=== JackyAlcine_ is now known as JackyAlcine [09:34] Good morning and happy Valentine's Day, everyone! :D [09:39] JamesTait, happy day 0 [12:01] good morning! [12:15] ralsina, morning! [12:15] good morning mandel! [12:15] mandel: So, how far did we get with your stack of branches? [12:15] ralsina, I need a review form you [12:16] mandel: hit me! [12:16] on it, let me find it [12:16] ralsina, do you understand this message: https://code.launchpad.net/~mandel/ubuntu-sso-client/creds-dialog-script/+merge/92533 [12:17] looking... [12:17] ralsina, I see now failures.. [12:17] mandel: launchpad is not working [12:17] ralsina, I got 'g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream returned 0 bytes on an async read (g-io-error-quark, 0). Exiting. [12:17] Terminated' [12:18] mandel: yes, I know that one [12:18] mandel: you just haveto retry [12:18] ralsina, ok [12:18] mandel: 2nd or 3rd time it will merge [12:18] ralsina, hm.. not very deterministic lol [12:18] ralsina, the MP: https://code.launchpad.net/~mandel/ubuntu-sso-client/load-creds-retry/+merge/92536 [12:18] mandel: it's pretty random. Happened to a branch from natalia on fridayand yesterday to one from gatox [12:19] ralsina, he [12:19] 4:1 test:code ratio :-) [12:20] ralsina, question, do you know if there is a way to retry a QNetworkRequest using a QNetworkAccessManager [12:20] mandel: this is just to have the credentials already in the entrie on retry? +1 then [12:20] ralsina, yes, that is all [12:20] mandel: I don't think so. Maybe using the copy constructor? [12:20] ralsina, in my little merges pipeline approach :) [12:21] mandel: approved but don't set it globally until you see the previous one merged [12:21] ralsina, yes, will do that.. but tarmac should be smart rnough to do that.. [12:22] mandel: it is not ;-) [12:22] mandel: if it gets two stacked branches on the same run, you get the weird "no merge candidate" messages we had yesterday. [12:23] hi! [12:23] ralsina, yes.. is weird.. it used to work [12:24] ralsina, I can get the reply from the request.. I suppose I can retyr using the request and get the verb from it [12:24] mandel: you can probably just do another QNetworkAccessManager.get with the same QNetworkRequest [12:24] mandel: but I have, of course,not tried it [12:24] ralsina, hm.. I can get the verb if I pass it to the error handler of the deferred, let see what happens :P [12:25] mandel: the request has the verb! [12:25] ralsina, where? [12:25] mandel: oh, no it doesn't [12:25] mandel: itcanonly contain a verb if it's not standard. Pfffft. [12:26] ralsina, yeah.. this part of Qt is a little dirtier than the rest hehe [12:26] mandel: the reply has it [12:26] mandel: QNetworkAccessManager::Operation QNetworkReply::operation () const [12:26] ralsina, nice! thx! [12:26] So the request has the reply, the reply has the operation [12:28] good morning gatox! [12:28] ralsina, buenas! :D [12:39] gatox, mandel: I need to go to the bank so I ... you know, get paid? Then pay taxes and a couple of errands, I will be back in about 90 minutes [12:39] ralsina, roger that [12:39] ralsina, ok, I'm going to have lunch a little earlier today (at my 2pm) so I'll see you more or less when you get back :) === eu is now known as Guest98507 [12:41] mandel, gatox:ok, see you guys then [12:42] ralsina, laters! [12:49] gatox, I'm off to lunch (a little earlier) [12:49] gatox, will be back once done :) [12:49] mandel, ack.... enjoy [12:49] gatox, is chinese.. so it won't be that great hehe [12:49] jejeej === dduffey_afk is now known as dduffey [14:05] hello! [14:05] ralsina: ping [14:06] dobey: pong [14:07] ralsina: so, this use qt in sso plan, means we need sso qt on the CD as well? [14:08] dobey: yes but we have a sort of catch 22 there [14:08] ralsina: and it's using qt webkit? [14:08] no webkit [14:08] dobey: it launches the URLs in the browser [14:08] oh? then why does it look like it's embedding some rather bland web pages? [14:08] dobey: it has a widget for bland html subset [14:10] dobey: nessita has the branch to make sso use qt by