[00:18] <ralsina> alecu: ?
[02:20] <alecu> ralsina, I was talking about "peer reviews"
[02:36] <leo-unglaub> hi, i just installed ubuntuone for the thunderbird contact sync and the ubuntu one icons is in the thunderbrid addressbook, but i can't create/copy entrys in there
[02:36] <leo-unglaub> is there something special todo?
[02:39] <dobey> https://lists.ubuntu.com/archives/ubuntu-desktop/2011-November/003474.html
[02:40] <dobey> i'd recommend not using the ubuntuone/couchdb integration for contacts sync in thunderbird right now, as it's not going to be supported for much longer
[02:41] <leo-unglaub> hmmmm, i know that using couchdb was one of the worst descissions of the ubuntuone team because the perfoemence sucks as hell, but is this really cancled?
[02:41] <leo-unglaub> i mean thats the only reasion i use ubntu one
[02:43] <dobey> we are building another system to replace it. it's not clear if there will be a good migration path on the client side though. i'm pretty sure that contacts sync is something we want to keep as a concept, but transition might be bumpy
[02:43] <leo-unglaub> dobey: are you one of the ubuntu one developers?
[02:43] <dobey> yes
[02:44] <leo-unglaub> ah, great...so if i finaly have one of the ubuntu one people live in the chat, maybe you can answer me this question
[02:45] <leo-unglaub> why are you trying using strange software for obuntu one if the solution is simple
[02:45] <leo-unglaub> sqlite
[02:45] <dobey> the solution isn't simple
[02:45] <leo-unglaub> good perfoemence, aproved for years
[02:46] <dobey> but the default storage backend for u1db on client side, for many of the platforms we support, will be sqlite
[02:47] <leo-unglaub> thats the great thing about sqlite...it's available everyware..
[02:47] <leo-unglaub> widows,linux,mac,android,ios
[02:48] <dobey> that doesn't mean it is the best option
[02:48] <dobey> sqlite is just an sql based data store. we still have to write a lot of code on top of it to do all the complex things we need
[02:48] <dobey> which is why we're building u1db, and nots simply just saying "we're using sqlite"
[02:49] <dobey> because it's not sqlite. the backend doesn't matter
[02:49] <dobey> it could be plain text .json files inside a directory structure if you wanted to be
[02:49] <dobey> or csv or anything else if you desire
[02:50] <leo-unglaub> if you support multiple storage backends or not is one thing, but if you write abstract controllers, it's no problem
[02:51] <leo-unglaub> anyway...back to my first question
[02:52] <dobey> u1db is an api and a schema
[02:52] <leo-unglaub> so you mean it's a bad idea use the u1 contact sync?
[02:53] <dobey> i wouldn't recommend it right now. there's a reason those packages are in universe instead of main :)
[02:55] <leo-unglaub> hmm, bad
[02:56] <leo-unglaub> contact sync is broken, tomboy notes sync broken...
[02:57] <leo-unglaub> you really make it hard for people to recomment u1 for other users :)
[02:58] <leo-unglaub> anyway dobey, thanks for the answers
[02:58] <dobey> tomboy notes sync isn't broken
[02:58] <leo-unglaub> it is :)
[02:59] <leo-unglaub> 70 of all notes is getting lost during every sync
[02:59] <dobey> have you filed a bug and/or support request?
[02:59] <leo-unglaub> sure
[02:59] <leo-unglaub> i even send a patch
[03:02] <dobey> ok
[03:03] <leo-unglaub> but i never got a response, so i wrote my own notes sync
[03:03] <leo-unglaub> and now it works :)
[03:03] <dobey> what bug #?
[03:04] <leo-unglaub> don't remember, it was in the irc
[03:04] <leo-unglaub> we had a discussion
[03:05] <leo-unglaub> i have to check the irc logs
[03:06] <dobey> well you didn't report the bug, according to launchpad :)
[03:06] <leo-unglaub> yep, definitly not in launchpad
[03:07] <dobey> then there is no bug report :)
[03:08] <leo-unglaub> maybe he dit the report himself, after we talked in the irc channel about it, don't know.
[03:10] <dobey> well, without a reference, i can't say anything about what it was. likely the issue has since been fixed. there are plenty of people syncing notes. :)
[03:10] <dobey> anyway, i need to sleep
[03:10] <leo-unglaub> :)
[03:11] <leo-unglaub> sleep well
[03:11] <dobey> thanks
[09:02] <JamesTait> Morning all! :)
[09:02] <mandel> morning all!
[11:17] <gatox> good morning!
[11:22] <mandel> gatox, morning!
[11:22] <gatox> mandel, hi!
[11:44] <nessita> hello everyone!
[11:45] <gatox> nessita, hi
[11:46] <nessita> hola gatox, how is it going?
[11:47] <gatox> nessita, fine..... i was able to submit the self performance review.... and now i'm close to fix the unicode issue (i hope :P)
[11:47] <gatox> nessita, you?
[11:48] <nessita> gatox: all good! apparently the house next door is being knock down to build a building, so there is a lot of "punching" around
[11:48] <gatox> nessita, bummer!
[12:04] <mandel> nessita, uh! that sounds terrible and is going to take a while..
[12:05] <nessita> mandel: yes... we know this was going to happen since may last year, but we were hoping it never happen
[12:06] <mandel> nessita, ouch! and you just moved there right? I mean, I remember you two move together not that long ago..
[12:06] <gatox> nessita, you can buy the headsets that block any sound!
[12:06] <nessita> mandel: we moved to this apartment april last year :-P
[12:06] <mandel> nessita, putadon!
[12:06] <nessita> gatox: and use that *to sleep*? :-P
[12:07] <gatox> nessita, ehhhhhhh i don't think they work at night
[12:07] <nessita> :'(
[12:09] <alecu> hello, all!
[12:09] <alecu> ralsina, ping
[12:09] <gatox> alecu, buenas
[12:09] <nessita> hola alecu
[12:09] <gatox> alecu, i think i found someone to bring me the ux31! \o/ (not you :P)
[12:10] <alecu> gatox, from the states?
[12:10] <alecu> gatox, stop buying stuff!!!
[12:10] <nessita> gatox: you're getting another laptop? with or without lights?
[12:11] <gatox> alecu, martin (from ninja)..... that is in usa since december, and is coming back in april :D
[12:11] <nessita> :-P
[12:11] <alecu> gatox, awesome.
[12:11] <gatox> yeyyyy
[12:12]  * alecu bought a modern videocard yesterday, and last night left downloading all the  humble bundles that 
[12:12] <alecu> *all the files from all humble bundles, I mean.
[12:12] <alecu> doh
[12:12] <alecu> too early and making mistakes already.
[12:15] <alecu> 11 Gb of Humble Bundles for Linux. Awesome.
[12:16] <gatox> wow!
[12:22] <nessita> gatox: would you be able to do a review for me? https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/always-in-.../+merge/99917
[12:22] <nessita> oh wait
[12:23] <nessita> gatox: no, no yet, I just saw I left a skip
[12:23] <gatox> ok
[12:25] <nessita> gatox: fixed and pushed to revno 303
[12:25] <gatox> nessita, ok
[12:25] <nessita> gatox: just FYI, next branch will be called "wait-for-it" :-P
[12:25] <gatox> jejejejeej
[12:32] <ralsina> good morning people!
