[07:40] morning all! [09:01] mornings [10:47] mandel: do you happen to have some code snippet that performs the same SSL verification as the SSO using Qt libraries? [11:01] rye, yes, give me a sec [11:02] rye, try with this: lp:~alecu/+junk/qt-ssl-error [11:02] gatox, buenos dias! [11:02] mandel, buenas === yofel_ is now known as yofel [11:14] erm [11:15] mandel: are we using qt4reactor in precise? [11:15] mandel: i see import qt4reactor in proxy dialog but found that my machine has no python-qt4reactor package installed at all [11:15] hm.. no, there we use the qt main loop [11:16] mandel: ah [11:16] mandel: but proxy dialog? [11:16] rye, we should not, let me check [11:17] rye, which file are you exactly looking at? [11:18] mandel: ah, if platform is win32, ok [11:18] rye, :) [12:04] lunch time here :) [12:04] * mandel lunch [12:29] brb.... need to do some errands === gatox is now known as gatox_brb [12:50] ralsina, could use a small review of https://code.launchpad.net/~thisfred/u1db/colored-squares/+merge/119376 por favor (we decided to not make the tag colors be the background color for the whole row, to increase readability, and reduce angry fruit salad, as aquarius puts it) [12:50] with thanks to rodney for the solution :) [12:51] heh [12:51] morning all! [12:51] beunos días alecu [12:53] damn, he's even putting the accents over the I's. We'll have to invite him to #ar soon... [12:53] alecu, unless #es makes me a better offer :P [12:54] :-) [12:55] I was going very fast for a while, and then I hit the past tense, and now I slowed down considerably, since it gets a little more irregular there === gatox_brb is now known as gatox [13:02] thisfred: will look! [13:02] muchas gracias [13:02] thisfred: noone cares about the past,even if it's tense [13:03] alecu, i have a u1-cp branch that keeps bouncing because of a lint issue that it says that a u1-client class doesn't has a"sync_menu" attribute.... but the u1-client branch that adds that already merge yesterday..... do you know what might be happening? [13:03] thisfred: also, one way to reduce angry fruit salad is to use a pastel color palette [13:03] ralsina, which is what I tried to do, but not to aquarius' liking [13:03] gatox: try branching trunk, and then merging that branch. And see if lint complains. [13:04] thisfred: fair enough [13:04] thisfred: BTW, pastel in spanish means pie. Just to avoid further confusion ;-) [13:04] gatox: merged != installed [13:05] dobey, ahhhhhhh that's the problem then! [13:05] dobey, is that going to take a lot? [13:06] gatox: should have happened automagically really. not sure why it isn't. not that familiar with jenkins/puppet oddities [13:06] ok..... i'll wait [13:06] gatox: it's updated now, so should work [13:06] trying again... [13:07] thisfred: can I ask you to do something that is really not branch-related? [13:07] ralsina, claro [13:07] thisfred: please get rid of simplejson or at least try to import json first, since it's in the stdlib [13:08] we need to get rid of simplejson [13:08] dobey: yes [13:08] ralsina, I have tried to do so, but to great resistance, since jam bookmarked them, and simplejson was still faster [13:08] dobey: but eliot promised to do a branch himself in the bug we have for that [13:08] thisfred: hmmmmm [13:08] ehm [13:08] benchmarked [13:09] thisfred: understood as that :-) [13:09] ralsina: eliot? [13:09] dobey: it's early ;-) [13:09] dobey: let me find the bug [13:09] Barry [13:09] thisfred: the problem is that there is no simplejson for py3; and we will have to support it soon with u1db as well [13:09] https://bugs.launchpad.net/ubuntuone-control-panel/+bug/1029094 [13:09] Ubuntu bug 1029094 in Ubuntu One Control Panel "Replace simplejson with json" [Undecided,New] [13:10] ah [13:10] thisfred: so, if you want, import simplejson first, but don't require it [13:10] dobey, oh well, that settles it, I guess :) [13:10] thisfred: consider it py3k prep work [13:10] ralsina, I'd much rather just have one, and I'm sure it wasn't a difference that will kill us. [13:11] thisfred: unless the user has like 200000000 todo items... [13:11] I will discuss with the u1db peeps first though [13:11] thisfred: and a celeron [13:11] thisfred: and is impatient [13:11] ralsina, well, it's not just cosas [13:11] thisfred: and on fire [13:11] users on fire are probably a use case we don't have to support [13:12] they