[11:16] good morning! [11:17] gatox, morning! [11:17] mandel, hi [11:17] gatox, I asked a question about the mp, can you let me know so I can approve :) [11:18] checking.. [11:20] mandel, are you seeing not implemented error when running the tests or just reading the code? [11:21] gatox, reading the code :) [11:21] gatox, there are rais not imple exceptions, why? [11:22] mandel, i did that, because i consider that they are some important tests, that the implementation is platform specific, so if you are going to extend from that class, i wanted to force to rewrite those tests and don't forget [11:25] gatox, so, is that test case ignored? or why I don't see any errors? [11:25] mandel, the darwin or windows implementation, depending of where are you running the tests, is rewriting those tests....... [11:26] the not implemented is in common [11:27] gatox, and those methods, cannot be implemented in common 'cause is all about ignored paths which is in common, right? [11:32] mandel, i didn't implement them in commonn because they have some "conceptual" difference in them, if you take a look in test_windows, they manage how to add the ignored path and generate fake events in one way, and for darwin in another way [11:33] you can take a look at test_darwin........ with the difference is really obvious [11:35] gatox, I'm looking at test_ignore_path looks so similar is a little concerning (at least in that tests) [11:36] gatox, the diffs are minimum yet there is a lot of lines that are the same [11:37] mmmmm maybe i can do a method to generate the fake event, and reimplement that on windows and darwin [11:37] gatox, seems more logical, but how are you in terms of time? [11:38] gatox, moved to the indicator already, if that is a problem I can take over it, no problem :) [11:39] mandel, the last i heard was that i can stay with fsevents a couple of day more if we need to close some stuff.... not indicator tasks assigned already (as far as i know) [11:39] so i can do that [11:39] gatox, ok then :) [11:39] gatox, that way I can focus on other changes [11:40] ok [11:40] gatox, how is this failing for you atm: https://code.launchpad.net/~mandel/ubuntuone-client/unify-processors/+merge/114906 [11:41] gatox, besides the path is ignored stuff [11:42] mandel, give me a couple of minutes and i'll check [11:42] gatox, thx! [11:54] gatox, for https://code.launchpad.net/~mandel/ubuntuone-client/unify-processors/+merge/114906 I have fixed the failing test and I have nor problem with running the tests in darwin [11:57] mandel, checking now [11:58] gatox, should be revno 1294 [12:01] mandel, when i try to run the tests for tests/platform/filesystem_notifications it says: no module named darwin...... which i pressume is from the daemon [12:02] gatox, ? can I see the full stack please [12:02] mandel, http://paste.ubuntu.com/1094762/ [12:04] gatox_mac, please add to the path the python code form lp:ubuntuone-fesevents-daemon [12:04] s/form/from [12:04] gatox, with that there everything works fine :) [12:09] mandel, no.... i added ubuntuone-fsevents-daemon to the path, and it fails in the same way [12:09] gatox, how did you add it to the path? [12:09] ahh i need to add the path to the python folder? [12:10] not the project? [12:10] ok, adding to the python folder it works [12:11] tests ok in darwin..... let me check in windows and linux, and i'll approve [12:11] gatox, of course, because the top dir is no a python package :) [12:11] gatox, I did it a number of times at the beginning too hehe [12:12] mandel, yes, i didn't even check that when i add it, then i saw the folder and i realize that :P [12:13] gatox, I'm off to by food and cook it (store is next home) let me know if everything is ok [12:13] ralsina, 1-1 after my lunch? [12:13] ok [12:15] ralsina, I'll assume that is a yes :) [12:17] * mandel small errands (buy food) + lunch [12:23] mandel: yes [12:29] and good morning [12:30] ralsina, hi [12:31] hi gatox, how's that fever doing? [12:31] gatox: obscure reference http://www.youtube.com/watch?v=Xg5HSoX4H18 [12:31] ralsina, naaa.... i'm fixed! jeje [12:32] ralsina, i rest ALL the weekend [12:32] ralsina: hey [12:33] hi dobey [12:43] alright. just mailed our list. see you all tomorrow (or later today if I manage to get everything finished up early) === zyga is now known as