[03:29] <mmcc> back for a while.
[05:44] <mmcc> Filed a few bugs, proposed a couple merges, wrote an email about bandwidth throttling that I'll revisit in the morning. That's not a bad night. See you all back in a little while...
[09:07] <thumper> hey folks
[09:07] <thumper> I'm trying to work out the best way to get my music onto U1
[09:08] <thumper> I want to have a subset of it downloaded to my laptop
[09:08] <thumper> but not all 50 gig of it
[09:08] <thumper> AFAICT if I sync a folder, I get it all ...
[09:23] <rye> thumper: hi, you can select certain folders within main music folder instead and select which folders to sync on your laptop
[09:24] <thumper> hmm... I don't really feel like adding 100 folders... :(
[09:59] <mandel> thumper, what about this (might not be the best idea) but we do not follow links, so what you can do is the following:
[09:59] <mandel> thumper, create a music folder, inside it add the folders you want to upload, for those that you do not want to upload, add symlinks
[10:00] <mandel> thumper, what would happen is that those that are links wont upload, those that aren't will and you will have a single folder to sync, also I think that banshee and rythimbox do follow symlinks so you can add that folder as the source for the music player lib
[10:00] <mandel> rye, is that ^ correct?
[10:02] <thumper> I was thinking perhaps just having two folders...
[10:04] <mandel> thumper, that works too :)
[10:04] <mandel> thumper, I went to the over complicated version maybe
[10:18] <rye> mandel: re: folders, well, your suggestion will work too
[10:20] <mandel> rye, a little hacky and if we support sync outside ~ we might as well start supporting symlinks
[10:20] <JamesTai1> Morning all!
[11:09] <gatox> good morning
[11:12] <mandel> gatox, morning!
[11:12] <mandel> gatox, how is everything going?
[11:13] <gatox> mandel, fine.... yesterday i jump from one problem to another trying to get u1-client tests to work jeje i'm finding everything that needs to be mac specific
[11:13] <gatox> mandel, you?
[11:14] <mandel> gatox, great! her things a looking very nice, I'm adding correct support for async operations in the twisted side when dealing with condig requests (using ^{} blocks as callbacks :P ) and will have the python code ready to be doing some really cool things
[11:15] <gatox> mandel, great!! i already want to see everything working :P
[11:15] <mandel> gatox, I might be able to be adding events to be processed in the main loop of the reactor which means we are 1/2 days way from having the daemon idea fully working with syncdaemon \o/
[11:15] <gatox> \o/
[11:39] <rye> verterok: i found a dumb way to break syncdaemon - truncating a data file to 0
[11:40] <rye> verterok: given that this is quite possible in real life (crash of the process before writing file, filesystem corruption due to power outage) can SD recover from this condition?
[11:40] <rye> joshuahoover: org.freedesktop.DBus.Error.UnknownMethod:
[11:40] <rye> joshuahoover: is it familiar ^ ? :)
[11:40] <verterok> rye, what's a "data file"?
[11:41] <rye> verterok: 1337008100237107.inactive.tritcask-v1.data
[11:41] <rye> verterok: so it is not a hint, but actual data
[11:42] <verterok> rye, how is that a read only file can be truncated?
[11:42] <verterok> rye, "crash of the process before writing file" can't happen in tritcask :)
[11:43] <rye> verterok: well... you remember time.time() returning NaN
[11:43] <verterok> rye, no...
[11:43] <verterok> rye, the idea of tritcask is that it only appends data to the "live"/active file
[11:44] <verterok> rye, only *append* is the important bit
[11:45] <verterok> rye, so an already inactive file can't be truncated by the SD process...it can only be done from outside
[11:46] <rye> verterok: does the file get renamed during its lifetime?
[11:48] <rye> verterok: the user has only these lines in the log - https://pastebin.canonical.com/66135/
[11:51] <verterok> rye, only it's renamed when marking it as inactive
[11:52] <verterok> rye, we only have a single "live" file, which is moved on startup to inactive and not touched again
[11:53] <verterok> rye, and 1336117336290348.inactive.tritcask-v1.data is a 0 size file?
[11:53] <rye> verterok: just tested writing garbage to a live file - SD recovers properly
[11:53] <rye> verterok: that's an artificially created truncated file, not on the user's machine, I am just guessing and asked the user to provide the information on metadata folders
[11:54] <verterok> rye, it should ignore all the data after it finds the garbage in a data file :)
[11:54] <verterok> rye, ah. ok
[11:54] <rye> verterok: for that user syncdaemon simply stops, control panel breaks, u1sdtool hangs...
[11:54] <verterok> rye, we could handle that case.
[11:55] <verterok> rye, do you get a traceback or sth?
[11:55] <rye> verterok: nope, nothing. dbus reply never comes back from SD
[11:55] <rye> verterok: i wonder whether the user has eCryptfs
[11:56] <rye> i remember we had fun with it
[11:56]  * rye migrates over to an encrypted filesystem
[11:58] <verterok> nothing in the logs?
[12:03] <rye>  verterok nope
[12:11] <rye> verterok: KeyError
[12:11] <rye> u'926c8fcf-70dd-47f6-94a8-0643c09d8b53'
[12:11] <rye> how sweet
[12:12] <rye> ralsina: enabling 2 folders in a row in control panel/installer after answering to "Merge?" result in an error message
[12:26] <rye> verterok: how does SD react to I/O Error on reading tritcask file?