don't tend to file a lot of bugs [13:12] thisfred: todo: lay down , roll [13:13] GTDOF [13:14] GDRAOF [13:15] thisfred: those squares look not pretty [13:15] thisfred: I have an idea to make them pretty but it will suck for an example app, so +1ing [13:16] ralsina, I'm doing the simplest thing first, and aquarius will tell me what's not acceptable. He's going to use it as an example for dev week. [13:17] thisfred: ack [13:22] alecu: btw, could you review https://code.launchpad.net/~dobey/ubuntuone-dev-tools/runner-tests/+merge/119436 please? :) [13:22] dobey: sure [13:22] * mandel back [13:24] alecu, gatox or ralsina, may I have an osx review for https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/sbjson-link [13:26] mandel: queued but it's like 4th [13:26] ralsina, sure, not in a huge hurry [13:26] cool.... i thought i will need to reassamble my mac jeje [13:27] dobey: runner-tests looks good to me [13:27] thanks [13:51] gatox, what happened to your mac? [13:52] mandel, jeje nothing serious.... but i have mini mac in this room..... and the keyboards, mouse, monitor in other room.... because i was using that to play [13:52] ah, ok :) [13:56] i hope alecu isn't having more problems with his isp again today :-/ [13:56] * alecu knocks on wood [14:26] dobey, ralsina any problem if I add an MIT licensed lib in the daemon project? and it seems that the recomended way is via copying the file :( [14:26] mandel: link to the thing? [14:27] ralsina, I can link to it, yes [14:27] mandel: as in "give me a link" sorry ;-) [14:27] ralsina, oh, ok, let me get that for you, one sec [14:28] ralsina, http://www.dribin.org/dave/blog/archives/2008/04/29/ddcli/ [14:28] ralsina, and a link to the code: http://www.dribin.org/dave/software/#ddcli [14:29] mandel: let me go back to work n my ray to SMITE ALL THOSE WHO LINK TO THE MIT LICENSE WITHOUT PUTTING THEIR NAMES ON IT [14:30] darn [14:30] ralsina, that is why I asked.. [14:30] Copyright 2007 __MyCompanyName__ [14:30] dadadadadarn [14:32] mandel: please ask the guy to include the license, with his name, in the sources :-( [14:32] ralsina, sure [14:33] lol [14:33] i love how /doc/ in the tarball just contains more code [14:33] ralsina, since you asked: https://code.launchpad.net/~thisfred/u1db/deprecate-simplejson/+merge/119555 [14:33] thisfred: cool, thanks [14:34] ralsina, the lib does have his name in the headers [14:34] mandel: that license needs to be edited to have his name in it [14:35] mandel: it also says the copyright holder is "__MyCmpanyName__" [14:35] unless his name is "copyright holder" ;-) [14:35] ralsina, in which file did you see that? [14:35] mandel: there is no license file in the sources. He just has a link to the generic MIT license [14:36] ralsina, so not only he has to add the license per file but also add a a license file in the soruces? [14:36] mandel: yes [14:37] mandel: or at least the license itself [14:37] ralsina, ok, I hate license problems.. [14:37] mandel: wait [14:37] mandel: is this ddcli? [14:37] ralsina, yes [14:37] mandel: and what you need is the contents of lib/ ? [14:38] ralsina, yes [14:38] mandel: then go ahead :-) [14:38] ralsina, I can link to the lib, right? [14:38] mandel: just remove the rest of it, and keep only lib/ [14:38] ralsina, so, do I copy the files or do I link against the lib? I can do either [14:39] the source files really need proper license/copyright information in EACH source file, in the comment headers [14:39] mandel: I would copy the sources [14:39] dobey: they do, in the library [14:39] dobey: just the example code is bad there [14:40] thisfred: what's the meaning of "# noqa" ? [14:40] ralsina, it tells flake8 to shut its trap [14:40] thisfred: cool, thanks [14:41] thisfred: good to know ;-) [14:41] which can't accept multiple imports like this :) [14:41] ralsina, I wish they'd made it clearer, but there you go [14:43] elopio: ping [14:44] thisfred: +1 [14:44] ralsina, thx [14:44] elopio: in bug #1036686, how are you checking that syncdaemon is connecting directly to the internet? [14:44] Launchpad bug 1036686 in Ubuntu One Client "Syncdaemon ignores the proxy for files and folders actions" [Undecided,Confirmed] https://launchpad.net/bugs/1036686 [14:46] alecu: I monitor the squid access log. [14:48] elopio: well, we are using one open connection for all the syncdaemon traffic, so you will not see many