zyga-food [13:30] * mandel back [13:30] ralsina, 1-1? [13:31] mandel: yes but on IRC, I have people sleeping [13:31] ralsina, sure, no problem :) [13:34] bug 1017450 [13:34] Launchpad bug 1017450 in Ubuntu One Client "Problems with special folders on Windows." [Undecided,Incomplete] https://launchpad.net/bugs/1017450 === zyga-food is now known as zyga-food] === zyga-food] is now known as zyga-food [13:54] very late but good morning to everybody! [13:55] good morning alecu! [13:55] alecu, hi! [13:55] gatox: I've been doing some research on how to approach the sync menu, please ping me when you have some time to mumble. [13:56] alecu, whenever you want.... i'm just refactoring some minor code [13:56] gatox: I've got the techleads meeting in 30', let's do it after your lunch. [13:56] alecu, ack === zyga-food is now known as zyga [14:01] alecu, hola, can you give me a review for: https://code.launchpad.net/~mandel/ubuntuone-client/fsevents-daemon/+merge/114836 [14:01] alecu, please... :) [14:02] mandel: it's over 1000 lines! why would I? [14:02] mandel: :-) [14:04] alecu, 'cause you like me? [14:04] alecu, and because is very easy and the length is due to a mv that just took a part of the file [14:05] alecu, and shows as a rm of the code an addito [14:05] addition [14:05] mandel: ack [14:06] alecu, what was it that convinced you, the 'because you like me' or the code issue? :P [14:06] mandel: the "like" issue. [14:06] alecu, hey! mandell has special treatment here?! jejee [14:06] gatox: don't be jealous! [14:06] jejeje [14:08] alecu, I knew it! hehehe [14:08] elopio, ping [14:25] alecu, facundobatista, gatox, when adding a new config for the ubuntuone-syncdaemon I need to provide a new conf file for darwin, although it sounds like something that should be in platform I think that putting it under syndaemon/config is better, what do you think?? [14:25] ralsina, cc ^ [14:26] mandel: what's darwin-specific about that config? [14:27] ralsina, in cofigglue you can add a new config file to add extra parameters to the command line, in this case is related to the use of the fsevents-daemon as the default monitor in darwin [14:27] ralsina, so that that extra parameter is not shown on linux or win becuase it makes no sense [14:27] mandel: why not an option that does nothing on windows/linux? [14:28] ralsina, 'cause is ugly? the amount of work is similar and this way we do not clutter the command line [14:29] ralsina, but I don't want to add it to platform because it does not make sense [14:29] mandel: you can put it in the same place the other config files are, and give it a darwinish name [14:29] ralsina, yes, that is the idea, but I also have to change the way the files are loaded under syncdaemom/config [14:30] mandel: hmmmmm ok, don't know [14:30] gatox, please can I have a simple review: https://code.launchpad.net/~mandel/ubuntuone-client/unify-filemonitors/+merge/115154 [14:30] mandel: you are talking about a "config option", not a config file, right? [14:30] ralsina, meh, I'll propose the code and people can later complain :) [14:30] mandel: a new config option, not a new config file, right? [14:30] mandel, yes, in a couple of minutes, i'm finishing with the refactor for tests-darwin [14:30] if the option has a default, you don't need a new file [14:31] mandel: that's what I call BDD! [14:31] alecu, ideally a diff config seems cleaner.. [14:31] alecu, BDD? [14:31] mandel: bitchin' based development! [14:31] mandel: sorry, "...driven..." [14:32] alecu, you mean, propose, people complain, adapt to complains? [14:32] mandel: right! [14:33] alecu, hehe could be [14:33] alecu, I like the idea of adding an if sys.platform == 'darwin': blah because is easy, I fear that it does not follow the design we have had so far in all the other packages/modules [14:34] alecu, I kind of like to be consistent.. but in this case seems to not be necessary [14:42] mandel: perhaps we can make this an option for every platform, like "--file-notification-backend=..." but the options for linux and windows are only one for now. [14:42] mandel: perhaps a bit shorter, even :-) [14:43] alecu, yes, that shoulds like a much better approach [14:43] alecu, specially if we ever have more backends [14:44] alecu, with the default being in the code per platform [14:45] mandel: and perhaps we