default, but that needs sso-qt on the cd,which needs qt on the cd, so it [14:10] 's approved but not merged [14:12] :-/ [14:35] I'm back! [15:00] me [15:00] me [15:00] me [15:00] mandel, i think noone is here :P [15:01] a 3! [15:01] the windows guys are here ;) [15:01] jejejeje [15:01] :) [15:01] ralsina, dobey, alecu ? [15:01] I am on call, skipping [15:01] notes! [15:01] me [15:02] mandel, go! [15:02] DONE: Merged all the branches from the creds dialog (inlucing review comments). Worked on re implementing the AuthProxy error in the Qt webclient implementation. [15:02] TODO: More qt webclient work. [15:02] BLOCKED: no, but I might ping alecu for some help [15:02] gatox, go go go [15:02] DONE: [15:02] Proposed several branches, all landed. Proposed one more branch today. [15:02] TODO: [15:02] Hunt people to approve my remaining branches. Keep working on some other issues. [15:02] BLOCKED: [15:02] No [15:02] briancurtin, go [15:02] DONE: work on bug #851810, somehow hosed my machine while upgrading to precise [15:02] TODO: 851810, get involved in more reviews [15:02] BLOCKED: none [15:02] NEXT: alecu [15:02] DONE: worked on branch for bug #929207 [15:02] TODO: more of the same [15:02] BLOCKED: no [15:02] NEXT: ralsina (when he comes back) [15:02] Launchpad bug 851810 in ubuntuone-windows-installer (and 2 other projects) "Notify clients when volumes info from server is ready (affects: 2) (dups: 1) (heat: 16)" [Medium,Triaged] https://launchpad.net/bugs/851810 [15:02] Launchpad bug 929207 in ubuntuone-client "Proxy "tunnel" for syncdaemon (affects: 1) (heat: 6)" [High,Confirmed] https://launchpad.net/bugs/929207 [15:03] mandel, how can I be of help? [15:03] briancurtin: I have hosed my machine upgrading to P... 6 times this cycle :-( [15:03] eek [15:03] alecu, let me finish a little and I'll push a branch with what I've done, I think I'm not handeling correctly the exception whithin and errback [15:05] briancurtin, do you have time for 2 really small reviews? :P [15:05] gatox: sure, send them over [15:05] briancurtin, thanks! https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/931452/+merge/93003 - https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/931577/+merge/92863 [15:07] what? [15:07] oh [15:08] meh [15:08] λ DONE: gwibber, decided to not move code around for now [15:08] λ TODO: releases, patchaes, packaging changes, twisted docs, move more code around [15:08] λ BLCK: none. [15:10] * dobey proposes his gwibber changes to trunk [15:13] alecu, can you take a look at lp:~mandel/ubuntu-sso-client/webclient-use-dialog [15:14] mandel, looking [15:14] alecu, I'm confuse with how to handle the exceptions in _handle_authentication within the qtnetwork webclient implementation [15:14] alecu, I'm sure you can get your head around it better than I can [15:14] mandel, you mean my head is bigger? [15:15] mandel, (have all dependencies on this branch landed?) [15:15] alecu, AFAIK yes [15:16] lol [15:16] alecu, and I mean, you have been dealing with twisted for longer :P [15:16] I thought you meant I was a "Cabezón" [15:17] alecu, is an expresion.. my head is really big too [15:19] lunch for me! brb [15:19] mandel, "that will return the appropiate values and we will peform the appropiate action" -> that sounds like noise [15:20] alecu, there is need to clean up there [15:20] ok, then I'm looking at the wrong place [15:20] mandel, is it in qtnetwork.py? [15:20] alecu, yes, run the test.. [15:21] alecu, you will see the errors [15:26] alecu, where did you find that sentence? [15:26] mandel, in the diff from your branch to trunk. [15:26] mandel, let me find the file [15:27] mandel, it was in webclient/common.py [15:28] mandel, so, these are the errors: http://pastebin.ubuntu.com/841819/ [15:29] alecu, yes, and the one I'm sure I'm not understanding to do correctly is the one of test_unauthorized [15:29] alecu, I though that using failure.trap in the errback would do the right thing.. [15:30] lookin [15:31] alecu, that code is messy, I wanted to get a working proof of concept and then clean it, but I got that