[04:49] hello, just joined to ask for some help. [04:50] It seems I may have lost an entire dir tree due to a failed sync with U1. [04:50] The folder structure is there, but no files. [04:51] u1sdtool status reports it is idle. [04:54] How it happened: I dragged a folder that was already a UDF and being synced with U1 into the 'Ubuntu One' folder in Nautilus. [04:55] Minutes later I reversed this, and copied the directory back to its original location. [04:55] End result: no files, just the directory tree. [04:57] I have checked online at my account on one.ubuntu.com, and same result. === zyga-afk is now known as zyga === teknico_ is now known as teknico [08:49] Good morning all! Happy Friday the 13th! >:-) [08:59] :( my quantal machine has been "uploading" some files for about a month now [09:00] so i removed all the files from the folder in nautilus just now.. [09:00] syncdaemon still seems to be chatting away about the files in its log [09:00] and sync is "in progress" [09:01] I stop and start and it still things it's uploading files which I moved out of that synced folder [10:42] popey, our staff is very very broken in Q atm [10:43] popey, sorry :( [10:43] popey, with the move to python3, the mac port etc.. we are overwhelmed atm but afaik ralsina is going to look into the problems [11:02] mandel, ok :) [11:09] good morning! [11:09] mandel, hi [11:09] gatox, morning! feeling any better? [11:10] mandel, MUCH, i slept all the day yesterday........ did you do anything with my branches yesterday? [11:11] gatox, I approved darwin4 I'm going to propose one with the daemon, I have move notify processor to its own package which removes circular importans and I have added a darwin only command line arg to choose if the deamon is used :) [11:12] gatox, my I have reviews: https://code.launchpad.net/~mandel/ubuntuone-client/fsevents-daemon/+merge/114836 :) [11:12] yes [11:13] mandel, 2163 lines?? and you complain about me...... shame on you! :P [11:14] gatox, wait, missing dep! [11:14] give me a sec [11:15] gatox, oh no, looks like a move is not considered a move.. [11:15] bad bzr.. [11:16] mandel, did you do bzr mv? [11:16] gatox, afaik yes, but in one of the pump when wrong and I have to resolve a conflict.. maybe that screwed up the diff [11:16] gatox, you can see there is a large delete and then the same code added later.. [11:17] ok [11:17] gatox, sorry for the PITA [11:18] :P i'm kidding [11:19] gatox, regarding the NotifyProcessor, in a later branch I have moved it to filesystem_notifications/notify_processor/ that was we can remove a circular import issue we had and make things cleaner [11:20] mandel, where are we having a circular import? [11:20] gatox, in you look in notify_processor you will see there is a XXX: circular import comment, I have fixed that [11:21] gatox, mainly because later I was having a circular import when trying to reuse the notify processor you added, the fsevents and the fsevents_daemon can share it [11:23] ahhhhhhh yes, the one in the __init__ i remember [11:24] gatox, exactly, so that is out, which I think is a good move :) [11:24] gatox, I'm starting to wonder (although won't change it) what is the need of having platform infront of everything, but that is a minor detail [11:24] gatox, the code is looking a lot better know, we might be able to port it to the sp3 ;) [11:25] sp3? ps3? [11:26] gatox, sorry ps3 [11:26] gatox, one hand is faster than the other.. :P [11:26] or was I thinking about xp sp3 [11:26] .. [11:26] jejej [11:26] you are such a windows lover..... maybe was that [11:27] maybe.. [11:40] gatox, early lunch for me, ok? [11:41] * mandel lunch [11:44] mandel, ack [11:56] * gatox realize that mandel's branch is going to cause conflicts in the refactor one.... fun not [11:58] mandel, mmmmm something don't look right..... shouldn't this be depending on darwin4?? i see some changes that are already in trunk [12:04] mmm i think there is some kind