=== IAmNotThatGuy is now known as M0hi [08:15] is the ubuntuone client that ships with xubuntu a bit borked? === smb` is now known as smb === vds` is now known as vds [09:27] Happy Leap Day, everyone! :D [09:28] JamesTait, :) [09:54] heh, russian search portal, yandex.ru changed the logo to say "The 0th of March" [09:54] i wondered why. Thanks JamesTait, now I know :) [09:54] :) [10:00] morning! [10:10] Good morning [10:16] Morning :) [11:20] good morning! [11:42] gatox, Daviey, czajkowski morning! [11:42] mandel, hi! [11:44] mandel: ola! [11:46] gatox, may I have a review for : https://code.launchpad.net/~mandel/ubuntu-sso-client/ssl-dialog/+merge/94012 [11:46] mandel, yes, on it [11:46] gatox, is Qt so you should be more that suited for this :) [11:57] mandel, ping [11:57] gatox, pong [11:58] mandel, after setting the direction of the arrow, are you repainting the widget? i'm at Arrow didn't see the rest.... [11:59] wouldn't be better to call self.repaint() from _set_direction? just asking.... [11:59] gatox, yes [11:59] gatox, I added it, I must have removed it when I was chaging the property from nattys comments, give me a sec [11:59] mandel, ok [12:03] gatox, push with the repaint [12:03] mandel, ok [12:06] buenos días! [12:07] nessita, buenas [12:07] hola gatox [12:11] mandel, i saw you were forcing the repaint in the expander, if you are adding the repaint to _set_direction, you should remove it from there [12:12] gatox, true.. [12:12] mandel, do you want me to add those comments or you are working on that right now? [12:12] gatox, fixing them as you speak :) [12:12] mandel, ok :P [12:16] nessita, i answer to your need info [12:17] gatox: I saw that, I'm still thinking about it [12:18] nessita, i think is better this way, and the code is cleaner [12:18] nessita, the ssl dialog branch is updated with no Qt fakes, care to review it when ever you have time? [12:19] gatox: I'm not convinced of that, but may be. Anyways, if we're leaving it like this, we definitely should rename the signal from showingOverlay to processingStarted/processingFinished [12:19] agree [12:19] gatox: giving that a signal called showingOverlay sounds like that the widget is actually showing it :-) [12:20] nessita, yes..... i wasn't really inspired to name things at that moment :P [12:20] hehe [12:35] mandel, let me know when the changes are apply to your branch, i finish the review, and except for those details its fine by me [12:36] gatox, already there [12:38] mandel, you forgot to remove this: self.label.repaint() [12:38] mandel: ack! (forgot to write it, did think about it) [12:38] mandel, if you are doing the repaint in the arrow now [12:39] gatox, let me check [12:45] gatox, you are right, pushing the changes [12:46] good morning ubuntuone! [12:46] ralsina, hi [12:46] does alt+tab work on your unitys today? [12:47] ralsina, alt+tab yes [12:47] ralsina, ctrl+alt-arrow no :( [12:47] ctrl+alt+arrow [12:47] it's really slow to change desktop right now..... i have tu use super+s arrows [12:48] gatox, hehe same here :) [12:48] gatox: I normally don't do multiple desktops. I have to switch apps using the launcher now :-( [12:50] mandel, i have 9 desktop..... and using super+s is really annoying! [12:50] or the launcher..... but sometimes doesn't work [12:50] gatox, I have 4x4 :P [12:50] wowwww [12:54] mandel, +1 to your branch [12:54] gatox, superb! [12:54] ralsina, let me know if the answer is correct: https://bugs.launchpad.net/ubuntu-sso-client/+bug/933729 [12:55] ralsina, also, do you have some time for a super fast chat? [12:55] mandel: looking... [12:57] mandel: +1 on the answer. I think they want bugs with larger scope because it's a bit of a flood. [12:57] mandel: and sure, I can chat! [12:57] ralsina, ok.. I just like little bugs than big ones hehe [13:09] hello! [13:09] alecu, hi [13:11] ok, lunch for me [13:12] * mandel lunch [13:12] hola alecu [13:13] mandel, ping [13:13] hi, peoples [13:13] oh, just missed him. [13:13] Chipaca, hi there megaboss! [13:14] alecu, ahg, I was leaving, after lunch maybe? [13:14] mandel, sure. I was just about to tell you about some ideas to test the ssl without a specially compiled squid. [13:14] alecu, +1000000000000000 to that, I'll ping you asap [13:15] mandel, have a good lunch [13:16] aha! So alt+tab does work, it just ignores chromium! [13:16] ralsina: unity 2d? [13:16] Chipaca: 3d [13:16] ralsina: had any alerts about bamfdaemon kicking the bucket? [13:16] Chipaca: Nope [13:17] ralsina: fun :) [13:17] ralsina: you told hr what you had to tell hr? [13:17] bamf? As in pulpfiction's "the wallet with bamf written in it"? [13:17] Chipaca: yessir! [13:17] ralsina: awesomeist [13:19] * ralsina is dragged kicking and screaming into firefox because not being able to switch to chromium using the kbd is really, really depressing [13:23] can everyone else connect to the canonical server? I seem to be unable to === duanedes1gn is now known as duanedesign [13:26] ralsina, my bip got kicked [13:26] ralsina, switch to another workspace and switch back [13:26] ralsina, bug #937898 [13:27] rye: cool [13:27] rye: is your bip at burst? They seem tobe havingsome routing problem to reach canonical [13:27] rye: yay, that fixed it! :-) [13:29] ralsina, interesting, can't get to irc.canonical.com via burst.net, but it is ok locally [13:30] rye: the traceroute stalls at canonical-gw.datahop.net [13:31] ralsina, i connected directly [13:31] rye: I am starting a second IRC client to do that, too [13:49] ralsina: hola! you have a slot for those kind of quick branches you usually manage to do? [13:49] nessita: of courses [13:49] ralsina: what about http://launchpad.net/bugs/939509 ? I know is not that fun, but the chicharra team really needs it [13:50] ralsina: can I assign that to you? [13:50] nessita: looking... [13:50] nessita: sure, no problem [13:50] ralsina: thanks! === yofel_ is now known as yofel [13:58] hmm [14:03] ralsina: can you hold on fixing that bug? it will conflict with a branch i pushed last night, that fixes the indicator to launch the installer (and thus qt panel) instead of gtk panel [14:04] nessita, this branch is ready for review: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/overlay-improves/+merge/94828 [14:05] gatox: can I have a review for https://code.launchpad.net/~ralsina/ubuntuone-control-panel/fix-scroll/+merge/95006 please? [14:05] gatox: show-erroes reviewed. Looking at that [14:06] ralsina, yes, on it [14:06] gatox: cool, thanks [14:07] nessita, do you have any bug in mind in particular for me, or i just keep killing the ones in my queue? [14:07] mandel: there is still a typo in the SSL constants: SSL_REMEBER_DECITION [14:07] gatox: let's have a mumble, shall we? [14:07] nessita, ok, now? [14:08] gatox: yes, give me 5 minutes [14:10] nessita, ok, i'm already in mumble [14:14] nessita, sorry, fixing right now [14:25] ralsina: I'm mumble for our 1-1 [14:26] hmmmm.... dobey, Iam setting up ubuntuone-developer-dependencies because it's my first u1-client branch in this machine, and it's failing with "epends: ubuntuone-database-dependencies (= 0.208~precise0) but it is not going to be installed" it seems that one depends on postgresql-8.4 that is not available on P? [14:26] nessita: going! [14:26] ralsina: FYI, ubuntuone-developer-dependencies are for the server bits "only" [14:26] (so, Lucid) [14:27] ralsina: you don't need that stuff [14:27] nessita: ok,makes sense. What's the easy way to get the client dev dependencies then? Because manually it seems to be a lot :-) [14:27] ralsina: you need to add the nightlies PPA, and do apt-get build-dep (ubuntuone-installer for example) [14:27] dobey: ack, thanks! [14:27] ralsina: apt-get build-dep ubuntuone-client [14:27] o what dobey said :-) [14:28] thanks, I had not done this since I joined the company :-) [14:40] moin [14:45] need to restart [14:46] nessita, is fixed [14:47] mandel: ack! [14:51] ralsina, briancurtin: Need to tidy up the U1 Music app. Got rejected by Apple and we need to make some small changes. [14:53] urbanape: so now that i have lion running, i take it brew is recommended? [14:53] Yeah, I'm a big fan [14:53] compared to older stuff like fink or macports [14:59] I suppose I ought to upgrade to Lion, really [15:00] me [15:00] me [15:01] me [15:01] ralsina, gatox, alecu, dobey, urbanape? [15:01] me [15:01] me [15:02] dobey, not dobey, ralsina, urbanape, standup? [15:02] mandel: wanna start, while the rets join us? [15:02] rest* [15:02] mw [15:02] me [15:03] mandel: go! [15:03] meh [15:03] DONE: Updated branchesaccording to comments from nessita nadgatox. Worked of the ssl tets although I'll be talking with alecu about the matter in a few mins. [15:03] TODO: ssl tests. Integrate the ssl dialgo with those tests and assert we do show the dialog when expected. [15:03] BLOCKED: no [15:03] briancurtin, please [15:03] DONE: ssh/screen/mumble with urbanape, install osx [15:03] TODO: more mac stuff, and i now have a functional mac, should probably do 1-1 with ralsina if he's less busy with calls :) [15:03] BLOCKED: nada [15:03] NEXT: nessita [15:03] DONE: control panel design review, reviews, catch up after holiday [15:03] TODO: propose branch for bug #933576, 1-1 with ralsina, some calls re: U1CP look and feel [15:03] BLOCKED: nopes [15:03] NEXT: gatox [15:03] DONE: [15:03] Proposed 2 branches on monday, tuesday holiday, today apply some improves to those branches. Review for mandel and ralsina. [15:03] TODO: [15:03] Mumble with nessita to talk about some ui issues. [15:03] BLOCKED: [15:03] No [15:03] alecu, go [15:03] DONE: proposed first two branches for bug #929207 that need reviews, a bit of debugging on two proxy related bugs. [15:03] TODO: work on tunnel process and SD integration, catch up with reviews [15:03] BLOCKED: no [15:03] REVIEWS NEEDED: [15:03] * https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-server/+merge/95075 [15:03] * https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-client/+merge/95077 [15:03] NEXT: urbanape [15:03] DONE: made some progress with briancurtin [15:03] TODO: recompile PyQt with the dbus stuff to get the mainloop [15:03] BLOCK: None [15:03] dobey: europe [15:03] λ DONE: fixed indicator support to work again, fixed indicator to launch installer (and qt panel) [15:03] λ TODO: rb extension bugs, look at twisted g/gtkapplication support/hanging issue [15:03] λ BLCK: none. [15:04] ralsina: you don't get out of it because you're a manager. [15:05] ralsina: TODO: attend the standup :-P [15:05] urbanape, briancurtin: you guys are probably aware that if getting dbus compiled on pyqt is troublesome you may be able to use the PerspectiveBroker based ipc instead, right? [15:06] urbanape, briancurtin, alecu: right, since we already have all the PB IPC in place [15:06] urbanape, we are using that on windows, and it should work on mac since it comes standard with twisted. [15:07] urbanape, I think I did write that in my crazy, full of typos, email, right? first approach, using twisted pb, later get something smarter [15:07] urbanape, briancurtin: perhaps you can try with that just to get things running, and then try with dbus to see if it works better. [15:07] yeah, PB is an option to try. ultimately there may be a more mac-specific thing, but PB may get us moving [15:08] briancurtin, the nice thing about pb may be that we won't be needing to compile pyqt ourselves each time a new version is released. [15:09] briancurtin, on windows at least we didn't like PB because it runs on a local tcp port, and that can give troubles with some firewalls [15:09] sorry, phone call! [15:09] briancurtin, but on mac I'm not sure if that's an issue. [15:10] DONE: design call, mgmt call, some code, misc things, nessita 1-1. TODO: standups! BLOCKED: no [15:10] alecu: good info, thanks [15:11] well, I doubt a sandboxed app can open a tcp server [15:11] but for theproof of concept? +1 to pb [15:13] alecu, briancurtin we wanted to be able to use the IOCP reactor to be listening to namepipes and not ports to work around that issue. The main problem with pb is that there is no C implementation for banana [15:13] which is a PITA and not a surprise.. [15:14] mmm, delicious pitas (sorry, kind of hungry) [15:18] re: PerspectiveBroker - how about third-party interaction with the code? [15:18] i mean with the service [15:21] i wonder how brew/etc will play out with mountain lion requiring signed everything [15:26] rye, that's a good point. With mandel we started implementing an experimental json over tcp ipc mechanism, as a freaky friday project. [15:26] rye, but we got sidetracked with other more urgent matters [15:28] rye, in any case, I agree that PB is not the best way to present an API [15:40] alecu: and DBUs is? :-) [15:41] this upgrade is killing my machine :S [15:41] gatox: lies! :-) [15:42] nessita, i wish [15:42] gatox: can I help [15:42] ? [15:42] nessita, no, unless you fix unity :P [15:42] gatox: I can fix that! [15:42] sudo apt-get remove unity [15:42] jejejej [15:43] yesterday unity was really broken so I spent 2 hours on openbox. Things were *fast*. And nothing had *menus*. [15:45] ralsina, I broke bash today, that was interesting :) [15:46] mandel, let's talk ssl around here [15:46] alecu, sure, so what was your idea :) [15:46] mandel, my idea is not using squid at all [15:47] mandel, for the ssl tests any ssl server would do [15:47] mandel, like a twisted http over ssl server [15:47] mandel, and make that server listen to CONNECT or other proxy methods [15:48] mandel, since the thing that you want to test is ssl exceptions the protocol that's tunneled thru ssl is not that much important anyway [15:48] mandel, what do you think? [15:49] alecu, I did consider that too, and I think is a good idea, the question know is: is that type of test general enough? if it is not we can simply add it to the sso trunk and forget about touching devtools [15:49] alecu, and I don't want to touch devtools since it will be an other ffe [15:49] mandel, right [15:50] mandel, then probably we can do this tests in sso. The ffe is not that strict for unittests, right? [15:51] alecu, I'll work on that on sso then since we have a ffe for that, and adding tests is not an exception :) [15:54] all, I'm going underground for a bit to focus in some hard reafactoring I need to do [15:54] will not pay that much attention to IRC, if you really need me please say my nickname :-) [15:58] * gatox lunch [16:06] jono: how's things? [16:07] Chipaca, good, man :-) [16:07] about to hop on a team call :-) [16:07] jono: hangouts still working for y'awl? [16:07] Chipaca, indeed :-) [16:07] we love 'em [16:07] google records them all :) === gatox is now known as gatox_lunch [16:30] dobey: ping [16:30] nessita: yes? [16:31] dobey: is gir1.2-indicate-0.7 a dependency in u1client? [16:31] yes [16:31] dobey: yesterday my laptop was broken since sd did not start on a importerror on that [16:31] in ubuntu [16:31] dobey: not in nightlies? [16:31] it's in Recommends [16:31] or i might have moved it to depends there too [16:31] dobey: right now the code requires it [16:32] dobey: when we fix it to be optional, yes, should be a recommends [16:32] nessita: yes, and there is a bug about it [16:32] are there issues not currently indicated on the status page? the app shows file sync in progress, --current-transfers shows uploads 0 downloads 0 and I know for a fact that not everything is synced [16:33] duanedesign, rye: ^^ any answer for cjohnston ? [16:35] cjohnston: probably need to look at the logs [16:37] cjohnston: since your last log folder was so big you want to try just the ~/.cache/ubuntuone/log/syncdaemon.log [16:38] nessita: quick question about theindicate bug. Messaging.show_message returns the Indicator. That only seems to be used in the tests. The easiest way I find to make indicate optional is to turn show_message into a noop if indicate can't be imported, but then I have nothing to return. Is that really really bad? [16:38] seem to be some 2012-02-29 10:15:13,345 - twisted - ERROR - Unhandled error in Deferred: [16:38] 2012-02-29 10:15:13,345 - twisted - ERROR - Unhandled Error [16:38] Traceback (most recent call last): [16:38] Failure: dbus.exceptions.DBusException: org.gnome.zeitgeist.EngineError.DatabaseError: Can't start transaction: 1, cannot start a transaction within a transaction [16:39] ralsina: hum, I'm not familiar with that code... did you check if dobey was modifying it as well? (I think he was?) [16:39] it only has thisfred name on it. I'll ask him instead. [16:40] whaddup [16:40] thisfred: have 2' to help me with something in the u1-client messaging? [16:40] sure [16:40] where do I look? [16:40] thisfred: give me a sec... [16:40] thisfred: https://code.launchpad.net/~ralsina/ubuntuone-client/may-indicate/+merge/95204 [16:41] cjohnston: is their anything in syncdaemon-exceptions.log ? [16:43] duanedesign: a ton of 2012-02-29 11:43:15,621 - twisted - ERROR - Unhandled error in Deferred: [16:43] 2012-02-29 11:43:15,622 - twisted - ERROR - Unhandled Error [16:43] Traceback (most recent call last): [16:43] Failure: dbus.exceptions.DBusException: org.gnome.zeitgeist.EngineError.DatabaseError: Can't start transaction: 1, cannot start a transaction within a transaction [16:44] ralsina: i guess you didn't see my request this morning then? [16:45] ralsina: I think the fact that it returns the indicator is a relic [16:45] dobey: apparently not [16:45] thisfred: it is used in the tests for something (have not looked) [16:45] ralsina: to prevent further confusion, I would make it not explicitly return anything, if we can rework the tests so they don't need it [16:45] thisfred: makes sense [16:45] dobey: what request? [16:46] ralsina: i asked you to hold off on that branch, as i did some work yesterday, which will conflict with fixing that bug [16:46] dobey: oops, no, missed it. Ok, holding :-) [16:46] ralsina: I'm sure I didn't build it that way just for the tests, I think my first attempt was just a mess, and this is a bit of that mess left over. [16:46] thisfred: ack, thanks dude! [16:46] i need to file a bug to link my branch to though, as nobody else filed it yet [16:47] right now, i really need to get some lunch [16:47] bbiab then will look at mess fixery [16:47] dobey: cool, buen provecho! [16:47] gracias [16:48] ralsina: two tests rely on it, but