[00:01] mandel, I think that with your fix we could remove the references to WebClient.proxy_instance [00:02] alecu, I honestly do not have the balls to make that change at 1 am when it works :P [00:02] :-) [00:02] alecu, sounds like an improvement we can do later hehe [00:02] I thought you were much braver :-) [00:03] alecu, you mean a lot more stupid, which I am, but I fail under peer pressure :P [00:03] good point :-) [00:04] alecu, if we get this on beta 2 we will have enough user to be sure it works, later we need to remove that instance thing and test it on windows etc.. [00:06] mandel, approved! [00:06] alecu, toma! [00:06] ralsina, ping! [00:10] alecu, lol: 'and code looks as good as if I had written half of it.' [00:10] alecu, there are a number of files around in the project that can match that, like windows file events [00:11] alecu: pong [00:12] ralsina, review review review: https://code.launchpad.net/~mandel/ubuntu-sso-client/fix-alecus-mess/+merge/98308 [00:12] mandel: hehe, ok, on it [00:13] BTW: sorry I disappeared, but had 2 doctor's appointments and could not get wifi in between [00:13] ralsina, I added the QApplication 'cause I saw it using libsoup.. just in case :) [00:13] mandel: haha [00:13] mandel: NEEDSFIXING! [00:13] mandel: nah [00:13] ralsina, no me undas!!! [00:14] mandel: line 27, str(domain)? [00:14] if domain comes from the cli, it could be unicode (but then the user is evil) [00:15] +1 [00:15] So, alecu, mandel, what did I miss? [00:16] ralsina, hm.. good point, but yes, he is evil setting a proxy with unicode [00:17] mandel: and it's going to be *really* obvious in the logs [00:17] ralsina, I was not paying much attention to irc to be hones, I was a little blocked trying to get that working with a factory.. this area of qt is buggy, and the bug is 3 years old O_o [00:18] mandel: ok, things happen [00:18] ralsina, I'm off to bed, catch you tom :) [00:18] mandel, alecu: did you guys find any clues about creds-qt on windows? [00:19] ralsina, will this land in the beta 2? [00:19] ralsina, I can take a look at that early in the morning tom :) [00:19] bye mandel [00:19] ralsina, no: no clues [00:20] alecu: ok [00:20] hahaha must be annoying to google a problem and get a page in your manager's blog, and he's not even around [00:21] mandel: we are probably going to do a release on windows that doesn't match exactly with ubuntu's [00:21] because we don't have time to merge the windows fixes before the ubuntu release [00:22] mandel: if you can take a look, and have nothing else, I would really really appreciate it [00:22] ralsina, ok, I'll take a look to see what is going on, can you send me an email with how I have to merge the branches for windows? [00:25] mandel: already did it yesterday ;-) [00:25] mandel: will resent just for you [00:26] ralsina, hmm if you sent it I have it.. I was updating my email client and there are some filter rules missing.. [00:26] mandel: I will resend just in case [00:26] ralsina, we get a ridicoulus about of mail per day.. [00:27] mandel: yes we do, I get 3 times what you get, too ;-) [00:27] mandel: resent [00:27] ralsina, which is stupid.. important ones get lost in the noise.. [00:28] ralsina, I'll take a look at that bug tom morning as well as getting jenkins set up to scream at us when we brake windows [00:28] ralsina, but with no emails, irc or twitter or something funny :P [00:28] mandel: cool, thanks [00:28] ralsina, like a chuck norris fact when you brake one of the platforms hehe [00:29] mandel, that already exists! [00:29] https://wiki.jenkins-ci.org/display/JENKINS/ChuckNorris+Plugin [00:29] I think I am going to be sane and just do one more review then rest [00:30] instead of the original plan of banging my head against windows for 3-4 hours [00:30] ralsina, lol [00:33] mandel, alecu: if this means proxy is done, great job, guys [00:33] * ralsina sees he wrote "done" and laughs histerically [00:33] let's say "shipped" ;-) [00:33] ralsina, there are some corners to make smoother, and the credits goes to alecu [00:34] ralsina, while I'll take the money :P [00:34] mandel: go to sleep! [00:34] * alecu should stop reading this article and get some food and sleep too. [00:39] ralsina, alecu I think I lost it: http://www.youtube.com/watch?v=bLr4pCg4PNk [00:39] lol, I'm going to bed :P [00:40] mandel: yes please ;-) [00:48] alecu, we