of problem merging darwin4.... first time i see this on launchpad: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin4-fsevents/+merge/114405 [12:19] good morning! [12:19] how are you today gatox? [12:19] ralsina, fine..... today was 100% rest :P [12:20] ralsina, do you know what it means "This proposal supersedes" here: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin4-fsevents/+merge/114405 [12:20] gatox: looking... [12:20] i'm trying to look for some doc or something, but i can't find anything clear enough [12:20] or how to fix it [12:20] gatox: that this branch was proposed twice and this is the newer one [12:20] that's not bad, AFAIK [12:20] ralsina, aja..... but why is not being merged? [12:21] ahhhhh [12:21] commit message [12:22] it seems i forget to include that when i repropose === zyga is now known as zyga-food === zyga-food is now known as zyga [12:47] hello! [12:51] gatox: I see that darwin4 has landed. Awesome! [12:51] gatox: I've created a bug for the deferred stop and the thread join fun. I'll probably be working on that. [12:53] alecu, yes, i saw it [12:53] alecu, ahhhh.... i thought it was for me :P [13:04] hrmm [13:05] no reviews on https://code.launchpad.net/~dobey/dirspec/python3/+merge/114713 :-/ === yofel_ is now known as yofel [13:19] alecu, ping [13:23] alecu, when you are back...... is this the bug you were talking about? https://bugs.launchpad.net/ubuntuone-client/+bug/1024102 i see that is assigned to me, and i was going to add it to the tests-refactoring branch [13:23] Ubuntu bug 1024102 in Ubuntu One Client "Deferred should be fired when a watch is stopped" [Undecided,New] [13:24] mandel, ping when you are back.... i've some questions about your branch [13:27] gatox: maybe you should review https://code.launchpad.net/~mikemc/ubuntuone-client/use-dirspec-get-cmdline/+merge/114720 [13:27] dobey, ack [13:35] mandel: 2163 lines?! [13:37] briancurtin: any more work on the py3-unicode branch? it's still needs fixing from alecu [13:38] * alecu is back [13:39] alecu, when you are back...... is this the bug you were talking about? https://bugs.launchpad.net/ubuntuone-client/+bug/1024102 i see that is assigned to me, and i was going to add it to the tests-refactoring [13:39] Ubuntu bug 1024102 in Ubuntu One Client "Deferred should be fired when a watch is stopped" [Undecided,New] [13:39] gatox: that's the bug, yes. I assigned it to you, but ralsina told us yesterday that you may start working on a different project [13:39] alecu, ahhhhh ok [13:39] gatox: that's why I said I might be working on fixing that. [13:40] ohhh i miss the meeting..... [13:40] dobey: i haven't gotten to that one. i'll take a look today (but i'll be gone for a while in the morning through late afternoon) [13:40] gatox: yes, I am planning on switching you to the menu/indicator thing starting next week [13:40] anyone know what i'm going to be doing? :P [13:40] s/late/mid [13:40] gatox: you are being moved to work on the u1 port to VMS. [13:40] ralsina, ohhhhh great!! \o/ [13:40] alecu, haah [13:40] gatox: or was it OS/2? [13:40] ralsina, cool! [13:40] gatox: see ⟰ [13:40] python 4 [13:41] dobey: writing it! [13:41] * ralsina liked VMS. Sniff. [13:41] it's the version of python where guido hits senility, and it's more like brainfuck than python [13:42] alecu, ok..... so..... if you are going to do that..... i think this is ready for review: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin-tests-refactoring/+merge/113289 (about the comments: we already talk about the assert_logs removal, and i removed the unnecesary setUp for TestNotifyProcessor) [13:43] gatox: ack [13:49] dobey: I'm running the tests for ~dobey/dirspec/python3, and I'm getting an error when running the python2 tests on Precise. [13:49] dobey: should I be running them on quantal? [13:50] alecu: python2 or python3? [13:52] alecu: the tests are passing for me on precise === teknico_ is now known as teknico [14:02] dobey: I'm getting these errors when running on precise: http://pastebin.ubuntu.com/1089886/ [14:02] dobey: I'm updating to see if I had something really old. [14:06] alecu: there is a testools in the nightlies PPA, that is needed [14:08] alecu: and you'll need python3-setuptools and python3-testtools [14:11] updated the description about that [14:14] gatox, shoot [14:15] dobey, yes, there was some code moves and turns out to be bigger than expected [14:15] dobey, is mainly a copy paste to a diff module to be honest [14:15] mandel, i'm seeing some lines that shouldn't be there because are changes from trunk..... or am i missing something? [14:15] bzr mv didn't work? :) [14:16] mandel, for example: class TestWatchManager(BaseTwistedTestCase): [14:16] 77 + """Test the watch manager.""" [14:16] and the tests in there [14:16] dobey, no because it was not the entiry module :( [14:16] gatox, which line of the diff? [14:16] dobey: how do you feel about uploading dirspec to pypi.python.org eventually [14:17] ralsina: "eventually" sounds fine. perhaps with the release on monday [14:17] mandel, 76.... also..... test_stop_multiple has no longer that implementation, we remove the time.sleep [14:17] gatox, hm.. there was probably a miss merge somewhere.. [14:18] * mandel hates having so many branches.. [14:18] dobey: yes, after release is ok [14:18] gatox, let me remerge and see what can be fixed [14:18] dobey: I've updated everything, installed python3-testtools, but I still get that error in python 2, and now the same error with python 3 [14:18] mandel, ack [14:19] alecu: weird [14:24] alecu: what does "ls -lhd ~/.config/x" show? [14:24] going to do my apartment thing, back hopefully soon [14:25] alecu: seems like you have a file or directory at ~/.config/x [14:26] alecu: indeed, if i touch ~/.config/x, the test then fails [14:26] wonder how you got that file :) [14:26] anyway, need to run an errand. brb [14:41] gatox, branch updated [14:41] mandel, ack [14:44] gatox, he, the diff lost 500 lines :) [14:44] :P [14:45] gatox, nad if the bzr mv had worked the diff would be smaller.. [14:57] mmcc, ping [14:57] gatox pong [14:58] mmcc, i put a need fixing here: https://code.launchpad.net/~mikemc/ubuntuone-client/use-dirspec-get-cmdline/+merge/114720 because that lint error is causing problems to run the tests on linux [14:59] gatox, ok thanks. wonder why that didn't show up when I ran them :\ [15:00] me [15:00] me [15:00] me [15:00] meh [15:01] me [15:03] alecu, thisfred: standup [15:03] ne [15:03] brian is away for a bit [15:03] me [15:03] me [15:03] so alecu is last, go mandel [15:03] DONE: Proposed fsevents daemon code for u1-client. Added a new branch that removes cicle imports from the NotifyProcessor code. Added a new command line arg just for darwin. [15:03] TODO: Split file system monitor for better import. Propose the branches. Be done with fsevents, although if gatox is moved there are a number of bugs that should be assigned to me next week regarding fsevents :( [15:03] BLOCKED: no [15:03] gatox, please [15:04] DONE: [15:04] Take the day off yesterday to reload the health battery. Reviews, small fix in tests-refactoring (NEED REVIEWS FOR THIS). Sent email to ralsina, alecu and lisette about the file publisher in the indicator menu, start reading the syncmenu wiki and X platform desktop indicator menus document. [15:04] TODO: [15:04] Finish with those documents. Keep reviewing branches and fixing fsevents related bugs. [15:04] BLOCKED: [15:04] No [15:04] mmcc, go [15:04] DONE: finished path finding, hacked on setup-mac [15:04] TODO: cleanup u1client branch, more setup-mac, reduce package size [15:04] BLCK: [15:04] NEXT: dobeh [15:04] DONE: team call, bug #1024003, bug #1016224, backported quantal software-center to nightlies to unbreak packages from sso-gtk removal [15:04] Launchpad bug 1024003 in dirspec (Ubuntu Quantal) "Unnecessarily requires twisted trial for running