[00:13] thanks duanedesign [01:56] duanedesign: were you typing drunk, or just off-by-one on home row? :) === IAmNotThatGuy is now known as C0nfus3d === Tm_K is now known as Tm_T === mthaddon` is now known as mthaddon [08:36] Happy St George's Day, everyone! :D [08:50] morning! [10:52] good morning! [11:04] aquarius:Any news on the Lan Sync feature? [11:05] tizza10, heya. LAN sync (for files) is on our roadmap, but it won't be done for a while yet. You're interested in it to save bandwidth? [11:08] Yeah, it is a very important feature for me...soon to be 45gb+ of data to download each time I install/test. Good to hear its still on the roadmap, cheers. [11:21] mandel, hi.... friday when i reach my eod , urbanape's branch was failing in windows... but i'm testing it now on windows and linux and it's working... although it has some lint issues... do you think that we can merge that, so it land and i'll propose a branch right away to fix the lint issues? [11:39] gatox, fix urbanapes one, and then we merge, I think is the best approach [11:39] gatox, If the mac work has to be later and does not brake windows I prefer it [11:39] mandel, fix and propose against that one you mean? [11:39] gatox, yes, if that is possible in lp, if not, do it, tell urbanape to pull from it and we re-review his [11:40] mandel, roger that [11:40] gatox, is just that i want windows and python be stable, so braken any of them is not an option :) [11:41] mandel, no, but this doesn't brake windows..... it's working now.... the only issues with this branch now is just pylint stuff [11:41] gatox, on linux? [11:42] in windows i can't run lint checks [11:42] but, i'll fix it and tell urbanape to merge with my branch so we can merge it [11:44] gatox, where are the lint error, if the are on windows we can ignore them [11:44] gatox, if they are on linux, the branch won't land because tarmac will stop it [11:45] gatox, atm I'm not on mac because I want to get the proxy tests running on windows to have full jenkins coverage so we do not brake proxy there [11:48] mandel, they are on linux...... so i'll fix it first [11:50] gatox, yep, then you are forced to fix them :) [12:02] mandel, i have power issues again..... so i'll be around until the battery of the laptop and phone run out..... [12:03] @ping [12:03] gatox, ouch! [12:03] gatox, pong! [12:04] mandel, yap... i don't know what happend.... the generator or i don't know what is breaking since the last week quake [12:04] hello! [12:04] alecu, hi [12:05] alecu, morning! [12:05] alecu, I'm about to have lunch, but can you later give me a hand with a problem I have with the domain sockets in the tests (just tests, rest works) [12:05] alecu, stopListening hangs forever and the tests gets stuck in the clean up [12:06] mandel, probably later. [12:06] mandel, I need to check on the ssl issue on windows. [12:07] alecu, sure, I have other stuff in my plate at the moment [12:11] OMG running the pylint is killing my battery! :S [12:11] s/the// [12:17] alecu, is this comment in the code ok? "# I have no idea what I'm doing LOL" :P [12:17] i found that in sso jeje [12:17] gatox, probably not :-( [12:17] i'll remove that [12:18] gatox, who added it? in what part of the code? [12:18] alecu, urbanape..... i'll ask in the MP which was the reason.. because i don't understand why is changing the height [12:19] gatox, bzr blame [12:19] mandel, it's part of the mac port [12:19] gatox, I can't find it anywhere [12:19] gatox, is it on a different branch? [12:19] alecu, it's not in trunk [12:19] alecu, in the initial_darwin_port branch [12:19] alecu, https://code.launchpad.net/~urbanape/ubuntu-sso-client/initial-darwin-port/+merge/101112 [12:19] gatox, oh, ok then. [12:20] * alecu reboots [12:21] brb...... power is back..... switching internet :P [12:32] good morning [12:32] ralsina, good morning [12:33] alecu, mandel: this was late on friday, but we had to pull the windows release because of SSL handshake errors :-( [12:33] So, we need to get a handle on that, and I can't reproduce it [12:34] ralsina, ok, I can reproduce it making some squid server on that [12:34] mandel: really? [12:34] ralsina, yup, I tried it on thursday and I was unable to reproduce it either. :-( [12:34] ralsina, or