connections when you add files. [14:48] elopio: I think you should block the ip route if you want to check that the proxy is used. [14:49] elopio: for instance, in my windows vm, when I'm checking the proxy support, I do something like this: [14:49] route delete 0.0.0.0 [14:49] route add 192.168.1.0 mask 255.255.255.0 10.0.2.2 [14:49] (from a cmd.exe with administrator permissions) [14:50] alecu: I should see at least one connection, right? [14:50] I'll give a try deleting the route [14:50] elopio: the only route that should be available is to the host machine that's running squid. [14:51] briancurtin, could you re-review this one? https://code.launchpad.net/~diegosarmentero/ubuntuone-control-panel/menu-desktop-services-actions/+merge/119370 i answered your need fixing [14:51] Hi folks! [14:51] mmcc, hi [14:51] elopio: right, and you will only see one open connection from SD thru the proxy. (well, more than one for some small bits that use a different channel, like: publishing a file, getting the list of public files, and sharing a folder) [14:52] elopio: those three operations are the only operations in SD that use a non-persistent connection. [14:52] ok, will confirm. [14:52] elopio: great, thanks. [14:52] thanks to you alecu. [14:53] elopio: please let me know if I can give any hints regarding how to set up the routes. [14:53] alecu: 10.0.2.2 is your proxy, right? [14:56] gatox: yep, looking now [14:56] briancurtin, thx [14:56] gatox: ohhh, approved [14:57] briancurtin, cool! thanks [14:58] mandel, gatox, briancurtin, mmcc, dobey, thisfred: standup in 2' do it without me, I have to be afk a little [14:58] ack [14:58] ok [14:59] drunk standup! [14:59] thisfred, can be easily be done [14:59] * mandel grabs the gin [15:00] thisfred, que tal va lo de aprender español? [15:00] elopio: 10.0.2.2 is the default gateway that the VirtualBox network has given to the windows vm. [15:00] me [15:00] me [15:00] me [15:00] elopio: 192.168.1.0 is my local network, where my squid is running. [15:01] me (writting notes) [15:01] me [15:01] mandel, muy bien :) [15:02] me [15:03] mandel: le está yendo más que bien. Si hasta le pone los acentos a las íes! [15:03] DONE: Python3 [15:03] TODO: finish bug #1036270 [15:03] NEXT: mandel [15:04] Launchpad bug 1036270 in Ubuntu Single Sign On Client "[nightlies] SSL errors crash ubuntu-sso-client and ubuntuone apps" [Undecided,Triaged] https://launchpad.net/bugs/1036270 [15:04] DONE: Got project o buil in lion correctly. Moved code to use ddcli to parse command line args specially to improve the test coverage. [15:04] TODO: More work on the main file of the daemon to add tests. Ask for reviews. [15:04] BLOCKED: no [15:04] thisfred, go! [15:04] DONE: cosas ui fixes, documentation1, simplejson dep removal TODO: documentation2...n for very large ns BLOCKED: no NEXT: alecu [15:04] alecu, thisfred, que es mucho mas de lo que hago yo, thisfred there is a missing one in mas :P [15:04] me [15:05] mandel, dónde? [15:05] DONE: some reviews, some proxy on windows debugging, finally got defeated by isp issues [15:05] TODO: mumble with briancurtin, lend him a hand or two [15:05] BLOCKED: stiff neck [15:05] NEXT: mmcc [15:05] thisfred, in 'mas', without it it means but, with it means more [15:06] ah [15:06] I thought pero was but [15:06] DONE: fixed buildout & env-mac, proposed storage-protocol branches [15:06] TODO: catch up on email, fix branches, more debug fsevents [15:06] BLCK: no [15:06] NEXT: gatox [15:06] DONE: [15:06] Land several control panel branches related to the sync menu feature, start working on the progress bar in the menu for current transfers (working on windows). [15:06] TODO: [15:06] Finish with this branch and propose. Look at the Shares menu feature. [15:06] BLOCKED: [15:06] No [15:06] dobey, go [15:06] DONE: u1db package pokery, start some reorg work for dropping installer, finish up bug #1032336 [15:06] Launchpad bug 1032336 in ubuntuone-dev-tools "New options parsing needs more test coverage" [Undecided,Fix committed] https://launchpad.net/bugs/1032336 [15:06] TODO: rush to get everything done for feature freeze, reorg to drop ubuntuone-installer, SRUs for lucid/natty/oneiric [15:06] BLCK: None. [15:06] cest fin [15:06] mandel, what project in lion? did you mean 10.6? [15:07] thisfred, also, mas is an old spanish thing you will probably not see it in any modern literature [15:07] * mmcc prefers numbers - large cats don't have a good ordering [15:07] mmcc, yes and no, I updated to xcode 4.4 and it changed all the targets to point to 10.8 so I had to fix that [15:07] mandel: ah. hooray xcode [15:07] mmcc, yes, exactly what I though [15:08] alecu: my gateway is 192.168.122.1. And squid is in 192.168.1.250. So, with this I can be sure that everything is going through the firewall? [15:08] route delete 0.0.0.0 [15:08] route add 192.168.1.0 mask 255.255.255.0 192.168.122.1 [15:08] oh also: today is my review day. send me reviews. [15:09] mandel, entendido [15:10] elopio: 192.168.122.1 is the gateway that "ipconfig" shows in windows, right? [15:10] alecu: yes. [15:10] mmcc, I have a few reviews for you :) [15:10] and someone pls review this buildout fix: https://code.launchpad.net/~mikemc/ubuntuone-windows-installer/improve-buildout/+merge/119195 -- maybe mandel, since you dislike the current buildout setup so much? [15:10] mmcc, sure! [15:11] elopio: then I think it should work. [15:11] mmcc, can you look at https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/sbjson-link (removes the lib to link it instead) [15:11] and https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/commserver-tests which adds extra tests [15:11] elopio: to make double sure, before and after running that, do a "ping 8.8.8.8" from the vm. [15:11] mandel, yeah, I see the list of reviews on +activereviews. looks like you have 7766 lines of reviews for me :\ [15:11] mandel: I'll get started :) [15:11] elopio: it should only work before that command. [15:12] mmcc, is mostly removing code :) [15:14] briancurtin: can we mumble for a bit? [15:14] alecu: sure, logging in [15:14] alecu: yes, that ping fails. [15:14] alecu: and files are still being uploaded. [15:14] So, actually tests passed. Nothing is left on the squid log. [15:18] elopio: so, it works? [15:18] yay! [15:18] ralsina: yes, it seems. [15:18] sorry about that. [15:18] elopio: np, it's tricky [15:18] oh, forgot something in standup - ralsina, the OSX Lion VM that IS set up works - I tested it out on friday. [15:20] mmcc: how was it? [15:21] ralsina: it was real, and it was spectacular [15:21] seriously though, it worked fine, and seemed pretty fast to me [15:21] the RT ticket has info on how to use it. I think they still have to set up all the accounts for it though [15:22] mmcc: I have one, mandel, did you get account data? [15:22] ralsina, account data? [15:22] ralsina, I need a little more context :) [15:22] briancurtin: apt-get build-dep ubuntu-sso-client [15:23] mandel for the lionvm that IS set up on that xserve. [15:23] ralsina, mmcc, oh, yes yes, for the vms, I have them [15:23] mandel, mmcc: awesome [15:24] ralsina, I have been looking at how to correctly set jenkins on that [15:25] alecu, briancurtin, joshuahoover: let's revive the RT for 3.0.2b :-) [15:25] elopio: awesome news! [15:26] ralsina, alecu, joshuahoover, briancurtin: I retried all the tests I marked as failed, and they passed. Changing the bug to invalid. [15:26] ralsina: we still have the issue where sd goes into a loop when the user enters his proxy credentials wrong [15:26] elopio: \o/ [15:26] alecu: yes [15:26] elopio, ralsina: that is good news :) [15:26] alecu: but that's not a regression, so let's mark it as high and have as 1st bugix after FF when brian is done with SSO-py3 [15:27] ralsina: should i reuse the same RT or create a new one? [15:27] briancurtin: i think you can use the same one since nothing has changed [15:27] yeah [15:27] alright, need to get lunch. bbiab [15:27] briancurtin: IS will simply think we're even more crazy than they already thought we were [15:27] nice [15:28] so, the only bug I got while running the smoke is bug #1000970 [15:28] Launchpad bug 1000970 in Ubuntu One Control Panel "AttributeError: "'NoneType' object has no attribute 'stop'" when registering an existing account" [Medium,Confirmed] https://launchpad.net/bugs/1000970 [15:28] joshuahoover: we'll never bow to public opinion [15:28] but that's not new. [15:28] elopio: I kinda have a fix for it in mind [15:28] elopio: it's not difficult, but not doing it until friday, probably [15:29] friday sounds good. [15:30] alecu: I'll add a test to the smoke to enter the wrong password. I didn't know about that. [15:36] * gatox lunch === gatox is now known as gatox_lunch === salgado is