hopefully they can be fixed by getting the indicator directly. [16:48] thisfred: right [16:49] ralsina: btw, I've moved back from accounts (I think, barring any disasters, and of course I'll still do reviews and maybe some small branches there) [16:49] thisfred: yay! === gatox_lunch is now known as gatox [16:50] thisfred: of course you are mostly on u1db, but welcome back to standups, weekly meeting and doing our reviews ;-) [16:50] I'll now be focusing on u1db, but when I'm up to speed on that in the next week or so, I may have some time for more general desktop things too [16:50] yep [16:50] thisfred: also, schedule a weekly 1-1 with me please [16:50] will do, what's a good day for you? [16:50] I usually have weekly 1-1s with everyone every two weeks or so, depending on how busy we are ;-) === IAmNotThatGuy is now known as M0hi [16:51] wednesday [16:51] ok, I have no time preference, considering I start at 8:30 EST [16:51] So after that, anything's good [16:51] thisfred: anytime I have free after that time is good. [16:52] ok [16:52] * ralsina notices neither wants to actuallyopen calendar and do it [16:53] I'm just trying to figure out how to view yours ;) [16:53] I have mine open [16:53] thisfred: haha, mine should be as well [16:54] thisfred: "find a time" is the trick IIRC [16:55] ralsina: I think I have yours [16:55] thisfred: let me do it [16:55] thisfred: ok :-) [16:57] * ralsina wonders how we managed this without technology. Oh, yes, with bits ofpaper and clocks. [16:58] ralsina: hmm, it defaulted to an hour, which is probably on the long side, but it won't let me edit anymore [17:01] ralsina: nm: I couldn't edit your copy of course, which was obscuring mine. [17:03] now, lunch [17:08] great, so pastebin.canonical.com, if you have the cookie from a "bad" account, will not let you login, even with a good account, until you remove the old cookie manually. Good UX there :-/ [17:10] briancurtin: building a new version of the U1 Music app for submission then lunch. Want to pick up after? [17:11] urbanape: sounds good [17:14] lunchtime! [17:15] mandel, hey!! Chipaca just told me....... now is: super+shift+arrows [17:15] gatox, cool, nice to know! [17:15] gatox, then there is a bug, I see the cheat window when I do it :P [17:16] I hate calling the windows key "super". It's at best "adequate". [17:16] mandel, here it's working fine..... just as ctrl+alt+arrow [17:16] gatox: mandel: shift+super, not super+shift [17:16] Chipaca: super-shift != shift-super???? AHHHHHHHHH [17:16] the order of the farols, in this case, does alter the albumrad [17:16] ah yes [17:17] Chipaca, hold down shift + super, I see the little window [17:17] mandel: shift-super and QUICKLY arrow [17:18] mandel, i don't see the little window until i release the super key [17:18] I don't see no little window, because i'm in unity 2d [17:18] lurve, lurve, unity 2d [17:18] ahhhh i'm in unity2d too [17:18] Chipaca, question :P [17:19] gatox: sarcastic answer [17:19] Chipaca, and how can i move a window from one workspace to the other?? it used to be ctrl+alt+shift+arrow [17:19] i tried a couple of combination, but no success [17:20] gatox: that one hasn't changed [17:20] mmmm so it's not working for me [17:20] gatox: ctrl+shift+alt+arrow? [17:21] Chipaca, nop [17:21] gatox: you suck [17:21] * Chipaca runs [17:21] jejejee [17:22] gatox: system settings -> keyboard -> shortcuts -> knock yourself out [17:22] davidcalle: [17:22] davidcalle: ping [17:22] Chipaca, yep.... i'll do that.... thanks [17:22] Chipaca, pong [17:22] davidcalle: ah! just remembered :) [17:23] davidcalle: was going to ask when trunk would be released :) [17:23] Chipaca, ahhhh it's alt+super+arrow now [17:23] Chipaca, after beta, yes :) [17:24] davidcalle: your latest commit is bad [17:24] davidcalle: *bad* bad [17:25] davidcalle: blow-up-in-your-face bad [17:25] * davidcalle is scared [17:25] Chipaca, I listen. [17:25] davidcalle: what happens if the results are, indeed, unicode, and include non-ascii characters? === Guest40451 is now known as zumbi [17:26] gatox, I think is the same as one of the combos from ryu.. [17:26] gatox, I cannot remember witch :) [17:26] mandel, i change it back to the old shortcuts [17:26] gatox, try Forward+HP -> C.LK -> Hadouken -> Shinkuu Hadouken [17:26] jejjejee [17:27] * ralsina tries the capcom code in unity, doesn't work [17:27] Chipaca, hmm, .encode("utf-8")? [17:27] davidcalle: what you need to do is first confirm whether they are always unicode (checked: they always come from json.loads, so they are unicode), and then if so, instead of str(thing), do thing.encode('utf-8') [17:27] davidcalle: if they weren't always unicode, you'd have ugly code, or a bug, or a lot of refactoring, or all three :) [17:28] gatox, we should totally have capcom commands for everything [17:28] Chipaca, got it. [17:28] mandel, shortcuts FTW [17:29] gatox, they do sound like a great addition for ninja [17:29] gatox, if you don't give me vim give me ryu! [17:30] mandel, well..... you actually have a lot of crazy shortcuts in ninja..... during the sprints the other developers find out about them..... i need to write documentation :P [17:37] hrmm [17:40] EOD here, catch you all tom! [17:46] mandel, bye === lifeless_ is now known as lifeless [18:00] karni, are you in twitter? [18:00] gatox, ralsina: philosophical question: (in the context of the SSO wizard) why do we have our custom Header that we insert in a QWizarPage layout instead of using the QWizarPage's title and QWizarPage's subTitle? [18:01] facundobatista: he is mkarnicki on there [18:01] dobey, gracias [18:01] nessita, it was because of the styling..... and how the wizard show (and in some cases don't show) the title and subtitle depending on the wizard style [18:02] gatox: can you give a little more details? as in, what we needed that wasn't possible [18:02] nessita, mmmm..... the font size for example [18:02] padding..... [18:02] nessita: you can't style the title and subtitle [18:02] gatox: couldn't we set that thru stylesheets? [18:02] tunning adding the ellipsis [18:02] nop [18:03] boo, thanks [18:03] Next question: [18:03] nessita: the wizard is really not-stylable [18:03] nessita, and some wizard styles don't show title directly [18:03] why defining this on every QWizardPage? [18:03] 205 def nextId(self): [18:03] 206 """Return the next page ID.""" [18:03] 207 return self.next [18:03] gatox: we are forcing the wizard style to modern (or something) anyway ;-) [18:03] nessita: because originally the pages did not inherit a common class and we never refactored it? [18:04] i think we are not using that anymore [18:04] ralsina: you lied; but i guess there's only one spot that will definitely conflict, in that branch [18:04] we shouldn't since you guys moved the logic to the wizard [18:04] dobey: I lied? How unsurprising. Where? [18:04] * briancurtin lunch [18:05] gatox, ralsina: so, my gut feeling that that code is not needed is correct, right? (unless we need a custom nextId to be set) [18:05] nessita: right [18:05] ack, thanks! [18:05] nessita, correct [18:05] ralsina: i asked you to hold off on that branch, and there's a proposal for it :P [18:05] dobey: sorry, I did the proposal before we talked about it [18:06] dobey: to show the diffs to others to clarify some points. Forgot to set it to WIP though [18:07] ralsina: and this wasn't enough to style a subtitle? "The subtitle may be plain text or HTML, depending on the value of the QWizard::subTitleFormat property." [18:07] (the same applies to the title) [18:07] ralsina: just trying to understand some code [18:07] nessita: that one I did not know ;-) [18:08] ralsina: would that have worked? [18:09] nessita: I don't know. It's not really modern HTML, but a subset plus CSS2 [18:09] nessita: probably yes [18:09] ack, thanks again! [18:11] sigh; google translate doesn't like unicode in urls at all :( [18:12] ralsina, gatox: one more question... how can I add a widget to a qwizardpage I'm creating "by hand"? [18:13] nessita, add the widget to the layout of the page [18:13] i'm not sure i understand the question [18:14] gatox: I'm reading the qwizardpage doc, and I'm not sure how I can set the page "main widget" [18:15] nessita, you can extend from QWizardPage and define the layout [18:15] nessita, i don't know how are you creating the page [18:15] gatox: I'm indeed extending QWizardPage [18:15] nessita, so, in the init you can do something like: [18:15] vbox = QVBoxLayout(self) [18:15] and add the widgets to that layout [18:16] gatox: there is no default layout in a QWizardPage? [18:16] yes.... i don't remember which kind of layout it is [18:16] people use icons in the weirdest ways, for the things which they aren't icons for [18:16] gatox: thanks, I'll research from here [18:16] nessita, you can ask for the layout with self.layout() [18:16] thanks! [18:16] and do: self.layout().addWidget [18:24] ralsina: were you asking about the messages indicator, because someone was not seeing share notifications in it, and was expecting to? [18:27] ralsina, gatox: is there any signal that gets emitted when a .close() is called in a widget ? [18:27] nessita, depends on the widget..... but you can overload the closeEvent method [18:28] gatox: I want to connect to a widget's "signal" when that widget calls .close() [18:28] destroyed! [18:28] that's what I want [18:29] nessita, mmmm i'm not sure about destroyed..... i personally wouldn't use that in python [18:29] gatox: why not? [18:30] nessita, i don't know what you want to do..... but in python with the garbage collector i don't know if that works ok....... maybe i'm wrong..... ralsina ^ [18:30] i think i had problems with that in the past....... [18:45] briancurtin: back [18:46] urbanape: on phone with USPS, but back in a few mins [18:46] k [18:48] urbanape: now im back [18:48] cool. mumble? [18:49] logging in [18:57] dobey: I lost track of the progress on the 'dark grey' bug. Are you doing the UIFe? [18:57] (yes I'm old and drop packages :-D) [18:59] nessita: sure, i can [18:59] dobey: wanna propose the branch for merge, so I can use that? [19:00] nessita: about to yeah. was filing bugs and proposing my client branch to fix the indicator [19:00] dobey: thanks [19:02] need to restart....... my computer is crashing all the time [19:03] gatox: lies! :-P [19:03] nessita, please believe me..... is not funny :( jeje [19:03] gatox: of course I believe you, I was kidding [19:04] Friends, [19:05] when I create a U1 contact list on thunderbird, the list is shown nowhere on the website, right? [19:05] hey now. let's keep it professional! :P [19:05] elopio: depends on what you mean exactly, but couchdb support is drifting away [19:06] elopio: so even if you did it the right way (tb UI can be confusing), and it doesn't work, we probably won't be able to get it fixed [19:06] ralsina: facundobatista: I replied to the tweet, thanks. 1) It depends how often picures are uploaded (obviously), which == radio 2) we'll be decreasing the memory footprint shortly, which can also help [19:07] gatox: was it the alienware? [19:07] karni, how reducing the memory footprint improves battery? [19:07] dobey, i'm having a lot of issues since the last upgrade (yesterday) [19:08] facundobatista: Decreases paging, which is I/O [19:08] gatox: maybe it just wants you to reinstall it again :P [19:08] facundobatista: Android prefers to have memory full, only then starts to swap pages. On the initial auto-upload, we're quite a memory hog (in-memory queues). [19:08] dobey, ohhhh fun..... not [19:08] jeje [19:10] karni, so probably it's the case [19:10] facundobatista: ralsina: FWIW, we use a very efficient way to detect new photos. We have a lightweight background service, so that's not the problem. [19:11] dobey: it's like a weird bug. But what you said is probably the best answer for the reporter. Thanks. [19:13] elopio: so tb has local lists and remote address books, and it can be confusing which one you've created [19:18] nessita: https://code.launchpad.net/~dobey/ubuntuone-control-panel/the-darkness/+merge/95239 btw [19:18] dobey: FINALLY :-D [19:19] like the name [19:19] I* like the name [19:19] ralsina: btw; any word on the product sprint thing? [19:21] dobey: this is a list of contacts on a U1 remote address book. I think... It's confusing. [19:21] dobey: btw, where did you get the dark grey color from? [19:21] dobey: I'm looking at the UBuntu paletter and it does not match [19:22] nessita: the titlebar on the window [19:22] Ubuntu palette* [19:22] ubuntu palette? [19:22] dobey: I guess we need to change it to the Ubuntu dark grey, which is listed in http://design.canonical.com/the-toolkit/ubuntu-logo-and-circle-of-friends/ [19:22] dobey: from http://design.canonical.com/brand/Tools/colour_palettes/UbuntuColours.png, I can tell the dark grey is: [19:23] elopio: don't know if desktopcouch supports that. or if the web ui for contacts does either. [19:23] elopio: i suspect not [19:23] dobey: #333333 [19:23] dobey: can you please change that? [19:23] nessita: why do we need to change it to that? [19:24] dobey: the agreement with the design team was to stick to the ubuntu palette for this change [19:24] dobey: so I promised we were gonna use the ubuntu dark grey [19:24] and I would like to keep my promise :-) [19:26] then it won't match [19:27] dobey: I think it will, since the window top bar is a gradient [19:28] nessita: it's a gradient, and i picked the color at the bottom of the gradient so it would match correctly. [19:29] and the window borders theme doesn't even use 333333 anywhere within it [19:29] that color doesn't even appear anywhere within the Ambiance theme at all, actually [19:30] dobey: the difference is extremely sutil [19:31] but i can see the edge [19:31] hum [19:31] dobey: then we need to run this by design again [19:31] granted, *i* can see the edge, and people can't even read the text on my screen [19:31] dobey: yes, you're right, I can also see the edge :-/ [19:32] hrm [19:32] ralsina: any opinion? ^ [19:32] nessita: but you can only see it because i said i can and you looked harder :P [19:32] dobey: yes, is very sutil [19:33] nessita: anyway, i pushed the change [19:33] I had to use xmag to see the edge [19:33] dobey: but you're right, it does not look good [19:33] I wish I had known this discrepancy this morning :-/ [19:33] should have used xmag :) [19:34] dobey: I assumed the dark grey in the windows was the ubuntu dark grey [19:34] well we should wait for the UIFe anyway [19:39] nessita: re: assumption: http://www.youtube.com/watch?v=miPPQu2iqts [19:53] oh, pitti gave the +1. so we just need to decide which color to use [19:53] :) [19:54] nessita: in the meantime… https://code.launchpad.net/~dobey/ubuntuone-control-panel/remove-gtk/+merge/95252 [19:55] dobey: you're tackling all the easy and fun branches, is not fair :-) [19:56] it's not fun :) [19:56] it's a very sad branch [20:05] ralsina: can I somehow connect a callback to the clicked event of a qwizard cancel button? [20:05] ralsina: getting dbus into PyQt4 should be as simple as adding a --dbus=/path/to/dbus/headers right? [20:06] to the call to configure.py [20:10] urbanape: should do, yeah [20:11] urbanape: i presume you have pkg-config working? [20:12] yeah, should be [20:14] urbanape: what does "pkg-config --cflags-only-I --libs dbus-1" output for you? [20:14] -I/usr/local/Cellar/d-bus/1.4.18/include/dbus-1.0 -I/usr/local/Cellar/d-bus/1.4.18/lib/dbus-1.0/include -L/usr/local/Cellar/d-bus/1.4.18/lib -ldbus-1 -lpthread [20:15] urbanape: and you have /usr/local/Cellar/d-bus/1.4.18/include/dbus-1.0/dbus/dbus-python.h ? [20:15] aha. [20:16] no, but it does exist at /usr/local/include/dbus-1.0/dbus/dbus-python.h [20:17] EOD here! see you tomorrow people! [20:19] urbanape: so it looks like dbus-python installed it to the wrong place. pyqt4 should be picking up the support automatically in configure.py. but that, and possibly the location of the dbus python package itself are causing that to fail [20:19] yup. since the dbus-python wasn't a homebrew recipe, it just installed it straight into /usr/local/{lib|include|&c} [20:19] ah ok [20:19] need to whip up a homebrew formula for it. [20:28] ugh, where did my postage stamps go :-/ [20:40] nessita: yes, get a reference to the button (QWizard.button, IIRC) and then use its clicked signal [20:40] nessita, http://developer.qt.nokia.com/doc/qt-4.8/qwizard.html#button [20:40] ralsina: I found it, is rejected [20:40] urbanape: in theory but I have done nothing like it in the last 10 years or so [20:40] the specific signal is rejected [20:40] nessita: oh, ok [20:41] ralsina: I think we figgered it out. [20:41] damn, I leave 1 hour and everyone solved everything! [20:41] homebrew segregates its stuff off and then links into /usr/local/* [20:41] and my dbus-python was installing into /usr/local/*, but pkg-config was not looking in the same place. [20:42] dobey: niggle about an older branch already merged, why not have both installer and control panel have the window class be "Ubuntu One" instead of "Ubuntu One Installer"? [20:42] urbanape: awesome [20:42] ralsina: so, I need some input from you [20:43] nessita: sure [20:43] ralsina: which one? http://ubuntuone.com/0H4mGL2S5HaF1PSN5UUW6p or http://ubuntuone.com/2fLrfst5ueA6MiAQEaFF0y [20:43] ralsina: I like neither :-/ [20:43] nessita: looking... [20:44] ralsina: if we could hide the ---------- bottom line, I would take the first one [20:44] nessita: http://ubuntuone.com/2fLrfst5ueA6MiAQEaFF0y has two "default buttons" so, on this choice, thefirst [20:44] ralsina: can fix that, no prob [20:44] nessita: we can! [20:44] ralsina: how???? [20:44] nessita: let me think a second ;-) [20:44] ralsina: well it's ubuntuone-installer, and because that would mean more work, and chanes to both, and more possibility to break [20:45] nessita: ok, that is the line above the buttons in a wizard... let me dig a little bit [20:45] dobey: yes, but it *is* weird, admit it ;-) [20:45] dobey: maybe a low bug for the indefinite future? [20:45] ralsina: well, i didn't design unity, so… ;) [20:46] ralsina: a) i wish we could just get rid of ubuntuone-installer. b) i wish the qt control panel didn't hard code everything. [20:47] dobey: I don't see how b) applies to this specific thing,but hey,whatever makes you