better, twisted one, sorry [12:34] mandel, were you able to reproduce it? with a proxy? [12:34] alecu: neither rick nor leo got it either [12:34] * ralsina hopes for a quick fix [12:35] alecu, no, I mean, I can try, but after lunch [12:35] ralsina, alecu, If you want I can take a look at that after lunch [12:35] ralsina, alecu, atm I'm just fixing the squid tests to work on windows and that is no high priority [12:35] ralsina, alecu, let me run to have lunch and fix it, is that ok? [12:36] * mandel lunch [12:37] mandel: ack [12:58] does anyone know if ubuntu one can be use at the command line? Like a backup script on a server? [12:59] Captain_Proton: yes, see http://rtg.in.ua/blog/2012/03/upload-to-ubuntu-one-using-curl/ and ask me if something is not clear === rye_ is now known as rye [13:01] rye, thanks that new I look before and did not find anything. [13:04] ralsina: could you rereview https://code.launchpad.net/~thisfred/u1db/u1todo-2/+merge/102770 (or at least lift the needs info): I added docstrings where missing. [13:04] thisfred: sure [13:04] thx [13:05] thisfred: +1 and I am *not* goint to mention that on classes, the PEP suggests an empty line above the docstring ;-) [13:05] thisfred: because it makes not a bit of difference ;-) [13:06] ralsina: huh? I thought an empty line *below* the docstring only [13:06] Captain_Proton: also you can run ubuntuone-syncdaemon on the server for backup. It will require ubuntu-desktop libs and can be set up according to https://wiki.ubuntu.com/UbuntuOne/Headless [13:06] thisfred: let me fid the quote [13:07] thisfred: "Insert a blank line before and after all docstrings (one-line or multi-line) that document a class" [13:07] thisfred: from PEP 257 [13:08] thisfred: but really do NOT fix it [13:08] ralsina: well I think we deviate from that everywhere, nor does pep8 the utility complain about that [13:08] so yeah [13:08] not gonna :) [13:08] thisfred: well, this was not on PEP8, so it's not unusual that pep8 doesn't complain ;-) [13:08] true [13:09] still, I do think an empty line after *sh [13:09] thisfred: we are very very far from pep257 compliance [13:09] ould* be the norm [13:09] "The docstring for a module should generally list the classes, exceptions and functions (and any other objects) that are exported by the module" [13:09] yeah I know [13:10] so, forget it. Until I form a PEP257 task force. [13:10] which I have in my agenda for 2057 [13:10] If we do that, we'll just have a bunch of documentation that's lies, after a week [13:11] it's all lies anyway [13:11] It's all busy work, and I get the idea that it's nice for introspection, but how often do you use that really? Just look at the code. [13:12] thisfred, dobey: which is why I don't care for it. I would care more for adding __all__ to our modules though [13:12] to avoid leaking API [13:13] that makes sense [13:13] we really should move most of our code out of the normal python path, really, as it's not libraries [13:13] have i mentioned before how much i detest various things in python? :P [13:13] dobey: well... syncdaemon has stuff u1cp uses and so on [13:14] dobey: occasinally, yes [13:14] dobey: you may have, at some point [13:14] ralsina: right. some code is actual library API stuff. which is why i said most and not all :) [13:14] dobey: I know that sometimes it feels a little like programming for Amiga where the whole OS was in a single namespace [13:26] gatox: about that comment and the height: 2 reasons. 1) That comment is mostly for Chipaca. and 2) One test fails on darwin with the UI height set the other way (it's a few pixels off - presumably a minor difference in Qt on Mac OS X) [13:26] sort of an in-joke for Chipaca [13:26] the what the who? [13:26] Chipaca: https://code.launchpad.net/~urbanape/ubuntu-sso-client/initial-darwin-port/+merge/101112/comments/222034 [13:27] but legitimately, there's a pixel difference when running on Mac OS X. [13:27] urbanape: HAH! :) [13:27] urbanape, ahhhh i understand now.... [13:28] urbanape, could you change that comment for: TODO: fix pixels difference in mac... so pylint will show us that... and i can fix it after this land branch and we start working in the ui stuff? [13:28] for you? Oh, sure. [13:28] gatox: see first comment