[11:11] good morning [12:19] /away [12:19] good morning! [12:20] ralsina, hi [13:03] is u1lint on osx being replaced with /bin/true or something? [13:05] dobey, hi! i'm having some problems with tihs branch: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/menu/+merge/118117 i don't know if tarmac is not working or something.... [13:05] it appears the -trunk tarmac is offline [13:08] hello dobey! [13:10] have no fear, i is here! [14:02] brb, need to reboot apparently. sound card didn't get picked up when i turned workstation back on after holidays [14:13] bah, stupid sound [14:15] hi everyone. welcome back, dobey! [14:17] hello mmcc [14:17] dobey: if you can't get the soundcard to work we can arrange something with phones [14:18] and skypes, and ponies [14:18] ralsina: is there a call today in about 40 min.? [14:18] joshuahoover: yes [14:18] ralsina: k, thanks [14:19] brb making coffee [14:20] ralsina: the internal sound card isn't getting picked up by pulseaudio for some reason. my external usb sound interface for my studio equipment works though, so i'll just have to use it [14:20] dobey: so we should get extra quality dobeysound! [14:23] eh [14:27] hey mmcc; thanks [14:31] dobey, is tarmac working now?? should i re-approve this branch or something? https://code.launchpad.net/~diegosarmentero/ubuntuone-client/menu/+merge/118117 [14:35] +SET IGNORE_MODULES="test_linux.py, test_libsoup.py, test_darwin.py" [14:36] huh, why are there spaces in there [14:36] gatox: it should be working now [14:36] all: it's my review day so let me know if you need a look at anything [14:36] alecu: https://code.launchpad.net/~brian.curtin/ubuntu-sso-client/remove-SyncTimestampChecker/+merge/118826 is updated, i think it should have the lint issues fixed [14:36] gatox: ah, something broke u1-client though [14:37] dobey, ahhhhhhh [14:37] i really need to figure out why lint on windows doesn't find any of the things alecu always responds with [14:37] briancurtin, if you have some time, please review this branch: https://code.launchpad.net/~diegosarmentero/ubuntuone-control-panel/refactor-sync-status/+merge/118809 is just refactoring some stuff [14:37] briancurtin: I don't have lint running on windows :P [14:37] briancurtin, for a future branch [14:37] briancurtin: https://code.launchpad.net/~alecu/ubuntuone-client/fix-download-finished/+merge/118759 [14:38] gatox, alecu: looking [14:38] gatox: next run should pick it up [14:38] dobey, great!! thanks! :D [14:38] in fact, both my windows and mac dev envs need some rebuilding [14:39] briancurtin, also this one please :D https://code.launchpad.net/~diegosarmentero/ubuntuone-client/ipcmenu/+merge/118621 [14:39] oi, am hungry [14:39] me too [14:40] dobey: that IGNORE_MODULES line, what's that from? looks like something I wrote but I don't know where. [14:41] mmcc: from https://code.launchpad.net/~mikemc/ubuntuone-control-panel/fix-1024623-install-confs/+merge/118404 [14:41] mmcc: it had spaces in it before your change though [14:41] although it is in quotes, so maybe not too bad [14:42] ah, right, the .bat file… [14:42] mmcc: though, did you see the comment i added on https://code.launchpad.net/~mikemc/ubuntuone-storage-protocol/fix-run-tests/+merge/118488 ? [14:43] dobey: yes I did. for some reason a pyflakes egg is in the buildout but not the environment it set up, so I never used it [14:44] I'll make the suggested change, no problem. [14:45] mmcc: right; i'm wondering how u1lint ever succeeded on osx/windows with your branch though, given the pylint brokenness there [14:47] dobey, it didn't. I should've made a note in the merge - there were so many lint errors that I figured that was an issue for later. Of course I didn't think about using pyflakes instead :| [14:49] right, we need pyflakes there :) [14:50] ralsina: are we going to do hangout or mumble? [14:50] dobey: mumble [14:50] my hangout is broken [14:50] call up google support...oh, wait [14:51] ralsina: do you use chromium or firefox? (or real chrome?) [14:51] dobey: chromium [14:52] ah. it surprisingly works ok for me in firefox [14:53] how do you tell a dialog in qt to be modal, programmatically? [14:54] briancurtin: http://pastebin.ubuntu.com/1137918/ [14:54] briancurtin: could you review https://code.launchpad.net/~dobey/ubuntuone-dev-tools/fix-coverage/+merge/118207 please? [14:55] seems to have gotten no reviews while i was gone [14:55] alecu: thanks, fixing [14:55] dobey: adding to my review queue, i'll look [14:55] briancurtin: if you can't make lint work on windows, please try running the tests on an ubuntu vm. [14:55] ralsina: ^^ also :) [14:55] dobey: got it [14:56] alecu: the weird thing is that lint on windows works, just that it doesn't seem to pick up as many things as it does for you. since this has been happening a bunch lately, i'll look into it [14:56] briancurtin: it used to pick up a milllion things too many in my case [14:57] ralsina: when i was putting together this branch alecu is mentioning, it said maybe five things and i fixed them, then alecu came back with like 10 more things. its weird because none of the things alecu found were platform specific, just bare imports that would happen everywhere [14:57] briancurtin: weird === zyga is now known as zya-afk [15:00] briancurtin: for the imports, I didn't rely on pylint; I just thought of them while reviewing the code. [15:00] alecu: good eyes/mind [15:00] :) [15:00] briancurtin: I think I didn't ran the tests at the time. [15:00] :-) [15:01] ralsina, thisfred, mmcc, briancurtin, mandel: mumble? [15:01] going [15:01] on the way [15:01] did i miss anyone? alecu/gatox were already there, as is joshuahoover [15:02] mandel is on vacation, so I think you missed noone [15:02] dobey: I think not [15:02] me [15:02] thisfred: mumble [15:02] ah [15:33] alecu: is it possible to unsubscribe the volume w/o actually firing up sd - manipulate the metadata [15:35] rye: there might be some way to manipulate the metadata with some byte twiddling, but we have no tool to do that. [15:40] so... I need to run some errands downtown. I've been putting this away the whole week because of the strike of the subway workers... but it can't wait anymore. [15:40] wish me luck! [15:40] alecu: good luck! [15:40] thanks! [15:40] subway workers strike? no subway service? [15:40] alecu is hopping up front to drive the train [15:40] rye: yup, since sunday. [15:41] briancurtin: lol [15:41] alecu: briancurtin http://xkcd.com/726/ [15:41] alecu: ever considered asking a motorcicle messenger service to deliver you? ;-) [15:41] ralsina: I'm on the "heavy packages" category. [15:42] That comic makes it look like the window is open in the cockpit [15:42] alecu: in paraguay you can take a moto-taxi like that :-) [15:43] ralsina: that would be an awesome service here [15:43] alecu: I have a friend who is motoquero, maybe he can take you [15:43] ralsina: even a two person bike where only the driver pedals.... we could be millonaires. [15:43] alecu: boke rickshaws like in london? [15:43] bike* [15:44] ralsina: it would have to be slimmer, to take advantage of our bike lanes. [15:44] a bike-rickshaw-train hybrid [15:44] given a pyflakes .egg, is there an easy way to turn that into a pyflakes runnable script? I'm hoping there's some command that'll do it for me [15:44] mmcc: pip? [15:44] * alecu should get a bike. [15:45] mmcc: pip install thisegghere.egg [15:45] ralsina: where does that put it? [15:45] mmcc: magical places [15:45] mmcc: no idea :-) [15:45] right. heh. "in my virtualenv" if I had one [15:46] I wanted to do it in a buildout compatible way, since the buildout includes the egg [15:47] ok, there's a buildout way to do this. I guess we just aren't doing it yet [15:52] no, the buildout should've installed scripts from the eggs it has... maybe pyflakes has a broken egg or something? [15:58] mmcc it's not in the buildout's bin? [15:58] ralsina: no, but there's a chance I've broken it. I'm testing it in a clean buildout. pyflakes' egg looks fine [16:01] gotta go banking before the banks bank no' mo' [16:01] perrito666: como kung fu? [16:01] oops, wrong channel [16:02] ralsina: was that a banking transaction? [16:06] * gatox luncch! === gatox is now known as gatox_lunch [16:08] ok, off to lunch, bbiab [16:13] Necesito más café [16:14] help? I'm getting this on 12.04: Traceback (most recent call last): [16:14] Failure: ubuntu_sso.utils.webclient.common.WebClientError: (u'SSL