=== spm` is now known as spm [01:18] alecu: I'll be happy if you would take a look at this for me lp:~karni/+junk/ubuntu-sso-java-api and let me know what you think/comments. It's very fresh. The API is pretty straight forward, but I'll make a sample android app anyway. As far as SSO is concerned, it'll be an Android library using that ubuntu-sso-java-api. [01:18] CardinalFang: Whenever you have time, I'll be happy to hear your thoughts on lp:~karni/+junk/ubuntu-sso-java-api [01:19] karni, I'll look tonight or after a sleep. [01:19] CardinalFang: Awesome. No rush. [01:20] Man time passes by fast. And MAN it's hot in here! [01:20] * karni opens the window winder [01:25] karni, looks very nice. [01:25] karni, one comment regarding pingUbuntuOne [01:26] perhaps it should be moved out of the SingleSignOnAPI [01:27] I think it should reside in a class that inherits from UbuntuSingleSignOnAPI, in a com.ubuntuone package [01:27] since it's very specific to our use case. [01:27] alecu: yes, I will move it out definitely [01:27] alecu: and it uses HMAC singer, whilst SSO uses Plaintext singer [01:27] signer, right. [01:27] alecu: btw thanks :) [01:27] haha [01:27] signer :) [01:28] at 2:30 AM I feel like singing! ;) [01:28] karni, I've only looked at the SingleSignOnAPI interface, though :-) [01:28] hahahaa [01:28] alecu: haha :) cool. take your time, I just wanted to update you I'm making progress with it. [01:29] karni, another thing: perhaps pingUbuntuOne should be called silently, by the validateEmail in the UbuntuOne class [01:29] alecu: upon success. good idea :) [01:29] karni, so it gets called automatically and the user of the library does not need to remember to do it. [01:29] yes. [01:29] alecu: however, it may be the case the email is properly validated, but the ping fails (say, U1 is down ;) ) [01:30] alecu: what then? should we throw? [01:30] karni, good point. [01:30] alecu: we have to be careful. I wanted to include the ping to make our life easier, but, like you notied, it's not integral part of SSO itself. [01:30] karni, perhaps we should ask nessita tomorrow for the details of it, to see how the ubuntu-sso-client does it. [01:31] alecu: sure. I referenced ubuntu-sso-client code quite a few times, but I don't recall when/how ping details. [01:31] alecu: validate_email() does not call ping directly [01:32] in ubuntu-sso-client [01:32] probably somewhere higher [01:32] higher as "not in accounts.py" :D [01:32] ok [01:34] alecu: it pings upon successful login ( credentials.py : _login_success_cb() ) [01:34] oh, right! [01:35] karni, you should definitely ask nessita about that, because she'll remember the details [01:35] alecu: ok :) [01:35] karni, I think it had to do with updating the sso pinged credentials on every login... or something. but it makes sense to do it there. [01:36] alecu: yeah, I'll definitely ask her to be sure [01:36] I don't know the details. [01:36] alecu: We're still caching SSO tokens in app servers, aren't we? [01:36] alecu: I believe this is what we want to fix (avoid) in the future. [01:37] alecu: But that doesn't concern us (us playing with Android apps) [01:37] karni, we've been doing it at least for the past six months, and we'll probably be doing it in the foreseeable future. [01:37] right === DanRabbit is now known as DanRabbit|spaces === DanRabbit|spaces is now known as DanRabbit === lifeless_ is now known as lifeless [03:49] hey guys. I'm having an issue with playlists on ubuntuone music for android. Every time I tap the playlist button, I get an error. "A network error occurred. Try again later." [03:49] everything else works. I can view my music library and stuff, and play music [03:50] any thoughts on what could be causing this? [05:16] tntc: Hi! [05:16] tntc: So, we're having problems with the database that backed this functionality. [05:16] tntc: The programmer who was looking into it left just a while ago [05:17] tntc: But I'll ask him tomorrow if there's anything new on the subject. [05:17] tntc: Sorry for any inconveniance you have due to this issue! [09:27] morning all [09:30] Is this a beta program? [09:33] JohnTeddy: some clients are, like the windows client. The Linux client is not. afaik. [10:00] * duanedesign wonders if mr. rye is around? [10:00] duanedesign, yes, today i am around [10:00] duanedesign: I mean server side. [10:01] I looked on the status page, there are service disruptions and service outages. [10:02] rye: hope all is well today. Trying to track the possible reason for an error. Thought I might ask if you had seen a common reason for this. http://ubuntuforums.org/showthread.php?t=1781451 [10:03] JohnTeddy: i read their was a problem with a database that backed a particular function causing a temporary outage. [10:29] brb [11:19] * mandel getting some coffee [11:30] * mandel back [11:39] * duanedesign waves at mandel [11:40] duanedesign: hello :) [12:22] * mandel walking dog [12:26] * fagan break [12:29] hello everyone! [12:29] yo nessita [12:30] fagan: you're awake! :-P [12:30] nessita: yeah for hours :) [12:30] lol [12:32] Hola nessita [12:32] hola facundobatista [12:51] rye: /5 [12:51] oops [12:53] duanedesign, so, i downgraded one of my system to natty default/updates and my evolution is working properly. I wonder whether the user is using nightlies [13:06] * mandel early lunch [13:18] hola [13:19] * ralsina is officially an old man. He fell down on the street last night [13:21] awh ralsina you ok? [13:22] Anyway, I didn't break my hip, but I twisted my knee. So I have a pressure bandage, taking some magic pain pill and can't walk. Yay. [13:22] fagan: worse case, I'll use a peg leg and live a pirate's life. Best case it will stop hurting in a few hours. [13:23] BTW: jumping up to my 1st floor home on one leg... not as easy as it sounds [13:23] pirate legs are cool [13:23] fagan: but they make you use those rubber thingies at the bottom soyou don't scratch the floor and kills all the tap dance effect [13:24] no fun :/ [13:27] rye: hmmm. Ok, thank you! All the bug reports I can find with this error are Fix Released... [13:28] ok, back to coding what I should have coded last night :-( [13:30] can I have a couple of reviews for https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/greet-and-report-quota/+merge/65598 ? [13:31] * thisfred is on it [13:31] ralsina: also, question. Is there any easy way to do transparent overlays in QT? (if you need to google, don't, I'm asking to the top of your head) [13:31] thisfred: thanks! [13:32] wow, we can handle YiBs! :) [13:36] nessita: yes there is, but I can't give you a recipe. Look for setMask in the docs [13:37] nessita: reviewing... [13:37] nessita: is it for the spinner thingy? I need one too for the wizard [13:38] nessita: OTOH I may not have understood what you want [13:39] ralsina: if you look lisete's windows share, go to: [13:39] U1_design_work / Windows / client / PSD : [13:39] and see: [13:39] win_client_008_please_wait.png [13:39] win_client_008_please_wait002.png [13:40] ralsina: she proposed those, and before agreeing to that, I wanted to see if there is a way to implement it [13:40] ralsina: I'll google [13:40] and see if I can create a widget [13:40] looks like I don't have lisette's only Ivanka and Claire's [13:40] lisette: ping [13:40] ralsina: did you subscribed the share? [13:40] ralsina: go to the linux control panel and subscribe it [13:40] nessita: I'll check [13:41] (maybe you have to accept it first?