in http://media.nada.chipaca.com/static/todo.js [13:28] this branch land :P [13:29] also you need two spaces before the # if it's on same line as code [13:29] also, [13:29] / some globals to keep things interesting [13:29] var WARMING = true; [13:29] dobey: that's fixed in gatox's branch [13:29] ah ok [13:29] Chipaca, jeje...... i didn't know the history :P [13:30] Chipaca, i have this: https://github.com/ninja-ide/ninja-ide/blob/master/ninja_ide/tools/completion/analyzer.py#L3 :P [13:30] gatox: I don't know if urbanape knows where that comment (which I then repeated on twitter) came from :) [13:31] nope, just from twitter [13:31] rye, quick ? I am trying the headless, but I am stuck at adding the key to syncdaemon.conf I do not have the file. I am assuming the running ubuntuone-sso-login.py creates it or do I have to create it? [13:32] Captain_Proton: you have to create it. ubuntu-sso-client uses keyring only [13:32] I dod have one at /etc/xdg/ubuntuone/syncdaemon.conf [13:33] gatox: merged, updated, and pushed. rev 963 [13:33] Captain_Proton: right, those are the defaults. you need to create one in ~/.config/ubuntuone/ though, with your oauth token information [13:33] urbanape, awesome! [13:33] ooh Cool [13:33] thx [13:33] Captain_Proton: in the correct config section and all, but i presume rye told you what already [13:34] just from the wiki so I need to create MAIN and under that add my key [13:35] Captain_Proton: something like that. i don't rmeember the exact config myself [13:37] ralsina, can you pass me the bug number of the ssl error? [13:37] Captain_Proton: the file should be created manually in ~/.config/ubuntuone/syncdaemon.conf [13:37] mandel: sure, let me dig it [13:37] ralsina, thx! [13:39] mandel: bug #985816 [13:39] Error: Launchpad bug 985816 could not be found [13:40] briancurtin: good morning! [13:40] cool, I got that part. I am trying to start the u1sdtool and I receive this : Autolaunch error: X11 initialization failed. I am missing a a package? [13:40] ralsina: hello [13:40] Captain_Proton: have you put the u1sdtool wrapper to your home bin directory? [13:40] briancurtin: rye has found some "interesting" stuff with the windows installer [13:40] yes [13:40] briancurtin: so maybe you can take a look [13:40] ralsina, rye: certainly [13:41] Captain_Proton: could you please try checking that it is indeed running - ~/bin/u1sdtool --help - does that work? [13:41] briancurtin: 1) since we don't ship windows-installer anymore, the old cached shortcut seems to still be there [13:41] briancurtin: 2) maybe we should do an uninstall as part of the upgrade if it is possible [13:41] briancurtin: 3) we seem to leave a lot of stuff in c:\program files\ubuntuone after uninstalling (leo reported this one) [13:42] alecu, do you know how to get all the active reviews for all u1 projects? [13:42] briancurtin: all 3 seem like parts of the same thing to me ;-) [13:42] here is the whole trace : http://pastebin.com/RpjeQRnh [13:42] ralsina: OR ship the stub file which will simply start the control panel in order not to break the shortcuts? [13:42] rye: true, specially since we also add one to the user's desktop [13:42] though this can create 2 icons then [13:42] because both will be valid [13:43] mandel, yes. I'll tell you privately. [13:43] rye: we should delete the icon from the global start menu, and keep the stub exe for desktop links [13:43] rye: since installing will only update the installer user's desktop [13:43] and yuck [13:43] ralsina: ack, looking. we should also add something to stop the current U1, since any of these changes shouldn't be done in place anyway (uninstall will "succeed" even if U1 remains running) [13:43] briancurtin: *and* that [13:44] ralsina: we can't delete that icon, it is put there by windows [13:44] rye: which one? [13:44] rye: the start menu one? We are putting it there. Windows doesn't add icons to the start menu at all. [13:44] ralsina: the one that in the start menu list that users tend to click on. the recently running apps [13:44] rye: the 'recently run' one will disappear if we remove the one we put there. I think. [13:45] sigh [13:45] rye, briancurtin: but anyway, plenty of crap to look at :-/ [13:45] stupid credit card companies [13:45] ralsina: yep. i take it this is #1 priority? [13:46] briancurtin: #0 priority is to find why we are failing at auth :) [13:46] briancurtin: well, I would love to get this fixed if/when we re-release 3.0.0 [13:46] rye: mandel has that one [13:46] ralsina: off the top of your [13:46] head, do you know the signal [13:46] rye: -1 priority is finding a way to replicate it :-) [13:46] that gets emitted when a list item is checked? [13:47] ralsina: i've spent 2 days installing/snapshoting windows virtual machines and nothing. I am angry [13:47] thisfred: itemChanged, IIRC [13:47] rye: same here [13:47] thx! [13:47] rye, yes --help runs. I check under htop I do not see any process for it [13:47] thisfred: is that a QListWidget? [13:47] rye, did you manage to get any info from the user that have it failing? [13:47] Captain_Proton: ok, let's switch to PM [13:47] ralsina: yep [13:47] thisfred: then yes [13:47] mandel: let me see [13:48] thisfred: not the most informative of signals, though: http://qt-project.org/doc/qt-4.8/qlistwidget.html#itemChanged [13:48] mandel: we have plenty of logs and it just freaking fails. No common elements found :-( [13:49] ralsina: that's probably ok, there is only one change that comes in from the ui. Or will this fire when code sets the data as well? [13:49] I guess I'll find out :) [13:49] thisfred: when code sets the data too, I think. But only after the item is added to the widget. [13:49] So you may want to disconnect it before doing updates. try it out ;-) [13:50] thisfred: you need [13:50] a new keyboard [13:50] :) [13:50] rye: bug #938862 may be caused by bad permissions on the config file, could you help the user delete the metadata & config and try again? [13:50] Launchpad bug 938862 in Ubuntu One Client "syncdaemon error on Windows client" [High,Incomplete] https://launchpad.net/bugs/938862 [13:50] ralsina, the longs are hot that great, I mean, they just throw at us the exception from qt, which is very crappy [13:50] On both of us it's windows firewall and avast antivirus free. French windows 7 64. [13:50] avast... [13:50] * rye goes there [13:51] mandel: "the longs are hot that great" is probably your best typo yet. [13:51] ralsina, he he [13:51] mandel: and yes, they are crap. [13:51] rye, you mean, in you managed to reproduce the bug? [13:52] mandel: no, that's from the user's bug - bug #985584 [13:52] Launchpad bug 985584 in Ubuntu One Client "Sign Up impossible because Captcha image doesn't load" [Undecided,Triaged] https://launchpad.net/bugs/985584 [13:52] rye, I don't like the french.. could be that my code doesn't either ;) [13:53] that is a joke, just in case :) [13:53] mandel: no, that's not specific to french windows version [13:54] mandel: also, just in case somebody here uses kvm and wants to try snapshots - careful. qemu-img can segfault in the middle of snapshot restore rendering the qcow2 file broken. Still trying to get this reproduced but be warned. [13:55] rye, ack [13:59] aquarius: may I poke you regarding the filesync api - renaming, should i PUT a file prior to renaming to it? How do I find out the real path then? Can I rename to an existing file? [14:02] rye, not sure I understand the question [14:02] ralsina, which version of pyqt did we use for that package? and what version of qt does it have? [14:03] briancurtin: ^ [14:03] looking [14:03] aquarius: https://one.ubuntu.com/developer/files/store_files/cloud - renaming, how do i find out the path i need to use to rename to [14:06] rye, GET the existing file metadata, change "path" in the JSON, PUT the JSON back to the original path [14:06] ralsina, mandel: i don't see where to get a better version number, but the docs i found in the installation are 4.9 [14:07] briancurtin: if you have designer you can look at the "about qt" in the menu [14:07] aquarius: ok, thanks [14:07] briancurtin: and there is no qt 4.9 so I guess you mean Pyqt 4.9 :-) [14:07] ralsina: that says "This program uses Qt version 4.8.0" [14:07] briancurtin, import pyqt and do the following: http://diotavelli.net/PyQtWiki/Getting%20the%20version%20numbers%20of%20Qt,%20SIP%20and%20PyQt [14:07] rye, so: json = GET https://one.ubuntu.com/api/file_storage/~/Ubuntu%20One/filename.orig; json["path"] = "/filename.new"; PUT