[12:32] <ralsina> alecu: ponga!
[12:33] <gatox> ralsina, hi
[12:33] <alecu> ralsina, hey there boss
[12:33] <alecu> ralsina, I've been doing some irl on windows last night, and I'm sure that not everything is working right on syncdaemon.
[12:33] <alecu> ralsina, right now I'm doing some debugging
[12:34] <alecu> ralsina, it seems that the proxy settings are not being used in the tunnel.
[12:34] <ralsina> alecu: sigh
[12:34] <ralsina> alecu: does it still work in non-proxy situations?
[12:35] <alecu> ralsina, it does
[12:35] <ralsina> alecu: ok, let me know what you find
[12:35] <mandel> alecu, how are you getting the proxy settings
[12:35] <alecu> mandel, on windows I'm using the system proxy, so it probably it's something similar to what you've found on -sso
[12:35] <mandel> alecu, are you doing a ProxyQuery, and if you are in the proxy settings on windows, there is a tiny evil checkbox that you have to uncheck for your settings to take effect
[12:36] <alecu> mandel, how is that checkbox called?
[12:36] <mandel> alecu, let me check
[12:37] <mandel> alecu, automatic detect settings
[12:37] <mandel> alecu, that hceckbox means that the automatic settings step on the manual ones..
[12:37] <alecu> mandel, I surely don't have that enabled.
[12:37] <alecu> mandel, I know. And automatic settings use a .pac file, which is a bit of javascript.
[12:38] <nessita> ralsina: I answered your email... let me know if that's ok, I proposed the branch at https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/always-in-.../+merge/99917
[12:38] <nessita> ralsina: also, this is from yesterday: https://code.launchpad.net/~nataliabidart/ubuntuone-client/851810-notify-on-volumes/+merge/99829
[12:38] <mandel> alecu, I wasted some time because I left that checked.. /me felt stupid
[12:38] <alecu> mandel, and that javascript is executed on every request, and it decides which proxy to use. We unfortunately do not support .pac autoconf.
[12:38] <alecu> mandel, but thanks for the idea to check on that :-)
[12:40] <mandel> alecu, better than looking deeper :)
[12:43] <gatox> nessita, +1
[12:44] <nessita> ralsina: got my messages above?
[12:44] <nessita> gatox: thanks!
[12:44] <ralsina> nessita: I really dislike the smaller label text :-(
[12:45] <ralsina> nessita: can we ask for design input on that?
[12:45] <nessita> ralsina: sure, we should show the same dialog with other translations, since that's my concern
[12:45] <ralsina> nessita: right, german and dutch are usually the longest
[12:46] <ralsina> nessita: will do the other review now
[12:49] <alecu> I thought the longest were rosarigasino and jeringozo.
[12:49] <alecu> rg_AR and jg_AR
[13:01] <alecu> ralsina, one very ugly thing I found in windows is that every windows is "popped under"
[13:01] <alecu> ralsina, like, the sso dialog. Or the proxy creds dialog.
[13:02] <alecu> this is on 7.
[13:02] <ralsina> alecu: known issue, doesn't happen for normal users
[13:02] <alecu> ralsina, how comes?
[13:02] <ralsina> alecu: basically, only child processes can pop-up
[13:02] <ralsina> alecu: usually, sso is started by u1cp
[13:02] <ralsina> alecu: but when you are debugging, you have an old sso and start new u1cps
[13:02] <alecu> ralsina, awesome then.
[13:04] <alecu> ralsina, so, correct me if I'm wrong: right now the only app in the windows "autostart list" is "u1cp --icon-only" ?
[13:04] <ralsina> alecu: right
[13:04] <alecu> awesome then,.
[13:04] <ralsina> alecu: BUT maybe I remember wrong. I will have to check
[13:06] <ralsina> alecu: we are also adding syncdaemon
[13:06] <ralsina> alecu: now, the thing is, id sd is the one that start sso, we will get popunders. Damn.
[13:07] <ralsina> alecu: so we need to change it so u1cp starts sd
[13:08] <alecu> ralsina, u1cp should already be starting sd to query for its state, I think.
[13:09] <ralsina> alecu: when started as an icon, I am not 100% sure
[13:09]  * ralsina checks. Luckily this can be checked on ubuntu
[13:09] <nessita> crap, internet went down
[13:12] <mandel> ok, lunch for me
[13:15] <dobey> anyone got time to do a quick second review on https://code.launchpad.net/~dobey/ubuntuone-client/hint32/+merge/99824 ?
[13:15] <ralsina> CRAP, the command line we are putting in autostart ( --minimized --with-icon) doesn't work
[13:21] <briancurtin> ralsina: i thought i noticed that yesterday but didn't have a chance for a second look. i know it works from that command line when you click on the desktop shortcut or the menu
[13:21] <briancurtin> but i thought maybe it was my shitty VM
[13:22] <ralsina> briancurtin: it should work if it's already running :-)
[13:22] <ralsina> briancurtin: but not if it is not
[13:24] <briancurtin> ralsina: i just went from nothing running (no u1 processes in taskmgr) to control panel (with icon), which started SD, which started SSO after clicking the desktop shortcut. i dont know why that command line wouldnt work on startup though, ugh. i will look into it
[13:31] <ralsina> briancurtin: that was doing --minimized --with-icon? If starting u1cp that way start sd, then we need to remove sd from autostart
[13:32] <briancurtin> ralsina: testing it now
[13:38] <briancurtin> ralsina: i'll propose a branch in a few minutes. SD being there does cause something minor (another instance already started msg) but it's not showing up because the "--minimized" flag
[13:38] <briancurtin> that gives an AttributeError. didn't do enough IRL testing :(
[13:56]  * mandel back from lunch
[13:56] <briancurtin> ralsina: https://code.launchpad.net/~brian.curtin/ubuntuone-control-panel/fix-windows-startup/+merge/99945 -- tested by changing the autostart keys to match this behavior, then restart, then have u1cp start up fine (it will replace the autostart keys when it opens, though, so you won't "see" the change via the registry after the fact)
[13:56] <briancurtin> i could do a full blown installer run of this if you want
[13:56] <ralsina> briancurtin: could do a test by, after logging in, removing credentials, then try to login again, check you don't get a pop-under?
[13:58] <briancurtin> ralsina: trying now
[14:03] <nessita> briancurtin: did you have any luck reviewing my branch yesterda?
[14:03] <briancurtin> nessita: i didn't get to it yesterday, will do it this morning after this u1cp autostart thing is wrapped up
[14:03] <nessita> ralsina: I would advice not to remove SD from autostart
[14:04] <ralsina> nessita: if we keep it there, and SD starts SSO, we will get popunders instead of popups in u1cp
[14:04] <nessita> briancurtin, ralsina: may I ask to prioritize the review of the u1client branch for the FFe? we need that ASAP in ubuntu due to freezes
[14:04] <ralsina> nessita: yes, I am doing that
[14:04] <briancurtin> nessita: on it now
[14:05] <nessita> ralsina: can you please give me some details about that?