handshake failed', u'') [16:15] can't connect or sync [16:31] ghoulmann: are you using any kind of a proxy? [16:32] rye: not at all [16:33] ralsina: ^ i don't recall this on Ubuntu [16:34] ghoulmann: the login screen constantly says that authentication failed, right? [16:36] rye it tells me nothing (syncing eternatally) until i go to account information. when I do that, i get WebClientError [16:36] u'SSL handshake failed' [16:37] ghoulmann: when you visit https://one.ubuntu.com - were you asked to add the site to the exceptions? [16:38] rye: not that I remember. I do not when I do it now. [16:38] i can login with no problem [16:38] on one.ubuntu.com [16:43] ghoulmann: may I ask you to run this script - http://people.canonical.com/~roman.yepishev/us/ubuntuone-ubuntu-syncdaemon-diag.sh - bash "ubuntuone-ubuntu-syncdaemon-diag.sh" - it will collect the log files and system/package info and will write an archive to the desktop folder. Please send the archive to ubuntuone-support@canonical.com, I will take a look [16:44] sure [16:48] rye : sent from rikgoldman@gmail.com [16:50] ghoulmann: got it and found the reason, switching to private chat [16:56] ralsina: us.archive.ubuntu.com does not have control panel 3.0.2, but ubuntuone-client is at 3.0.2 causing failures, going to #is [17:02] so, I can't figure out why buildout isn't installing pyflakes' script -- if anyone is more familiar with buildout and wants to take a look, here's a very verbose trace of it re-installing pyflakes after I deleted the egg: http://paste.ubuntu.com/1138110/ [17:07] hrmm [17:07] ralsina: cancelling, there's no 3.0.2 released, sorry [17:08] rye: i think cp 3.0.2 had no changes from the previous version [17:08] rye: so i didn't waste my time uploading it to precise yet [17:10] ghoulmann: what version of python-ubuntuone-storageprotocol do you have installed? [17:11] mmcc: odd [17:18] I think the problem is that pyflakes doesn't have an entry_points.txt in its EGG_INFO [17:25] i'm not sure that's it [17:25] why would it have one? [17:27] because buildout apparently uses that to decide what scripts to put in bin/ [17:28] weird [17:28] i would think it would do setup.py install and be done with it [17:28] or rather, it seems to be doing easy_install [17:28] which i would think also uses setup.py install [17:28] and pyflakes clearly lists bin/pyflakes in scripts= argument [17:30] but it's installing from the egg, there's no setup.py [17:30] easy_install is installing from an egg? [17:30] buildout is [17:31] then what is this mention of easy_install? [17:31] Unpacking pyflakes-0.5.0 to /var/folders/sn/knffnnd96s9c237ktkb4llxm0000gn/T/easy_install-5gXYNo/pyflakes-0.5.0 [17:31] like that [17:32] ok, it runs easy_install - see line 137 of the paste [17:32] it seems to build an egg in that directory [17:33] mmcc: are we using python in /usr/local/Cellar, or are we supposed to be using a python from the buildout dir? [17:35] dobey: for this part we should be using /usr/local/Cellar -- the buildout generated python shouldn't be generated until after the egg dependencies are installed, I think [17:36] the generated python is just "sys.path = [list-ofpaths-to-installed-eggs] , then code.interact(), basically" [17:36] creating build/bdist.macosx-10.7-x86_64/egg/EGG-INFO/scripts [17:36] mmcc: ^^ what's that file contain? [17:37] oh i guess that's a directory [17:37] copying build/scripts-2.7/pyflakes -> build/bdist.macosx-10.7-x86_64/egg/EGG-INFO/scripts [17:37] and pyflakes should be in it [17:38] that's in a temp, but the final buildout-dir/devsetup/eggs/pyflake.egg/EGG-INFO/scripts has the pyflakes script, yees [17:39] the problem is, I want buildout to put that in the same bin/ as e.g. the pep8 script, and its wrapped python [17:39] so it is there [17:39] yeah, buildout just doesn't know about it [17:39] mmcc: where is u1lint in the ubuntuone-dev-tools egg, as a point of comparison? [17:40] dobey, also at eggs/ubuntuone_dev_tools-3.0.0-py2.7.egg//EGG-INFO/scripts/u1lint [17:40] and also not installed correctly by buildout [17:41] it is; it's just that pyflakes and dev-tools don't use pkg_resources entry points to run the scripts [17:41] (this is why the env-mac is setting $u1lint) [17:41] and