[12:32] <mandel> ok, lunch for me
[12:32]  * mandel lunch
[12:34] <gatox> mandel, mmcc i added a new instruction in the Ubuntu One Port document, because in MAC OS you need to download protobuf and build it following the readme instructions, they provide protoc already built only for windows
[12:35] <mandel> gatox, ok!
[12:35] <mandel> gatox, is stupid the amount of work that has to be done ;)
[12:35] <gatox> jeje yep
[12:35] <mandel> gatox, but I think we might get a working version a lot earlier than what we had for windows :D
[12:35] <mandel> now, really having lunch
[12:36] <gatox> enjoy
[12:40] <verterok> rye, while reading? opening? :)
[12:41] <rye> verterok: while opening
[12:41] <verterok> rye, it catch the error and marks the file as "broken"
[12:41] <verterok> rye, that's done renaming the file
[12:41] <rye> verterok: in case of ecryptfs i managed to make it fail on opening, not reading...
[12:41] <verterok> s/(live|inactive)/broken
[12:42] <verterok> thats the rename ^
[12:57] <gatox> ralsina, mandel mmcc I CAN RUN THE TEST!!!! FINALLY!!!
[12:57] <gatox> so excited! :P
[12:57] <ralsina> rye: could you file the bug about "merge?" and assign to me, please?
[12:57] <ralsina> gatox: awesome!
[12:57] <ralsina> And good morning everyone!
[12:57] <gatox> ralsina, good morning!! :P
[12:57] <gatox> now i only need to fix the tests, and fill the empty implementations!
[12:57] <gatox> :P
[12:58] <ralsina> gatox: ONLY!
[12:58] <gatox> jejeje
[12:58] <ralsina> gatox: so just implement it and test it? Yay! ;-)
[12:58] <gatox> ralsina, but now i have a more clear picture of what need to be done mac specific
[12:59] <ralsina> gatox: "a sync client"? ;-)
[12:59] <gatox> jejeje
[12:59] <gatox> you are ruining my happy moment!
[12:59] <ralsina> gatox: but yes, congrats, I know it was a ton of boring work
[12:59] <gatox> jejee
[12:59]  * ralsina is back on the no-cofee wagon and has to go to facundobatista's house so, a bit grumpy
[13:00] <gatox> ahhhhhhh i understand.... it happens the saame here when i need to leave my home :P
[13:42]  * mandel back
[13:42] <mandel> gatox, hurray!!!!
[13:42] <mandel> gatox, now.. do they pass ;)
[13:42] <mandel> briancurtin, ping
[13:42] <briancurtin> mandel: pong
[13:43] <gatox> mandel, some of them yes..... the new implementation of os_helper for move_to_trash obviusly not..... but i'm not running all the tests....... i fix a lot of import tihngs, but a lot have empty implementations, but now we know where to look
[13:43] <mandel> briancurtin, it seems that the jenkins tests fail because is not using the right dev tools which are set up by the buildout is that right?
[13:43] <mandel> briancurtin, or do we have to update the system installation?
[13:44] <briancurtin> mandel: i think we have to update the system installation because the jenkins machine doesnt have an environment setup by buildout
[13:44] <briancurtin> and i thought it was going to update itself manually once nightlies caught up, or something like that
[13:44] <mandel> briancurtin, ok, then let me take care of that :)
[13:44] <briancurtin> s/itself manually/itself automatically/
[13:44] <mandel> briancurtin, we have not been that smart on windows just yet, although we could
[13:45] <mandel> briancurtin, give me 5 mins and I'll make sure we get jenkins on track!
[13:45] <briancurtin> awesome
[13:47] <mandel> briancurtin, afaik you changed the jenkins config to add the env variable that is needed by the skip, right?
[13:47] <briancurtin> mandel: yep, that's in there. it sets JENKINS=1
[13:49] <mandel> briancurtin, was there an other branch for the control panel tests that fail on jenkins?
[13:50] <briancurtin> mandel: you mean that needed any jenkins skipping?
[13:50] <mandel> briancurtin, yes
[13:51] <ralsina> mandel: why doesn't jenkins use the buildout? It's easiest! :-)
[13:51] <briancurtin> mandel: i haven't seen what failures occur on the CP branch on jenkins. if the site will ever load for me i'll get them corrected. last time i looked at jenkins it was claiming that branch was never even run
[13:52] <briancurtin> s/that branch/that project
[13:52] <mandel> briancurtin, ok, I've launched a new build for u1-client then I can take a quick look at control panel and will let you know
[13:53] <mandel> ralsina, right now best answer: 'because'
[13:53] <mandel> ralsina, by the way, I'm going to create the new project for the fsevents code, have you guys had the chance to try it in your macs?
[13:54] <briancurtin> mandel: i see the failures there now - there's two we should skip
[13:54] <mandel> ralsina, I have some license questions, can we use MIT libs?