should get a tecno viking plug in, is scarier that chuck :P [00:48] ok, of to bed :) [00:48] see you all tom! [00:49] bye mandel! [00:57] annnnnnnd I am gone too, will be working early tomorrow. [01:03] bye too! === JanC_ is now known as JanC [09:21] morning all! [09:32] Good morning all! :D [09:47] Good morning! :) === jml` is now known as jml === mpt_ is now known as mpt [11:14] good morning! [11:16] gatox, morning! [11:17] mandel, hi! how are you? [11:17] gatox, did you watch the season finale then? (Walking dead) [11:17] mandel, yes... i said EOD and run to watch the episode! :P [11:17] gatox, tired.. ended up working 'til late yesterday and now sorting out the windows code to make sure proxy works there :) [11:18] gatox, it is no an ending.. I cannot wait for season 3! [11:18] mandel, me too........ [11:18] mandel, yes, i saw your twit about the review last night [11:19] gatox, hehe [11:20] gatox, have you watch the show in english? [11:20] mandel, yes, of course, i download it in english and 720 :D [11:21] gatox, then be ready for the wtf: https://www.youtube.com/watch?v=ZblD1BNudAk [11:21] gatox, I cannot believe his is from... I'll let you watch it :P [11:22] mandel, let me see [11:36] buenos días! [11:37] nessita, buenas [11:37] hola gatox [11:37] nessita, buenos dias! [11:37] hola mandel! [11:56] ralsina: when you can, could you please fix the conflict in https://code.launchpad.net/~ralsina/ubuntuone-control-panel/tweaks/+merge/97244 ? [12:41] nessita: did you see my comment re: string formattun [12:41] formatting, even [12:41] thisfred: I did, I was not aware that .format is much slower [12:41] thisfred: have a link I can read? :-) [12:42] good morning! [12:42] nessita: it was a thread recently on python-dev, let me look [12:42] nessita: of course! [12:42] ralsina: thanks [12:42] thisfred: *how* slow are we talking here? ;-) [12:43] nessita: the more imortant point was it ties us to a recent python, thou [12:43] ralsina: in practice I'm sure it won't matter [12:43] thisfred: and how "recent" is that? [12:43] thisfred: I guessed so [12:43] ralsina: 2 weeks or so? [12:43] thisfred: no, I meant how recent a python :-) [12:44] ah, I think it may be 2.7 [12:44] mandel, urbanape: how nice would it be to have this? http://blog.rburchell.com/2012/03/qt-51-aka-when-qfilesystemwatcher-might.html [12:44] thisfred: yes, I also read that, but I'm not concerned about that... it ties us to 2.6, right? [12:44] ralsina: also since the old formatting is never going away, and already used everywhere [12:45] thisfred: yeah. But the new formatting is much less error prone [12:45] ralsina, really nice! the actual qfilesystemwatcher of qt sucks right now [12:45] thisfred: although since we d have the plan to go back in time up t maverick (hopefully) we should not do stuff that's not available in maverick's python [12:45] gatox: yes it does [12:46] ralsina: also we make it harder than necessary to port to legacy distros ; [12:46] ) [12:47] ralsina, thisfred .format is available since 2.6 ;-) [12:47] and lucid has 2.6, so no worries on that front [12:47] ralsina, thisfred: http://docs.python.org/library/stdtypes.html#str.format [12:47] yes 2.6 is good enough [12:47] nessita: yep 2.6, so not so big a pro [12:47] blem [12:48] ralsina, looking [12:48] with legacy distros I was referring to red hat which I think is still stuck on 2.5 [12:48] not enough of an incentive maybe [12:48] thisfred: let me dig deep in my sould to see how much I care... [12:48] I'm sure we already require 2.6 for other stuff [12:49] thisfred: I know it's there somewhere [12:49] heh [12:49] also, we'll probably be going to 3 soon [12:49] ralsina, it states the problem we already have with move on linux.. is a delete created if you are looking at the creating dir.. [12:49] thisfred: fedora is 2.7, RHEL is so not our target [12:49] * mandel hates file systems.. [12:50] nessita: ok, so critique retracted ;) but then I approved anyway [12:50] thisfred: thanks :-) [12:51] ralsina, mandel: I still think Apple made the right choice with iCloud. But it's orthogonal to a model like Dropbox. [12:52] Simpler, but less room for fuckupery on the part of the filesystem [12:52] urbanape: since they own the freaking FS they have little excuse [12:52] urbanape: but yes, there is a diminishing returns thing on how much effort you can put there [12:53] There's also the belief that hierarchical approach is outdated/doesn't fit the users' minds so much. [12:53] It would be hard to get Ubuntu apps to play along without a (more?) common underlying framework behind the apps, too. [12:54] urbanape: yep [12:55] ralsina, urbanape I wonder what I they using instead of ReadDirectoryChangesW in qt [12:56] and it uses kqueue for the time being [12:56] mandel: why not ask the guy? [12:58] nessita, for completeness' sake: http://python.6.n6.nabble.com/Status-regarding-Old-vs-Advanced-String-Formating-td4503327.html [12:59] ralsina, so, I fixed the issue with the dialog not working on windows and... then I hit this bug: http://twistedmatrix.com/trac/ticket/2226 [12:59] nessita, how much control do we have over patched to the qt reactor? [12:59] mandel: looking... [12:59] also: http://pyvideo.org/video/948/pragmatic-unicode-or-how-do-i-stop-the-pain [12:59] ralsina, python keeps running so do int a sys.exit(return_code) does not return the correct code.. bummer [12:59] thisfred: thanks! [13:00] mandel: not sure what you mean... we can propose patches to upstream thru github [13:00] nessita, ok.. [13:00] mandel: ugh [13:00] * mandel goes to github [13:00] mandel: why? [13:00] mandel: on windows we can build qtreactor from whatever we want [13:01] mandel: and we don't use it on linux [13:01] ralsina: we use it to run tests [13:01] mandel: we are literally building it out of a branch tarball on github [13:01] (on both) [13:01] nessita: ok, but that won't be affected by this, because it's a bug on exiting the process [13:03] ralsina, nessita the issue i the following, all the thread that are used by deferToThread with qt reactor get glocked if you use exec_ for a QDialog (one bug in the reactor) [13:03] nessita: while fixing the conflict in tweaks, I think the gotowebbutton __init__ is kinda broken. Specifically, self.legend is never going to be not None [13:03] ralsina: why not? [13:03] ralsina, nessita my work around was to use show and connect to finish which will get the return code, then do a sys.exit(return_code) and reactor.stop() but that does not work (an other reactor bug) [13:03] nessita: oh, right, subclasses, nevermind [13:04] mandel: I would seriously consider bringing those to upstream... [13:04] ralsina, nessita a third way to fix this is do a callFromThread in the windows keyring so that we do not use a thread but execute the call in the main loop of Qt, then return a deferred that gets fired when done, which does work and does not require to fix the reactor [13:05] nessita, sure, I can loop at what is being done in the reactor.. [13:06] mandel: did you talk with alece about this? he is our master when it comes to reactor stuff [13:06] nessita, will do when he wakes up.. we work 'til late yesterday [13:07] nessita, is this upstream => https://github.com/ghtdak/qtreactor [13:07] ?? [13:08] mandel: yes [13:08] nessita, thx [13:09] and alecu is in uruguay [13:13] ralsina: I got a little lost regarding your tweaks branch... is it ready? [13:14] nessita: yes, pushed the conflicts fix [13:14] ack [13:16] * gatox brb..... goes to buy fruit salad to try to unblock the brain with the backend issue :S [13:16] if fruit salad doesn't work.... i'm going to buy a gun [13:18] gatox: you may also try getting help. You know. Professional help ;-) [13:21] ralsina, how well does the python threading and the pyqt interact? I'm looking at http://twistedmatrix.com/trac/browser/tags/releases/twisted-8.1.0/twisted/python/threadpool.py#L27 which is what is used as a threadpool in twisted and I wonder how well does that work [13:21] mandel: not very well AFAIK [13:22] ralsina, he, then I guess there is some way to easy improve this :) [13:22] mandel: or rather, python threads and pyqt work, as long as you don t do anything pyqt-ish in the non-main threads [13:23] back! [13:27] hello #ubuntuone! [13:27] hello alecu! [13:27] hola alecu [13:28] alecu, buenas [13:28] ralsina, mandel, alecu, gatox: isn't the issue in windows, related to the keyring using deferToThread and not working, the thing that is making the 'Sign In' button not being enabled? it makes sense [13:29] since that would mean that ussoc can not progress and can't tell if there is no credentials in the system [13:30] is there any branch about that, that i could test? [13:30] nessita, yes, that looks