can make the "..." be an id in a dict of module names that we end up importing. [14:45] mandel: and then, only that dict should be in the platform module... [14:46] alecu, yes in the __init__ of the monitors package which is in platform, is simple and nice because we can easily update it if we wanted :) [14:48] alecu, yet, the id has to be passed to the platform code to decide if it is the correct implementation to use, if the daemon is not running lets use the user base one an ignore the command (maybe log it for extra info) [14:49] hiya [14:50] gatox, can I bug you about this - you pointed out a lint error I've since fixed here: https://code.launchpad.net/~mikemc/ubuntuone-client/use-dirspec-get-cmdline/+merge/114720 [14:50] mmcc, ack [14:51] and I wanted to ask when the ubuntu-sso-ssl-certificate-qt dialog is used. AFAICT it shouldn't launch correctly on windows with the code in trunk [14:51] background in my notes here: https://code.launchpad.net/~mikemc/ubuntu-sso-client/use-dirspec-get-program-path/+merge/114690 [14:52] ^^ ralsina, alecu, mandel, anyone else ? [14:52] also, thanks gatox :) [14:54] Sorry am on a call [14:54] np, take your time :) [14:56] mmcc, as soon as I'm done with some code I'll take a look :) [14:56] * mandel is not great at leaving code in the middle [14:58] mandel, thanks, no problem. it's not necessarily blocking, just wanted to ask while everyone was around :) [15:01] me [15:01] oops, typing [15:02] alecu, ralsina mandel dobey mmcc thisfred standup' [15:02] me [15:02] me [15:02] otp, so no notes yet [15:03] otp2 [15:03] so you guys go ahead [15:03] dobey is not around today [15:03] ok, mandel last [15:03] DONE: [15:03] Reviews, improve tests in darwin-tests-refactoring, starts reading the docs for sync menu and desing draft. [15:03] TODO: [15:03] Fix remaining fsevents issues. mumble with alecu about sync-menu. Keep looking into sync-menu stuff. [15:03] BLOCKED: [15:03] No [15:03] briancurtin, go [15:03] me [15:03] DONE: a few reviews, looking into how to do some of the unicode branches better based on review comments [15:03] TODO: finish up unicode branches [15:03] NEXT: thisfred [15:03] DONE: fix integration tests TODO: u1db documentation update BLOCKED: no NEXT: mmcc [15:03] oops, still typing [15:04] me [15:04] DONE: windows path problems, hopefully [15:04] TODO: land branches, fix setup-mac [15:04] BLCK: none [15:04] NOTE: off thursday, friday and next monday to move [15:04] mandel [15:05] DONE: proposed severla branches to clena fs events a little and merge the daemon code. Made changes to the option parse to add a new option to the ubuntuone-syndaemon bin. [15:05] TODO: finish the extra arg parsing. [15:05] BLOCKEd: no [15:05] COMMENTS: reviews are welcome :) [15:06] ah..... todo for me also: more reviews [15:17] brb [15:21] b [15:26] mandel, you can re-review this branch now with the changes you ask for: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin-tests-refactoring/+merge/113289 and i'll reviews your branches and mmcc 's after lunch [15:27] gatox, ok, awesome :) [15:45] mmcc, do you think you could do a review for https://code.launchpad.net/~mandel/ubuntuone-client/unify-filemonitors/+merge/115154 ? [15:46] sure mandel, I'll look at it soon [15:46] mmcc, thx! [15:52] ok, lunch for me === gatox is now known as gatox_lunch [16:15] so mandel, where does ubuntuone.darwin.fsevents come from again? is that the daemon in your separate project? [16:16] oh wait, I think you said it in the irc backlog... [16:16] onwards and lunchwards [16:18] mmcc, yes :) [16:18] mmcc, is the python code from the daemon [16:18] mandel, ack. I just re-read your conversation with gatox === salgado is now known as salgado-lunch [16:26] mandel, sorry - needsfixing - test_darwin still imports common from the old place... [16:29] mmcc, sure, no problem let me check :) [16:31] mmcc, he, side effects of using pipes, switching from one to other leaves the .pyc behind.. [16:40] mandel: ah, good to know. need pipe-cleaners [16:43] mmcc, yes, is a PITA cause things work even though it should 'not' be there [16:43] mmcc, fixing now [16:51] mmcc, I just pushed the new version with the fix [16:51] ok mandel, looking now [16:53] mandel, which tests should I be running for this? they're going further but don't all pass [16:54] mmcc, just the