[14:05] <nessita> "popunders instead of popups in u1c" (was not aware)
[14:05] <ralsina> nessita: only child processes can popup windows over its parents
[14:05] <ralsina> nessita: (on windows)
[14:05] <nessita> ralsina: and doing tcp activation means generating a "child" process? :-/
[14:05] <ralsina> nessita: if a non-child process opens a window, it will be *below* u1cp. So we have to ensure all processes are started by u1cp
[14:06] <ralsina> nessita: yes
[14:06] <nessita> ralsina: ok then (but feels like puaj :-/)
[14:06] <ralsina> nessita: but if SSO is started before u1cp, then when u1cp asks sso to show credentials, they will pop below u1cp
[14:06] <nessita> I understand
[14:07] <ralsina> nessita: platform-specific weirdness
[14:07] <briancurtin> now i see why everyone hates windows
[14:07] <ralsina> nessita: which lucklily we can fix in a platform-specific way this time ;-)
[14:07] <nessita> ralsina: right
[14:07] <nessita> briancurtin: do you blame us? :-P
[14:07] <briancurtin> i was brainwashed all this time
[14:08] <nessita> briancurtin: do not wake up, since *you* are our windows guy :-P
[14:08] <briancurtin> haha
[14:08] <nessita> we can not afford converting you
[14:09] <ralsina> briancurtin: linux in the desktop is for geeks, man ;-)
[14:09] <ralsina> briancurtin: windows 8 will be awesome!
[14:09]  * ralsina goes wash his mouth with soap
[14:10] <gatox> the moment when you feel you are doing something with the code that nessita will disapprove...... is the moment when you realize that probably you are doing something nassty :P
[14:11] <nessita> I'm not sure if fell flattered or hurt :-P
[14:11] <gatox> jejee
[14:11] <gatox> flattered
[14:12] <ralsina> gatox: you need a virtual nessita plugin for ninja ide :-)
[14:12] <nessita> LOL
[14:12] <gatox> ralsina, jjejeje
[14:12] <ralsina> like the virtual RMS for packaging they have in debian
[14:12] <gatox> ralsina, actually...... perrito suggest to do something like that! jejeje like clipper but with nessita telling you: "YOU ARE DOING SOMETHING WRONG! SHAME ON YOU!"
[14:12] <gatox> jeejej
[14:13]  * nessita pings perrito666
[14:13] <ralsina> gatox: and have it popup every N minutes randomly. It's bound to be right sooner than later.
[14:13] <nessita> ralsina: you're doing something wrong, shame on you!
[14:13] <gatox> jejje
[14:14] <ralsina> nessita: +1
[14:14] <gatox> ok.... i'll try to reach to a cleaner solution about this unicode thing.... it's weird
[14:14] <ralsina> nessita: on your branch. And yes, I am :-)
[14:14] <nessita> jejeje
[14:15] <ralsina> gatox: could also shout NEEDS TEST on audio before every save of a file not called test*
[14:15] <briancurtin> nessita: i think the branch looks good, running the tests now
[14:15] <gatox> ralsina, LOL!
[14:15] <nessita> briancurtin: thanks!
[14:31] <briancurtin> mandel: do you have a few minutes to mumble?
[14:31] <mandel> briancurtin, yes!
[14:31] <mandel> briancurtin, let me launch the app :)
[14:32]  * alecu on the other hand, installed win 8 preview and thinks metro looks awesome, and very useable.
[14:33] <alecu> I don't think it will be my default at all, but it's very very nice.
[14:33] <nessita> briancurtin: sorry to bother (and to insist) but were you able to finish the review?
[14:34] <mandel> ralsina, ping!
[14:35] <mandel> ralsina, you fancy to talk with a awesome guy and briancurtin on mumble :P
[14:35] <briancurtin> nessita: the test *just* finished 10 seconds ago, and it's all good, and the review is all good. marking approved now
[14:35] <ralsina> mandel: pong
[14:35] <ralsina> mandel: what about?
[14:35] <nessita> briancurtin: sorry, and sorry for making you hating me :-)
[14:35] <mandel> ralsina, windows things and planning
[14:35] <briancurtin> ralsina: the mac/windows split we talked about
[14:35] <mandel> ralsina, should be super quicl
[14:36] <mandel> s/quicl/quick
[14:36] <ralsina> briancurtin, mandel: in 15'?
[14:36] <briancurtin> that's fine with me
[14:36] <mandel> ralsina, sure
[14:46] <ralsina> mandel, briancurtin: quick mumble now before team weekly call?
[14:46] <briancurtin> ralsina: sure, join in
[14:46] <mandel> ralsina, go!
[14:52] <nessita> will reboot before the weekly call
[14:55] <ralsina> briancurtin: did you get to check the popunders test I asked?
[14:56] <briancurtin> ralsina: sorry, got occupied with the review and then the call. i'll check it out now
[14:56] <ralsina> briancurtin: thanks
[14:59] <dobey> and now is the time on sprockets when we dance!
[14:59] <briancurtin> lol
[15:00] <ralsina> dobey, briancurtin, thisfred, alecu, nessita, gatox, mandel: to the bat-mumble!
[15:00] <mandel> na na na na na na na na naaaaa
[15:00] <dobey> ralsina: also urbanape ?
[15:01] <urbanape> me
[15:01]  * ralsina keeps dropping people
[15:01] <urbanape> omw
[15:01] <ralsina> urbanape: mumbles on thursdays
[15:01] <urbanape> yup yup
[15:01] <dobey> mandel: are you playing katamari?
[15:01] <mandel> dobey, that was the batman song!
[15:02] <mandel> dobey, maybe I miss a na or two
[15:02] <dobey> *kapow*
[15:04] <mandel> cataplan!
[15:04] <mandel> or something :P
[15:06] <briancurtin> ralsina: nope, no pop-unders
[15:24] <nessita> gatox: please note that you have this bug in High https://bugs.launchpad.net/ubuntuone-client/+bug/917222 (for when you finish what you're doing)
[15:33] <nessita> gatox: I changed the code here https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/always-in-.../+merge/99917 due to new suggestions from ralsina, would you like to confirm or deny your vote?
[15:33] <gatox> nessita, ko
[15:33] <gatox> ok
[15:35] <gatox> nessita, i can only see a string change
[15:35] <gatox> is that ok?
[15:35] <nessita> gatox: and a test was added
[15:35] <nessita> and a commnet explaining why :-)
[15:36] <gatox> nessita, seems ok for me
[15:41] <mandel> nessita, so, I move to the autoupdate!!! \o/
[15:41] <nessita> mandel: yes!
[15:41] <nessita> gatox: thanks
[15:41] <mandel> nessita, is there a bug number for that? where shall it land (project)?
[15:42] <ralsina> nessita: if that's not the sneakiest fix I have ever proposed, it's close :-)
[15:42] <nessita> mandel: not sure about bug #, let me check. Target is controlpanel.