tests" [Undecided,Triaged] https://launchpad.net/bugs/1024003 [15:04] TODO: reviews, investigate twisted-less dev-tools, work on some reorg to drop ubuntuone-installer for good, SRU verification poking [15:04] Launchpad bug 1016224 in dirspec trunk "Does not work with Python 3.x" [Undecided,In progress] https://launchpad.net/bugs/1016224 [15:04] BLCK: None. [15:04] ralsina [15:04] DONE: team call, reviews, VM haggling, talked a lot with lisette, contractor wrangling, some reviews, little coding TODO: more of the same, try to debug lists on quantal BLOCKED: coughing a whole lot. NOTE: mandel, we can delay gatox's switch a couple of days more if needed NEXT thisfred [15:04] DONE: worry about txids in u1db TODO: write better integration tests that do the worrying for me BLOCKED: no, but I could use a few clones or a time machine NEXT: alecu [15:06] mandel, i've right now 2 issues related to non-root fsevents, do you have any other in mind? [15:06] * alecu is writting notes [15:08] DONE: many reviews, 1-1 and team meetings, rebuilt my broken thunderbird setup [15:08] TODO: catch up with briancurtin and py3k [15:08] BLOCKED: no [15:10] EOM? [15:10] NEXT: briancurtin (when he returns) [15:11] alecu: did you see my replies about my branch and the failing test? [15:11] dobey: yes, I've just ran the tests after deleting .config/x, and all tests passed. [15:11] ah ok [15:13] dobey: are we going to use "from __future__ import unicode_literals, print_function" in a given module even though we don't use print there? [15:13] dobey: I think it's a good convention, so +1 to i. [15:13] it [15:14] alecu: i think we should, in case anyone adds a print it will complain even on python2 [15:14] it's good to keep the code from regressing in that syntax [15:15] yup [15:15] gatox, mandel - darwin4-fsevents doesn't fix the tests, right? just checking if I should be expecting passes [15:16] mmcc: i guess according to jenkins, it doesn't [15:16] mmcc, all the tests inside filesystem_notifications [15:17] dobey, that is windows work that was broken by someone else. I have it in my radar to fix it asap [15:17] dobey, or pass the ball ;) [15:17] ok, well fix it :) [15:18] alright. [15:18] dobey, yes yes I will but I need longer days hehe [15:18] heh [15:24] dobey, regarding your comment about adding ubuntuone/utils.py - I'm glad you noticed. I was going to ask for specific feedback on that in the MP but forgot. I really didn't know where to put it, but I guess syncdaemon/utils.py would work, as long as there aren't any circular imports created... [15:25] well, the circular imports would exist regardless of where it lives, if there are any. that's a separate problem to where the module itself lives [15:26] dobey: in assert_utf8_bytes the value.decode(...) was changed to value.encode(...). I can't understand why that change makes sense. [15:26] alecu: python3 doesn't have decode on the data type that value is; only encode [15:27] dobey, I meant because utils is used in platform/tools and platform/ipc, which are imported by syncdaemon, so if I move utils into syncdaemon, then platform <---imports--> syncdaemon, but i think it'll work out anyway [15:27] dobey: in assert_utf8_bytes, value should be a sequence of bytes, so it only has "decode", not encode. [15:28] mmcc: ah, but syncdaemon.utils itself doesn't import those, so it's not circular. also i think we should move all the other code to be under syncdaemon at some point as well [15:29] dobey: also unicode_path() has the same issue. [15:29] alecu: i think that was broken before then [15:30] dobey: righto - it's ok as long as syncdaemon.