now known as salgado-lunch [16:07] alecu: thanks for that review [16:11] dobey: is there a file that has the version number of the u1 client in ubuntu? (strange request, i know) [16:14] joshuahoover: the package itself [16:14] dobey: ah, right...and since i pay no attention to such details, where might one find that :) [16:15] joshuahoover: dpkg -l $package [16:15] ubuntuone-client ubuntuone-control-panel ubuntu-sso-client python-ubuntuone-client python-ubuntuone-storageprotocol and on and on [16:16] dobey: right, but i have a guy who can't run any commands, just has access to the filesystem and needs to find the version info [16:16] dobey: (don't ask) [16:17] how is he going to do that if he can't run any commands? [16:17] 'vi' 'less' 'cat' are commands :) [16:18] i guess he's in windows? [16:18] and mounting the fs or something === cheese is now known as Guest65807 === Guest65807 is now known as DanimalCracker [16:19] dobey: yeah, that's my understanding [16:21] joshuahoover: he can look for the package versions in /var/lib/dpkg/status i guess. as long as he doesn't modify that file at all [16:21] he'll have to search through it though [16:21] dobey: k, thanks! he'll enjoy that === gatox_lunch is now known as gatox [16:25] EOD for me, catch you all tom! [16:26] bye mandel === salgado-lunch is now known as salgado [16:38] mmcc: if you have time for a review, https://code.launchpad.net/~brian.curtin/ubuntu-sso-client/remove-SyncTimestampChecker/+merge/118826 should be up to date. it's almost all -'es [16:39] briancurtin: ok, I'll put that next in line. [17:17] briancurtin: are you using pylint or pyflakes on that branch? [17:18] I tested it with my new buildout, which uses pyflakes, and it had a few complaints, a couple of which might be relevant to your changes [17:18] mmcc: pylint but for whatever reason it doesnt pick up nearly half of what everyone else sees. i always fix the very few things that pop up, but others always see more warnings. if you have something that i havent fixed can you pastebin it? [17:18] briancurtin here you go http://paste.ubuntu.com/1147216/ [17:19] that's pyflakes. [17:19] i still need to stop and get this whole ordreal straightened out...its getting to the point where every branch i produce ends up being crap [17:19] mmcc: ubuntu-sso-client uses pylint [17:19] mmcc: afaik there's no reason it should be using pyflakes on windows or osx instead of pylint [17:20] unless briancurtin has USE_PYFLAKES set? [17:20] i dont even think i have pyflakes installed. my pylint just sucks [17:20] or windows sucks [17:20] dobey: hmmm [17:20] or something [17:20] pylint on windows does suck a lot [17:20] briancurtin: the fixed buildout will install pyflakes. you do have the egg, it just didn't ever get the script from it [17:21] I never got a decent run out of it (but instead I got bazillion fake warnings) [17:21] USE_PYFLAKES is set in run-mac-tests for sso client [17:21] oh crap, that's only in my working tree :) [17:22] ralsina: s/on windows// [17:22] dobey: it sucks differently ;-) [17:23] I clearly don't understand what colo does, I thought creating a new colo branch would leave uncommitted changes in the old branch [17:23] so dobey, we should use pylint instead of pyflakes for sso, even on mac? [17:24] pyflakes is so much faster for me [17:24] mmcc: well, we should use the same thing on every platform, for each project. i'd like it to be pyflakes for everything, but there are still some issues with that [17:26] so what was the reason for using pyflakes on the other project (iirc, storage-protocol was what started my buildout fixing episode) [17:27] btw, the difference on my system is 30 sec for pylint and .9 sec for pyflakes, on sso [17:27] mmcc: we started with pylint, so there's a lot of inertia [17:27] mmcc: specially since pyflakes detects less and/or more depending on the wind [17:29] * gatox was happy with the progress bar in the menu... but it seems that u1-client is broken and need to go back to fix that :S.... crap! [17:30] ralsina: ok, well since I have a couple of branches to propose that clean up the mac test scripts, it'd be nice to know where we need to use pyflakes and where we should use pylint [17:30] for my part, I'm always using pyflakes since it's fast enough to use with flymake in emacs [17:31] mmcc: you'd have to check the run-tests scripts [17:31] * ralsina doesn't have