[15:42]  * gatox quick lunch
[15:42] <ralsina> mandel: you had a branch. That branch should have a bug attached. If it doesn't: your fault ;-)
[15:43] <mandel> nessita, ralsina this bug 845659
[15:43] <ralsina> mandel: yes
[15:45] <mandel> briancurtin, lets postpose a little the 'get jenkins working' stuff until I'm done with the autoupdating and we go there at full speed
[15:45] <nessita> ralsina: would you please review, when you can https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/always-in-.../+merge/99917
[15:45] <mandel> briancurtin, do you have work on the mac side? do you need anything from me to get something started in the jenkins world?
[15:46] <ralsina> nessita: sure!
[15:48] <mandel> alecu, is there anything wrong with using a looping call for the autoupdate periodic checks?
[15:48] <briancurtin> mandel: i have nothing pending on the mac side, just helping out where needed. i think what i'll start with before even involving jenkins itself is getting Windows tests more solid - the readonly/readwrite problem being fixed is needed first, which i started looking into yesterday afternoon
[15:49] <mandel> briancurtin, superb! let me know how that goes
[15:49] <alecu> mandel, not for the periodic checks, no.
[15:49] <alecu> mandel, it was wrong to use a looping call for some places where a "callLater" made more sense.
[15:49] <ralsina> lunch time for me. Nessita, +1
[15:50] <mandel> alecu, where was that? just to understand the issue
[15:50] <alecu> nessita, your branch is named with three dots? weird!
[15:50] <alecu> mandel, I don't recall. There was a bug opened for that I believe.
[15:51] <alecu> mandel, https://bugs.launchpad.net/ubuntu-sso-client/+bug/815889
[15:51] <mandel> alecu, yes, I saw it, that is the reason I was asking you.. I remember it was either created by you or assigned to you
[15:51] <nessita> alecu: :-)
[15:51] <nessita> mandel, briancurtin: would you confirm which version of the _get_update_path is the correct? http://pastebin.ubuntu.com/905801/
[15:51] <nessita> mandel, briancurtin: do we need one call to dirname() or two?
[15:52] <mandel> nessita, two AFAIK
[15:53] <mandel> nessita, that was to fix a bug that we found on vista, apparently there was a dll that we ship with the python code that was loaded by the autoupdater and will cause vista autoupdates no to work
[15:53] <mandel> nessita, ;et me see if I can find the bug number of the vista issue
[15:53] <gatox> end of lunch
[15:53] <alecu> mandel, oh, crap. We are doing the same stupid thing in qt:
[15:54] <mandel> alecu, ein?
[15:55] <alecu> we are dancing around QTimer start and stop, and keeping a reference to it, instead of just using QTimer.singleShot() which is much simpler
[15:55] <mandel> nessita, here is the reason: 875283
[15:55] <mandel> alecu, when/where did I/we do that?
[15:55] <mandel> nessita, bug 875283
[15:56] <dobey> thisfred: care to review https://code.launchpad.net/~dobey/ubuntuone-client/hint32/+merge/99824 real quick?
[15:56] <alecu> mandel, we are doing the most complicated thing in both places. A full QTimer, and a LoopingCall instead of a callLater.
[15:57] <alecu> mandel, ubuntu_sso/main/qt.py:def timeout_func(interval, callback, *a, **kw)
[15:57] <mandel> alecu, I suppose that our brains were wired to do it that way, as long as we can change it :)
[15:57] <nessita> mandel: I'm sorry but the bug report is not answering my question
[15:57] <nessita> mandel: we have almost the same code written differently in 2 projects
[15:57] <mandel> nessita, the one with the two two basedir :)
[15:58] <nessita> mandel: ok, will that be the same case for the uninstall.exe?
[15:58] <mandel> nessita, one is buggy certainly, so I'll fix that in the branch I'm working on
[15:58] <mandel> nessita, no, uninstall was not affected by it AFAIK
[15:58] <nessita> mandel: no worries, I m fixing it
[15:58] <nessita> mandel: since I need to provide the uninstall
[15:58] <alecu> mandel, hmmm....
[15:58] <nessita> and will re use the same function
[15:58] <alecu> mandel, it seems they are being used right. It's the naming of our function that's wrong.
[15:59] <nessita> briancurtin: ping
[15:59] <briancurtin> nessita: pong
[15:59] <nessita> briancurtin: did you get the question before? I'm pinging cause I got a follow up question for you
[15:59] <mandel> nessita, ok,  so you have to make sure that the bitrock xml puts the .exe in the correct path (one level higher that the dist folder) and we should be safe
[15:59] <alecu> mandel, puaj
[16:00] <mandel> alecu, 2*puaj
[16:00] <alecu> I can't understand a thing.
[16:00] <nessita> mandel: no idea what that means :-), you should let briancurtin knows that
[16:00] <mandel> nessita, ok will do the following, ping me for the review and I'll talk with briancurtin to make a bug to work correctly in the .exe
[16:00] <nessita> mandel: ok, but I have a wider question
[16:01] <nessita> mandel: will the uninstall.exe, autoupdater.exe, syncdaemon.exe etc be all in the same dir?
[16:01] <briancurtin> nessita: i got it but i don't really know the context. the second looks more correct
[16:01] <alecu> mandel, the function is named right, but it's used to fire a different function named "shutdown"
[16:01] <alecu> sorry, the function is named wrong.
[16:02] <mandel> nessita, they should be, if they are not, is a bug in the bitrock installer
[16:02] <briancurtin> nessita: except for the hasattr usage should become getattr, but that's another story
[16:02] <alecu> it's being used to fire a function named "shutdown", that does not shutdown most of the time.
[16:02] <alecu> so, ugh.
[16:02] <nessita> briancurtin: so, let me ask differently (I know this is complex). Would you please show me the hierarchy of folders that an installed U1 in windows generates?
[16:03] <nessita> briancurtin: because, if the uninstall.exe is in the same dir as the ubuntuone-syncdaemon.exe, we have bugs in our code
[16:03] <briancurtin> nessita: sure, i will lay that out, will respond in a few mins
[16:03] <mandel> briancurtin, nessita sper quick mumble about this and I'll explain the issue?
[16:03] <briancurtin> logging in
[16:03] <urbanape> alecu: were you able to get past the errors you had building qt? I'm looking into it from the homebrew side.
[16:03] <nessita> briancurtin: as far as I understand, those exes would be in the same dir as the autoupdate-windows.exe. If that's the case, we can re-use the same "path builder" func
[16:03] <mandel> nessita, briancurtin, that way we are in the same page :)
[16:04] <ralsina> nessita: the commit message in your branch is wrong
[16:04] <nessita> mandel: can I get a summary of what you guys talk? I'm cooking right now :-/ (sorry, is 1pm)
[16:04] <nessita> ralsina: which branch? (has more than one in review(
[16:04] <nessita> ))
[16:04] <mandel> nessita, ok
[16:04] <ralsina> nessita: https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/always-in-.../+merge/99917
[16:05] <nessita> ralsina: will update
[16:06] <nessita> ralsina: fixed
[16:08] <alecu> urbanape, I was able to make it work with a fix for the brew qt recipe.