__init__ doesn't import them. which it doesn't. and I agree, someone ought to do that. maybe when manuel gets his longer days or borrows eric's time machine [15:30] MINE [15:32] this one I have to share because is just funny, the spanish government just defined funerals as 'entertainment' to increase the VAT on them.. hehehe [15:33] alecu: happy to fix it to be the other way around, but not sure exactly how to do that at the moment. [15:33] and i need to get lunch [15:34] mandel: that's a killer tax! [15:34] alecu, lol [15:36] dobey, which all other code? [15:37] killer party man! [15:37] mandel: everything under ubuntuone/ in ubuntuone-client, which is not already under ubuntuone/syncdaemon/ [15:38] mandel: don't worry about it though [15:38] dobey, you mean getting out of platform and moving it to syncdaemon? [15:38] dobey, it does make sense because it is a lot cleaner know, is a matter of getting time like mmcc said :) [15:38] dobey, I'm had that in my mind for some time [15:38] mandel: i mean, i really haven't thought about how to do it exactly yet, but it seems like everything should be under there, that is in ubuntuone-client [15:39] anywya [15:39] need to get lunch [15:39] bbiab :) [15:52] lunch for me! === gatox is now known as gatox_lunch [16:10] gatox_lunch, ralsina ping! [16:10] mandel: pong [16:11] ralsina, can you run the windows tests from trunk and let me know if all but one pass? [16:11] mandel: not right now [16:11] mandel: that was the case last week [16:11] ralsina, i have some timeout failures but it might well be that I have a slow machine.. [16:11] mandel: jenkins seems to say that's still true [16:12] ok [16:13] mandel, which windows tests? u1-client? [16:14] mmcc, yes [16:14] mandel, I am about to run those, I'll let you know. it was the case yesterday - only one failure, having to do with unicode paths [16:15] mmcc, awesome, please let me know [16:17] really would it have killed them to make 'cd E:\' work like it does everywhere else? [16:18] mmcc, probably the did it so support some old dos program.. [16:18] well, I get the *reason* - it's that E: is a 'drive' not a 'directory', and they're really different things, but seriously. [16:19] and the fact that the right way is just to type the drive name with no command? who came up with that [16:19] [16:25] and for some reason python.exe wasn't on my path anymore? winDOOOWSS! [16:25] * mmcc shakes fist [16:27] ugh, 2 hours and only saw one apartment. at least it was a decent place === gatox_lunch is now known as gatox [16:31] mandel, pong [16:31] gatox, I was going to ask you to run the windows tests but mmcc is on that :) [16:31] ok! [16:32] gatox, I have an other branch for review, is refactoring and some tiny changes but since there is no bzr mv --line 22-50 the diff looks bigger than the number of changes [16:32] gatox, is https://code.launchpad.net/~mandel/ubuntuone-client/unify-processors/+merge/114906 [16:33] mandel, ok..... i'll finish with the reviews no [16:33] now [16:35] gatox, haha I though you said no :P [16:35] gatox, I imagined you as joda hahaha [16:36] jaajajaj [16:36] LOL [16:39] mandel, also.... i need your review here: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin-tests-refactoring/+merge/113289 [16:40] gatox, we are the masters of the large reviews lol [16:40] hmm [16:41] mandel, jejeje it's a valuable skill jeje [16:41] gatox, so that branch, just changes the darwin tests, is that right? [16:41] mandel, no.... windows too...... move all the common code to common.py [16:42] doesn't add or remove anything.... just move things [16:42] gatox, dobey reviews requested for this guy: https://code.launchpad.net/~mikemc/ubuntuone-windows-installer/setup-mac-add-sd/+merge/114749 [16:42] well...... it removes repeated code, but you know what i mean [16:42] mmcc, ack [16:42] mmcc, let me know when you fix the other one [16:43] gatox, ok, a couple minutes. made larger changes due to dobey's comments, testing on win/linux now [16:43] mmcc, ack [16:43] gatox, ack [16:44] gatox, it will have conflicts with mine.. but I'll fix those later :) [16:44] does everyone just always set TRIAL_TEMP_DIR to something short on windows, or am I the only unlucky one with long paths to branches? should we just set it in env.bat? [16:44] mandel, yes, your