ones under filesystem_monitor [16:54] mandel: ack [16:54] mmcc, I need to do some black magic in the future to let you run all u1-client tests.. [16:54] mmcc, as with windows, getting the tests to run is harder than getting the software to run because they are integrations tests O_o [16:54] but I'll get there :) === gatox_lunch is now known as gatox [17:12] gatox, some extra comments :) [17:12] and with that EOD for me [17:12] mandel, wait [17:13] waiting... [17:13] mandel, i don't understand your comments, this is made in this way for previous requests [17:13] s/is/was [17:14] gatox, what do you mean? the idea was to reuse as much as possible, that is great, yet setting a modules variable from a diff module for it to work is wrong [17:14] gatox, same happened with the filesytem_notifications/common.py remember? [17:14] mandel, sorry right as you're leaving - did you mean just the tests under filesystem_notifications? those passed then just hung (in cleanup?) for > 10 minutes for me :( [17:15] gatox, if it is a PITA we can land it like that and later change it [17:15] mandel, can you show me an example of this being done in another way..... as far as i remember this was discuss it, and this was the solution that we decided [17:15] mmcc, yes, tests hang there due to a thread not being cleaned but that is fixed in the branch that gatox is trying to land, so that is 'correct' :) [17:16] gatox, possibly the same way we did under ubuntuone/platform/filesystem_notifications/common.py where you import the platform bits that you do not know how are implemented [17:17] gatox, but not to worry, we can always use alecu to vote and decide :) [17:17] that is the good thing of having to reviewers, no hay empates :) [17:17] mandel, ah, ok. next time would you mind saying a little more in the notes for a merge about this kind of stuff? ie, how to run tests, what to test, what should / shouldn't be broken? it'll make it easier to review things after you leave for the day. [17:17] i'll take a look, i'm not sure what are you talking about exactly [17:17] mmcc, yes, I should sorry I though you already knew all this little issues [17:18] gatox, not to worry, ask alecu for his point of view and if he is ok with the current implementation I have no problem in leaving it like it is [17:19] mandel, no prob really. It just makes it smoother, esp. if you're not around to bug :) === salgado-lunch is now known as salgado [17:19] besides mandel, I had to learn how to build the windows exe on Friday, so that completely flooded my mental cache :) [17:19] mmcc, I know, I should have done it, sorry again won't happen again if I remember :) [17:20] mmcc, lol yes, if something can be said about this code is that it makes you learn a lot of stuff [17:20] ok, EOD, gatos let me know what alecu says and I'll approve as fast as lighting :) [17:21] mandel, ahhh i think i understand... you mean doing: if sys.platform, etc, etc..... i'll change it..... but i'm pretty sure that this was alreadu discuss it [17:21] gatox, got a bad memory or I was becoming japanese :P [17:21] gatox, again, don't do work if it was agreed, ask alecu first don't waste brain cycles :) [17:22] all, see you tom! [17:22] have a great evening! [17:24] ralsina: vista user is getting this error which prevents him from completing setup: RemoteError 'remote_register_to_signals() takes exactly 2 arguments (3 given)' - i know i've seen it before but can't remember more than that, any ideas? [17:24] alecu, are you around? [17:25] joshuahoover: i haven't seen that particular error but we know of one similar to it (the 2 arguments, 3 given part) [17:26] briancurtin: yeah, maybe i'm thinking of a different error, i know i've seen the 2 args, 3 given thing in logs before [17:26] btw ralsina, if you're reviewing https://code.launchpad.net/~mandel/ubuntuone-client/unify-filemonitors/+merge/115154 today, look at my approve comment for a summary of the caveats to run the tests on darwin. [17:29] hola gatox, I was afk. [17:29] alecu, can you take a look at mandel 's comment: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin-tests-refactoring/+merge/113289 ? [17:30] gatox: sure [17:31] alecu, can you please tell me if we should fix that inn common using if sys.platform: blah blah blah....... or as it is now, where if you are running