[13:54] <mandel> briancurtin, if you add a bug and add the skip I'll approve as fast as you can say 'un pelo tiene mi barba'
[13:55] <mandel> briancurtin, dev-tools issue in jenkins fixed: https://jenkins.errormessaging.com/view/Windows/job/ubuntuone-client-windows-test/61/console
[13:55] <briancurtin> mandel: cool, adding the skips right now
[13:56] <mandel> briancurtin, let me know the branch, I don't think we need two +1 for a simple branch like that so we can land it asap
[13:57] <ralsina> mandel: AFAIK yes, ask dobey just in case
[13:57] <mandel> dobey, can we use MIT licenses? and are you the person that I have to ask to create a new project? can I do it?
[13:58] <dobey> use mit licenses for what?
[13:58] <dobey> and anyone can create a project on lp
[14:00] <mandel> dobey, I want to use a mit licensed lib that simplifies arg parsing with objective-c and I believe that our code is all GPL so I wondered if there are license issues
[14:00] <mandel> dobey, about the project, are there any considerations I have to take? ownership etc..
[14:02] <dobey> it is safe for GPL things to link/load MIT things
[14:03] <mandel> dobey, superb, thx!
[14:03] <dobey> main consideration about the project is license
[14:08] <mandel> dobey, I think I'll use a c lib from bsd and will ignore the arg parsing lib, makes more sense to have less deps
[14:09] <dobey> right, less deps is better, generally\
[14:09] <dobey> i meant the license of the project you're creating though :)
[14:13] <mandel> dobey, yeah, I understood that one :) I'll remove the dep and will do a push to my junk branch and I'll annoy you a little with the license of the project, I'm a little worried about a sjon lib dep I have (but I think I can easy find a diff lib easily)
[14:30] <mmcc> morning everyone -- catching up
[14:33] <mmcc> hi mandel, I'd go with the C getopt for the daemon, too - it won't usually be run with any arguments, so there's no need for that interface to be particularly nice. Besides, getopt isn't all *that* bad :)
[14:34] <mandel> mmcc, yep, changing to getopt_long as we speak :)
[14:42] <dobey> oi, not feeling so great today
[14:52] <ralsina> dobey: remember, I don't want sick people working. At least not sicker-than-usual so if you are feeling bad, let's take care of it.
[14:53] <ralsina> dobey, briancurtin, mmcc, gatox, thisfred: standup in 8'
[14:53] <ralsina> alecu is coming in a bit
[14:59] <dobey> not ill
[15:00] <gatox> me
[15:00] <gatox> ralsina, mandel dobey briancurtin mmcc alecu thisfred standup?
[15:00] <briancurtin> me
[15:01] <mandel> me
[15:01] <mmcc> me
[15:02] <dobey> me
[15:02] <dobey> thisfred is sprinting
[15:02] <dobey> alecu, ralsina ?
[15:04] <gatox> well..... go for me?
[15:05] <alecu_> go gatox go
[15:05] <alecu_> me
[15:05] <gatox> DONE:
[15:05] <gatox> Get the test for os_helper running on MAC OS, modify the document of the mac port to add the instructions to have protobuf working on mac, Fix several platform imports in ubuntuone-client, add a couple of Dummy implementations for some packages that need to be implemented now.
[15:05] <gatox> TODO:
[15:05] <gatox> Finish running tests in all platforms and propose, go back to filesystem notifications, and the dummy modules.
[15:05] <gatox> BLOCKED:
[15:05] <gatox> No
[15:05] <gatox> briancurtin, go
[15:05] <briancurtin> DONE: spent most of the day trying to setup a Server 2008 domain controller to test autoupdate inside a domain, a few small reviews, some quick research on another autoupdate issue
[15:05] <briancurtin> TODO: fix this remaining jenkins issue, get installer automation going live on jenkins
[15:05] <briancurtin> BLOCKED: no
[15:05] <briancurtin> NEXT: mandel
[15:06] <mandel> DONE: twsited script and root daemon have two way communication which makes twsited protocol return deferreds that get callback or errback according to the result of the op in the daemon side. Added args support to launch daemon with or without launchd. Updated domain sockets branch after alecus review.
[15:06] <mandel> TODO: Add tests!!! Create project.
[15:06] <mandel> BLOCKED: no
[15:06] <mandel> mmcc: please!
[15:06] <mmcc> * DONE: discussion and wrote email about bw throttling.
[15:06] <mmcc> - filed a few UI/UX bugs for SSO client and controlpanel based on my poking around (bug 1000056, bug 1000059 and bug 1000061)
[15:06] <mmcc> - fixed a trivial controlpanel UI bug, learned some QT UI in the process (XML, hooray)
[15:06] <mmcc> - got new xcode, added to apple dev team by urbanape
[15:06] <mmcc> * TODO: finally merge my network state branches
[15:06] <mmcc> - work on packaging plan
[15:06] <mmcc> - check mandel's code on 10.6 with new XCode
[15:06] <mmcc> - Need reviews for network state branches, I added alecu and mandel as reviewers because they asked for it. The branches are:
[15:06] <mmcc> https://code.launchpad.net/~mikemc/ubuntu-sso-client/no-darwin-in-linuxnwindows-tests/+merge/105924
[15:06] <mmcc> and
[15:06] <mmcc> https://code.launchpad.net/~mikemc/ubuntu-sso-client/fix-networkstate-darwin/+merge/105926
[15:06] <mmcc> * BLCK: none
[15:06] <dobey> too long
[15:06] <mmcc> * NEXT: dobey
[15:06] <dobey> λ DONE: uds, swap days
[15:06] <dobey> λ TODO: fix bug #853060, 1-1 review meeting, 3.0.1 release/SRUs
[15:06] <dobey> λ BLCK: none.
