[07:56] morning [09:01] Good morning all! :) === zyga is now known as zyga-afk [11:52] morning, all! === zyga-afk is now known as zyga [12:30] alecu, morning! [12:31] just got be leaving for lunch, alecu want to talk when I'm back about the fsevents ancestors bug? [12:31] mandel: sure [12:31] alecu, ok, I'll ping you when I'm back :) [12:31] * mandel lunch [12:51] morning! [13:21] mandel: don't set my update-2-0 branch back to Approved please. it won't land in tarmac. [13:34] dobey, ok [13:34] * mandel back [13:35] dobey, so we should reject it, right? or is it going to be there waiting? [13:36] alecu, I'm back, we can talk about the ancestors stuff whenever you have the time [13:39] mandel: no, i think i'll have to land it by hand [13:40] dobey, oh, ok, if you have to do that, can you also take a look at https://code.launchpad.net/~alecu/ubuntuone-client/timestamp-autofix-1-6/+merge/82754 [13:41] dobey, do we have to land that? [13:42] we should. i will have to look at the older branches another day though. i need to do stable-4-0 releases today [13:42] mandel: you should only worry about branches that have 0 or 1 reviews really. :) [13:43] dobey, I know, but I see them and I want to get the things fix, is annoying to see that branch always there [13:44] alecu, can I have a review for https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/remove-arc-fsevent/+merge/116433 ? [13:44] dobey, may I have a review for the following https://code.launchpad.net/~mandel/ubuntuone-fsevents-daemon/long-domain-socket/+merge/116269 [13:45] hello back, mandel. [13:46] let's mumble! [13:46] alecu, ok, launching the thing :) [13:50] good morning everyone [13:52] alecu, bug #1012709 [13:53] Launchpad bug 1012709 in Ubuntu One Client "Add Watches to udf ancestors" [High,In progress] https://launchpad.net/bugs/1012709 [14:04] brb, coffee [14:09] ralsina, 1-1? [14:09] mandel: we had one like 3 days ago :-) [14:10] ralsina, yes, but since we usually have then on monday.. that is why I asked, we can skip it, only thing diff Iwas going to say is that I'll take a few days off to visit the family in mallorca [14:10] mandel: ack [14:10] mandel: yes, let's skip [14:11] ralsina, ok, I'll do the paper work on canonical admin later this night [14:11] mandel: ok [14:20] mmcc, the branches regarding the arc have been updated [14:20] ok mandel, I'll take a look today [14:21] mmcc, superb, thx! [15:00] me [15:00] me [15:01] me [15:01] me [15:02] me [15:03] alecu? [15:03] dobey? [15:03] meh [15:04] ok, alecu is last, go thisfred [15:04] DONE: https://code.launchpad.net/~thisfred/u1db/ergo-sum/+merge/117186 | https://code.launchpad.net/~thisfred/u1db/iterate-over-list-of-dicts/+merge/117120 TODO: wrap up cosas example app | documentation BLOCKED: no NEXT: mandel. [15:04] DONE: Update all arc MPs to follow reviews. Refactored communication code so that it can be reused for the jenkins testing daemon. [15:04] TODO: Finish the jenkins testing daemon so that we do not need to run tests as root. Canonical admin paper work. [15:04] BLOCKED: no [15:04] ralsina, please [15:04] no notes yet, sorry [15:04] DONE: chased contractors, catchup, weekly team call [15:04] TODO: start actual work, contractor chasing [15:04] BLOCKED: no [15:04] briancurtin: go [15:04] DONE: team call, unicode [15:04] TODO: unicode [15:04] NEXT: mmcc [15:05] DONE: fixed new reactor import, gave up on test for that [15:05] TODO: darwin cert location, darwin dirspec paths, first-run code [15:05] BLCK: none [15:05] NEXT: dobey [15:05] DONE: finish dev-tools refactor, reviews [15:05] TODO: releases, fix review issues in refactor branch, reorg to drop ubuntuone-installer for good, SRUs for lucid/natty/oneiric [15:05] BLCK: None. [15:05] alecu [15:08] Comments? [15:08] since alecu seems to be away [15:08] me! [15:08] writting notes! [15:08] ok, let's wait [15:09] * ralsina taps fingers [15:09] * ralsina looks at his wrist where a watch could be [15:09] * ralsina whistles [15:09] DONE: team meeting, debugging of #1029450, reviews TODO: reviews for mandel, back to py3k [15:09] ok, I'll stop now ;-) [15:09] BLOCKED: no. EOM [15:09] EOM it is [15:11] so FYI ralsina, I gave up on that 'fail if we import reactor in the wrong place' test, because it looks (at least in this case) like the code under test is imported before the test module is imported, so it's not clear how to hook it in… [15:11] mmcc: hmmm ok [15:11] er, "not clear how to add a hook to catch the import, since it's already imported. " [15:11] mmcc: I suspect there is no simple way to test that since the tests use a reactor too [15:12] ralsina, mmcc , exactly, u1trial does the reactor import so you have a problem there [15:12] So to test this it would have to be a script that does what the windows main does, not a "test" [15:12] ralsina: yes. If the test script didn't use reactor we could just import and check sys.modules. I thought about running a separate process inside the test to just try importing, but it was getting late. I'm happy to revisit that today if you think it'd be worth the time [15:13] mmcc: probably not worth it :-( [15:13] ralsina: no, all it has to do is "import platform", then check sys.modules for reactor [15:13] mmcc: how about a script that does exactly that and running it at the end of run-tests ? [15:13] or, really, import 'platform.everything-but-filesystem_notifications'… [15:14] ralsina: I'm fine with that. I even know where to add it to the makefile [15:14] mmcc: :-) [15:14] mmcc: get dobey's or alecu's opinion for confirmation [15:15] mmcc: but I would really like to have a test for that since we have broken it, like, 4 times already [15:15] I really like the idea of a script that tests that the reactor was not imported at module level in any of our code. [15:15] mmcc: +1 [15:15] ok to use 'banner' to print the error message for whoever breaks it? [15:16] mmcc: then we'd need to include a "banner.exe" in our source tree :) [15:16] mmcc: and use ferret-over-ip to steal their cookies [15:16] alecu: worth it! [15:16] mmcc: you know the problems about finding binaries cross-platform ;-) [15:17] * mmcc googled "native python banner", found nothing. [15:18] ok, extra test script coming up [15:20] mmcc: http://www.noah.org/wiki/banner [15:21] cowsay [15:22] ralsina: nice! :) [15:22] eww [15:22] "# This font storage format is not very efficient" [15:22] no kidding [15:22] dobey: I am sure it could be worse, but will not be easy to find HOW [15:23] also, there's no real need to have such a thing here [15:23] also [15:23] YAY COPYRIGHT WITH NO LICENSE [15:24] dobey: this conversation is a joke. We are not putting a copy of banner in devtools [15:24] dobey: don't panic! [15:24] dobey: no, no need for a banner. Especially once we get those remote controlled nerf rockets [15:24] i'm not [15:28] mmcc, can I have a review for : https://code.launchpad.net/~mandel/ubuntuone-client/clean-imports/+merge/116256 [15:28] dobey, ralsina, super easy review: https://code.launchpad.net/~mandel/ubuntuone-client/correct-namespace/+merge/116264 do you have time to take a look? [15:28] mandel: sure [15:29] ralsina, thx [15:29] mandel: yes [15:29] mmcc, thx to you too :) [15:37] mandel: looks good to me, I suspect there is no way to write tests for this :-) [15:38] ralsina: required imports are self-testing :) [15:38] dobey: exactly [15:39] and i need to get lunch and ship back my doa raid enclosure. :-/ [15:39] bbiab [15:39] ralsina, install the new version and run the already present tests ? :) [15:39] mandel: ok, +1 ;-) [16:29] mandel: +1 on your clean-imports branch [16:29] * briancurtin lunch [16:36] mmcc, awesome, thx! [17:04] ok, EOD for me [17:04] see you all tom! [17:04] mmcc, tom we should talk about @property (retain) which allows you use autorelease in the init (I was doing some tests and found a mem leak) [17:05] ok mandel, sounds good. [17:10] lunch and bank time for me. [18:09] before i set https://code.launchpad.net/~brian.curtin/ubuntu-sso-client/py3-unicode-part-2/+merge/116764 to "needs review" does anyone know why it's showing changes from the prerequisite branch inside the diff? i suspect that's going to be a PITA to anyone reviewing. it