test_windows.py, that module will ensure that the proper values are set in common in order to run the tests [17:40] gatox: I don't understand why this is needed in common.py: ..."filesystem_notifications = None" [17:41] alecu, that can be actually in test_darwin.py and extend those tests to add the patch [17:42] gatox: I'm asking about alll the global assignments starting in line 54 of the diff [17:45] alecu, as we are unifying the tests, there are some stuff or functions (to do things one way or another) depending on the platform..... the thing is.... should we do those assignments using if sys.platform..... etc, etc..... of leaving those assignment like that, and because of test_darwin and test_windows are extending common.py, just leave the responsability to set those values to the specific tests as it is now [17:45] gatox: and also the stuff about common_tests starting in line 839 [17:45] same answer [17:46] gatox: Use sys.platform? no. Global assignment? does not look right either. [17:46] gatox: there must be some better way. [17:47] gatox: I'm guessing the "template method" pattern, but I might be missing something. [17:48] alecu, i don't understand which is the problem right now.... we are not suppose to run common.py by its own, so when we run test_windows.py for example, that modules sets the proper values.... but maybe i'm not seeing things clearly [17:49] that sounds ok for me [17:49] gatox: the issue is that all those global assignments are aweful. [17:50] gatox: I remember nessita found a better way to do them, but I'm not sure if it was on the test classes [17:57] alecu, ok..... i'll try to think another way to implement this..... [17:59] but..... i'm not really convinced [17:59] alecu, lost his connection [18:00] (that message was for everyone ^) [18:13] ralsina: know anything about bug #1025070 that leo found on q? [18:14] Launchpad bug 1025070 in ubuntuone-installer "Ubuntu one installer shows the package names and doesn't install " [Undecided,Confirmed] https://launchpad.net/bugs/1025070 [18:14] joshuahoover: not yet, have not seen it :-/ [18:14] ralsina: well now you have ;) [18:14] joshuahoover: and I installed it early this morning, too [18:14] interesting [18:14] joshuahoover: we are getting rid of installer anyway [18:14] joshuahoover: soonish [18:15] ah, ok [18:16] joshuahoover: I'll assign to dobey anyway [18:16] ralsina: thanks [18:29] so it looks like jenkins has an out-of-date dirspec - how can I help fix this? [18:30] mmcc: ping sidnei [18:31] mmcc: since dobey is out [18:31] ralsina: ok. [18:35] lunch now, finally [19:49] alecu, please ping me when you are back [19:50] (if you receive this message jeje) [19:50] ralsina, mmcc jenkins fixed [19:51] mandel: woohoo [19:51] ralsina, well, fix, it runs the tests.. which mmcc had problems with [19:52] mandel: baby steps! [19:52] ralsina, I think the config issue is easy but I'm not looking at it just now :) [19:52] mandel: it's late for you anyway. Go have a beer. [19:52] ralsina, will do :) [20:26] ok...... eod for me... i'll review my suggestions tomorrow with alecu [20:26] bye! [20:36] gatox! too late :P [20:45] alecu: missed him by this much! === yofel_ is now known as yofel === alecu is now known as alecu86 === alecu86 is now known as alecu [21:10] EOD for me. See you all tomorrow! [21:11] mmcc: did you ever get dirspec tests running on windows? [21:11] briancurtin: no, but it sounded like others were able to do it last week [21:14] briancurtin: why do you ask? [21:15] mmcc: making it work via buildout is, for some reason, challenging [21:16] briancurtin: for automated testing? or just getting the current trunk using buildout? [21:22] mmcc: getting trunk's run-tests.bat to work in a buildout env. not specifically for jenkins, just for me being able to run it [21:22] mmcc: maybe i'll check tomorrow with whoever did get it working in their env and see what the deal is [21:23] briancurtin: ah, ok. yeah, I had a real hard time too… dobey told me who did it, let me see if I have it in the scrollback === salgado is now known as salgado-afk [21:29] briancurtin: according to this: https://code.launchpad.net/~dobey/dirspec/python3/+merge/114713 -- gatox is the last person to run the windows tests. [21:29] ah, i think they have the non-buildout way setup [22:52] ok, dinner...