[15:06] <dobey> alecu
[15:07] <mmcc> yeah, sorry about that - forgot about the bugbot
[15:07] <alecu_> DONE: lots of reviews,  catch up with email, got back to security patches
[15:07] <alecu_> TODO: small sprint with ralsina and facundobatista
[15:07] <alecu_> NEXT: ?
[15:07] <ralsina> me (sorry)
[15:07] <ralsina> DONE: fixed bandwidth settings, a cuple minor bugs, TODO: minisprint BLOCKED: no
[15:08] <mandel> ralsina, mini-sprint?
[15:09] <mmcc> what's the right place to send an email following up on the bw throttling discussion we had here yesterday?
[15:10] <dobey> mmcc: while the bug bot might spam a bit, it's nice to try and keep the entries for done/todo short and simple, and a single line for the whole set of each
[15:11] <dobey> rather than one item per line
[15:11] <mmcc> dobey: noted. will do
[15:11] <dobey> mmcc: i think the internal list is probably good for that
[15:11] <mmcc> dobey: is thatubunet-discuss?
[15:11] <dobey> yes
[15:11] <mmcc> thanks
[15:24]  * gatox lunch
[15:28] <dobey> ralsina: should we do the 1-1 review meeting after lunch i guess?
[15:39] <dobey> i guess that's a yes. need to take a slightly longer lunch today. bbiab :)
[16:10] <mandel> ralsina, can you tell alecu to do a re-review of my domain sockets branch?
[16:29] <briancurtin> does CP pass on linux?
[16:36] <mandel> briancurtin, should, want be to check?
[16:37] <briancurtin> mandel: if you have a chance. im looking at that failure from jenkins and i dont know what im missing here, but im also not sure it would be passing on linux either
[16:37] <mandel> briancurtin, running tests from trunk atm
[16:38] <mandel> briancurtin, all green with revno 327 in a linux running P
[16:39] <gatox> mandel, mmcc if you can please, i would be nice if you can review this branch: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/darwin-os-helper/+merge/106014
[16:41] <mmcc> gatox, sure. looking now
[16:41] <gatox> mmcc, thanks..... please let me know if you find anything that can be improved
[16:42] <ralsina> mandel: sure, I can. We were having lunch.
[16:47] <mandel> ralsina, thx
[16:48] <alecu_> hey mandel!
[16:48] <mandel> alecu_, hello!
[16:48] <alecu_> mandel, can you send me the link to the branch I need to rereview?
[16:48] <mandel> alecu_, sure, give me a sec
[16:48] <mmcc> gatox, what happened to run-tests.bat in that branch? did it get switched from DOS line endings or something? it's all one big diff...
[16:48] <mandel> alecu_, https://code.launchpad.net/~mandel/ubuntu-sso-client/use-unix-domain/+merge/104244
[16:49] <gatox> mmcc, yep..... line endings...... i'll try to revert it and edit it on windows
[16:50] <mmcc> gatox: was it edited on macos?
[16:50] <gatox> mmcc, i don't remember if it was on mac os or linux
[16:51] <mmcc> gatox, np. just was going to suggest editors on mac that'd do the right thing
[16:51] <mmcc> gatox or http://waterlan.home.xs4all.nl/dos2unix.html
[16:53] <mandel> gatox, vim!!!
[16:53] <gatox> mandel, naaaa..... i like ninja-ide jeje
[16:53] <mandel> vim vim vim
[16:54] <mmcc> on mac, maybe http://code.google.com/p/macvim/  ?
[16:55] <alecu> mandel, approved!
[16:55] <mandel> alecu, thx!
[16:57]  * mmcc goes to get lunch, almost manages to not say "emacs"
[16:57] <gatox> mmcc, mandel .bat file updated
[17:05] <dobey> heh
[17:17] <dobey> hrmm, need to install this win7 vm too
[17:18]  * briancurtin lunch
[17:19] <ralsina> briancurtin: while you are eating, I am about to assign to you a bunch of windows-installer bugs ;-)
[17:19] <mandel> ok, I need to go an walk the dog I'll probably be back later
[17:21] <dobey> ralsina: can you reliably reproduce the u1-installer GError bug, while having network available?
[17:25] <ralsina> dobey: yes
[17:25] <ralsina> dobey: at least I could a while ago
[17:26] <ralsina> dobey: OTOH "GError"? I reproduced the "it doesn't install stuff" bug ;-)
[17:26] <dobey> ralsina: resulting in a "GError: cannot execute program blah blah" no?
[17:27] <ralsina> dobey: yes
[17:28] <dobey> ralsina: right, i can't seem to make it happen, while i have network available. makes it hard to figure out how to write a unit test for it :)
[17:28] <dobey> and hard to know if it's fixed, if i can't write a test
[17:28] <ralsina> dobey: I only made it fail by having outdated apt-get data, so again, external causes
[17:28] <ralsina> dobey: you could LD_PRELOAD a broken gethostbyname
[17:29] <dobey> well i can fake netowrk breakage easy enough
[17:29] <dobey> but that doesn't help me fix it for the cases where network isn't broken
[17:30] <dobey> at least, after talking with people last week, i think i know how to fix it. i just don't know how to test that the fix works, since i can't make it break, outside of just not having network
[17:32] <dobey> also, when the heck does win7 prompt you for the registration key?