[16:09] <thisfred> dobey: on it
[16:09] <alecu> urbanape, last thing I stumbled with was trying to get buildout installed.
[16:09] <alecu> urbanape, easy_install is not finding buildout; do you remember how you installed it?
[16:11] <urbanape> I had it installed in the system Python (/usr/lib/python…), but yeah, the buildout bootstrap wasn't finding it.
[16:12] <ralsina> urbanape,alecu: bootstrap.py is not supposed to require buildout. That's the whole point of it.
[16:12] <alecu> ralsina, oh, awesome then.
[16:13] <alecu> urbanape, I'll give it another go after I finish with this windows bug.
[16:13] <dobey> thisfred: thanks
[16:14] <briancurtin> nessita: the update and uninstaller exes are in the top level of the install for privilege reasons (UAC stuff), then the U1 exes are inside the dist folder. this is how the layout looks: https://pastebin.canonical.com/63368/ - it has to be in this way especially due to how things work in vista
[16:14] <nessita> mandel: briancurtin: ack, perfect, thanks
[16:19] <dobey> ok, have an appointment and need to get lunch. bbiab
[16:28] <urbanape> alecu: what change did you make to qt.rb?
[16:29] <urbanape> hard code it as x86?
[16:31] <alecu> urbanape, it was this small fix: https://github.com/tmf/homebrew/commit/f34897104399f48b36e1e86a776c70dd2d882c5a
[16:31] <alecu> I've added it to the google doc.
[16:32] <alecu> gatox, when you want to setup your mac for u1 dev, we've got a google doc with some instructions.
[16:32] <urbanape> danke
[16:33] <gatox> alecu, awesome..... please share it.... i want to close 2 issues first, but maybe depending on the needs in that moment i'll do that
[16:33] <gatox> but then maybe*
[16:33] <alecu> Nichts zu danken
[16:57] <gatox> alecu, do you think this need tests? http://paste.ubuntu.com/905894/
[16:58] <alecu> gatox, I have not seen it, but yes.
[16:58] <gatox> ¬¬
[16:58] <gatox> alecu, can you take a look please?
[16:58] <alecu> gatox, and after seeing it, I say "yes" again :-)
[16:59] <alecu> gatox, delete the line 21 in the paste. Make a test for it, see that it breaks.
[16:59] <alecu> gatox, then add line 21. See that it was fixed.
[16:59] <alecu> gatox, done!
[16:59] <nessita> gatox, alecu: also, line 21 should not be there
[17:00] <nessita> gatox, alecu: the API of  default_folders should always return unicode
[17:00] <nessita> if it does not, that's a bug
[17:00] <nessita> gatox: so we should add tests and fix for the  default_folders API
[17:00] <gatox> nessita, do you mean i should modified the function: validate_path_for_folder ?
[17:01] <nessita> gatox: decode gives you  unicode, no?
[17:01] <gatox> nessita, yes
[17:01] <nessita> gatox: so default_folders should give you unicode as well
[17:01] <gatox> nessita, i wanted to avoid changing the api because this was working for everyone else
[17:01] <nessita> gatox: if is not giving you that, that is the bug
[17:02] <nessita> gatox: default_folders should return unicode, always
[17:02] <nessita> gatox: if it's not, then is a bug
[17:02] <alecu> gatox, so, the only line in your paste should be line 8+9?
[17:02] <nessita> food is ready, need to step aside for some minutes
[17:02] <alecu> hmmm... foooood
[17:03] <gatox> nessita, alecu, ok...... i'll check volumes_info how it's not returning unicode.......
[17:03] <alecu> gatox, awesome.
[17:09] <nessita> gatox: is not volumes_info, no?
[17:10] <nessita> gatox: isn't the method you need to check  default_folders?
[17:10] <gatox> nessita, i'll need to check which one.... folders is getting the info from volumes_info.......
[17:10] <gatox> i need to see if volumes_info is handling unicode there or the problem came from somewhere else
[17:18] <gatox> nessita, ahhhh you were right..... i was analyzing another thing.... sorry
[17:18] <gatox> default_folders
[17:19] <gatox> and the problem was pretty obvious
[17:23] <ralsina> briancurtin, nessita, gatox: small branch -- https://code.launchpad.net/~ralsina/ubuntuone-control-panel/fix-968369/+merge/99982
[17:23] <gatox> ralsina, ack
[17:23] <briancurtin> ralsina: ack, will review and test now
[17:24] <ralsina> gatox, briancurtin: thanks
[17:30] <gatox> ralsina, +1
[17:32] <nessita> ralsina, briancurtin2: when you can, would you review the counterpart of the bug #851810 for controlpanel? https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/wait-for-it/+merge/99983
[17:32] <briancurtin2> nessita: yep, will look next
[17:32] <nessita> ralsina, briancurtin2: please note that this require latest-latest nightlies on linux to work (so syncdaemon actually has the newer API)
[17:33] <nessita> in windows, running syncdaemon from trunk will be enough
[17:35] <ralsina> nessita: ack, will start the review in 30' once updates for today are installed
[17:35] <nessita> ralsina: thanks
[17:50] <mandel> EOD here, see you all tom!
[17:51] <ralsina> bye mandel!
[17:59] <nessita> ralsina: you on windows atm?
[17:59] <ralsina> nessita: not yet
[17:59] <ralsina> nessita: need windows tests, I can switch
[17:59] <nessita> ralsina: I do, was considering bugging you or briancurtin2, whoever I interrupt the less
[18:00] <ralsina> I think him,  I am in the middle of your review
[18:00] <nessita> ralsina: I can't acces my Vm right now, external drive will not mount
[18:00] <briancurtin2> nessita: i can do something, i'm currently running your "wait-for-it" tests
[18:00] <briancurtin2> and doing the review
[18:00] <nessita> briancurtin2: nice, branch is:
[18:01] <nessita> lp:~nataliabidart/ubuntuone-control-panel/uninstall
[18:01] <briancurtin2> nessita: does that depend on anything else i'm reviewing for you, or is fine just by itself?
[18:01] <nessita> briancurtin2: "should" be fine in its own :-D
[18:01] <nessita> let me know if that's not the case
[18:02] <briancurtin2> nessita: cool, will look in a few minutes
[18:02] <nessita> I have this feeling I'm opening this usd drive in the wrong way :-/
[18:07] <ralsina> nessita: opening as in duble clicking or opening as in sticking a knife and popping the cover?
[18:08] <nessita> ralsina: the second, and it was a scissor instead of a knife (kitchen is too far away)
[18:08] <nessita> USD drive opened, still not mounting
[18:08] <ralsina> nessita: jejeje I have a butter knife reserved for delicate electronics work
[18:09] <nessita> and can't attach to the SATA cable either, since the USB artifact is "soldado"
[18:10] <alecu> nessita, what happened?
[18:11] <gatox> nessita, alecu review please: (turns out to be really small) https://code.launchpad.net/~diegosarmentero/ubuntuone-control-panel/unicode-syncing-error/+merge/99990
[18:11] <nessita> alecu: the USB drive is not mounting
[18:11] <ralsina> nessita: looks like it's an ex-drive. It's deceased. It's not pining for the fjords.