issue. ideally the retyr should not break a thing.. [15:32] mandel, so, basically the "exception = ProxyUnauthorizedError(error_string, content); d.errback(exception)" is not being trapped, right? [15:33] alecu, yes, or at least that is what I think [15:33] mandel, but the error I'm seeing is not the proxy throwing a 407, but instead the web server throwing a 401 [15:34] mandel, that means, that the *server* is not authorizing, and that it should be retried. [15:34] mandel, the proxy seems to be crossed just fine if that error is thrown by the server, no? [15:35] if you get an error from the server, the proxy worked, yes [15:36] alecu, let me check [15:37] mandel, also, I think we have a bug in common.py: raise UnauthorizedError('User did not provide proxy credentials.') [15:37] alecu, that could be.. [15:37] mandel, ^ that should be the other exception [15:38] alecu, the error number is a 205 that is a ContentReSendError [15:38] mandel, hmm... or perhaps not. It may be right if the user cancels the dialog not to ask him again. [15:39] alecu, looks like I'm getting the error when trying to resend the request [15:39] alecu, sorry, my bad is a 204 [15:39] AuthenticationRequiredError [15:40] mandel, that's perfect. On trunk that's already being handled by the existing code. [15:41] alecu, I just found the error.. [15:41] mandel, tell me [15:41] alecu, the tech reason is: soy gilipollas [15:41] alecu, look at the if statement [15:41] mandel, sounds like a reasonable assertion ^ [15:41] mandel, looking [15:42] mandel, I've found a few ifs... which one do you mean? [15:42] alecu, 159 [15:42] alecu, more or less [15:42] I should have used and elif [15:42] * mandel facepalm [15:43] mandel, right! [15:43] mandel, tdd ftw [15:44] alecu, indeed, I just have on more failing test and I'm done [15:44] I need to do an end to end test but we should have all of the thing working by then :) [15:50] alecu, all green \o/ [15:50] mandel, váaaamos! [15:51] alecu, it seems that we also have to listen to the authenticationRequired for when we go through the proxy and get the 401 from the server, which i think is wrong.. but makes the tests pass [15:52] alecu, so we might have a concepts issue there [15:52] mandel, I don't get it. [15:52] mandel, let me try an example [15:52] A tries to connect to C using a proxy B [15:53] alecu, let me restart my machine so that mumble works and get talk about it [15:53] alecu, I might be wrong :P [15:53] mandel, wait [15:53] alecu, is the other machine, I'm not leaving you :) [15:54] mandel, sorry, it was the doorbell. [15:54] mandel, so, back to the example} [15:54] alecu, ah, I though I had to wait [15:55] mandel, my mumble machine is somewhere else too... that's why I want to try the example on IRC [15:55] so, again: [15:55] A tries to connect to C using a proxy B [15:55] B says "407: proxy creds required!" [15:56] alecu, that will result in a 105 within qt [15:56] A tries to connecto to C using proxy B passing proxy credentials [15:56] now it's C turn to say "401: webserver credentials required!" [15:57] A tries to connect to C passing server credentials, using proxy B passing proxy credentials [15:57] alecu, then authenticationRequired is raised [15:57] alecu, aauthenticationRequired is for ' A tries to connecto to C using proxy B passing proxy credentials' [15:57] mandel, right: both are raised: on the first try proxyAuthRequired. On the second try authRequired. On the third try it works. [15:58] alecu, no, on the first QNetworkReply.ProxyAuthenticationRequiredError is returned in the error of the QtNetwrokReply [15:58] alecu, yet proxyAuthRequired is NOT raised [15:58] mandel, oh, yes you told me that was a QT issue, right? [15:59] alecu, yes, ralsina and I found that on friday [15:59] mandel, but we are able to catch the 105 (meaning 407) and process it just like a signal was thrown, right? [15:59] alecu, then from that point on (although I have to write a test for that) after the QNetworkReply.ProxyAuthenticationRequiredError the theory states that authenticationRequired is raised [16:00] right [16:00] mandel, but? [16:00] alecu, yes, we grab the 105 and raise an exception that is catch by an errback [16:00] great. But something goes wrong, right? [16:01] yes, the issue is as follows, authenticationRequired can be raised when we pass through the proxy and we get a 407 from the server [16:01] alecu, ^ [16:01] alecu, and at the moment our webclient treats it as a proxy auth issue when is not [16:02] alecu, in trunk, look at _handle_authentication [16:02] mandel, we should not get a 407 from the server... only from the proxy, right? [16:02] alecu, you mean in the ubuntu one case, or in general? [16:02] mandel, in general [16:03] mandel, http://en.wikipedia.org/wiki/List_of_HTTP_status_codes#4xx_Client_Error [16:04] mandel, from a server we could get a 401 (or some other error), but never a 407 [16:04] I know that the symbols for "one" and "seven" look similar, but a server should not confuse them! [16:07] alecu, ok, let me push the code that passes all test and we look at the results of the tests, ok? [16:07] alecu, that way we are both in the same page [16:09] alecu, can you pull lp:~mandel/ubuntu-sso-client/webclient-use-dialog [16:10] mandel, sure! can you pull this finger meanwhile? [16:10] alecu, sure, the smell will stay in buenos aires.. [16:10] alecu, talking about this, I have to change my dogs food, I bought a diff one and know he is a pedorro.. [16:11] no good when you are watching tele [16:22] dobey: is this known?: http://pastebin.ubuntu.com/841882/ (libuntuoneui fails to install/upgrade) [16:24] sure, blame the dog [16:24] wtf [16:24] oh [16:25] yes, but webkit was broken so the fixed build failed on amd64 [16:25] * dobey tells it to rebuild [16:25] cool [16:28] thisfred: to workaround you can apt-get remove --purge the old package, then install the new [16:28] kthx [16:29] dobey: which is the old one? [16:30] libubuntuone-1.0-1 [16:31] same error when I try that [16:31] eh? [16:32] maybe you have to apt-get -f install first [16:32] then remove old [16:32] did that [16:32] keeps saying it's trying to overwrite the locale [16:32] well screw apt then [16:33] indeed, I'll just wait for the fix [16:33] it's nonsense === m_conley_away is now known as m_conley [16:34] dobey: to increase the fun bamfdaemon keeps crashing, then coming up and showing every icon on the launcher twice. Which wouldn't matter but that means they show up twice in the alt-tab list as well, making alt-tab *really* unusable. [16:35] oh [16:35] well what's fun is that it thinks i have two rhythmbox windows on one screen [16:35] mandel, alecu briancurtin is anyone having this issue: http://paste.ubuntu.com/841903/ ? [16:36] holy crap [16:36] alt+tab really hurts something [16:37] and makes jack quite unhappy [16:37] why does compiz use like 30% cpu for alt+tab?! [16:37] grr [16:39] i guess i will just have to screw around with audio toys in windows or something :( [16:40] gatox, give me a few mins to boot the vm [16:40] gatox: i don't see that IOError [16:40] mandel, briancurtin did you know if the way to run sso in linux has changed?? it is necessary to set the USE_QT_MAINLOOP or something there? [16:41] i remove the setup.py line and i have sso working [16:42] no idea. i was actually about to ask what's required of all of these projects to run IRL - the things i've done have all been through the tests [16:42] User is having trouble installing client on Windows 7. here is the installer.log. not sure what the error means? https://pastebin.canonical.com/60166/ [16:43] a goat, an altar, and a sacrificial dagger. [16:43] duanedesign: looks like syncdaemon isn't starting up [16:43] duanedesign: looking... [16:44] duanedesign: yes, syncdaemon is not starting [16:46] gatox, try "U1_DEBUG" instead of "DEBUG" [16:47] gatox, is that ussoc trunk? [16:47] alecu, yes, i already changed that and sso is workiing.... but the setup.py build i have to run it from linux and remove that line from my windows script [16:48] gatox, weird! [16:48] alecu, yep [16:48] gatox, can it be that the VM disk is mounted from