similar to the problem I see in the proxy creds dialog (which is a smaller simpler app) [13:30] nessita, deferToThread on the qt reactor is not work at the moment, nor is callInThread [13:30] gatox: apparently there is no fix yet, AFAIK [13:30] nessita: may be! [13:30] nessita, yest callFromThread does (which means, q in the main loop) [13:30] mandel: how come this worked before? [13:30] right now i'm seeing like the app get stuck when trying to do: reactor.connectTCP and ask if it's listening [13:31] mandel: it may be the exact same thing i the -ui-qt script! [13:31] nessita: this never worked on windows with that on a separate process [13:31] ralsina: aaaaahhhhhhh uuuuhhhhhhh craaaaaaaap [13:31] nessita, I'd need to look at all the changes, I've noticed that when a dialog uses exec_ it does not work, but if you use show it does [13:32] nessita: before it was either in u1cp r in -installer, so it worked [13:32] mandel: exec_ starts a secondary event loop, that may break reactor [13:32] mandel: we're displaying the "main" sso UI with show.... [13:32] mandel: no, I lie, it uses exec_ [13:32] mandel, ralsina: but is a separated process... why would that interfere with the "first" mainloop? [13:33] nessita: lost me there. May not be the same thing then. [13:33] in any case, I'm incline to think that the "backend" not ready bug is related to this [13:34] nessita, who is working on this? [13:34] gatox: a couple of tests you could do on Linux, is to lock your keyring, and then try to sign in [13:34] nessita, in a diff process it should not be an issue.. I now exec_ brakes things I have not looked any deeper atm [13:34] gatox: mandel I guess, but not sure. Ideally we'd need advice from alecu_, being him our reactor master [13:35] nessita, sorry: the reactor master you are looking for is dobey :-) [13:35] alecu_: can you mumble? [13:35] nessita, I know how to *use* reactors, not how they work internally. [13:35] que pasa? [13:35] could we get rid of deferToThread and just block on those calls? The windows keyring should be quick anyway [13:35] alecu_: I have some ideas I would like to run by you [13:35] nessita, I can try mumbling, but not sure if mumble will work for me today, since I'm not working from home. [13:35] ralsina: what if it's locked, like the linux keyring can be? [13:35] nessita: you can't lock the windows keyring [13:36] nessita: AFAIK at least [13:36] alecu_: let's try... mandel, gatox, can you do a quick mumble? [13:36] nessita, yes! [13:36] ralsina: will research. I know you're in a meeting, but if you want/can, mumble? [13:36] nessita, sure [13:36] on mumble [13:36] I am at the design call, in mumble so no [13:36] :-( [13:37] gatox: would you drag me to the channel you're in? I can't [13:38] alecu_: we're in mumble if you can [13:44] nessita, this connection is very flaky, so no mumble today. [13:45] alecu_: :-/ [13:45] nessita, so, how can I help? [13:46] nessita, lp:~mandel/ubuntu-sso-client/proxy-creds-windows [13:46] alecu_: we're mumbling, will get back to you with concrete questions [13:46] thanks! [13:46] ack [13:48] nessita, here is the 'interesting' changes: http://bazaar.launchpad.net/~mandel/ubuntu-sso-client/proxy-creds-windows/view/head:/ubuntu_sso/qt/proxy_dialog.py#L158 [13:50] ralsina: i don't think you saw it yesterday, but syncing with a proxy works fine. still trying to figure out what the on_upload_finished issue is with getting the wrong number of arguments [13:50] briancurtin: awesome! [13:50] briancurtin: if you can pop into mumble, they are covering some windows-related bugs now [13:50] ralsina: logging in now [13:51] briancurtin, did anybody else got that error on XP? [13:52] alecu_ it turns out that same thing happens on Win7, just way way faster (i noticed it on XP since its a really slow VM) [13:52] so its all windows, apparently [13:53] briancurtin, so, it's a relief knowing it's not a platform dependent bug. [13:53] yep [13:53] makes it sound a lot less scary :-) [13:53] yeah, if that was XP only, it was caused by martians [13:54] ralsina, martians that should be extinct long ago! [13:54] EOLife'd martians, yes [13:55] "Your race is doomed!!!". 