pep8 does [17:41] right. but buildout uses those to decide to generate scripts in the bin/ directory [17:41] well env-mac is setting u1lint because we need to use the development version, not the old one we're installing as an egg [17:42] then buildout is broken [17:42] dobey - env-mac just does a 'find . -name u1lint' -- to get this to work, I need to delete the dev-tools egg :) [17:43] buildout is definitely broken [17:43] I should be able to add entry points definitions to the buildout.cfg to get it to install all these things [17:43] trying that now [17:44] well, we should also stop pulling dirspec/dev-tools as eggs [17:44] i think most of us do that but it's not easily automate-able, at least as far as i can see [17:45] briancurtin: do what? [17:45] mmcc, alecu: also, i see you had mentioned me, and the options parsing changes to u1trial on monday. is there a problem there? [17:45] mmcc: use dirspec/dev-tools from a checkout rather than the outdated eggs [17:45] briancurtin: oh right [17:46] briancurtin: I think there is a way to fix that… [17:46] mmcc: i looked for a while and couldnt find anything that shows how to get something from a checkout to be inserted into the sys.path. it automatically does that for the eggs [17:47] dobey, briancurtin FYI in a call last week about darwin stuff I volunteered to un-break buildout, so this is happening unless ralsina tells me to stop… the goal is to get darwin working without so many fiddly steps [17:47] briancurtin: ah ok. I'll see if I get lucky and notice a way [17:47] briancurtin: by the way, how 'bout those buildout docs? [17:47] mmcc: moving away from buildout is also an option. i once looked into virtualenv/pip but moved on to other tasks [17:47] ugh [17:48] * mmcc seconds that ugh [17:49] also, gotta love silently ignoring config file entries that aren't right. that'll teach me! [17:58] … that'll teach me to be editing the right file. [17:58] * mmcc feels shame [18:07] anyway, buildout makes a bin/pyflakes now, moving on to u1lint and u1trial [18:22] mmcc: go ahead and unbreak [18:22] mmcc: and/or switch [18:22] ralsina: I'm on it boss. I think unbreaking will be faster than switching [18:23] mmcc: I just got a message telling me "INSTALLBUILDER 8.5 NOW AVAILABLE, INCLUDES MAC OS X MOUNTAIN LION SUPPORT" [18:24] ralsina: is that what we use on windows? [18:24] mmcc: yep [18:24] we do? [18:24] dobey: yes [18:25] pinky swear! [18:25] I mean, not that version, but yes that thing [18:25] ah, so we do [18:25] i thought it was just called "bitrock installer" [18:25] bad memory [18:26] "bitrock installbuilder qt edition" I think [18:26] with a "professional" somewhere in there [18:26] right [18:27] now what broke in quantal :-/ [18:28] what broke in quantal in general or in our code? [18:30] in general. just pulling a patch to fix a bug in ubuntuone-installer, so we can ship it out to precise for 12.04.1, but tests failing to run on quantal now when i try to build the package [18:33] oh [18:33] i see what happened :( [18:34] and crap [18:35] ugh; new aptdaemon broke the tests in ubuntuone-installer, incidentally [18:38] briancurtin, alecu: this looks interesting for the future: Qt + metro http://www.digia.com/en/Blogs/Qt-blog/Sami-Makkonen/Dates/2011/11/Deploying-on-Windows-8-Tablets-with-Qt-Commercial/ [18:42] cool. im interested to see what happens with the python side of things. i think martin (one of the other windows core devs) has python ready to run inside metro apps, or something like that. i havent looked into it too much [18:46] briancurtin: it looks like there is a proof of concept QML/Metro thing that exists already, and (separately) a Metro style for Qt widget-based apps. [19:35] brb [19:35] * mmcc -> lunch [20:00] back [20:00] I just walked 5km... not used to this. [20:05] ok....... eod here, and i need to go to buy mass effect 2 and 3 jeje so.... bye! :D [20:06] ralsina: qt on metro sounds interesting... [20:07] ralsina: and I really like the video. The guy sounds quite the salesman... [20:07] an undead one [20:07] haha [20:07] and that company is the one that bought Qt from Nokia yesterday [20:08] which video?? [20:08] * gatox thinks he misses part of the chat cause of internet