[17:32] <ralsina> dobey: first reboot, IIRC
[17:32] <ralsina> dobey: I can IRL the test, at least
[17:33] <ralsina> dobey: I could reproduce it by havig nightlies enabled, having something get into the repo, then trying to install without running apt-get update first. Does that break for you too?
[17:33] <dobey> also, in fixing the bug, what should we do when install fails exactly? just do the "OMG EXCEPTION!" thing?
[17:35] <dobey> ralsina: no, because the installer does the same as "apt-get update" before doing the install :-/
[17:35] <ralsina> dobey: and I think the "finish" of that call is what you are catching, and the you are not doing the real install
[17:35] <ralsina> dobey: at least that's why it ended in 3 seconds and never downloaded anything
[17:36] <dobey> ralsina: but *i* can't make it do that :)
[17:36] <dobey> for me it always works :-/
[17:36] <ralsina> dobey: your network is too fast
[17:36] <ralsina> dobey: race condition
[17:36] <dobey> heh
[17:37] <dobey> even at uds?
[17:37] <dobey> i guess if it was actually pulling off the local mirror, maybe
[17:37] <ralsina> dobey: could be
[17:37] <ralsina> dobey: time to try at a bad gas station or something :-)
[17:38] <dobey> though ppas don't get mirrored to the uds conference host
[17:38] <ralsina> dobey: I can give you ssh if you want to debug it remotely
[17:38] <dobey> but maybe i only tried at times when the bandwidth wasn't totally saturated by 500 nerds
[17:38] <ralsina> dobey: although not today
[17:42] <dobey> i guess i'll just write a test for the network problem case
[18:00] <ralsina> dobey: could you do whatever launchpad needs so that https://launchpad.net/ubuntuone-control-panel is not full of lies?
[18:00] <ralsina> dobey: as in, we don't want people to use version 2.0.0 anymore? ;-)
[18:02] <dobey> i can, but it's a fair bit of tedious work to do.
[18:02] <dobey> hrmm, not too bad though perhaps
[18:04] <dobey> certainly not as bad as -client
[18:05] <ralsina> dobey: is it automatable?
[18:05] <ralsina> dobey: just so that I am not burdening you with an infinite amount of future work
[18:05] <dobey> it isn't easily
[18:06] <dobey> and it's a one-time thing now
[18:06] <dobey> not an infinite amount of future work
[18:06] <ralsina> well, a bit every 6 months, right?
[18:06] <dobey> nope
[18:06] <dobey> not with the new process of making releases from stable branches only
[18:07] <dobey> we only have to fix up the ones that we weren't doing that for, before
[18:07] <ralsina> dobey: even better then
[18:07] <dobey> and any ones we screw up from now on, but that should be manageable :)
[18:08] <ralsina> dobey: well, noone complained about this, and it has been broken for about a year
[18:08] <ralsina> dobey: but I suppose noone ever actually reads launchpad project pages
[18:10] <ralsina> briancurtin: you already committed the fix for bug #954450 right?
[18:11] <briancurtin> ah, yeah, i forgot i entered that one. its covered in the installer automation branch
[18:11] <dobey> i guess win7 doesn't care about the registration key
[18:12] <briancurtin> dobey: it will eventually care in a few days. it'll ask you for the key to activate "windows genuine advantage" or something stupid
[18:13] <dobey> briancurtin: perhaps. but i don't see a way to enter it *now* if i have the key, and want to
[18:14] <ralsina> briancurtin: ok, I'll mark as fix-committed but you will have to mark it fix-released manually in the next release
[18:15] <ralsina> dobey: bug #925398 is about including changelogs on windows release. Since we have changelogs for each individual project, any ideas on how to create a global one?
[18:15] <dobey> ugh
[18:15] <dobey> also, we don't have changelogs for each individual project
[18:16] <dobey> well, outside of "bzr log" anyway
[18:17] <dobey> ralsina: u1cp page shows 3.0.0 as latest now.
[18:18] <ralsina> dobey: the branch that merges to the stable branch has a decent changelog
[18:18] <ralsina> dobey: thanks on the u1cp page fix!
[18:20] <mmcc> mandel and alecu, when you get a chance, can you look at the two merges I had that depended on mandel's unix-domain merge? ping me for the URL if you need it
[18:20] <alecu> mmcc, ping for the url :-)
[18:20] <mmcc> https://code.launchpad.net/~mikemc/ubuntu-sso-client/no-darwin-in-linuxnwindows-tests/+merge/105924 and https://code.launchpad.net/~mikemc/ubuntu-sso-client/fix-networkstate-darwin/+merge/105926
[18:20] <ralsina> briancurtin: could you put in your todo list checking if bug #921043 still happens after the Visual C++ runtime change?
[18:21] <dobey> ralsina: it's not really that decent, and only exists sometimes by virtue of multiple changes being merged in at the same time. could do some automation perhaps, but it's not an entirely trivial set of work
[18:21] <ralsina> dobey: ack
[18:21] <ralsina> dobey: so, we don't have changelogs on ubuntu either?