'No, we *will* prevail' it said while climbing back in the saucer. [14:01] briancurtin, nessita, gatox very important regarding spawing a ui that uses the qt4 reactor and expecting the return code: http://twistedmatrix.com/trac/ticket/2226 [14:02] sys.exit(3) reactor.stop() with the qtreactor does not work.. :( [14:16] mandel: so, let me know when you come back from lunch and when the branch is ready for inspection :) [14:19] so, good news? [14:29] ralsina: kinda... I'm very confident gatox's backend issue is caused by the Qt SSO UI not running the qt4reactor in windows [14:29] ralsina: regarding mandel's issue, need to see his branch [14:29] nessita: that is good news [14:30] the stuff about "the app doesn't end', we have seen before [14:30] or at least very similar things [14:32] nessita, yes, it was reproducible from sources also..... i'm adding the qtreactor now [14:32] gatox: yey! [14:36] alecu_, ralsina, mandel: there are like 4 branches from mandel that will not make it for this release.... were you aware of that? [14:37] nessita, can you please point me in the direction of the qtreactor on u1-cp? [14:37] gatox: yes, first look at the bin/ubuntuone-control-panel-qt [14:37] ahhhh ok [14:37] 25 if sys.platform == 'win32': [14:37] 26 import qt4reactor [14:37] 27 qt4reactor.install() [14:37] 28 [14:37] 29 from ubuntuone.controlpanel.gui.qt import main [14:38] gatox: then, go to ubuntuone.controlpanel.gui.qt.main [14:38] nessita: I piged him about one yesterday but it has a broken prerequisite [14:38] thanks! [14:38] gatox: in there, see the windows.py file :-) [14:38] gatox: I advice doing something similar to the qt sso ui [14:38] nessita, will do [14:54] * mandel back [14:54] nessita, yes, we know is not a problem, is a feature that we did that won't be used in our case [14:55] nessita, I'm making the changes to the branch to see the different issues [14:56] mandel: got lost, what thing is not a problem? [14:56] nessita, branches not landing, we know they wont and we are not in a hurry for those [14:57] nessita, and they should have merge issues.. [14:57] mandel: ack then [15:00] me [15:00] me [15:00] me [15:01] me [15:01] me [15:01] ralsina, alecu, dobey, thisfred? [15:01] meh [15:01] me [15:01] me(no notes yet) [15:01] briancurtin: go! [15:01] DONE: debugging, proxy IRL testing [15:01] TODO: hooking up VS to debug on_download_finished TypeError now that i know it's not XP specific, get any other windows stuff out of the way to get back with urbanape for a bit [15:01] BLOCKED: XP kinda sucks [15:01] NEXT: urbanape [15:01] DONE: looked into /dev/fsevents reading for file-level event granularity [15:01] TODO: Tests, land branch, for god's sake! [15:01] BLOCK: my brain. [15:01] gatox, go. [15:01] DONE: [15:01] Propose the unicode fixes branch for sso and fixed the test_main suit. Investigate the backend issue. [15:01] TODO: [15:01] Finish with the backend problem. [15:01] BLOCKED: [15:01] No [15:02] nessita, go [15:02] DONE: proposed computer-to-cloud-page u1cp branch, reviews [15:02] TODO: land branches for release, help with stucked apps to mandel and gatox, leave early due to teaching duties [15:02] BLOCKED: nopes [15:02] NEXT: mandel [15:02] * mandel finishing notes [15:03] DONE: We got the webclient working correctly by fixing bug (late night work). Looked at issues with prxoy creds dialog on windows. [15:03] TODO: more creds dialog debugging. Look at jenkins. Write sso script example for aquarius. [15:03] BLOCKED: no [15:03] dobey, please [15:03] λ DONE: askubuntu, bug #932103 (pinging right people, and testing) [15:03] λ TODO: bug #953119, releases, uploads [15:03] λ BLCK: none. [15:03] me? [15:03] Launchpad bug 932103 in gstreamer0.10-fluendo-plugins-partner (Ubuntu Precise) "Cannot install MP3 playback support; silently fails" [High,Fix released] https://launchpad.net/bugs/932103 [15:03] Launchpad bug 953119 in ubuntuone-client-gnome (Ubuntu) "valgrind invalid read error" [Undecided,Triaged] https://launchpad.net/bugs/953119 [15:03] alecu_: go! [15:03] ALL: is there any branch that *needs* to be in tomorrow's release that hasn't landed yet? [15:04] nessita: I don't think so, if we don't mind the windows release being done from another tarball [15:04] DONE: got a few windows-specific branches merged and the focus tweaks branch, reviews, tech leads call, mgmt call (in progress), design call, missed part of the day because of doctor TODO: help with windws fixes, start windows release process BLOCKED: not really [15:05] gatox, briancurtin, alecu_, mandel, dobey, thisfred, urbanape: is there