turn to fix now! [16:45] mmcc, we all set trial_temp_dir, that was the reason we added it :) [16:45] mandel, right but do you set it manually or what [16:46] mmcc, you can set it as env var for windows [16:46] mmcc, look for environment variable in the control center on windows 7 and will show you where to add it [16:47] mandel, ok that's what I was looking for, thanks. [16:47] np [16:52] gatox, needs fixing, lint issue! [16:52] bloody hell [16:53] gatox, lol [16:55] heh [17:01] fu**, my income taxes just went up 140%.. hijos de la gran pu.. [17:02] * mandel looks for a shot-gun to go to the demonstration [17:03] mandel, ouch [17:03] mandel, before you go to the revolution, I'm sure you'll want to know that u1client trunk still has just that one windows failing test. [17:03] mmcc, yes, in a single day.. also VAT went up 160% [17:03] mandel, also http://paste.ubuntu.com/1090200/ [17:03] mmcc, awesome, so is my slow vm :) [17:04] ugh [17:04] mandel, how slow? mine's awfully slow too - note the 617 seconds those tests take [17:04] mmcc, slower.. I need to see what is wrong with the setup [17:05] mandel, press the turbo button on the front of your minitower [17:05] python standard lib is so… conflicted [17:07] hahs and the social security tax went up a 250%!!! [17:07] mmcc, lol I loved those turbo buttons :) [17:11] * ralsina suspects mandel needs a refresher in percentages [17:11] as in: going from 10 to 14 is not a 140% raise [17:15] ralsina, some things have gone in the VAT from 8% to 21% which is a 160% and all the other changes are the same (smaller diff) [17:16] mandel: he, here's 21% for everything. Ecept phones, there's 37.5% [17:16] ralsina, then I'm move from a 18% to a 21% plus an extra 21% on top of that one + social secury has gone from around 100 eur to 350 [17:17] ralsina, but did they increase it in a single day? oh, while they have 30.000 million eurs to the banks... [17:17] s/have/gave [17:18] see? plenty of money there! ;-) [17:18] welcome to america [17:19] only 350eur for social security? Man, our republican conservatives would love to live in Spain I bet! [17:20] dobey, well, most of madrid is on strikes round now blocking streets etc.. [17:20] dobey, I bet they wouldn't: http://en.wikipedia.org/wiki/Spanish_National_Health_System [17:21] mandel: good thing i didn't take a holiday to madrid then [17:22] mmcc, yes, we are waaaay to the left [17:22] dobey, never know, might be fun, did you see the video of the miners? [17:23] dobey: we need to package pyasn1 for python3 for Q sometime in the next few weeks, to upload to main [17:23] ok folks, lunchtime for me, u1 client path branch tests are running now, I'll check in and ping for a re review in a few mintues [17:23] or minweds, minthurs , you know, whenever [17:23] ralsina: huh? [17:24] dobey: missing dependency for twisted pythn3 [17:24] oh, twisted.conch uses it apparently [17:25] dobey: yep [17:25] dobey: forget about it, barry will do it [17:25] ok [17:27] uh, linux u1-client tests failed with this error I don't recognize, in filesystem_notifications: http://paste.ubuntu.com/1090236 anyone recognize? [17:27] * mmcc lunchtime for real === salgado is now known as salgado-lunch [17:31] os.path methods do not return bytes :( [17:31] dobey: they should if they get bytes [17:31] dobey, python3? [17:40] gatox: both 2 and 3 [17:40] dobey, for python3 i don't know..... but for 2 it should be as ralsina says [17:41] for python3 it's the same thing [17:41] if you give it bytes you get bytes [17:41] dobey: what specific function? [17:51] ok, eod for me [17:51] see you all on monday! [17:54] mandel, bye [17:54] mandel, also.... lint issue fixed here: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin-tests-refactoring/+merge/113289 [17:54] and +1 to your branch... the one with the 1500 lines diff [17:57] ralsina: some methods only accept strings (like os.path.join) [17:57] dobey: nope [17:58] dobey, [17:58] >>> os.path.join(u'asd', 