[18:21] <mmcc> Note - this is my first branch with new tests I wrote, so if anyone wants to look and chime in on my tests, I'd appreciate the feedback
[18:22] <briancurtin> ralsina: i'm 99% sure that doesn't happen, but i'll double check
[18:22] <ralsina> briancurtin: awesome, update accordingly :-)
[18:29] <dobey> finally found the win7 "type your key here" bit
[18:29] <gatox_mac> fsevents on mac os: FAILED (skips=2, failures=3, errors=53, successes=39)
[18:30] <gatox_mac> keep fixing :P
[18:49] <dobey> hmm, don't know how to write a test for this :-/
[19:02]  * mandel reviewing
[19:03] <mandel> mmcc, is approved, setting to be merge now
[19:04] <mandel> mmcc, well, alecu did... but is not yet merged, I'm reviewing yours now
[19:04] <mmcc> right, thanks mandel
[19:08] <mandel> mmcc, is there a bug for this guy: https://code.launchpad.net/~mikemc/ubuntu-sso-client/no-darwin-in-linuxnwindows-tests
[19:08] <mandel> ??
[19:08] <dobey> eh, i'll make changes without tests to start
[19:08] <mmcc> mandel - no
[19:09] <mmcc> mandel: should there be? it was necessary to test the other changes but seemed like it should be a different branch...
[19:09] <mandel> dobey, <sarcasm>nooooooo how can you</sarcasm>
[19:10] <mandel> mmcc, no need, just checking in case you forgot to link it :)
[19:10] <mmcc> mandel, ok cool
[19:11] <dobey> heh
[19:14] <mandel> mmcc, in lp:~mikemc/ubuntu-sso-client/no-darwin-in-linuxnwindows-tests how do you feel about doing something like IGNORED_TESTS="test_windows.py, test_darwin.py" etc and then call XVFB_CMDLINE u1trial --reactor=gi --gui -p "$QT_TESTS_PATH" -i "IGNORED_TESTS" "$MODULE"
[19:14] <mandel> mmcc, that way we just need to edit a single line for the ignore paths and not touch the command that runs the tests
[19:14] <dobey> oh, ffs
[19:15] <dobey> trial doesn't handle self.skipTest() either
[19:16] <mandel> mmcc, you could do something smart for example, have GTK_IGNORED_PATHS and PLATFORM_IGNORED_PATHS so that you can do something equally smart for the step of running the qt tests
[19:16] <mandel> dobey, look at how we skip tests in u1-dev-tools, I've had wtf already
[19:16] <dobey> IGNORE_IGNORE_HITLER
[19:16] <dobey> mandel: yeah, actually, that's pretty close to what the code in python unittest does now
[19:16] <mandel> dobey, don't say anything bad about the germans, we need their money! ;)
[19:20] <dobey> mandel: http://ignorehitler.tumblr.com/
[19:20] <mandel> lol
[19:20] <dobey> http://24.media.tumblr.com/tumblr_m3xnd4jtku1rvxreno1_1280.jpg
[19:20] <dobey> specifically that one :P
[19:25] <mandel> politically correct :)
[19:25] <mmcc> gatox, done reviewing your branch, needs fixing but should be really fast if you're OK using some code I wrote. I fixed up the badly broken code for move_to_trash() that I wrote a couple of weeks ago and sent you. Now it works and there's a link to a paste in the comment.
[19:26] <gatox> mmcc, awesome!! thx!
[19:29] <mmcc> gatox, you're welcome. glad you didn't mind the review that says "no, use my code instead" :)
[19:30] <gatox> mmcc, jeje na, no problem! better to do everything as good as we can now... or we are going to suffer in the future :P jeje
[19:30] <mmcc> dobey, thanks for Ignore Hitler. :D
[19:30] <dobey> heh
[19:30] <mandel> mmcc, next time 'no, use my code instead, puto!' gatox likes rough love
[19:30] <mandel> :P
[19:31] <gatox> mandel, jejejejee
[19:31] <mmcc> mandel, I'd try but I am constitutionally incapable of swearing on the internet.
[19:32] <mandel> mmcc, puto is not bad perse, is a weird word..
[19:33] <mmcc> anyway, mandel - I'm fine with your suggestion for an env var to simplify the run cmd. however, the recent refactoring means that the ignore paths are probably not going to change in the future, right? ie, they're all named "test_$platform.py", so the run scripts should be OK for a while... ?
[19:34] <mandel> mmcc, yes, but as gatox said, lets do it as best as we can, otherwise we have 'tech debt' although in this case is minimum
[19:34] <mandel> mmcc, gatox, some stupid song about puta and its possible meanings: http://www.youtube.com/watch?v=TFtt3Tgw8VY based on http://www.youtube.com/watch?v=MQc7Vcd5xTo (sorry, is in spanish)
[19:35]  * mmcc reading wikipedia about 'puto/a'. 