linux? [16:48] gatox, perhaps it's vbox acting up? [16:48] gatox, in any case, try rebooting the vm. [16:48] alecu, i don't know.... it's the same configuration as always....... [16:48] alecu, i already try that :P [16:49] gatox, then, no idea. [16:51] alecu, and did you know this one? :P http://paste.ubuntu.com/841922/ [16:51] alecu, for control panel [16:52] gatox, yes, I knew about that one. Is your branch merged with trunk? [16:52] yes..... i'm trying to run control panel from trunk [16:52] alecu, ^ [16:52] gatox, if it still happens in trunk, then it means that it's a same problem we had last week in a branch with nessita, and she decided to fix it for windows after the current linux releases. [16:53] alecu, mmmmm so....... the short answer is that control panel can not be executed in windows now? [16:54] gatox, probably :-P [16:54] gatox, I don't have the exact answer, but that sounds likely [16:54] alecu, ok..... thanks [16:54] mandel, so, I've been taking a look at the branch, and everything seems reasonable. [16:54] * gatox start looking for another bug :P [16:54] mandel, what was the conceptual issue you mentioned? [16:55] hey all [16:55] mandel, (if you want to mumble, let me know and I'll get the laptop) [16:55] hi dpm! [16:55] buenas alecu :) [16:55] dpm, como andás, bienvenido por acá. Ponete cómodo :-) [16:56] alecu, bien,gracias, gracias... :-) [16:56] * alecu is reminded to get some pants on. [16:56] lol, I wasn't thinking of getting _that_ comfortable [16:57] alecu, let me see if it works [16:57] dpm, uh, pero que haces por aqui? [16:57] heh [16:57] mandel, nada, una pregunta sobre u1, que estoy seguro que me lo habeis roto :P [16:57] dpm, I know you are one of the few that undertands how funny this is: http://www.youtube.com/watch?feature=player_embedded&v=FhJR6OO1X8Y [16:58] dpm, could be hehe [16:59] mandel, ok, looking at libsoup.py I see something that looks off... [16:59] alecu, let me fix some pylint issues and we can work in the code that passes everything [16:59] alecu, give me 5 min and we mumble [16:59] mandel, yeah, I had seen the video and had a good laugh :) [16:59] mandel, for instance, _on_authenticate is used for "server" authentication. But it's calling the "request_auth_credentials" that are used for proxy. [16:59] mandel, so, yes: there's a conceptual issue there. [17:00] mandel, we should not be calling the "proxy credentials" as "auth credentials" because that will lead to errors like this on our part. [17:01] mandel, we should be very strict with which credentials are for proxies, and which are for the final server. [17:01] alecu, yes, that is what I was thinking og [17:01] s/og/of [17:01] alecu, exactly, lets talk about that to make it very very strict.. [17:02] alecu, the scary thing, tests pass [17:02] dpm, I tried to play it to my father, yet he does no speack catalan.. [17:02] choni de castefa, classic [17:03] I think there was a Spanish version somewhere? wait... [17:03] mandel, right... and now I can see some other issues in qtnetwork.py [17:03] alecu, I'm in mumble, shall we chat [17:04] mandel, let me grab the laptop. [17:04] alecu, sure [17:04] dpm, the nice thing is that you can download the song for free.. my friends in madrid are loving it ;) [17:06] mandel, here's the complementary version - http://caballe.cat/wp/oh-montserrat-la-jenifer-dels-catarres-a-linreves/ [17:09] hm, hoping onto a call now, will have to leave my actual question about u1 for later... [17:11] dpm, lol [17:13] alecu, http://developer.qt.nokia.com/doc/qt-4.8/qnetworkaccessmanager.html#proxyAuthenticationRequired [17:16] meh, come on system upgrade; and ups guy [17:18] dobey: you remember what was the suggested way to update from O to P? [17:19] update-manager -d? [17:20] dobey: right, thanks [17:20] * ralsina crosses a bunch of fingers [17:21] ohh i didn't remember that... i'll try to upgrade now [17:24] slackers [17:28] speaking of being updated, now that i have this precise box setup, is there any wiki or doc for setting up a development environment on linux? [17:30] or, ralsina - should i make this buildout stuff work there as well? [17:30] briancurtin: we have a doc for that somewhere in the wiki [17:30] * ralsina goes wiki-hunting [17:32] mandel, mumble was much better. [17:32] briancurtin: https://wiki.canonical.com/UbuntuOne/DeveloperSetup [17:32] * alecu loves when a plan comes together. [17:32] ralsina: ah thanks, was looking on wiki.ubuntu instead [17:33] alecu, certainly, too many username and passwords around.. [17:34] mandel, can you review this really trivial branch? https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/931577/+merge/92863 [17:34] mandel, if you are busy maybe i can bother ralsina :P [17:35] and this one too (if someone can): https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/931452 [17:36] gatox, I can [17:36] gatox, I hate, re-do everything!!! in latin!!! [17:37] mandel, what did i miss?? what about re-do? [17:37] gatox, how evil will it be asking why there are no tests when you changed code ;) [17:37] gatox, the re-do was a bad joke hehe [17:37] mandel, hey there is a test! [17:38] gatox: I have a hard day to do reviews today [17:38] ralsina, no problem! [17:38] mandel, ahhh do you mean ubuntu_sso_wizard.py ? [17:38] gatox, :) [17:39] gatox, but you are welcome to tell me to frack off, the change is simple enough [17:39] mandel, ok, review the other branch, i'll use this branch to enhance the test for that file that are really poor [17:39] gatox, he, superb :) [17:39] mandel, naaaa... that file lacks of several tests..... i'll do it in this branch [17:40] gatox, which one is the other branch? [17:40] the MP I mean [17:40] mandel, https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/931452 [17:40] on it [17:41] so I've got a folder containing some mp3 files. It's marked as sync'ed on 2 PCs, and I can also see the files on the web. However, on one of the PCs the folder is actually empty, so it seems that the files did not get synchronised. Running 'u1sdtool --waiting' on the PC with the empty folder does not show me any activity. Is there anything I can do to get that PC to download the files from U1? [17:43] dpm: tried reconnecting? [17:44] dobey, I've since then restarted the computer a couple of times. I guess that implies reconnecting, but I can try to explicitly reconnect, just a sec... [17:45] if you had it connected while the computer was off, i'd love to know how ;) [17:47] alecu, https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/931452/+merge/93003 looks good, nice use of partial! [17:47] dpm: you can also try to unsubscribe from it, and then resubscribe [17:47] i really need to get lunch [17:47] agreed on the good use of partial :) [17:47] bbiab [17:49] dobey, how do I subscribe? I guess using the --subscribe-folder=FOLDER_ID option on u1sdtool, but how do I find out the FOLDER_ID? [17:49] briancurtin, :) [17:50] dpm: --list-folders; or you can use the control panel to do it :) [17:52] dobey, ah, I see what you mean. I don't have a subscription for this particular folder, but rather for its top folder. Let me try to unsubscribe and subscribe from the control panel... [17:54] gatox, +1 on the 931452 [17:55] dobey, yeah, that seemed to sort it, thanks! (or at least I see some download activity now) [17:59] ralsina, gatox, dobey, briancurtin EOD in Spain, so its rugby time, catch you all tom! [17:59] enjoy [17:59] hello everyone! [17:59] mandel, bye! [17:59] just in time to say bye to mandel (?) [17:59] nessita, hi [17:59] hola [18:00] nessita, yes, see you tom hehe [18:00] nessita, short yet intense ;) [18:01] hello nessita! [18:01] hola ralsina! got my sms? [18:01] nessita: nope! [18:01] nessita: oh, wait, yes! [18:01] bu [18:01] damn silent phone ;-) [18:01] ah! [18:02] Ok, I am going to have a quick lunch before my next mac-head-interview [18:02] ack [18:02] if anyone needs anything, put it here [18:02] ralsina, mmmm una hamaca paraguaya :P [18:03] ponela ahi :-) [18:07] dobey: hola! have you started with releases? [18:09] * briancurtin lunch + buying flowers === verterok` is now known as verterok [18:40]