that matrix in memory [17:31] mmcc: we have to use pyflakes for protocol/client because pylint does not deal with some crazy stuff twisted/protobuf does [17:31] but really, pylint needs to diaf [17:32] dobey: ack. so flakes for those two and lint for the others. gotcha [17:33] well, dirspec is using pyflakes too [17:33] ok, too hungry to think [17:33] lunch! [17:33] i should probably make dev-tools use both [17:37] anyway, briancurtin if you want to experiment with pyflakes, take a look at https://code.launchpad.net/~mikemc/ubuntuone-windows-installer/improve-buildout/+merge/119195 -- it includes a change to env.bat, so someone who knows bat files would be welcome [17:37] and it installs pyflakes in the buildout/bin [17:38] the only people who know bat files have really, really long beards [17:38] but i can take a look instead [17:45] * mmcc had a suitable beard once: http://ubuntuone.com/7carFI0DvBy3GwdYCAKO3K [17:46] that's not long enough to have a section of your geocities page dedicated to windows batch scripting [17:46] ha, nope. [17:47] briancurtin, ohai: the pycon blog post you did says pycon 2012 instead of 2013 ;) [17:47] crap [17:48] especially confusing since there is also still an advertisement for pycon 2012 up on the site with different dates in March [17:49] yeah we need to get someone to update that image [17:53] briancurtin, here is the pylint output from darwin for your branch: http://paste.ubuntu.com/1147273/ [17:54] not sure why it's complaining about dirspec and devtools though, those are explicitly put on sys.path in the u1lint wrapper script :( [17:54] mmcc: that's because you have an unpatched logilab-common i think [17:54] is there a way to make buildout apply a patch when installing something? [17:55] dobey, hrm. probably, but it might be easier to just put a patched version somewhere for buildout to grab - there are a couple of things that are doing the tnow [17:55] doing that now [17:55] eg, pykeyring, pycrypto, etc [17:56] do you have a patched egg laying around somewhere we can just throw into the find-links section? [17:56] not really; just have the patched ubuntu packages [17:56] no idea how we've dealt with this on win in the past [17:57] someone made a patched egg and ralsina put it on u1.to/ralsina/blah.egg [17:58] for the pylint stuff? [17:59] for keyring, crypto, com types, even windows twisted apparently [17:59] right [17:59] but doesn't explain why nobody has complained about this same issue before, with pylint on win for example :) [18:04] mmcc: http://pastebin.ubuntu.com/1147294/ is the patch against logilab-common that we need; i guess you can apply it against the installed version on your buildout to test if you wish [18:04] ok, thanks dobey [18:31] there is a buildout patch applying recipe, works cleanly [18:31] * briancurtin lunch [18:32] however, after applying that patch I get a lot more spew from pylint: http://paste.ubuntu.com/1147338/ [18:32] 2200 lines more [18:33] yeah, that looks like it just broke lots more imports, since it still isn't finding devtools or dirspec [18:33] hrm [18:40] mmcc: wow [18:41] that doesn't seem right [18:41] nope… [18:41] it's failing to import pretty much everything there [18:42] I wonder if it's having some interaction with the wrapping script that sets path [18:44] maybe [18:47] dobey: I am about to make you a very happy man [18:47] dobey: kill pylint [18:47] rly? [18:47] dobey: rly [18:48] YAY! [18:48] dobey: thought about it in my lunch, and decided we were just being too conservative [18:48] dobey: so instead of wasting time fixing it, let's kill it [18:48] will do [18:48] mmcc: is PATHSEP on mac ":" ? [18:49] mmcc: because on windows pylint doesn't parse PYTHONPATH correctly [18:49] ralsina: the wrapper script sets sys.path directly [18:49] mmcc: ok [18:50] but just for fun, os.path.sep == '/', and os.pathsep == ':' [18:50] dobey: let's do something like the USE_PYFLAKES flag so that pyflakes is used by default and we can try pylint on demand === lifeless_ is now known as lifeless [18:52] ralsina: yeah. plan is to switch everything to run u1lint with USE_PYFLAKES (or to pyflakes directly where appropriate), get all the issues fixed that pop up, then switch dev-tools to use it by default; maybe changing the env var to be something like U1_PYLINT_CMD=[pyflakes|pylint] and have it do appropriate things based on that [18:52]