_ [13:41] ) [13:41] ralsina: hi [13:41] lisette, I don't seem to have your share on Ubuntu One, could you re-invite me? [13:41] nessita: nope, no email about it either [13:42] * ralsina may have lost it [13:42] ralsina: sure, i will uninvite and reinvite you :) [13:42] lisette: cool, thx [13:43] lisette: got it, thanks! [13:43] ralsina: done [13:44] nessita, dobey, I fixed the problem of ubuntuone-client in Oneiric, the one from bug #798979 [13:44] Launchpad bug 798979 in ubuntuone-client (Ubuntu Oneiric) (and 2 other projects) "ubuntuone-syncdaemon crashed with KeyError in add_watch(): '/home/caravena/Ubuntu One' (affects: 7) (dups: 7) (heat: 68)" [High,Triaged] https://launchpad.net/bugs/798979 [13:45] facundobatista: yey? [13:45] nessita: the second is really easy [13:45] ralsina: from my pov both are the same [13:45] nessita, dobey, note that after the quick fix it now fails with NoAccessToken: CredentialsNotFound [13:46] nessita: basically, disable the whole tab, add a child widget to the tab (the square), place it in the middle, that's it [13:46] nessita: the first one is the same, but don't disable, make the child widget cover the whole tab [13:46] ralsina: ack, I'll try. I'll build the widget so you can later reuse [13:47] The child widget can be just a plain QLabel with a QMovie in it [13:47] ralsina: so we don't duplicate work [13:47] nessita: here's a C++ example http://www.developer.nokia.com/Community/Wiki/CS001434_-_Creating_a_loading_animation_with_GIF,_QMovie,_and_QLabel [13:48] facundobatista: that's fine. i'm working on a fix for that :) [13:49] nessita: but that seems quite complicated for what it does, though [13:50] standup in 10' people [13:51] * mandel back [13:51] nessita: +1 on greet-and-report-quota [13:52] thanks! [13:53] dobey, great, thanks [13:57] standup in 3 [13:59] hello! [13:59] yo alecu [13:59] moi [14:00] Hola alecu [14:01] me [14:01] mandel, thisfred, dobey, ralsina, alecu standup [14:01] me [14:02] me [14:02] alecu, dobey? [14:02] thisfred: ? [14:02] me [14:02] me [14:02] me [14:02] ok go me [14:02] WORK IN PROGRESS [14:02] rst port of the FAQ pages [14:02] Blocked [14:02] * nope [14:02] nessita: go [14:02] DONE: bug #800683, failed attempts to have syncdaemon running on win XP due to SSO failures, meetings [14:02] TODO: Review Lisette's proposal for "loading" messages, try to have XP working with SSO? nail some other UI bug? [14:02] BLOCKED: freaking out [14:02] NEXT: ralsina [14:02] Launchpad bug 800683 in ubuntuone-control-panel "Storage use and user greetins should be shown in the QT control panel (affects: 1) (heat: 6)" [High,Fix committed] https://launchpad.net/bugs/800683 [14:03] DONE: Windows catchup call, Windows+design call, bunch'o'reviews, fixed bug #800381 (needs 2 reviews), fixed #800715 (needs 2 windows reviews), did most of bug #800359 (SSO on wizard) [14:03] TODO: Finish the SSO on wizard integration, start syncdaemon on wizard stuff, hopefully get that mostly done, then whatever I can fit in. [14:03] BLOCKED: not yet, but if syncdaemon doesn't work on windows today I will be [14:03] NEXT: mandel [14:03] Launchpad bug 800381 in ubuntuone-windows-installer "Add "sync now or later" page to the installer (affects: 1) (heat: 6)" [Undecided,In progress] https://launchpad.net/bugs/800381 [14:03] Launchpad bug 800359 in ubuntuone-windows-installer "Add sign in page to the installer (affects: 1) (heat: 6)" [Undecided,Triaged] https://launchpad.net/bugs/800359 [14:03] DONE: Partially complited all tests for window_sdtool. Looked at my windows XP vm and updated the setup env to later test txnamedpipes on it. [14:03] TODO: Complete the test and provide them as an example for nessite to use sdtool. [14:03] BLOCKED: no [14:03] COMMENT: Will be here longer than usually so I can help [14:03] dobey: go [14:03] λ DONE: update for amz on server, more work on bug #800294, closed old sso milestones [14:03] Launchpad bug 800294 in ubuntu-sso-client (Ubuntu) (and 2 other projects) "Support new fdo secrets API property names (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/800294 === mbarnett` is now known as mbarnett [14:03] λ TODO: finish bug #800294, fix more stuff, chase people [14:03] λ BLCK: None. [14:04] thisfred: your turn [14:04] la la la [14:05] DONE: Rethinking the launcher and doing some experiments with counters/quicklists TODO: turn that thinking into bugs and branches BLOCKED: nope NEXT: alecu [14:05] DONE: calls and more calls, worked on bug #800669, found bug #800887, a few reviews [14:05] TODO: calls, debug CredentialsManagementTool [14:05] BLOCKED: no kinder still; but less planned interruptions than yesterday [14:05] NEXT! [14:05] sry [14:05] Launchpad bug 800669 in ubuntu-sso-client "InsecureJelly exception thrown when SD requests credentials (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/800669 [14:05] Launchpad bug 800887 in ubuntu-sso-client "UbuntuSSOClient fails when starting the SSO process (affects: 1) (heat: 6)" [High,Triaged] https://launchpad.net/bugs/800887 [14:05] thisfred, "doing some experiments with counters/quicklists" <- cooool! [14:06] alecu: yeah, for new udfs I think counters may make more sense than the messaging menu for instance [14:06] thisfred: nice! [14:06] and the quicklists for all subscribed folders, I'm thinking [14:07] any closing comments? [14:07] nope [14:08] Be excellent to each other! [14:08] looking for closure? [14:08] ok dont be nice be excellent [14:08] dobey, thisfred we need to talk about keeping u1 working on oniric, could you guys give me a status update on that by email? [14:08] ralsina: its broken on my machine in 11.10 [14:08] ralsina: uh ok: I have not been able to install O yet [14:09] will try again on the laptop [14:09] I mean talk between you then send me a short list of things that are completely broken. Ok, thisfred, then just dobey I guess ;-) [14:09] or if you can get it installed today, send me the report late today [14:09] ralsina: uh, after i fix these tests again in my branch, and it gets approved/landed, we should be good to go if facundobatista's fix has already landed [14:09] sure [14:10] dobey: cool then, it should be a very short report [14:10] dobey: also we need to squeeze "port everything that stays gtk to gtk3" somewhere in the next 4 or 5 months :-( [14:11] and from pygtk to the (currently semibroken, right) gtk-via-introspection-thing [14:11] which is mostly not really an issue at this point [14:11] couple of minor things to change, but we should be good for the most part === m_conley_away is now known as m_conley [14:12] dobey: cool, not bad there then. Hopefully things won't diverge too quickly [14:13] the only big thing with that is the banshee/rhythmbox issue [14:13] because banshee is not gtk3 yet, and rbox is [14:13] ralsina, mandel, alecu, Chipaca: mumble in 2 minutes? [14:13] ralsina: 4 or 5 months? I thought we had like 3 more weeks or so :) [14:13] nessita: ack [14:13] nessita: no can do, got the web thing in 2 minutes [14:13] thisfred: well... that's one way to look at it. gtk2 is leaving the CD n the next cycle [14:13] nessita: at 15 pass, sure [14:13] thisfred: well, more like 2 months [14:13] Chipaca: ah, ok. How long