body=json https://one.ubuntu.com/api/file_storage/~/Ubuntu%20One/filename.orig [14:08] rye, will rename ~/Ubuntu One/filename.orig to ~/Ubuntu One/filename.new [14:08] and ralsina yeah, it's PyQt 4.9 with Qt 4.8.0 [14:08] rye, obviously that's pseudocode ;) [14:08] briancurtin, check with the script, just to make super sure [14:08] mandel: check what script [14:08] briancurtin, using http://diotavelli.net/PyQtWiki/Getting%20the%20version%20numbers%20of%20Qt,%20SIP%20and%20PyQt [14:09] aquarius: uh-huh, i understood. Will check how this works [14:09] mandel, ralsina: Qt 4.8.0 - SIP 4.13.1 - PyQt 4.9 [14:09] briancurtin, that is good thx! [14:10] briancurtin: that is very slightly outdated [14:10] briancurtin, mandel, alecu: should we try to use the same PyQt/Qt versions in ubuntu and windows? [14:11] ralsina, 1000000+ [14:11] ralsina, mainly to ensure that we don't have diff framework bugs [14:11] mandel: exactly [14:11] ralsina, we'll probably have to support more than one pyqt/qt versions. [14:11] alecu: yes, once we backport [14:11] alecu: but we don't need to add one more in windows just for fun :-) [14:12] ralsina, right. I agree we should use the same version we are shipping in Precise. [14:12] briancurtin: we could try to build a version with exact same Precise versions, and send to users with the problem to see if that's it [14:13] briancurtin: that one could be shipped unsigned just for testing [14:13] ralsina: that's fine by me. what are the versions needed? [14:14] pyqt 4.9.1 [14:14] qt 4.8.1 [14:14] that's remarkably current, BTW [14:15] i'll upgrade now [14:16] ralsina: should 4.9.1/4.8.1 just become "the" version to use, or should i only use it for this one-off? [14:16] briancurtin: let's use that from now on [14:16] briancurtin: and before releases we have to check what the current version is in Precise [14:17] ralsina: sounds good [14:22] ralsina, is the old package in the web? [14:22] mandel: let me check [14:23] mandel: not in a public place, but it should be in the RT [14:23] mandel: yes [14:23] mandel: https://one.ubuntu.com/windows/ [14:23] ralsina: also ^ [14:24] ralsina: i have mixed feelings regarding that folder [14:24] hello disk i/o, my old friend. [14:24] rye: thanks! [14:24] rye: why mixed? Since we are not linking it anywhere I think it's cool [14:24] dobey: i offloaded my disk io to the server [14:25] ralsina: no icons :( [14:25] rye: oh [14:29] dobey, alecu, mandel, briancurtin, thisfred, gatox, urbanape: let me introduce you to mmc_, also known as Michael McCracken [14:29] he's our new mac developer, and is starting today [14:29] Hi everyone! [14:29] mmc_, hi!! o/ [14:29] mmc_, hello! welcome! [14:29] hi mmc_! Welcome to the team! [14:30] heya, mmc_ -- welcome to the show that never ends. :) [14:30] :) thanks [14:30] hi mmc_. i'm the windows guy, and that means its a great time to mention i'll be rebooting. brb [14:30] heya, mmc_ [14:30] mmc_: I expect today you will be mostly reading docs and setting up accounts and installing stuff [14:30] welcome! [14:31] rye, one more ? I think I have it running but do I need to tell it to sync [14:31] ralsina: ALARM [14:31] ralsina: Traceback (most recent call last): [14:31] File "/usr/lib/python2.7/dist-packages/ubuntu-sso-client/ubuntu_sso/account.py", line 274, in request_password_reset_token [14:31] hi mmc_ [14:31] result = yield restful_client.restcall(operation, email=email) [14:31] WebClientError: (u'SSL handshake failed', u'') [14:31] ralsina: this means this is not Windows-only [14:31] mmc_: we have a standup in 30 minutes, which is everyone giving a very short status report, so you are welcome to look to see how we do it [14:31] ralsina: http://pastebin.ca/2139113 [14:31] ralsina, ok, sounds good [14:31] rye: we have occasional SSL handshake failures on linux, don't we? [14:31] ralsina, mmc_ just in case....... i'm writing a script for the buildout that will allow to use python, u1lint, run tests, etc.... from the buildout easier in mac [14:32] gatox: cool [14:32] ralsina: occasional, but not constant, based on the pastebin, these are not temporary [14:32] mmc_: the people who have been working on mac lately are urbanape, mandel and gatox [14:32] rye: interesting [14:32] rye: and weirder