[18:11] <alecu> nessita, is dmesg showing anything? or lsusb?
[18:11] <nessita> gatox: did you confirm the liux side behaves properly?
[18:11] <gatox> nessita, yes
[18:11] <nessita> gatox: both implementation should return unicode :-)
[18:12] <nessita> alecu: dmesg says [11540.941217] hub 1-1:1.0: unable to enumerate USB device on port 5
[18:12] <nessita> ralsina: but badblocks give nothing wrong with the disk
[18:12] <gatox> nessita, yes
[18:12] <nessita> ralsina: my guess is that the USB-adaptor-plaque is crazy
[18:12] <nessita> but the disk itself may be sill good
[18:12] <ralsina> nessita: ack, maybe careful unwelding can remove it
[18:12] <alecu> nessita, :-(
[18:13] <nessita> I have all my (more than 10) VMs there :-/
[18:13] <ralsina> ugh
[18:13] <nessita> gatox: that needs tests on the windows side!!! :-D
[18:13] <nessita> gatox: to ensure we return unicode
[18:13] <gatox> nessita, eh??
[18:13] <ralsina> nessita: wait for it... wait for it.... +1 on wait for it!
[18:14] <nessita> gatox: ah, sorry, LP diff did not showed the change in the test file
[18:14] <nessita> now is there
[18:14] <gatox> nessita, is that ok?
[18:14] <gatox> nessita, you scare me
[18:14] <nessita> yeah
[18:14] <nessita> gatox: sorry
[18:14] <ralsina> whoa merge proposal 99990!
[18:14] <nessita> blame the USB drive (?)
[18:14] <gatox> ralsina, did i win anything?
[18:15] <ralsina> gatox: yes, 2 reviews!
[18:15] <gatox> ralsina, awesome!! just what i wanted!
[18:15] <gatox> and for desert..... OH! LOOK AT THAT! another unicode issue :P
[18:15] <ralsina> gatox: that was second prize, and you won it too!
[18:15] <gatox> jejejeeje
[18:16] <gatox> i'm not feeling really luck
[18:16] <gatox> lucky
[18:16] <gatox> jeje
[18:18] <briancurtin2> nessita: i'm guessing "exceptions.AttributeError: 'FakedSyncDaemonTool' object has no attribute 'connect_signal'" -- is because the ubuntuone-client isn't matching? i just did a whole environment clean from trunk
[18:18] <ralsina> gatox: just don't ask for the third prize.
[18:19] <gatox> ok
[18:19] <gatox> :P
[18:19] <briancurtin2> ralsina: i hope that is a reference to Glengarry Glen Ross
[18:19] <briancurtin2> (great movie)
[18:20] <nessita> briancurtin2: hum, no, that's a bug in the tests, out fake should provide that. Have a trace handy?
[18:20] <briancurtin2> nessita: https://pastebin.canonical.com/63391/
[18:20] <nessita> briancurtin2: thanks, will get back to you on this
[18:23] <nessita> briancurtin2: reproduced on linux, thanks!!! (we had an skip on linux that was preventing this to explode)
[18:24] <dobey> hrmm
[18:29] <nessita> briancurtin2: fixed and pushed to revno 304
[18:34] <gatox> brb...... need to reboot
[18:40] <ralsina> briancurtin2: indeed
[18:40] <ralsina> briancurtin2: big mamet fan, am I
[18:41] <ralsina> briancurtin2, gatox: here is the 3rd prize thing http://www.youtube.com/watch?v=wVQPY4LlbJ4
[18:42] <gatox> ralsina, ok..... thats' not good news.....
[18:42] <ralsina> gatox: just kidding
[18:42] <ralsina> gatox: and you are not getting a cadillac eldorado either. Or a set of steak knifes.
[18:42] <gatox> :P
[18:43] <briancurtin2> feature request: if you break the jenkins windows build, the "coffee is for closers only" clip plays on your screen
[18:46] <ralsina> briancurtin2: hahaha
[18:47] <briancurtin2> nessita: your "uninstall" branch has some failures, looking into it
[18:48] <nessita> briancurtin2: just share the pastebin with the errors, I will fix
[18:50] <briancurtin2> nessita: https://pastebin.canonical.com/63396/
[18:51] <nessita> briancurtin2: I can fix all but this one ubuntuone.controlpanel.utils.tests.test_windows.AutostartTestCase.test_add_syncdae mon_to_autostart (is not "mine")
[18:51] <nessita> ralsina: perhapos your latest branch has something to do with that? ^
[18:51] <ralsina> nessita: looking...
[18:52] <ralsina> nessita: doesn't seem likely
[18:53] <briancurtin2> hm, his branch didn't change that
[18:53] <nessita> briancurtin2: are tests passing in trunk OK?
[18:53] <briancurtin2> nessita: yep
[18:53] <nessita> briancurtin2: then will re-review my branch
[18:53] <ralsina> nessita: no, there was a branch by brin I think that added --with-icon in that call and the tests probably never ran on windows
[18:53] <briancurtin2> or, well, let me check. i don't know if something landed recently which might have undone that --with-icon change
[18:54] <ralsina> briancurtin2: you added or removed something there, right?
[18:54]  * ralsina is fuzzy on details
[18:54] <briancurtin2> i added --with-icon yesterday, i think
[18:54] <ralsina> briancurtin2: did you merge trunk with nessita's branch?
[18:54] <briancurtin2> ralsina: ah, i just pulled nessita's branch rather than pull trunk and then merge hers
[18:54] <briancurtin2> maybe that's it
[18:55] <ralsina> briancurtin2: I think so yes, nessita is probably starting from an older trunk or something
[18:55] <briancurtin2> yeah, and the test got left out of trunk but i added in later, so that's probably it
[18:55] <briancurtin2> i'll re-run right now
[18:57] <dobey> oh i know what we need to do
[19:03] <briancurtin2> nessita: ok, here's a couple of failures after i did a better setup: https://pastebin.canonical.com/63401/
[19:03] <nessita> briancurtin2: will fix!
[19:05] <alecu> gatox, nessita: did you guys had to add a reactor to sso on windows, because the keyring methods needed it?
[19:05] <gatox> alecu, yes
[19:05] <nessita> alecu: and because of client side IPC activation
[19:06] <alecu> gatox, crap. That was what was breaking the tunnel.
[19:06] <gatox> :S
[19:06] <alecu> nessita, gatox: thanks./
[19:06] <nessita> alecu: bah, no activation, but PC communication
[19:06] <nessita> alecu: the UI needs to communicate with the sso backend using PB, which requires a reactor on each end
[19:06] <nessita> alecu, gatox: the UI never access the keyring, FYI, it just access the backend
[19:07] <alecu> FML. There's a lot of defertoThread in windows keyring.
[19:07] <alecu> frakkkkkkkk
[19:07] <nessita> so, strictly speaking, correct answer to initial alecu's question is "no"
[19:08] <ralsina> alecu: we could block on the keyring on windows since it can't be locked, should always return "fast" (briancurtin2, am I lying?)
[19:08] <alecu> ralsina, that sounds tempting, because otherwise I would have to use the qtreactor in the tunnel too.