then? we can reschedule? [14:14] I've got the checkpoint on my calendar at :30 [14:14] that's a bit tight, i'd rather :45 [14:14] thisfred: OTOH the nautilus plugin probably will have to be ported in three weeks yes [14:14] Chipaca: ok, :45 it is [14:14] and there's that whole *other* issue to deal with [14:14] mandel, alecu, ralsina: mumble at :45 (in 30 minutes) [14:14] ralsina: the nautilus plug-in already works, it just gets installed to the wrong directory, i think [14:14] dobey: you mean the shiv? ;) [14:14] nessita: ack^2 [14:15] :-) [14:15] thisfred: yes [14:15] ack [14:15] nessita: sure [14:15] shivs and shanks [14:15] dobey: well, that is good news [14:15] * ralsina should get oneiric installed too :-( [14:16] nessita, dobey, https://code.launchpad.net/~facundo/ubuntuone-client/avoid-double-watch/+merge/65661 (feel free to review) [14:19] facundobatista: interesting diff. :) are there no unit tests for that block of code? [14:19] dobey, it's just a removal of a duplicated line... but the duplicated line is not even in that file [14:20] (this is just dirtiness left ages ago, when VM took the responsibility on setting up 'root') [14:20] right [14:20] mandel, any ideas on this? bug #801145 [14:20] Launchpad bug 801145 in ubuntu-sso-client "I won't stop right on windows/txnp (affects: 1) (heat: 6)" [Critical,Confirmed] https://launchpad.net/bugs/801145 [14:21] facundobatista: approved [14:21] alecu: let me take a look [14:22] ralsina: ^^ one branch down for oneiric :) [14:22] dobey: yay! [14:23] dobey, thanks [14:24] alecu: the bug is in line 56 of bin/windows-ubuntu-sso it is telling the reactor to stop when it should be telling that to the QApplication and not the reactor [14:24] alecu: shutdown function here: http://bazaar.launchpad.net/~ubuntuone-control-tower/ubuntu-sso-client/trunk/view/head:/bin/windows-ubuntu-sso-login [14:24] mandel, nice, thanks [14:24] alecu: np [14:27] mandel: I am integrating the SSO into the wizard and it is working, but I have the strange feeling that I am using it completely wrong. Could you take a quick look at https://code.launchpad.net/~ralsina/ubuntuone-windows-installer/fix_800359 ? [14:28] ralsina: sure,, on it [14:29] mandel, basically http://bazaar.launchpad.net/~ralsina/ubuntuone-windows-installer/fix_800359/view/head:/ubuntuone_installer/gui/qt/gui.py [14:31] ralsina: looks good to me, you are passing the controller to the ui so that it hooks it up, that is the idea [14:32] mandel: ok [14:32] mandel: I am not really happy because I am putting lots of logic in something that depends on Qt but well, it's much harder the other way around. [14:32] ralsina: what do you mean? [14:33] oh, wait, all the controller stuff in sso is qt-specfic! Nevermind :-) [14:35] dobey, "There are additional revisions which have not been approved in review. Please seek review and approval of these new revisions." [14:35] dobey, and that's only because nessita gave me an approve after you approved the MP [14:36] dobey, I'm approving the MP again... but it seems that Otto is wrong there [14:37] * nessita wanted to help [14:37] ralsina: that is the entire point of the controller, they have no notion of the framework, they just want to add cb to the signals :) [14:38] mandel: cool, hacking it some more then ;-) [14:39] facundobatista: no i think what happened was that the empty commit you added for --fixes showed up after i'd approved the branch (so the old revision was approved, and not the new), and launchpad was being a bit slow, and i wasn't expecting another revision to show up :) [14:39] nessita: had nothing to do with your +1 [14:39] ok [14:40] dobey, nop, my empty commit is 22 minutes ago, and your approve is 18 minutes ago [14:40] facundobatista: yes, but that has nothing to do with when launchpad actually completed the rescanning [14:41] dobey, mmm... and Otto checks when the scanning was done, and not when the commit was done? [14:41] facundobatista: if the lp page had no idea of the second commit at the time, then when i approved, it couldn't have chose the later revno, but only the latest one it knew at the time. [14:42] facundobatista: tarmac just tries to merge the revision that is approved in launchpad. if that revision is not the tip of that branch, it complains [14:42] dobey, ok [14:43] facundobatista: and launchpad can be slow at times, and it doesn't help when unexpected revisions pop up :) [14:43] HUZZAH! [14:43] PASSED (skips=1, successes=494) [14:46] facundobatista: also, there are two rescan processes. the rescan of the branch for the bracnh data page in lp, and the rescan/regeneration of the diff for the merge proposal. and the merge proposal one happens after the branch one, and is the one we have to wait for, so that the revisions are all up to date there. [14:47] nessita: can you re-review https://code.launchpad.net/~dobey/ubuntu-sso-client/new-kr-props/+merge/65419 please? :) [14:47] nessita, mandel, ralsina: I see that Chipaca is already on our mumble channel [14:47] shall we? [14:47] rye: hi [14:48] ok, let's! [14:48] it's :45 :) [14:48] bah, it was when i mvoed in [14:48] can you help me with this https://wiki.ubuntu.com/UbuntuOne/RunClientFromTrunk ? [14:48] adorilson, hi [14:49] adorilson, hm, this is first time i see this page but make install will most likely not work [14:49] adorilson, it can be ran w/o installation, that's what chicharra team does, actually [14:50] hrmm, yeah, i'd avoid the make install bit there [14:50] Thanks to Roman??? Hmmmm [14:51] rye: someone told me you had written this page. did not you? [14:51] maybe he was talking about Pilot [14:51] i don't think rye wrote that page :) [14:52] at least, i don't think he's crazy enough yet to be thanking himself on his own wiki pages [14:52] hmm, update-manager fails wil an OSError (No such file or directory) ... [14:53] adorilson, i am listed there as "Thank you" but i don't recall having the discussion of running the client from trunk, other than that it is really simple - bzr fetch for protocol and client, build client specifying the protocol location and then run the client specifying correct PYTHONPATH [14:54] and now, my laptop is AGAIN in limbo between natty and oneiric [14:54] facundobatista, by the way, i have now a test client (natty, latest updates, no nightlies) that cannot upload flv file because it claims ubuntuone.SyncDaemon.StorageClient - TRACE - start - processMessage: id: 13 type: ERROR error { type: UPLOAD_CORRUPT comment: "Deflated size does not match hint." } [14:55] rye, size? [14:59] facundobatista, 96380073 bytes, BUT, this is completely different to my current issue of 180Mb file which keeps being uploaded on 100% and hangs, this just fails with upload corrupt and requires client restart to try uploading again [14:59] rye, and it always happen? [15:00] verterok, ^ we did have an issue with persistent corruptness in resumable, is it fixed now? [15:00] rye, see this http://paste.ubuntu.com/631200/ please [15:00] facundobatista: yes, we have...and isn't fixed yet [15:01] rye: could you add the node details to bug #778805 [15:01] verterok: Bug 778805 on http://launchpad.net/bugs/778805 is private [15:01] ? [15:01] verterok, ok, thanks [15:01] rye: upload_id, share_id, node_id and hash please :) === karni_ is now known as karni [15:04] verterok, added [15:04] rye: thanks! [15:05] adorilson, apt-get install gobject-introspection [15:07] hmm [15:15] * fagan break for an hour [15:16] joshuahoover: Looks like ill take a long day to get this done. I have a few done now but there are good few pages [15:16] fagan: ok, thanks for the update...can you make sure you commit what you have done to your branch by eod (for you)? [15:17] joshuahoover: sure will do, ill hang around till around 9 UTC id say. [15:17] thanks :) [15:17] alecu: fix the lint warnings before setting back to approved! [15:19] rye: installing. but I need to tell a secret: I'm on Fedora 15 [15:20] adorilson: it may be prudent to make RPMs in proper order :) [15:21] adorilson, another thing is that you will need to have ubuntu-sso package too since syncdaemon depends on it to get the credentials [15:21] well, will need a lot of things [15:21] dobey: I searched, but I dont found ubuntuone's RPMs [15:22] adorilson, or, if you like hack-driven approach to make it work first and think how make it better later you may want to use the command line script to get the credentials from Ubuntu SSO and put them to oauth= config variable [15:22] adorilson: right. keyword in my statement was "make" :) [15:25] dobey: sure. but I dont have a minimal idea how to make RPMs [15:26] rye: gobject-introspection is already installed. [15:27] rye: any ideia? [15:29] adorilson, do you have pkg-config? try pkg-config --print-errors "gobject-introspection-1.0" --modversion [15:31] rye [15:31] $ pkg-config --print-errors "gobject-introspection-1.0" --modversion [15:31] Package gobject-introspection-1.0 was not found in the pkg-config search path. [15:31] Perhaps you should add the directory containing `gobject-introspection-1.0.pc' [15:31] to the PKG_CONFIG_PATH environment variable [15:31] No package 'gobject-introspection-1.0' found [15:32] dobey, great, thanks. [15:32] rye: one way to workaround the upload corrupt bug, is to move the file out of u1, wait for the client to delete the file on the server, and move it back [15:32] verterok, this file has not been uploaded to the server for this account [15:33] rye: the file was created, but not uploaded [15:33] rye: the "node" exists in the server, but with no content [15:33] rye: gobject-introspection version is 0.10.8 [15:33] rye: moving the file out of u1 will trigger a unlink of the node, so the stalled upload will be cleanedup [15:33] adorilson: you need the -devel packages of things [15:35] dobey: installing gobject-introspection-devel... [15:37] verterok, RROR error { type: UPLOAD_CORRUPT comment: "Deflated size does not match hint." [15:38] verterok, the file is already uploaded under a different user [15:38] hrmm [15:38] rye: if the same file is already uploaded it shouldn't upload it again [15:38] rye: magic uploads! [15:39] rye: so, it might be in a different shard [15:39] verterok, the original file was uploaded this february, before magic [15:39] rye: ah, ok [15:39] * alecu will be afk for 30' [15:40] rye: try moving it out of u1, wait for the client to delete/unlink it, and move it back [15:41] verterok, for the original account? [15:43] rye: from the account you'r getting the upload corrupt error [15:43] rye, dobey: what is lndir ? [15:44] verterok, already done that, unlink sent to the server, sd got IDLE, moved the file back in... UPLOAD_CORRUPT [15:45] rye: it tried to resume the upload? [15:45] rye: look for BEGIN_UPLOAD message in the logs, offset should be 0 [15:45] adorilson: install xutils-devel [15:46] verterok, BEGIN_CONTENT begin_content { offset: 0 [15:46] although we can probably remove that at this point [15:46] rye: ok, so this isn't the same bug...something else is going on [15:47] facundobatista: ^ isn't resuming the upload...so isn't the stalled resumable uploads issue :/ [16:04] can someone on natty send me some screenshots of the contacts picker? [16:05] rodrigo_, contact picker from ubuntuone? [16:05] joshuahoover: the links will probably be wrong is that ok [16:05] rye, yes [16:06] fagan: links to? [16:06] joshuahoover: the other pages [16:06] alecu: ping [16:07] fagan: maybe muffinresearch has ideas on how to handle links (if there is any way to handle it better than just leaving them be wrong for now) [16:07] joshuahoover: yeah he would be the one to ask [16:07] muffinresearch: ^ [16:07] oh and I just finished the first 2 sections [16:08] fagan: do you have examples? [16:08] muffinresearch: sec [16:08] rodrigo_, http://ubuntuone.com/p/10om [16:09] muffinresearch: https://wiki.ubuntu.com/UbuntuOne/Tutorials/Setup/Maverick#Already%20have%20an%20Ubuntu%20One%20account [16:09] rye: ugh, contacts picker -- [16:09] taht kind of thing, should I just leave it or should I guess what the link would be or something [16:10] rye, thanks, but don't you have more contacts? It's for showing other people what it does :) [16:11] rodrigo_: i think he was pointing out the disconnect between the search field contents, and that there are any results at all :) [16:11] dobey, :) [16:11] fagan: the faq itself will be at /help/faq/ - The table of contents as shown on that page should be auto generated by ReST using the directive for that. http://docutils.sourceforge.net/docs/ref/rst/directives.html#table-of-contents [16:12] muffinresearch: hmmmm well in the interest of speed ill leave them and sort out all the links tomorrow [16:13] ill give it a shot anyway and if they are wrong you guys can sort it since its a small thing to fix [16:15] rodrigo_, well, i can create more contacts [16:15] rye, please, if it's not too much hassle :) [16:15] rye, although I'd prefer to get contacts with photos [16:15] dobey, do you have? [16:16] rodrigo_: i have no friends dude :) [16:16] heh [16:16] no even ennemies with photos in your addressbook? :D [16:16] so, anyone else please? [16:16] rodrigo_, hm, the photo part is tricky, replication is disabled for my shard [16:16] no [16:16] rye, can you send me the logs of that session? thanks [16:16] but i don't see why it matters [16:17] the bug i'm seeing in that screenshot has nothing to do with photos [16:17] rye, don't worry then, I'll try to find someone with contact photos [16:17] dobey, it's for showing people designing a contact picker for gnome how ours look like [16:17] so yes, I'd prefer photos [16:18] I have photos myself for lots of contacts, but u1 doesn't work on oneiric yet [16:18] joshuahoover might have it [16:18] joshuahoover, do you have photos for contacts that are displayed in the contacts picker in nautilus? [16:19] rodrigo_: so rye wasn't showing you a bug in that screenshot? [16:19] dobey, nope [16:19] oh, because all i saw was "look at this bug" :) [16:19] rodrigo_: let me check [16:19] dobey, no, not a bug [16:20] rodrigo_: doesn't look like it...hmmm [16:20] ok [16:21] rodrigo_: the thing is, almost nobody has photos of contacts in evolution [16:21] dobey, I have, they get imported from facebook [16:21] not for u1 contacts though [16:21] rodrigo_: the only people that do are people who added photos to demo the fact that photos can be used in evo :) [16:21] :) [16:22] rodrigo_: how did you get fb contacts in evolution if not through u1? [16:22] joshuahoover, ok, so can you take a screenshot please? [16:22] dobey, via pidgin plugin, iirc [16:23] oh, weird [16:23] rodrigo_: i don't have any photos or contacts here :( [16:23] joshuahoover, I'll replace it with one with photos as soon as u1 works here [16:23] rodrigo_: only via the web [16:23] joshuahoover, doesn't mind, if you have more than 1 contact, it's enough :) [16:23] rodrigo_: ah, ok...let me get that for you then [16:23] rodrigo_: didn't the design mock-up have fake contacts with photos? [16:23] joshuahoover, cool, thanks [16:24] dobey, the design mockups are different from what we implemented [16:24] and they are not public, afaik [16:24] aren't they? [16:24] i thought they were, but no idea [16:25] *shrug* [16:25] i'm surprised gnome isn't just using the OSX design [16:25] https://live.gnome.org/action/subscribe/Design/Apps/Contacts <- this is what's up for now [16:25] although, i'm sure they will settle on that, but it will have the gtk3 theme [16:26] yeah i've seen that [16:26] but i never did like os x [16:26] me neither :) [16:28] what happened to nessita? [16:28] rodrigo_: http://ubuntuone.com/p/10ow/ [16:28] dobey: I'm here, why? [16:29] nessita: can you re-review https://code.launchpad.net/~dobey/ubuntu-sso-client/new-kr-props/+merge/65419 please? [16:29] dobey: sure, in a couple of minutes [16:29] joshuahoover, thanks! do you have any problem in making that screenshot more public? that is, adding it to https://live.gnome.org/action/subscribe/Design/Apps/Contacts ? [16:29] nessita: you've been silent for the last hour. it is unlike you :) [16:29] nessita: thanks [16:29] rodrigo_: no, that's fine...just names shown there [16:29] dobey: I'm coding like a monkey [16:29] joshuahoover, ok! [16:32] lunch time, bbiab [16:33] dobey: yum dont found xutils-devel. And I can't found in another place. :-/ [16:35] runtime directory /run/udev not writeable. That sounds ungood. [16:36] verterok:do you know what is the bug number for 'persistent corruptness in resumable'? I have a couple bugs with that issue and have been looking for a bug to dupe them too. [16:36] duanedesign: depends on what 'persistent corruptness in resumable' means :), looks like we have 2 different issues [16:36] verterok: aha [16:36] verterok: glad you said that. i was thinking we might looking at the conversation [16:37] duanedesign: if the client get a BEGIN_CONTENT with offset > 0 and the upload fails with UPLOAD_CORRUPT, it's bug #778805 [16:37] verterok: Bug 778805 on http://launchpad.net/bugs/778805 is private [16:37] duanedesign: you need TRACE logs to see the BEGIN_CONTENT msg [16:39] ahh. you predicted my next question :) [16:40] thank you for info [16:40] heh, np :) [16:44] nessita: running the test on linux and will ask you to re-review [16:44] mandel: ack [16:47] ok, my laptop is severly screwy now. can anyone with a working oneiric send me the output of dpkg --get-selections ? [16:52] mandel, pong. sorry, was afk [16:55] alecu: did the change in the shutdown function work? [16:55] mandel, I'm just working on that right now. I'll let you know in 15' [16:56] alecu: cool, the idea is that you just get the QApp instance and close it there, and there should be no need to tell the reactor to do it at all [16:57] mandel, cool, I understood. [16:57] :) [16:59] ralsina: ping [16:59] nessita: pong [17:00] ralsina: in GTK I can connect a signal to a widget and pass extra arguments like this: button.connect('clicked', my_callback, data1, data2, data3, ...) and then, when the button is clicked, my_callback will be called with the follwoing params: button, data1, data2, data3, ... [17:01] ralsina: can I do something like that in PyQt? I've tried with QtCore.QObject.connect(button, ... but I can't pass extra args [17:01] no, no extra args [17:01] you can use a lambda [17:01] ralsina: hum... no I can't... [17:02] button.clicked.connect(lambda : recipient(bla, foo, bar)) [17:02] I'm inside a loop, and I need the extra params (bla, foo, bar) be those from the loop [17:02] so, I have code that looks like: [17:02] for bla, foo, bar in data: [17:02] QtCore.QObject.connect(button, QtCore.SIGNAL('clicked()'), lambda: my_callback(bla, foo, bar)) [17:02] that will not work [17:03] bla, foo, bar will always be the last ones from the loop run [17:03] nessita: pass them as arguments to the lambda: lambda foo=foo, bar=bar: recipient(foo, bar) [17:03] that way, the foo inside the lambda is the name of the arg, which is bound to the current value in the loop [17:04] trying that, but is EXTREMELY DISGUSSING :-) [17:04] disgusting * [17:04] nessita [17:05] yes? [17:05] nessita, at least it's not lambda bar=foo, foo=bar: f(foo,bar) ;-) [17:05] does not seems to work... and I can see why [17:05] ah no, wait [17:06] there are cleaner solutions using signalmapper but it's really verbose [17:06] it works [17:06] thanks [17:06] * ralsina remembers doing this [17:06] is uglier than mestre [17:06] glad to help! [17:07] it's poor man's python closures [17:08] dobey: reviewing [17:09] mandel, so, after calling app.quit it just wont finish [17:10] mandel, I even tried calling a shutdown function that calls app.quit and then reactor.stop, just in case. [17:11] alecu: bummer… that looks like there is a thread being evil somewhere... [17:11] mandel, absolutely. It must be the threaded reactor thread. [17:11] mandel, but for now, we'll raise the value of the sso timeout to infinity-1 [17:12] mandel, so the sso client will keep running forever and that's it. [17:12] mandel, we'll fix this properly later. [17:13] alecu: sure.. it should not be a major issue for the time being [17:25] nessita, alecu: please can you look at https://code.launchpad.net/~mandel/ubuntuone-client/windows_sdtool/+merge/65493 [17:26] yes [17:29] dobey: branch works very good, approved [17:29] dobey: I would like alecu's doing the second review [17:29] nessita: great, thanks [17:29] sure [17:29] alecu: can you review https://code.launchpad.net/~dobey/ubuntu-sso-client/new-kr-props/+merge/65419 please? [17:29] sure [17:33] dobey, why do you abbreviate collection as CLXN? [17:35] alecu: PEP8 [17:35] mandel, I'll review your megabranch after lunch. [17:35] alecu: sure, np [17:36] mandel: instead of raise Exception('Not implemented.') [17:36] mandel: can you please raise NotImplementedError? [17:36] sure [17:38] mandel: and class ErrorSignal(Exception): needs a docstring (and then remove the "pass") [17:39] nessita: can you write all this in a comment so that we can keep track? [17:39] mandel: I could, but adding a needs fixing and waiting for your fixes use to take too much roundtrip [17:39] mandel: if I stare at you while you fix this it goes faster :-D [17:40] dobey, I don't see anything on pep 8 regarding that kind of abbreviations. But I find it confusing that in some constants it's spelled COLLECTION, and on others CLXN [17:40] nessita: ok [17:40] dobey, would you mind changing the CLXN to COLLECTION? [17:41] alecu: line length limit [17:42] mandel: let me know when those 2 are fixed and pushed [17:42] mandel: should be quick, right? [17:42] alecu: having to do 'FOO = \\\n"some long string"' is lame [17:43] alecu: and there isn't really any raeson for those to be used outside of txsecrets.py anyway, aside from the fact that the tests also use them [17:43] nessita: yes, just done it and pushing, I'll ping you when done [17:43] thanks! [17:43] I'll have lunch after that [17:45] nessita: done, r1061 [17:49] nessita: ran into some trouble making SSO show the right