issues [20:09] i read that qt5 is going to work on the integration with windows8 ui [20:10] * gatox feels ignored [20:18] gatox: http://www.digia.com/en/Blogs/Qt-blog/Sami-Makkonen/Dates/2011/11/Deploying-on-Windows-8-Tablets-with-Qt-Commercial/ [20:19] briancurtin, thanks brian!!! you are a really good person for not ignoring me! [20:19] jejejeej [20:19] :) [20:19] gatox: that's why you should use quassel or bip ;-) [20:19] gatox: I can give you an quassel account, someone else can get you a bip one [20:20] ralsina, i don't know what happend i have been having problems with internet..... i realize after a long time....... when the chat does timeout [20:20] ralsina, but i like to use xchat :( [20:20] ralsina: https://code.launchpad.net/~dobey/ubuntuone-dev-tools/fix-coverage/+merge/118207 please? [20:20] now..... i can leave in peace :P [20:20] bye! [20:23] dobey: sure [20:29] dobey: why the print('') ? [20:30] ralsina: because print() was printing a literal '()' on the console for some reason; no idea why [20:30] python == crazy [20:31] dobey: I would just tack a \n at the end of the previous print, or if you want to get fancy, add end='\n\n' ;-) [20:32] dobey: it was probably printing the literal before you added __future__.print_function? [20:32] briancurtin: maybe [20:33] dobey: which makes total sense [20:33] I can confirm that with print_function, print() does the right thing [20:33] should i revert that line then? [20:34] dobey: or remove the print completely [20:34] or replace it with a \n instead [20:36] yeah i put a \n in the prior print statement [20:38] dobey: ok, +1ing it [21:01] EOD for me, will be back late at night, ping me for reviews [21:01] bye! [21:14] ralsina: https://code.launchpad.net/~dobey/ubuntu-sso-client/fix-glib-main/+merge/119045 if you could [21:14] briancurtin: ^^ also you please :) [21:14] note, it's stable-3-0, not trunk [21:15] dobey: will look shortly [21:24] briancurtin: so, regarding the {} vs dict issue, I see that {} is usually faster. Is there any other reason I'm missing? [21:24] alecu: not really. it just stuck out because i saw other places in the code that were using the literal, but that one used the dict constructor [21:25] i dont think its speed is enough to matter. it just looked different so i mentioned. not a big enough deal to hold anything up, though [21:25] briancurtin: my reasoning is that dict(...) looks cleaner to me when doing dicts that will be used as kwargs... [21:25] but I don't have a clear explanation, other than "I like it like that" :P [21:25] that works for me [21:30] briancurtin: ok, I'll leave it like that then. Thanks for bringing it up, though :-) [21:32] alecu, you're gone tomorrow right? do you have a sec to think about the syncdaemon bug on darwin now? I'm not sure if it's a simple fix or not… [21:35] On darwin, creating a file by copying it in or by "echo foo > text.txt" results in a single IN_CREATE event, but as far as I can tell it looks like the code is expecting a create to be followed by a close, and it won't be on darwin. [21:36] I looked at the windows docs, and it looked like the same is the case on windows, so I'm not sure if I have the right root cause yet [22:08] have a good evening all [22:13] mmcc: yes, I'm gone tomorrow. And yes, I think on Windows we don't have the close events either [22:13] ralsina: from what i can see having run 2to3 again, pending the few branches i just proposed, the syntax should be ready for us to start running [22:14] briancurtin: go ahead and run it!!!!! :-) [22:15] ralsina: the next/__next__ thing is not really an issue at the moment, Qt is just shadowing the built-in "next" name. we'll have to add some lint enable/disable comments around tests where we define the same name, but thats it [22:15] alecu: hmmm, ok. [22:15] alecu: i'm about to take off for the day, and don't currently have a 3.x environment setup for it with all of the dependencies. i'm off tomorrow so i can start with that first thing on monday [22:16] briancurtin: oh, right. It may make sense to start running this in a Quantal vm [22:17] alecu: good call, i'll do that [22:17] briancurtin: have a good weekend! [22:17] alecu and all: you too! [23:36] ooook, EOW here!