[19:35] <dobey> lol
[19:35] <mandel> dobey, ^ in case your spanish is better than I think ^
[19:35]  * gatox is afraid how mandel knows all those videos
[19:36] <mandel> gatox, is a good one :)
[19:36] <mmcc> mandel, I see what you mean about multiple meanings. Still. I have crippling paranoia that someone important and humorless from middle america will one day read my chat logs and become offended :p
[19:37] <mmcc> mandel - anyway, so do you want me to make a change to that branch and simplify the test scripts?
[19:38] <dobey> http://ubuntuone.com/6oD7q9VXAqBUnLzYfC4FRx
[19:38] <dobey> fun times
[19:39] <mmcc> dobey, what language are those menus in?
[19:39] <mandel> mmcc, yes please :)
[19:40] <dobey> mmcc: sq_AL
[19:42] <mmcc> dobey: do you speak albanian? ( yes I had to google that )
[19:43] <dobey> a bit, yes
[19:44] <mmcc> so is that to keep it fresh or to debug things? I tried setting things to spanish for a while to see if I could work that way... I couldn't :(
[19:44] <dobey> both
[19:45] <mandel> mmcc, needs fixing https://code.launchpad.net/~mikemc/ubuntu-sso-client/fix-networkstate-darwin/+merge/105926  seems like a module that should be ignored is not
[19:46] <mandel> ok, I need to have dinner, mmcc I'll re-review after dinner or I'll do it early in my morning :)
[19:46] <mmcc> mandel, that error is why that branch depends on the earlier one.
[19:47] <mandel> mmcc, oh, but they are not merged?
[19:47] <mandel> mmcc, trying again
[19:47] <mmcc> I marked the no-linuxnwindows- branch as a prereq of the fix-networkstate branch...
[19:47] <mmcc> ah, no they might not be merged. ugh
[19:48] <mandel> mmcc, can you merge and push?
[19:48] <mmcc> yes, sorry
[19:48] <mandel> I need to go I'll review in a couple of hours
[19:48] <mandel> mmcc, no worries :)
[19:48] <mmcc> ok, later
[19:49] <briancurtin> ill be back in a bit, i need to get away from this stupid computer.
[20:05] <gatox> eod here! see you tomorrow people!
[20:27] <mmcc> Wow, just got my email to ubunet-discuss from 5 hours ago
[20:30] <salgado> mmcc, yeah, just got that as well. (am also waiting for one I sent ~2h ago)
[20:30] <salgado> I'm also checking with the Launchpad folks to see if there's anything wrong there
[20:31] <mmcc> salgado: cool, thanks
[20:38] <joshuahoover> dobey: bug #978903 appears to be ready for an sru in precise...look ready to you?
[20:40] <dobey> joshuahoover: yes; need to get the installer bug fixed as well, then we should be good to do the releases/srus i think
[20:41] <joshuahoover> dobey: do you have the # for the installer bug?
[20:41] <dobey> joshuahoover: i think there are a couple more u1client bugs that probably need to go in sru as well
[20:41] <dobey> joshuahoover: bug #853060
[20:42] <joshuahoover> dobey: yeah, i was looking but having a hard time finding the other potential candidates, if you see them, please let me know...i want to make sure we get them tested asap once they're in proposed
[20:45] <dobey> joshuahoover: bug #883252 and bug #984407
[20:45] <joshuahoover> thanks
[20:46] <dobey> and bug #993574
[20:46] <joshuahoover> dobey: for 984407, that applies to ubuntu as well as windows?
[20:46] <dobey> no, but it needs to be in the 3.0.1 tarball
[20:46] <joshuahoover> ah, ok
[20:47] <dobey> we are trying to maintain consistent and equal support for all supported platforms :)
[20:48] <dobey> hrmm, i suppose i could do the SRU bits first, actually, since ubuntuone-installer isn't on windows
[21:07] <ralsina> joshuahoover, dobey; we want to do a 3.0.1 for windows too, with the "don't run as root" fix and others, too.
[21:08] <joshuahoover> ralsina: right
[21:08] <joshuahoover> i've heard that's an issue for some windows users ;)
[21:09] <dobey> ralsina: right, that's why i listed it as a thing we need to include :)
[21:12] <ralsina> joshuahoover: yes, in fact, I suspect a bunch of the "u1cp is stuck with the overlay blocking stuff" is that.
[21:27] <ralsina> joshuahoover, briancurtin: here's a strange thing https://bugs.launchpad.net/bugs/855428
[21:27] <ralsina> apparently users are running the installer and/or the updater as NOT admin
[21:27] <joshuahoover> ralsina: interesting
[21:28] <ralsina> I thought windows would ask for prov. elevation before trying to do the install!
[21:28] <joshuahoover> ralsina: uhhh...so did i
[21:28] <ralsina> joshuahoover: it used to!
[21:28] <joshuahoover> ralsina: hmmm
[21:28] <dobey> ralsina: huh? isn't that error because syncdaemon isn't running?