buttons on the wizard, will take another 60, 90 minutes [17:50] ralsina: OH NO! what are we doing now! NOOOOOO :-P [17:50] ralsina: ok, let me know :-) [17:50] ralsina: need help? [17:50] nessita: sure. I am trying to make it good enough not to have to work on it again tomorrow after design sees it ;-) [17:50] good choice [17:50] nessita: Every time we authenticate the user (and then save oauth tokens, naturally) we add some, name it, host token - "Ubuntu One @ foobar". Is there a way to authenticate a user and reuse such token instead of creating new oauth tokens if the user, say, uninstalls U1 software, and then installs it again? [17:51] dobey, why did you change the out_signature of SecretServiceMock.CreateCollection ? [17:51] nessita: I'm trying to be precise, but it's a tricky question that blajk just asked. [17:51] mandel: approving. When I get a windows env I'll test it IRL, and I'll file bugs if necessary [17:52] karni: nopes [17:52] nessita: Say, we authenticated with "Ubuntu One @ foobar" token once. We remove the software completely. Then we install it and re-authenticate - can we request previously generated oauth tokens or we need to create new once? [17:52] nessita: ok, I'm writing the script you asked atm [17:52] nessita: So a new one is neccessary, right? [17:52] nessita: That's what I thought. Thanks! [17:53] karni: you're welcome! [17:53] :) [17:53] dobey, the code looks fine otherwise. The only doubt I have is with that out_signature change. [17:53] * alecu will have lunch now. [17:53] alecu: one second my firefox went nuts [17:54] joshuahoover: did a big push of a few pages [17:54] joshuahoover: about halfway there [17:54] fagan: cool, thank you! [17:55] joshuahoover: you know where the branch is? [17:56] fagan: yep [17:56] joshuahoover: cool ill do some more then and should be almost done for tomorrow morning [17:56] alecu: hrmm, no idea. changed it back, and pushed [17:56] fagan: excellent [17:57] ok changing location brb [17:57] ]/quit [17:59] alecu: this is new: http://paste.ubuntu.com/631291/ is it on purpose? [17:59] * mandel walks dog, will me back in 30 min [18:11] ok, lunchtime! [18:18] ok, no luch yet, firs, merge proposal: https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/explore-folder/+merge/65697 [18:18] can I have a couple of reviews for? ^ please [18:18] now yes, lunchtime! [18:24] * thisfred reviews [18:26] i'm not good at doing reviews. rebuking on the other hand, i am ace at. [18:32] dobey, there are a few error messages shown while running the tests that are not shown on trunk: http://pastebin.ubuntu.com/631316/ [18:33] dobey, the tests pass fine though. [18:33] alecu: maybe that's why i changed it to v. no idea why those are happening, or why they're not causing failures [18:35] hrmm, though those must be coming from a different call [18:35] mandel: ping [18:35] ralsina: pong [18:36] mandel: not for now, not for soon, but using buttons inside the pages for controlling the wizard is not really good for integrating it with other code [18:36] ralsina: I agree, the issue here is the design.. [18:37] for example, the same effect can be achieed (as long as you hae less than three "action" buttons, using the custom buttons of the wizard view [18:37] ralsina: the problem is the location, not the number of buttons, that I know how to do [18:38] ralsina: the actual best thing for you to do if it is a problem is to reuse the controllers and just provide the same interface for the UI [18:38] mandel: well, the thing is ... I prefer them to be located in the wizard :-) [18:38] mandel: yes, I will probably do that eventually [18:39] ralsina: the problem is that certain buttons they placed in the UI cannot be in the wizard, but have to be in the page, remember the wireframes? [18:39] mandel: yes for some, but for example "create account" on the signup page is clearly a "Next" in disguise [18:40] anyway, I think it's looking ok, een if it doesn't quite follow the wireframes to the letter [18:40] ralsina: if you but it as a next, how would you ensure that is placed where they wanted? [18:41] ralsina: you would have to play with the button layout of the wizard that can be everything but flexible [18:41] it is at the bottom right corner, where the next button usually is. I mean in the page with the captcha [18:42] but yes, the "choose sign in" page's wireframes mean no wizard buttons there [18:42] dobey, it looks like it's the "Get" dbus methods on the Mock, being called with the wrong propname. They don't match the propname, and end up returning None, which can't be converted to a dbus value, and this error is caught by your new code and ignored (and that's fine). Please open a bug for this, and assign it to me. [18:42] ralsina: exactly, so we have to fight between design and the QWizard implementation [18:42] mandel: exactly [18:43] which is why, I will probably (later) sit down with design and move a few buttons around :-) [18:43] alecu: ah right. guess I should make it raise an exception [18:44] ralsina: feel free to do so :) [18:44] mandel: hahaha [18:44] dobey, right, rising a new "NotFoundException" would be the right thing there. [18:46] alecu: well i was thinking a custom "InvalidProperty" exception [18:46] alecu: ping [18:46] or DBusException("Invalid properties") or something [18:47] dobey, "InvalidPropery" sounds fine. It will be translated by dbus automatically into a DBusException on the other side. [18:47] right [18:47] * dobey fixes that [18:49] dobey, approved [18:50] mandel, still need the huuuuuge review? [18:50] alecu: yes, but is not for that, go a question, when I try to run sd on my machine, which used to work, I now get the following: http://paste.ubuntu.com/631291/ [18:51] alecu: looks like credentials related, which is strange.. sd should be able to run even without the credentials, the state machine does include the state (with no user, with network) [18:51] mandel, oh, right. [18:52] alecu: great, thanks [18:53] alecu: oh, right? suena peligroso ;) [18:53] mandel, "peligroso" is my middle name. [18:54] alecu, mandel: syncdaemon has a setting which is autoconnect [18:54] so it will ask the keyring for the credentials [18:54] mandel: do you have SSO working? [18:54] running, I meant [18:55] nessita: no, it was not running, but I wanted to check if the sd would work from trunk, it should be able no matter if there are creds or not [18:55] * mandel tries with sso [18:55] alecu: I know what's mandel issue, he updated his syncdaemon to the latest, which now handles credentials "properly" (after your fix) [18:56] mandel: the thing is that before alecu's branch, the connect() method was buggy and was not doing what is should, which is, seek for credentials [18:56] nessita: so why do we have a state in the state machine that allows to have no user with network, it does not make much sense.. [18:57] does it? [18:57] mandel: so, it was working before for you because connect() was doing kinda nothing [18:57] mandel: it does not make sense, but you can not connect by setting autoconnect to False in the conf file [18:57] I does* make sense, I mean :-D [18:58] alecu, nessita: ok, so now I get the jelly serialization error.. whic I'm guessing we have to fix [18:59] mandel: please do not overlap work with alecu. I think you should work on having the start() method