandel: I just grepped on controlpanel, there is only this: [18:00] nessita, I have not touched the controlpanel code, I know it is there due to a very annoying debugging process :) [18:00] ubuntuone/controlpanel/gui/qt/main/windows.py:37: from twisted.internet import reactor [18:01] which is inside a function call [18:01] so that code should never be executed on Linux [18:01] nessita, maybe is this guy: ubuntu_sso/utils/tcpactivation.py:21:from twisted.internet import defer, protocol, reactor [18:02] nessita, it is not in a function and is not ignored on linux [18:02] mandel: but that is not imported from controlpanel... afaik. And if that were interfering, ussoc will not work on trunk right now (and it does) [18:03] mandel: I guess your solution can work as a workaround, but please file a bug about this [18:03] nessita, will do, so is it ok if I add that in my branch? [18:03] and add the bug in a comment where you implement the workaround, ok? [18:03] mandel: I would do a separated branch [18:04] nessita, sure, should I then file a bug, propose the branch with the work around and a couple of tests to ensure it works? [18:05] mandel: no need for tests, this is not testable I think since we always have a reactor running in the tests [18:05] nessita, we can patch sys.modules, try without, then sys.modules with it and patch .running [18:06] nessita, but if you say I don't need to, I wont ;) [18:06] mandel: no need for now [18:06] nessita, great! [18:09] meh, pyqt4 supports python3 it seems [18:11] dobey: has supported it for 2 years or so [18:12] dobey: the same source package builds for python 2 and 3 and for Qt 2 to Qt 4 which is completely insane [18:12] yeah i know [18:12] well, this is only qt4 afaict [18:18] ralsina: can I have a review? https://code.launchpad.net/~nataliabidart/ubuntu-sso-client/use-qtmainloop-by-default/+merge/93472 [18:18] nessita: of courses [18:19] dobey: would you know how to have gettext returning unicode translations? [18:19] nessita: uh, it does already [18:19] dobey: hum, on ussoc is giving me bytesstrings [18:19] sorry, "bytes" is what I meant [18:19] nessita: they should be utf-8 byte strings though [18:20] dobey: right, I want unicode, not bytes [18:20] (which is what makes sense for words) [18:20] nessita: then you'll have to do the python thing to convert it [18:20] dobey: I think there is way of telling gettext to return unicode [18:20] instead of doing the decode by hand [18:20] will google! [18:20] nessita: afaik the api we're using is a straight wrapper to the C api [18:21] nessita: and C does not have "unicode". it has arrays of chars [18:21] dobey: ins't this API? http://docs.python.org/library/gettext.html [18:21] nessita: yes [18:21] dobey: there it says "gettext defines a “translations” class which implements the parsing of GNU .mo format files, and has methods for returning either standard 8-bit strings or Unicode strings" [18:22] you can set unicode=True when you call install [18:22] we can pass a switch to unicode [18:22] briancurtin: yeah! you won :-) [18:22] nessita: that is a completely different thing [18:22] briancurtin: thing is that we're not installing it anywhere [18:22] than what we are using [18:22] dobey: it is? [18:22] we're not using th class based api [18:23] converting to it would be a pretty big change i think [18:23] dobey: any reason not ro? [18:23] ah [18:23] dobey: how big? [18:23] hmm, maybe not that big [18:24] isn't just calling install() and not importing _ anymore? [18:24] well i don't know how it affects the other functions we need [18:24] particularly ngettext() or dngettext() [18:25] dobey: ack. Though we only use those in u1client [18:25] maybe in sso it's not an issue [18:25] right [18:25] since i guess there aren't any plurals [18:25] and we're having a nasty error ATM because of mixing unicode and bytes [18:26] or just switch to python3 [18:26] where everything is unicode [18:33] nessita: +1 and looks trivial to me [18:33] ish, but yes [18:34] nessita: launchpad is bouncing my +1 I am retrying [18:34] bu, any reason why is bouncing? [18:34] "the following error: [OK]" [18:34] now it's in [18:34] ok, so +1 for reals now [18:34] thanks@! [18:35] hmm [18:35] ralsina, so, the bug where when you get a share, the messaging icon goes blue without telling you why [18:36] is it targeted for 12.04? [18:36] beuno: we didn't get to it [18:36] beuno: if we have a few spare cycles, we may yet [18:36] beuno: but it's looking unlikely [18:37] while following https://wiki.canonical.com/UbuntuOne/DeveloperSetup on precise, i'm getting trouble at installing ubuntuone-developer-dependencies (complaint from database deps about postgres versions) [18:37] beuno: unless you buy dobey some rhum and convince him to work on saturday ;-) [18:38] I can do that [18:40] https://pastebin.canonical.com/60447/ is what i'm seeing while apt-get install'ing (pardon the potentially dumb question, i'm a windows guy :) [18:41] briancurtin, so, before I even look [18:41] our server is only currently guaranteed to work in Lucid [18:41] we're all using LXCs for that [18:42] any reason you didn't choose to do that? [18:44] beuno: i just got a new P box setup and was just going through the steps on that wiki. basically i'd like to be able to test the stuff i'm doing on windows also on linux, rather than proposing branches and asking people to test for me [18:44] briancurtin, right, so lets get you set up in an LXC [18:46] beuno: that might actually be fixed now [18:46] briancurtin, https://wiki.canonical.com/UbuntuOne/DeveloperSetup/LXC [18:46] beuno: on precise, anywya [18:47] dobey, you mean the envelope thing? [18:48] it's not, just happened to me with an updated system [18:49] ok [18:52] nessita, when you have a moment: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/sso-styling/+merge/93475 [18:52] nessita, i'm going to work in the captcha thing now [18:52] gatox: ack! [18:53] ralsina, also, do you know that the existing P control panel tells you to install desktopcouch? [18:53] I'm assuming that's going to get smashed by the QT one? [18:53] beuno: yes [18:54] cool cool [18:54] beuno: and the answer is "it depends" ;-) [18:54] i think they put extra triptophan in my chicken [18:54] But yes, the desktopcouch dependency makes no sense anymore [18:55] nessita's going to build a gtk3 control panel tomorrow, and we're going to get a ffe/uife to get it in ;) [18:56] dobey: yeah, and I also loved when you pull me up back in .ar (?) [18:56] lol [18:57] speaking of which.... mate and medialunas! [18:57] hehe [18:58] * nessita is hungry [18:59] ralsina: would you have at hand a screenshot of the current first screen of the windows wizard? the one that has a U1 specific image [19:00] ralsina: I would need that to finish filing the FFe [19:01] nessita: sure, in 1' [19:02] nessita: http://screencast.com/t/Zui8Y0Io19 [19:04] ok, 2 hours [19:09] ralsina: may I ask some more screenshots? I would need those from the wizard pages that are not sso-related, to fill the other FFe [19:10] nessita: sure, doing one of each [19:10] thanks! [19:10] nessita: you only need the non-sso ones? [19:16] i think nessita will have to buy me mate and medialunas for touching pyqt [19:16] ralsina: yes [19:16] dobey: you can count on that [19:17] dobey: consider it part of your yearly bonus [19:18] hah [19:24] nessita: http://screencast.com/t/xzPusNP1 http://screencast.com/t/0EIZHQ1PmXt am I missing any you need? [19:25] ralsina: bandwidth setttings? [19:25] ah no, sorry! [19:25] ralsina: computer to cloud [19:25] nessita: ok, trying to trigger that one now [19:28] nessita: http://screencast.com/t/je9ffgCMF5uo [19:28] ralsina: thanks! [19:29] nessita: de nada [19:36] ralsina, MIR approved. [19:36]