[21:28] <briancurtin> ralsina: it has never asked me for it, but i kind of assumed it was supposed to be that way
[21:28] <ralsina> dobey: yes, but it's not running because it's not properly installed
[21:29] <ralsina> briancurtin: so, there's some stuf to check, at least
[21:29] <briancurtin> ill see if theres anything in there that can help reproduce that get_rootdir exception because i can never get it to happen. i think i actually saw it *once* while using it but couldnt get it to happen again
[21:29] <dobey> hmm
[21:29] <ralsina> briancurtin: the get_rootdir ust means "syncdaemon is not running"
[21:30] <ralsina> briancurtin: the interesting part is why it's not running. A few of these may be our friend "don't run as root"
[21:30] <briancurtin> ralsina: ha, well then that makes sense why i cant reproduce it because im always starting it
[21:30] <ralsina> gotta EOD
[21:30] <briancurtin> yep, i would bet its probably the dont run as root
[21:30]  * ralsina uses this bug as a ninja smoke bomb and vanishes
[21:36] <dobey> well i guess we'll know soon enough
[22:13]  * dobey suspects it's too late to beg for reviews any more today
[22:13] <dobey> especially since it's alecu and ralsina's review day today
[22:15] <dobey> so on that note, i'm out. but 3 branches proposed for client stable-3-0 so we can do some releases tomorrow
[22:15] <dobey> later!
[22:20] <Monotoko> hiya guys... having a problem with Ubuntu One 3.0.0 for Windows (college issued laptop) - it fails after install with the following error: "AttributeError 'NoneType' object has no attribute 'get_rootdir' "
[22:20] <Monotoko> I've tried to google for it... however all it's come back with is the likely cause is non-ascii characters... not the case on my machine
[22:20] <Monotoko> however, my "Users" folder is on a seperate partiton to the "Windows" folder... could that have anything to do with it?
[22:21] <Monotoko> (Windows 7 64bit btw)
[22:23] <mmcc> hi Monotoko, I believe this is a known issue that's being worked on (and maybe is already solved for the next release.) briancurtin - is that right?
[22:24] <Monotoko> mmcc, thank you.. do you know what the issue is? As far as I could tell from launchpad, it's mostly non-ascii characters... my path doesn't have any
[22:24] <Monotoko> was my guess right? :D
[22:24] <briancurtin> Monotoko: if you open up task manager, do you have a ubuntuone-syncdaemon.exe process?
[22:25] <briancurtin> that's the likely cause of the get_rootdir problem
[22:25] <Monotoko> briancurtin, no I don't
[22:25] <Monotoko> (sorry - took me a moment... I haven't used Windows in years!)
[22:26] <Monotoko> I did see that... but no way to kick start the process, any ideas?
[22:27] <briancurtin> Monotoko: i think i know what the problem is going to be. let me find the file to look for, but i think you're going to see a message in the logs about not being able to run as root (admin)
[22:27] <briancurtin> i'm not on my work computer, brb with the location of that log
[22:29] <Monotoko> briancurtin, okay... I am an admin and I did try running U1 as an administrator... but it didn't seem to like that
[22:29] <Monotoko> (as in, right clicked and specifically made sure it was running as administator)
[22:30] <briancurtin> Monotoko: that's the problem - we have a check that is broken that won't let you run as admin
[22:30] <briancurtin> see if C:\Users\brian\AppData\Local\xdg\cache\ubuntuone\log\syncdaemon.log says something like "don't run syncdaemon as root" on the last line or last few lines
[22:30] <briancurtin> (replacing my name with yours)
[22:31] <Monotoko> gotcha.. let me go have a look
[22:31] <briancurtin> Monotoko: we have a fix that isn't released yet that just lets it work on admin and non-admin. the check is really more for linux, but we tried to be too fancy on windows and it bit us. it worked for me (the one who wrote it), but it doesnt work for a lot of people :/
[22:32] <Monotoko> briancurtin, confirmed the problem... yea it doesn't like being run as admin... will creating a limited user and running it under that work?
[22:33] <Monotoko> *yeah
[22:33] <briancurtin> Monotoko: i think that should work, but id hate to have you set it up and have it still be broken
[22:34] <briancurtin> Monotoko: did you just install U1 and see this problem, or were you previously using version 2.0.3?
[22:34] <Monotoko> briancurtin, I just installed it... it's a brand new machine ^^
[22:35] <briancurtin> Monotoko: i wonder if you would have better luck using the older version for now
[22:35] <Monotoko> briancurtin, maybe... where can I download it?
[22:36] <briancurtin> https://one.ubuntu.com/windows/ubuntuone-2.0.3-windows-installer.exe -- you should first stop U1, then uninstall it, then run that 2.0.3 installer
[22:36] <Monotoko> briancurtin, it's asking for a reboot.. do I need to?
[22:37] <briancurtin> Monotoko: you probably should, just to be sure all of the U1 executables are no longer in use before carrying on with the new installer
[22:38] <Monotoko> alright will do, il be right back :P
[22:47] <Monotoko> worked perfectly, thank you briancurtin :)
[22:47] <briancurtin> Monotoko: you're welcome. one thing you'll soon notice is that you're going to be prompted to upgrade to the new version, which is the one you just uninstalled
[22:48] <briancurtin> Monotoko: if you want to turn off the check for updates, at least until we put out a version that will work for you, edit C:\Program Files (x86)\ubuntuone\update.ini and there should be some option (can't remember what it's called right now) about checking for updates. it'll be set to 1, turn it to 0 to disable
[22:49] <briancurtin> Monotoko: you will have to edit that file as administrator in order to save it
[22:53] <Monotoko> briancurtin, got it... thanks ^_^
[22:59] <mmcc> ok, heading out for the day. see you all tomorrow...