implemented [19:00] alecu: you're dealing with the jellyfish, right? :-) [19:00] I am [19:00] I am the walrus [19:00] great [19:00] choo-choo-choo-choo [19:00] mandel: so, you work now on the SDTool.start() [19:00] yes? [19:01] nessita: yes, I wanted to test sdtool agains trunk, but seems impossible… [19:01] mandel: WELCOME TO MY WORLD :-P [19:01] and alecu's [19:02] * mandel cries 'but it was working….' [19:03] mandel: using hard code credentials does not mean it was working ;-) [19:04] coded* [19:04] or hard core credentials, as well [19:07] details, details... [19:07] heh [19:07] mandel, I found the solution to the jelly, but the branch is missing some bits still [19:08] mandel, anyway, I'm reviewing your megabranch right now. [19:08] alecu: awesome, what was the issue for the jelly thing? [19:09] mandel, self.sso_proxy.find_credentials(APP_NAME, {}, reply_handler=reply_handler, error_handler=error_handler) [19:10] mandel, both reply_handler and error_handler were set to NO_OP [19:10] mandel, but sso_proxy was a pb proxy, not a dbus one [19:10] mandel, alecu, nessita: Can I get two kind reviews for https://code.launchpad.net/~ralsina/ubuntuone-windows-installer/fix_800359/+merge/65703 [19:10] mandel, so both NO_OP ended being sent thru pb as keyword parameters [19:10] alecu: yes, but there is a decorator that should ensure that the callbacks are Referenceable objects, right? [19:11] ralsina: I'll get one [19:11] mandel, well, they were being *sent* thru pb [19:11] mandel: let's trade https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/explore-folder/+merge/65697 [19:11] nessita: there is no test for the constants because the tests don't work on any platform. On Linux SSO doesn't install teh qt stuff, and on windows I get a reactor problem when trying to use u1trial [19:12] mandel, instead, they should not be sent thru pb, but instead the corresponding callbacks are already being connected to the remote signals. [19:12] mandel, so, some bits are missing there, and a bit more debugging. [19:12] mandel, but please let me finish your megareview first. [19:12] ralsina: the tests DO work on linux, we ran them yesterday [19:12] alecu: true, the pb returns a deferred [19:12] ralsina: what error you mean by "On Linux SSO doesn't install teh qt stuff"? [19:12] mandel, pb returning a deferred has nothing to do with this! [19:12] nessita: importing sso qt stuff breaks them [19:12] ralsina., nessita: I'll take a look at your proposals [19:13] ralsina: let's set a pythonpath [19:13] ralsina: and on windows, what reactor problem are you having? [19:13] nessita: since ubuntu-sso-client doesn't install the SSO things, tests don't run. It doesn't even compile the .ui files unless you hack setup.py [19:13] mandel, that deferred returned by pb only means that the find_credentials was correctly handled on the other side. [19:13] mandel, it is not the return value of the find_credentials. [19:13] nessita: will show you in 2' [19:13] oops/SSO things/qt things/ [19:14] mandel, the return value of find_credentials is sent thru a signal; but that signal should be "subscribed" first with "register_to_signals". [19:14] nessita: I get "The specified reactor is not supported." on windows. With both qt4 and txnp reactors [19:14] alecu: yes, you are right [19:14] mandel, and btw: register_to_signals/unregister_to_signals is braindead to me as an API. [19:14] ralsina: are you u1devtools updated? [19:15] ralsina: you do not have them installed, you are getting an import error and is printing that [19:15] mandel, so, bits and pieces missing. I'll keep working on it later. [19:15] nessita: trunk [19:15] ralsina: are you pulling stuff from wrong PYTHONPATH? [19:15] alecu: ok, I'll get back to reviews and sdtool [19:15] ralsina: do you have them installed, like mandel says? [19:15] mandel: ok, will look into the tests then [19:15] ralsina: right, no test is not an option (still! :-)) [19:15] ralsina: also, as dobey said, check your path is ok [19:15] I am pretty sure I have the qt4 one installed since it works ;-) [19:16] but yes, could be PYTHONPATH crazyness [19:17] I would like some IRL testing to know if it only works for me though :-) [19:17] ralsina: what thing may work only for you? [19:18] nessita: the whole ubuntuone-windows-installer thing [19:19] ralsina: I'll do an IRL, but I would like you to work on adding tests, and having them running. Since the next work is going to be more complex and we need to have tests for those for sure. [19:19] nessita: yep, tests the rest of the day, in principle [19:20] and if sdtool is looking good, start the syncdaemon/cntrolpanel integration tonight [19:21] ralsina: well, ideally I would expect you to have the tests done before you eod, and then start integrating u1cp widgets. You think that's doable? [19:21] nessita: what I said [19:21] ralsina: ah, I understood something else [19:21] I will have the tests on before EOD, and then in my couple of late hours will look at u1cp widget integration [19:22] * ralsina has two eods a day now ;-) [19:22] grr lp [19:22] ralsina: ah, thanks for the clarifiaction, I misunderstood before [19:22] ralsina: sounds great then [19:22] yay finally [19:23] anyone need a second review on something? I dn'twant to code for 3 0 minutes :-) [19:24] ralsina: did you do mine? :-) (LP does not show any review from you) [19:25] nessita: probably got an idea on how to fix something and forgot. Link? [19:26] https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/explore-folder/+merge/65697 [19:26] nessita: ok, on it for real this time [19:27] thisfred: I share your concerns from the review, but all those will be cleared up with the design team when they see this [19:27] nessita: I just finished this : https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/explore-folder/+merge/65697 [19:27] nessita: that's what I figured, so I approved [19:28] mandel: oh thanks [19:28] ralsina: no need then [19:29] nessita: oooook then. [19:29] how can I have a cmd.exe on w7? start -> run command is gone and I can't find it in less than 3 seconds [19:29] (so I better ask the experts here) [19:29] nessita: yes, windows key -> r and type cmd [19:30] or type cmd in the search line and press enter. [19:30] great [19:30] worked [19:30] Any other important reviews pending? [19:30] ralsina: line 253 in https://code.launchpad.net/~ralsina/ubuntuone-windows-installer/fix_800359/+merge/65703 is that print on purpose? [19:30] mandel: surely not. [19:30] nessita, if the windows key is captured by unity, try using "Ctrl-Esc" [19:31] alecu: is not, but thanks! [19:31] ralsina: also, line 268, if you do nothing with the exception, at least log it so that we know it occurs, right? [19:31] ralsina: most of the news docstrings have no ending dot, can you please add those [19:32] mandel: ok. The exception is a bug in PyQt when nothing is connected to the signal [19:32] nessita: ok. [19:33] ralsina: oh, then we should have a comment so that the next guy knows about it [19:33] mandel: it's ok, I will log it and add a comment [19:33] ralsina: lines 434 and 441, why did you comment out the imports? [19:34] I mean, if you do not need them then dd, right? [19:34] ralsina: going back to have the tests running on linux, is there a chance to have the running if we set PYTHONPATH to the proper ussoc