any branch that *needs* to be in tomorrow's release that hasn't landed yet? [15:05] DONE: reviews | gave up on segfault for now TODO: u1db filter functions in C | u1db and other reviews | turn 40 | go to Savannah BLOCKED: no NEXT: ralsina! [15:05] nessita: not from me [15:05] nessita, not from me AFAIK [15:05] nessita: is there a branch that fixes everything to be how i wante it to be? :) [15:05] nessita: windows specific, but we need this one in whenever the windows release happens (1 approval already): https://code.launchpad.net/~brian.curtin/ubuntuone-client/930398-windows-is_root/+merge/96224 [15:05] nessita, i don't have branches for review.... working in the qtreactor thing right now [15:05] thisfred: happy birthday! (you turn 40 today? :-)) [15:05] nessita: tomorrow :) [15:06] ack to all the branch info, thanks [15:06] thisfred: will try remember saying happy bday tomorrow then :-) [15:06] ALL: so I won't be here the rest of the week [15:06] dobey: no :-) [15:06] nessita: I'll be in Georgia :) [15:06] thisfred: sorry, I shot first. Like Han Solo [15:06] np [15:07] the state, or the country? [15:07] thisfred, 40!?! If you where a british northern girl you could be my mother! [15:07] I'd have given you up for adoption for sure [15:08] mandel: your mother gave birth to you at the age of 13? [15:08] lol [15:08] explains a lot [15:08] dobey, that is why I said 'british northern girl' my mother is spanish :P [15:08] that and the heavy drinking she did during the pregnancy [15:09] heh [15:09] lol === alecu_ is now known as alecu [15:10] * gatox lunch! [15:14] nessita: we have a new comment in the color chage bug from robert, is that good enough to move forward/ [15:14] ? [15:14] ralsina: looooooking === gatox is now known as gatox_lunch [15:16] ralsina: may be enough.... I guess we can start the process. Asking joshuahoover for confirmation is the best thing to do right now [15:16] nessita: ack [15:16] ralsina: also, I proposed https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/license-page/+merge/98432 so we can add the strings before releasing (though the license page is never shown, for now) [15:17] nessita: looking [15:17] nessita: yeah, that should be good...i'll move it forward as soon as i get off this current call i'm on [15:17] joshuahoover: thanks!!! [15:17] nessita: I can do the --installer branch later myself if you want [15:17] nessita: since I have been hacking on command lines lately [15:17] ralsina: you can try, but it will not be trivial [15:18] nessita: sure, I can try, if I fail I'll cry and let you know ;-) [15:18] ralsina: since the wizard is created 'automatically' from the setupUi, and thus I don't know how to pass custom arguments to custom widgets [15:18] nessita: I'm siure there is a way ;-) [15:19] briancurtin: your last note in the standup reminded me: Should I get W7 for my VMware? [15:19] I'm currently running XP [15:19] ralsina: nice [15:22] nessita, btw: was it useful, the info on the package dependencies I provided yesterday ? [15:22] nessita: +1 on that [15:22] alecu: yes, the -proxy package is already available (I mentioned this at noon yesteday). Have you try it? :-) [15:23] nessita: do you want to do the release of ubuntuone-client as well? [15:24] dobey: I can if you teach me :-) [15:24] dobey: make dist I guess and then upload that tarball? [15:24] nessita, I updated lp:~mandel/ubuntu-sso-client/proxy-creds-windows I have added a new arg for testing, if you use --use-exec the diaog will be launched using exec_ otherwise a show + finish signal solution is used (which 'works') [15:24] ralsina, ^ [15:24] oh, right. it's a bit more complicated than that, because of gtk-doc [15:24] ralsina, and if you have time we can take a look to see if I did something stupid [15:25] mandel: I will be available in about 15 minutes [15:25] mandel: so, I just confirmed that the reactor has to be installed *before* the qapp is created [15:25] nessita: i'll do the tarball release, and you can do the upload [15:25] dobey: sounds great [15:26] nessita, before? then the doc is a lie: https://github.com/ghtdak/qtreactor [15:26] urbanape: yeah i think you might as well have it [15:26] mandel: that stinks of "QDialog.exec_()" breaks the qtreactor to me [15:26] mandel: so I am ok with using show() [15:26] nessita, from the docs in github: '= Using the Qt4Reactor = [15:26]