[19:09] <alecu> ralsina, (at least while on windows)
[19:10] <alecu> We need a "from ubuntu_sso.platform import deferToThread", that uses either twisted, or qt or glib.
[19:10] <briancurtin2> ralsina: not too familiar with keyring, not sure
[19:11] <alecu> anyway, it seems a much smaller fix to add the qtreactor to the tunnel than changing it to use the keyring directly.
[19:11] <gatox> nessita, ralsina another review please: https://code.launchpad.net/~diegosarmentero/ubuntuone-client/unicode-info/+merge/100007
[19:11] <gatox> really small
[19:11] <ralsina> gatox: sure. Oh, you missed the 100000 by so close on both sides ;-)
[19:11] <gatox> ohhhhh
[19:11] <alecu> ok, kindertime.
[19:12] <ralsina> alecu: hey, me too! :-)
[19:13]  * alecu_kindertime will be back in a while. And very likely in a nicer mood.
[19:14] <nessita> briancurtin2: any chance you re-run the suite for lp: ~nataliabidart/ubuntuone-control-panel/uninstall with revno 306
[19:14] <briancurtin2> nessita: yep, will do
[19:15] <nessita> thanks!
[19:15] <nessita> briancurtin2: do no get scared by the last commit message, please
[19:15] <nessita> :-P
[19:22] <briancurtin2> nessita: one failure but it's an unclean reactor in another area, so your changes are fine
[19:23] <nessita> briancurtin2: my idea is to propose for merge, and I would need a IRL test to confirm the app is uninstalled when not accepting the license. ANy chance you can do that review? (today/tomorrow before noon is fine)
[19:30] <briancurtin> nessita: sorry, got disconnected. yep, i can check that out and do a full installer test if you want
[19:31] <nessita> briancurtin: that would be the best
[19:33] <nessita> briancurtin: merge proposal is https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/uninstall/+merge/100012
[19:38] <briancurtin> nessita: ack, looking
[19:46] <ralsina> briancurtin: about fix-windows-startup... we do want it to start minimized, since we don't want to spam people's desktops every login
[19:46] <ralsina> briancurtin: but I am guessing that makes syncdaemon not start, right?
[19:46] <briancurtin> ralsina: yeah i misunderstood that issue at first. i took the proposal back and abandoned the branch
[19:46] <ralsina> briancurtin: ok then
[19:47] <ralsina> briancurtin: launchpad mail seems to be lagging quite a bit
[19:49] <ralsina> gatox: about unicode-info ... unicode(somebytes) ? really? That only works for ascii paths
[19:49] <gatox> ralsina, i'm changing that
[19:49] <ralsina> gatox: I would have expected somebytes.decode(someencoding)
[19:49] <ralsina> gatox: ok!
[19:49] <gatox> ralsina, i'm using sys.getfilesystemencoding
[19:50] <gatox> but that works for linux..... i'm testing on windows
[19:51] <ralsina> gatox: on windows paths should already be unicode by default
[19:52] <ralsina> gatox: but who knows
[19:52] <gatox> ralsina, usually..... with getfilesystemencoding you get mbcs..... that is not fully unicode
[19:59] <nessita> ralsina: that depends on the optparse implementation, no?
[19:59] <nessita> ralsina: if optparse always gives bytes, I guess we should expect bytes in windows, no?
[20:05] <ralsina> nessita: right, arguments
[20:05] <ralsina> nessita: I would not bet one way or the other, really :-/
[20:06] <nessita> me neither
[20:17] <briancurtin> nessita: IRL the uninstall branch doesn't work :/ https://pastebin.canonical.com/63408/ happened after hitting "yes i want to cancel"
[20:17] <nessita> briancurtin: looking
[20:17] <nessita> briancurtin: how can I fix? the code to uninstall is the same from the windows-installer
[20:18] <nessita> ralsina: would you know? ^
[20:18] <nessita> briancurtin: are you an admin?
[20:18] <ralsina> looking...
[20:18] <briancurtin> nessita: i'll look into it
[20:18] <nessita> briancurtin: perhaps admin rights are needed... not sure honestly :-/
[20:18] <briancurtin> nessita: i'm on an account that has admin rights, yes
[20:18] <nessita> hum
[20:18] <nessita> weird
[20:18] <ralsina> briancurtin: the uninstaller has the "needs run as admin" bit
[20:19] <briancurtin> ah
[20:19] <ralsina> briancurtin: can you confirm that the uninstaller path is correct? and that it works from the command line?
[20:19] <nessita> ralsina: the code is:
[20:19] <nessita>         win32api.ShellExecute(None, 'runas', uninstall_path,
[20:19] <nessita>                               '--mode win32', '', 0)
[20:19] <nessita> ralsina: and the path exists, since the call     uninstall_path = get_exe_path(exe_name=UNINSTALL_EXE_NAME) will return None if the path does not exist
[20:20] <ralsina> gatox did that ^
[20:20] <nessita> ralsina: the whole method is:https://pastebin.canonical.com/63409/
[20:20] <briancurtin> i just ran "uninstall.exe --mode win32" and it uninstalled properly, so something isn't hooked up properly
[20:21] <gatox> yes...... i did that
[20:23] <ralsina> may have to put "" in the string with the path
[20:26] <briancurtin> get_exe_path has to change for uninstall.exe - it's up one directory from sys.executable
[20:27] <briancurtin> hm
[20:28] <nessita> briancurtin: heh
[20:28] <nessita> briancurtin: you said otherwise, I think :-)
[20:28] <nessita> you said both autoupdater and install were 2 up from sys.executable
[20:28] <nessita> no?
[20:28] <briancurtin> one up
[20:28] <gatox> i'm having some kind of problem with the reactor trying to run u1sdtool on windows
[20:29] <briancurtin> C:\program files\ubuntuone\uninstall.exe and then theres things like C:\program files\ubuntuone\dist\ubuntuone-syncdaemon.exe
[20:29] <nessita> briancurtin: did you have the talk in mumble with mandel abuot that?
[20:29] <briancurtin> nessita: yes, that's what i sent you along with the diagram of the directory
[20:29] <nessita> briancurtin: yeah... so, hum, I understand you wrong?
[20:29] <nessita> briancurtin: you said https://pastebin.canonical.com/63368/
[20:30] <nessita> briancurtin: is control panel at the same level as syncdaemon.exe?