should only be around a 500 line diff and it depends on a 500 line branch, but they're both shown together [18:09] (i know i've complained about this before...haven't found any solution) [18:12] i also suspect that's going to be a PITA to complete the actual merge... [18:13] briancurtin: no idea really, let me check [18:14] briancurtin: it shows as a 500 line branch to me [18:14] briancurtin: 594 lines (+106/-78) 13 files modified to be specific [18:14] ralsina: the diff i'm seeing on the MP is 1042 [18:14] briancurtin: reload :-) [18:15] wtfffff [18:15] surprise! It seems lunchpad recalculated at some point [18:16] heh [18:16] i wonder if it did the recalculation when i entered a commit message and description, because it was showing 1042 after opening the page in a fresh browser like 10 minutes ago [18:16] briancurtin: did you push a new revision about 15 min ago? [18:17] dobey: i pushed about 15 minutes ago but there was no changeset to push [18:18] er, no revision to push [18:18] hrmm [18:24] ralsina: https://code.launchpad.net/~dobey/dirspec/update-4-0/+merge/117317 easy peasy review [18:24] dobey: looking... [18:27] dobey: +1 because why not [18:42] awesome [19:36] test script for reactor imports: https://code.launchpad.net/~mikemc/ubuntuone-client/add-no-reactor-test/+merge/117328 [19:39] * mmcc lunchtime === dduffey_afk is now known as dduffey [21:39] dobey, can you be a little more specific about which certain things you're referring to in https://code.launchpad.net/~mikemc/ubuntuone-client/add-no-reactor-test/+merge/117328/comments/252306 ? [21:40] I've added the unicode_literals import and fixed the print statement to be a function, and used the new string formatting syntax, but I'm not sure what else you might be referring to [21:43] mmcc: i don't think you pushed the unicode_literals or print_function changes [21:44] briancurtin: no, not yet. Was just going to discuss first, but I'll go do that. [21:44] done [21:45] ish [21:48] mmcc: oh, you can use %-notation for string formatting [21:48] mmcc: so primarily the future imports is what i was talking about [21:48] interestingly, python -3 is complaining about that code, giving the DeprecationWarning about CObject, on the line where I call real_import (line 63 of the script). real_import is from earlier: real_import == __import__ [21:49] mmcc: have you tried to run it under python3? [21:50] the test will obviously fail to import the syncdaemon stuff, but if there's anything wrong with the script itself, it should expose that [21:50] because i don't know what the deprecation warning means there :) [21:51] dobey: I did run it under 3, but there's the issue of builtins vs __builtin__, what's the done thing there? should I do an 'try import builtins, except ImportError, import __builtin__? [21:51] in other words, does it need to run in both 2 and 3? [21:51] it doesn't need to run in both now [21:52] but i think we should stick to at least print_function in new code, and unicode_literals where possible (since unicode is a huge and lumbering pain in python) [21:58] ok, sounds reasonable. so, should I leave in the import of unicode_literals even though I'm not actually using bytes objects anywhere? [21:58] * mmcc was reading through the PEP again to try to be sure I'm right about that [22:01] i think so [22:01] well, i guess it's not about bytes objects [22:01] bytes objects are still bytes in 3 [22:02] but str is not bytes in 3 [22:05] right, and since I'm using str here, I don't need to write bytes literals. so I don't really need to import unicode_literals, I guess [22:07] well, if anyone else ever changes that code and has to do unicode stuff in there for some reason, having unicode_literals will catch problems [22:08] and if you aren't doing unicode stuff right now, it doesn't hurt anything [22:09] ok, no prob. also, python3 didn't complain about the CObject deprecation when I ran it just now, after s/__builtin__/builtins [22:10] ah, just saw your re-review. thanks. [22:11] mmcc: approved [22:12] and with that, i'm out. later everyone [22:15] later all, time to go!