happy, dude. [20:48] nessita: I am sure it can be done, but it may involve some 15 minutes of experimentation. Stay tuned ;-) [20:49] ralsina: in your personal opinion, which one is better? [20:49] ralsina: the icon [20:49] which i don't understand why it's not using the same icon as on linux, but hey [20:49] ralsina: I updated the second screenshot with one that has only one default button [20:50] "the same icon as on linux"? The icon is set on designer, it's not hardcoded. [20:50] dobey: ^ [20:51] ralsina: set how? it's embedded in the binary as i understand, and if not, then the full path must be listed. because windows doesn't do the Icon Theme Spec. [20:52] ralsina: and we are using a different icon for the winodws version, than we are on linux. [20:52] it's not using the ubuntu one icon we use everywhere [20:54] nessita: to get rid of that line: the parent() of any QWIzardPage (after they are added) is a QFrame. do frame.setFrameShape(frame.NoFrame) [20:54] dobey: well, it's using the icon we are telling it to use. If on Ubuntu we should use a different one, it's trivial to fix... [20:54] ralsina: can we do that with stylesheets? [20:54] nessita: maybe with some weird selector. gatox may know [20:55] nessita: but then there will be no line ever (even if we do have buttons) so maybe the second one is a better idea anyway [20:56] ralsina: you IRL tried it? it did not work here [20:57] nessita: nope, I can try to IRL it though [20:57] ralsina: no rush, I'm leaving. Which one do you like better? [20:57] The second [20:58] ack then! [20:59] Ha, pdb really hates it if you have an object called p. p.whatever() gives a syntax error! [21:06] nessita, ping [21:06] brb. gotta run a couple quick errands [21:06] nessita, in your spawn-related code, how did you end up getting the cross platform path to the executable being spawned? [21:07] alecu: good question! [21:07] nessita, I need something similar for the tunnel process [21:07] alecu: the path is multiplatform, but it points to the python script, which in windows, strictly speaking, is not executable [21:08] alecu: understand my point? [21:08] nessita, so, via __file__ ? [21:08] alecu: no, please look at the get_bin_dir in ubuntu_sso/utils/__init__.py [21:08] looking [21:08] alecu: the same way we grab the data files [21:09] dobey: what's the "windows top bar" grey? I'm writting an email to robert [21:11] nessita: you want to learn about kcolorchooser :-) [21:12] nessita: #3C3B37 [21:13] nessita, so, dirname(__file__)/../../bin [21:13] ralsina: I know kcolorchooser and xmag, but I'm also very lazy at this time of the day [21:13] nessita, does that work at all inside py2exe? [21:13] alecu: that does not work on install [21:14] alecu: not sure. But surely what you propose does not work on ubuntu installs [21:14] nessita: there, doneitfor you then :-) [21:14] ralsina: thanks! though I recall there was an 'A' in it [21:14] ralsina: you sure you grabbed the last pixel from the bottom? [21:14] nessita, right. I see that ubuntu_sso.constants is used for installs [21:15] alecu: yeap [21:15] alecu: I was counting on tweaking that if py2exe was a pain [21:15] nessita: fromthe bar inside chromium [21:15] nessita: same thing for the tabbar inside thunderbird [21:15] ralsina: ack then! [21:21] briancurtin: I'm EODing for now, though I typically get back on once Lex goes to sleep. I'll take a whack at brewing up the dbus-python package, rather than trying to piggy-back it on the dbus brew install location. [21:27] urbanape: the intertubes broke, im back on mumble [21:27] briancurtin: I'm EODing for now, though I typically get back on once Lex goes to sleep. I'll take a whack at brewing up the dbus-python package, rather than trying to piggy-back it on the dbus brew install location. [21:27] urbanape: sounds good [21:36] dobey: remove-gtk reviewed [21:37] ok, I'm off crowd [21:38] bye all! [21:38] bye nessita! [21:38] dobey: do you have any pending reviews? [21:42] nessita: the bottom of the gradient is the color my branch was using before the 333333 change [21:44] nessita: #3c3b37 [21:46] ralsina: https://code.launchpad.net/~dobey/ubuntuone-control-panel/remove-gtk/+merge/95252 [21:46] ralsina: https://code.launchpad.net/~dobey/ubuntuone-control-panel/the-darkness/+merge/95239 [21:47] ralsina: https://code.launchpad.net/~dobey/ubuntuone-client/installer-not-panel/+merge/95237 [21:47] nessita: ^^ you might want to look at that last one too [21:47] dobey: ack! [21:47] but tomorrow, I'm gone now [21:47] ralsina: that last one is the one that will conflict with your branch [21:51] dobey: ok, will take a look at one or two [21:53] dobey: why import glib as GLib and right away glib = GLib? in installer-not-panel? [21:53] dobey: and the same with pyindicate and indicate [22:03] ralsina: to avoid the lint warnings [22:03] "redefinition of previously blah blah" [22:18] Question.. I think its been a few weeks now that U1 periodically says "Downloading XXXXXXXX and 199 other files to your computer" .. I haven't checked.. but.. is U1 broken on my system? Whats the deal? [22:18] BTW, XXXXXXX is always the same file [22:19] (and I'm running precise) [22:21] SpamapS: hm! [22:21] verterok: is that a you, or is it a nessita? [22:22] Chipaca: ? [22:22] verterok: or was that alecu [22:22] Chipaca: notifications? I think it's alecu/thisfred [22:22] probably [22:23] verterok: i know the notification issue was addressed, but i also know chicharra was involed in the fix :) [22:23] Chipaca: ah, ok. I didn't land the fix for it [22:23] I remember seeing a bug report for that... [22:25] So, is this a stuck, untrue notification? [22:25] or am I missing files that I should have? [22:26] 3022 clint 20 0 435m 64m 5520 S 7 1.6 3:46.80 ubuntuone-syncd [22:26] It always seems to be taking 5 - 8 % of the CPU .. so I assume its trying to do something [22:36] hmm... this is the bug I remembered: https://bugs.launchpad.net/ubuntuone-client/+bug/807005 [22:38] there's a different report [22:38] SpamapS: probably it's a lot more than 199 files and it just takes a long time [22:39] https://bugs.launchpad.net/ubuntu/+source/ubuntuone-client/+bug/939483 [22:39] alecu: would be a different issue than that filename reset bug i think [22:40] dobey: it certainly looks so [22:40] alecu, dobey: I think beuno mentioned this bug a few days ago [22:41] ralsina, dobey just pasted the url [22:41] he, and he reported it, of course [22:41] ralsina: its always 199 .. and always the same file. [22:41] SpamapS: yes, the number is a lie [22:41] ... like the cake.. [22:42] SpamapS: what does "ls -lh ~/.cache/ubuntuone/log/*.log" list you as having? [22:43] ralsina, that number is taken from the events that went thru syncdaemon. Probably something changed with the changes that facundobatista was doing regarding storing some part of the queue on disk. [22:43] "something changed with the changes" -> doh [22:43] alecu: right, and we are only getting the size of the not-in-memory queue or something [22:43] dobey: http://paste.ubuntu.com/862649/ [22:44] the 199 is because there is a limit of 200 items in the queue [22:44] memory_pool_limit.default = 200 [22:44] ralsina, dobey: data/syncdaemon.conf:memory_pool_limit.default = 200 [22:44] heh [22:44] ditto :-) [22:44] I see downloads completing.. [22:44] and more being added [22:44] so it probably is updating the notification every time a file completes, and a new item gets added to the queue [22:45] now the interesting question is "can we know the size of the on-disk queue?" [22:45] ralsina: i think that's a question for facundo [22:45] dobey: indeed. facundobatista? ^ [22:46] I brought this up a while back in discussions. I'd much rather just see the queue length, and any errors. I don't really care about U1 doing its job normally. [22:46] SpamapS: how many files are you syncing exactly? [22:46] dobey: I'm not sure [22:47] AFAICT, these files were always on this particular machine, I'm not sure why they're being Downloaded again. [22:48] ralsina, it seems that it's very easy to find out the size of the queue with len(OffloadQueue), but that info is not useful in order to show the same data we were showing. [22:49] SpamapS: so the notifications do in fact, work the way you want (though i guess we don't have notify-osd notifications for errors); but a change in the underlying behavior of syncdaemon caused this particular issue to occur [22:49] alecu: I won't pretend to understand what you said ;-) [22:50] dobey: SpamapS proposed on u1-discuss we got rid of most superfluous notifications altogether :-) [22:50] ralsina, sorry :-) [22:50] alecu: yes, we did that [22:50] dobey: we did a small part of that... [22:50] dobey: what I don't understand is that the files I see in the queue in magicicada, and the one in the notification, are already on disk. [22:51] SpamapS: i can't tell you why they're queued [22:51] so, is it just verifying hashes? [22:51] alecu: I mean, I don't understand what "the same data we were showing" means in this context. Would weneed to examine the items before they are offloaded to disk and count something? [22:51] ralsina, well: we know the queue len, but we don't know how many items in the queue are files uploading, how many are files downloading, and how many are random stuff that syncdaemon needs to do. [22:52] ralsina, not in a clean way... [22:52] SpamapS: no, they wouldn't be in the content queue if that was the case. i have no idea why it's doanloading them again [22:52] if they're already there, that is [22:52] dobey: I have a theory [22:52] SpamapS: "the mayans did it" ? [22:52] http://paste.ubuntu.com/862655/ [22:52] because I copied these files from my mac hard drive.. [22:53] as root.. they're all owned by uids that aren't 'clint' [22:53] SpamapS: ah well, yeah, permissions problems are not fun [22:53] fix it :) [22:53] dobey: U1 may have been spinning for *months* on this [22:53] ralsina, the notification "aggregator" is loosely coupled with syncdaemon: it listens on a syncdaemon event bus. And the events are fired on that bus when they exit the "on-disk" queue. [22:54] $ find . -uid 501 -group dialout -type f | wc -l [22:54] 3556 [22:54] that'd be a lot of files indeed [22:54] ralsina, we *could* do a kind of chanchada and send events only to the aggregator before they are sent to the on-disk queue... but it would probably break somewhere else. [22:55] Ok, chowned them all.. lets see what happens [22:55] well the exceptions stopped [22:55] SpamapS: you might get quite a lot of notifications still, but it should spin down to a stop fairly quickly [22:56] restarting syncdaemon might make it happen faster [22:56] dobey: yeah, even with it spinning on these forever, the notifications weren't that annoying. I just realized that it was the same file.. over and over.. and same length.. [22:56] dobey: I would love to see the full queue length there. :) [22:56] 2012-02-29 14:56:52,930 - ubuntuone.SyncDaemon.MuteFilter - DEBUG - Blocking FS_FILE_CLOSE_WRITE {'path': '/home/clint/Pictures/FromMac/iPhoto Library/Originals/2009/Dec 9, 2009/2009-12-09 11.17.27.jpg'} (2051 left) [22:57] 2051 left .. that looks like a queue length [22:57] SpamapS: yeah, it used to be there. a change in syncdaemon to limit the in-memory queue length broke that though. i've just done some triage on the bug, and hopefully we can get it fixed for precise [22:57] dobey: *cool* [22:58] ralsina: facundobatista: FWIW, I just saw this review on the Android Market (re: the question from twitter) "Android/Ubuntu One/Xbmc = sweet setup, can instantly share my photos with the family @ home on the TV's... I have no issues with battery drain or sluggish performance, I don't get it, it works flawlessly on my end. Do your research people, this is a great app..." [22:58] I do not say there are no battery issues, we will always try to improve in that area. [23:05] karni: stop writing reviews for your own apps, it doesn't look good [23:05] :D [23:05] * karni thought he would trick people.. [23:06] Chipaca: I'm the person to know all shortcomings of the app, I wouldn't write such a review :) Now, back to building for the marketz [23:09] alright, i shouldn't still be here :) [23:09] later! [23:10] bye dobey [23:52] bye bye, all!