[20:30] <briancurtin> yes, all U1 executables are within dist, then uninstaller and autoupdate are up one directory in C:\Program Files\ubuntuone
[20:30] <ralsina> nessita: : that looks like what briancurtin just said
[20:31] <briancurtin> this function seems to give the right path when i hardcoded the u1cp path as exec_path
[20:31] <briancurtin> i'll step through and see where it's going wrong
[20:33] <nessita> briancurtin: the thing is that get_exe_path will give you None if the path does not exist, and the uninstall_application has the guard:
[20:33] <nessita>     125     uninstall_path = get_exe_path(exe_name=UNINSTALL_EXE_NAME)
[20:33] <nessita>     126     if uninstall_path is not None:
[20:33] <nessita>     127         win32api.ShellExecute(None, 'runas', uninstall_path,
[20:33] <nessita>     128                               '--mode win32', '', 0)
[20:33] <nessita> briancurtin: so there is an uninstall.exe 2 directories up from the syncdaemon.exe
[20:33] <briancurtin> one
[20:33] <nessita> briancurtin: the code you're running goes 2 up
[20:34]  * nessita confirms
[20:35] <nessita> ok, is 2 calls to dirname, 1 dir up
[20:35] <nessita> briancurtin: so you are right, 1 dir up
[20:35] <briancurtin> which is correct, that's where it's at
[20:35] <nessita> briancurtin: right
[20:36] <nessita> briancurtin: so, back to where we started, no idea why is failing? :-)
[20:36] <briancurtin> and if i can just run "uninstall.exe --mode win32" fine, i don't get it either
[20:36] <nessita> briancurtin: without using the absolute path?
[20:36] <briancurtin> let me try not using win32api.ShellExecute, just putting in a temporary subprocess.Popen or something
[20:36] <nessita> ok
[20:36] <gatox> u1sdtool is  working for you in windows??..... can i execute it from sources? or i need to create the .exe or something like that?
[20:37] <gatox> i get "the system cannot find the file specified" even from trunk
[20:38] <gatox> usiing it with --info
[20:38] <nessita> gatox: but u1sdtool works?
[20:38] <nessita> u1sdtool -s works?
[20:38] <ralsina> Ok, got visits, will have to take a break and do some late night work as usual
[20:38] <ralsina> So, mail me requests
[20:38] <gatox> nessita, u1sdtool get executed..... execute the "main" function, but always fails on: yield run(options, sync_daemon_tool, out)
[20:39] <gatox> and the exception is executed
[20:39] <nessita> gatox: how are you running it?
[20:39] <gatox> ralsina, ack
[20:39] <nessita> gatox: and, is syncdaemon running in another terminal?
[20:41] <gatox> nessita, yes....... and no...... i tested both ways..... u1sdtool open a syncdaemon instance if it's not running
[20:41] <nessita> (05:39:09 PM) nessita: gatox: how are you running it?
[20:41] <gatox> nessita, python bin\u1sdtool --info "%USERPROFILE%\Ubuntu One"
[20:41] <gatox> and different variations of the path
[20:41] <nessita> gatox: pythonpath is set?
[20:41] <gatox> nessita, yes
[20:41] <nessita> u1sdtool -s what gives you?
[20:42] <gatox> that works
[20:42] <nessita> gatox: ok, then --info needs fixing in windows
[20:43] <nessita> I don t think anyone used it before
[20:43] <gatox> nessita, ok...... i'll fix that in this branch
[20:43] <gatox> but it's a really weirdddd crash
[20:46] <gatox> eod here...... i'll keep fixing this tomorrow
[20:46] <gatox> bye!
[20:58] <briancurtin> nessita: Popen([uninstall_path, "--mode", "win32"])
[20:58] <briancurtin> replace that with the win32api.ShellExecute
[20:58] <briancurtin> and it works :)
[20:59] <briancurtin> (er, other way around: replace shellexecute with Popen)
[20:59] <nessita> briancurtin: you sure?
[20:59] <briancurtin> i just did it
[20:59] <nessita> briancurtin: why were we using win32api.ShellExecute before then?
[21:00] <briancurtin> i have an installer built with that and it just properly uninstalled after not agreeing to the license
[21:00] <briancurtin> nessita: no idea. i'm guessing it was done that way in the past because maybe the process required admin rights before?
[21:01] <nessita> briancurtin: and why it does not require admin now?
[21:01] <nessita> briancurtin: will the popen fail for non admin users?
[21:01] <briancurtin> nessita: let me try it on another user account and see what it does
[21:02] <nessita> briancurtin: you know I know nothing about windows, so ultimately I will trust you, but replacing something that was (supposedly) working with other thing just because it works for you makes me doubtful (not personal, of course)
[21:02] <nessita> briancurtin: I know it does not work for you right now, but perhaps we're missing some trivial thing
[21:02] <briancurtin> nessita: understood on that, but i won't say i understand why it is that these things are working in this way. it seems weird
[21:23] <briancurtin> ahhhhh wtf how is this installer finding creds, ugh
[21:27] <nessita> briancurtin, ralsina: would you please review this trivialish? https://code.launchpad.net/~nataliabidart/ubuntuone-client/valid-typo/+merge/100039
[21:28] <nessita> briancurtin: and yes, I remembered your mom on that one :-P
[21:28] <ralsina> nessita: +1
[21:28] <nessita> not because of the typo, but because of the missing test
[21:30] <briancurtin> nessita: approved visually but i can't test at the moment
[21:30] <nessita> briancurtin: no problem, thanks!
[21:31] <briancurtin> its a good thing this XP is a VM otherwise i would throw it out the window, so instead i will have to pretend
[21:41] <briancurtin> is there any secret place creds are stored on windows because there is no keyring on this machine, no registry keys, and the installer doesn't find creds when running as admin...but suddenly when i run as a non-admin user it just logs me in
[21:41] <briancurtin> this makes absolutely no sense
[21:42] <nessita> briancurtin: gah
[21:42] <briancurtin> ah, figured it out
[21:42] <nessita> briancurtin: no idea about that :-/ ... how are
[21:42] <nessita> what was it?
[21:43] <briancurtin> on a non-admin account, the creds are stored in someting like "C:\Documents and Settings\limited\Application Data\Identity"
[21:43] <briancurtin> but in admin they seem to show up in the registry
[21:44] <nessita> briancurtin: oh really? :-.
[21:44] <nessita> that's... odd
[21:44] <briancurtin> nessita: probably because non-admins have less access to registry stuff, i guess. i havent looked into how the keyring stuff is implemented (yet)
[21:45] <nessita> briancurtin: good luck :-D
[21:45] <nessita> anyways, I'm eoding now
[21:45] <nessita> briancurtin: if you think I should do some changes to the uninstall branch, let's talk about that tomorrow, yes?
[21:46] <nessita> briancurtin: I would like to run by ralsina and mandel any change in that front (ie use popen instead of what we currently have)
[21:46] <briancurtin> nessita: sounds good. i will try a few more cases here and see if i can figure out why Popen works but not ShellExecute. if i dont figure it out, i'll get started again first thing tomorrow
[21:46] <nessita> briancurtin: thanks a lot
[21:46] <nessita> bye!
[21:46] <briancurtin> bye nessita
[21:55] <briancurtin> ha, that identify folder wasn't from U1 i guess because it still works with ghost credentials or something...
[21:59] <alecu> this is *just*. awesome.
[22:00] <alecu> ubuntuone.logger imports platform.xdg_base_directory, which imports platform.windows, which imports the filesystem_notifications, which in turn import the reactor.
[22:01] <alecu> so, when I try to import the logger from the non-reactored tunnel, it ends up importing a reactor nonetheless.
[22:01] <alecu> just frakkin awesome.
[22:01]  * alecu takes a long walk.
[22:02] <briancurtin> im also ready for a good long walk, and/or a few drinks
[22:13] <dobey> later all