source code, where all the QT stuff is available? [19:34] mandel: yes, will delete [19:35] nessita: possibly, but you need to edit setup.py to make it compile the .ui files [19:35] ralsina: the title passed to the controllers should be translated via getted, lines 469 and 483 [19:36] ralsina: but... ./setup.py is al;ready building the qt ui for sso. What do you mean exactly? [19:36] nessita: last I checked, on Linux, it didn't [19:37] nessita: I may be remembering wrong, of course [19:37] mandel, def is_running(bus=None): does not use bus for anything, right? [19:37] mandel, ralsina: if you use the proper constant from u1client no translation should be needed, right? [19:37] alecu: no, left it to be compatible with linux, no other reson [19:38] nessita: true, but is the header of the page, I don't know if we have that in the contansts… probably we do [19:38] mandel, oh, so the api is the same, right_ [19:38] ? [19:38] nessita: there is no constant for this, I think [19:38] ralsina: which string is it? [19:38] "Sign In" [19:39] alecu: yes, that is the idea, so that nessita does not have to do weird things with it [19:39] mandel, but start() in (the huge and perhaps wrongly named) SyncDaemonTool sends a self.bus to it. Which is not used in any other place... [19:39] mandel, (it should have been named SyncDaemonSwissTool) [19:40] alecu: hahaha [19:40] thank you, thank you. I'll be here for the remaining of the week. [19:41] alecu: ein? [19:42] mandel, SyncDaemonTool.start: "if not is_running(self.bus)"... but self.bus is not defined anywhere! [19:43] mandel, nessita: added periods, removed commented code, added comment for exception catch, translated "Sign In". [19:43] alecu: oh, that is a bug, let me check 'cause I though that was fixed... [19:43] dog is acting weird I think she needs to go, bbiab [19:46] mandel, another thing: reply_handler inside get_current_downloads already has access to "d" (thru the closure), so no need to receive it as an arg, and no need for the lambda that calls reply_handler [19:47] alecu: ok, let me check [19:48] alecu: true, that is ugly, I would remove the use of the lambda, what do you think? [19:48] mandel, that sounds perfect. [19:49] mandel, also, inside get_current_downloads too: since no error handlers are attached to downloads_d, any error would be lost. [19:49] mandel, it would be *much* safer to rewrite all this using inlineCallbacks. [19:49] mandel, but perhaps nessita will agree that we should better do this as a refactor later. [19:50] alecu: if those are 15-20 minutes changes, let s do them now [19:50] alecu: I agree with your suggestions, they will ease tracking busg later if needed (let's hope not!) [19:51] mandel, same thing with the lambdas and inlineCallbacks in get_current_uploads [19:51] mandel, and in pretty much every function in that module. [19:52] alecu: ok, I'll move all the possible methods to use the inlineCallbacks decorator [19:52] well, not every function :P [19:52] alecu: well, you understood what I meant :P [19:52] right! [19:54] mandel, so, every function that's creating an returning a Deferred is candidate to rewrite with inlineCallbacks, and it would be much safer written that way. [19:55] mandel, please let me know if you need help with that. And if you consider that it's too much trouble let's file a bug and do it next week. [19:57] alecu: then all those that are using the pb code directly to, like all the config ones [19:57] that is why there are lots of them [19:57] A few minutes ago I tried to stand up and my knee didn't let me. I am going to see another doctor now, I should be back in IRC in 30 minutes or so from the cell phone on the waiting room, and back at coding in 90 minutes. [19:58] ralsina: looking your branch [19:59] nessita: cool, thanks. [19:59] ralsina: good luck in the doctor [19:59] nessita: even more thanks :-) [20:04] ralsina, you should only stand up at 10AM sharp! [20:05] alecu: seems doing it twice a day is too much formy old bones ;-) [20:05] * ralsina will live in a bath tub from now on [20:06] BTW: quasseldroid rocks! [20:13] mandel: ping [20:13] nessita: pong [20:13] mandel: I'm trying to install bzr, but for that I need to install its dependencies, but for that I need pycurl-ssl-bla, but when trying to install I'm getting an error: [20:14] back [20:14] "The program can't be started becasue MSVCR71.dll is mising from your computer" [20:14] mandel: what the heck is that? [20:14] nessita: nest thing is to install the bundle version, the complain is about a dll that comes with vs [20:15] mandel: ok, the wiki should said so, I'll add it [20:15] nessita: yes, mea culpa [20:15] is ok, we'll improve it together [20:20] nessita, alecu, ralsina: I'm going to call it a EOD since its 21:20 here.. I might be back later, but I'm not sure [20:20] laters! [20:21] mandel: question [20:21] mandel: did you finish the branch alecu was reviewing? [20:22] mandel: c u tomorow! [20:23] trivial review on https://code.launchpad.net/~dobey/ubuntuone-client/fix-headers/+merge/65714 please? [20:33] on it [20:33] dobey: done [20:34] thanks [20:34] dobey, approved. [20:35] ping rye [21:12] ok, gt x-rays, turns out I have what the doctor calls "heck of a bump" in the knee, and it's swelling, and it's probably nothing, except it's going to hurt like crazy for a few days. [21:13] ralsina: morphine ftw [21:13] dobey: good band! [21:13] ;-) [21:19] what program can I use to tar xzv? [21:20] 7zip! [21:27] also have https://code.launchpad.net/~dobey/ubuntuone-client/fix-headers-1-6/+merge/65723 and https://code.launchpad.net/~dobey/ubuntuone-client/fix-headers-1-4/+merge/65728 that are trivial if anyone can take a quick look and approve :) [21:30] dobey: on it! [21:30] and thisfred needs another review on https://code.launchpad.net/~thisfred/ubuntuone-client/better-progress-bar/+merge/64887 [21:31] not urgent though, any windows work should go first [21:33] well, would be nice to put it in the release for oneiric [21:33] thisfred: I really want to use Linux for 5' so, reviewing ;-) [21:34] dobey +1 on the two trivial ones [21:34] awesome! [21:34] thanks! [21:35] thisfred: so this makes the progressbar go by bytes instead of files? [21:36] ralsina: exactly [21:36] thisfred: nice! [21:37] thisfred: works too! ;-) [21:37] yay! [21:38] thisfred: +1'd [21:38] thx! [21:39] thisfred: now if the unity folks make a visible progressbar we are good with it! ;-) [21:39] heh [21:39] I made a zoom in mine, it has 4 pixels of border and 2 pixels of actual progress thingie between them. [21:42] noice [21:53] InsecureJelly exception? Some developers should really joke less and give descriptive names more :-( [22:00] sounds like the bad old twisted days. They had error msges like: "Your opulent throne will be your undoing." [22:01] Slightly less informative than "guru meditation #00000004.0000AAC0" [22:02] also: http://www.youtube.com/watch_popup?v=EVwlMVYqMu4&vq=medium#t=125 [22:07] alright, i'm out. later all! [22:07] later dobey [22:26] thisfred: that error is from twisted, so it's still the bad old days over there ;-) [22:31] he [22:32] ok, gonna walk the dogs, may be back later, ping me if anyone needs reviews or anything [22:46] ok, first EOD, tests are almost there, will resume work in a couple of hours === m_conley is now known as m_conley_away [22:54] bye all