'foo') [17:58] u'asd/foo' [17:58] dobey: https://pastebin.canonical.com/70055/ [17:58] ralsina, we are really creative with the examples :P [17:59] os.path.join(b"C:\\lol", "rofl".encode("utf8")) [17:59] dobey, gatox: however, it's a bad idea to pass one unicode and one bytes because there will be an implicit str.decode(ascii) there somewhere [17:59] briancurtin: "rofl".encode("utf8") is just evil :-) [17:59] unless it's python 3 of course [17:59] but you can do it! [18:00] just to add to the examples [18:00] you can also put ketchup on fries. Doesn't mean it's right ;-) [18:00] sriracha goes on fries [18:00] you have my attention [18:00] +1 on sriracha on fries -1 on encoding bytes ;-) [18:00] what sriracha is? [18:00] gatox: hot sauce [18:00] ahhh [18:01] it's like a hot ketchup. more of a paste-like thing than liquidy hot sauce [18:01] also...... cheddar (melt) with fries..... it's awesome! [18:01] 'r'.encode().decode().encode().encode().decode() [18:01] cheddar cheese [18:01] python2 is fun! [18:01] i guess dirspec should enforce bytes everywhere though [18:01] being a file path thing [18:02] alecu, ralsina, briancurtin: does that sound right? === salgado-lunch is now known as salgado [18:08] dobey: "should enforce bytes" <- I agree for python 2. I'm still not sure it's the proper api for python 3 libraries, but since we'll be using it mostly from u1, bytes on python 3 sounds right. [18:09] dobey, alecu: whatever, as long as it's consistent [18:11] alecu: well, filesystem paths are bytes. and it's basically impossible to enforce bytes on one version and strings on another. either we enforce unicode strings everywhere, or bytes everywhere [18:11] on py3 where there's no automatic conversion, ad thus no risk of mixing them up, it doesn't matter as much [18:11] previously, the code was doing a mix, which is why this diff is a bit large [18:11] dobey: "filesystem paths are bytes" only on linux. [18:12] but yes, and bytes everywhere sounds good for our usecase. [18:12] alecu: they're bytes everywhere. on windows they're just a much smaller subset of possible bytes, and it's easier to get a reliable encoding for the bytes they're in :) [18:13] ok [18:14] i don't care either way, but we have to enforce either bytes or unicode strings, and not dilly dally somewhere in the middle :) [18:18] alecu: https://code.launchpad.net/~dobey/dirspec/python3/+merge/114713 is updated and changes stuff to be bytes everywhere [18:20] gatox: ^^ if you could please review that too [18:20] dobey, ack [18:31] * briancurtin lunch + doctor run [18:36] oh, lunch sounds like a cool concept [18:36] * ralsina investigates [18:37] ralsina: given the temperature there, maybe you should eat a hot lunch [18:40] dobey, i'm getting this running the tests for dirspec-python3 http://paste.ubuntu.com/1090342/ [18:41] do you know what i'm missing? [18:41] i have python3-(testtools|setuptools) [18:43] so did anyone recognize this: http://paste.ubuntu.com/1090236 test_zip_release_lock_compression_error failing on linux? [18:46] gatox: do you have python-testtools installed? that's python 2.7 that's failing there [18:47] mmcc: i've seen it a couple times now, but not sure why it happens [18:47] mmcc: i'm tempted to add a skip for that test and file a bug if it keeps happening (and a bug isn't already filed) [18:47] oh, it's intermittent? I"ll rerun and see if I see it again... [18:48] dobey, i have it for python3 but not for 2 :P [18:49] mmcc: yeah; it might be system load related too, not sure [18:50] dobey: interesting. that'd make sense. this was the first time I'd tried running the tests in both VMs at once on this poor old mac mini [18:50] ah [18:50] dobey, +1 https://code.launchpad.net/~dobey/dirspec/python3/+merge/114713 [18:51] gatox: thanks. though i just realized, i bet the tests are broken on mac/win now :) [18:52] actually, no. windows is still using trial [18:52] forgot about that [18:59]