at the same time :-( [14:32] ralsina: at least that proves that we can get it in Ubuntu too [14:32] rye: since we have 10x more users on linux [14:32] rye: yes [14:33] mmc_, o/ [14:33] ralsina: but not on Precise yet [14:33] rye: yes, lost perspective there [14:33] rye: where did that came from? [14:34] ralsina, ok. what's the best way to get familiar with the code / build system, etc? Is there a wiki page or other docs I can dig into? [14:34] rye, ralsina: SSL errors in ubuntu-sso-client code are most likely coming from login.ubuntu.com [14:34] mandel, sorry, IRC newb here. o/ = ? [14:34] mmc_: it's ok :-) [14:34] mmc_: there is a doc, urbanape can you share it with mmc_? [14:35] yup [14:35] rye, ralsina: you should get someone to poke at logs from that server i guess. other option is that it /could/ be the ping url on one.ubuntu.com which gets loaded [14:35] it's available to all at Canonical: https://docs.google.com/a/canonical.com/document/d/1f7xaDT-hblCKIXrKXZjpKtFRsBRnionDix5NojwdgmE/edit?pli=1 [14:35] ralsina: bug #942312 but the user sent them directly to me [14:35] Launchpad bug 942312 in ubuntuone-client (Ubuntu) "Cannot resolve proxy hostname " [Undecided,Confirmed] https://launchpad.net/bugs/942312 [14:35] mmc_, that is a little guy saying hi! [14:35] mmc_: o/ == wave [14:35] o/~ == music note [14:35] dobey: could be. Wonder who I can ping for sso server things? [14:35] \o/ == yay [14:36] dobey: these fail for captcha urls and for password recovery urls too, does not seem to be the ping only [14:36] ralsina: web-ps i would guess [14:36] mandel, aha, ok - Clearly I have a lot to learn :) [14:36] ralsina: is there a way to connect signals that will give me an error? The signal's not being fired, but apparently I can pass in arbitrary strings to QtCore.SIGNAL() and it won't complain? [14:36] Is there anything more rigid? [14:36] thisfred: use the new notation: object.signal.connect(callable) [14:36] awesome! [14:36] mmc_, he, no worries :) [14:37] urbanape, if you're the owner of that googledoc, you're about to get a request for access from michael.mccracken@gmail -- that's me. (surprise) [14:38] mmc_: k. Also, once your canonical.com address is set up, you'll be able to log into our Google Apps area for this stuff, too. [14:38] joshuahoover, rye and I'd like to get a little more info about the ssl issues that made us roll back, is this happening to all 3.0 users, if not, are all the systems win 7 or other, are the all in the same location (USA, France, etc..) + anyother things [14:38] urbanape, ok thanks [14:38] mmc_: hi there! [14:38] mmc_: you are about to become painfully familiar with google's so-called "multiaccount" support [14:38] hi Chipaca [14:39] ralsina, yes, I heard *sad trombone* very clearly just now [14:39] joshuahoover: just for fun - one user complained about these ssl errors on ubuntu, so this seems to be not windows-specific [14:39] mandel: not all 3.0.0 users, no...we had probably around 25 requests come in last week related to being unable to authenticate...most were windows 7, a few were vista [14:40] rye: that may be, we didn't see any requests for ubuntu users coming with these types of issues but had plenty from the windows 3.0.0 users [14:41] rye, joshuahoover, it 'should' be happening in both unless is something which can be happening with the env [14:41] mandel: right, and it may, it's just that we haven't seen them come into support like we did with the windows 3.0.0 release [14:41] mandel: now that we have an ubuntu user with this we may try to get more info by crafting a script which will replicate whatever we are doing in that webclient [14:41] and just for good measure, i never saw any auth issues with proxy or no proxy on Win7 and WinXP, both proxy and no proxy [14:41] mandel: and every user who ran into this problem was able to fix it by rolling back to the 2.0.3 release [14:42] so that's fun [14:42] * ralsina has rarely been so pissed off by a bug [14:42] Or pissed on. It depends. [14:43] briancurtin: maybe they're all Vista users? :P [14:43] let me think about it, I'm sure is something stupid and we are all going to say WTF (or in my case, me cago en el puto windows de mierda) :) [14:43] haha [14:44] ralsina: one user said that happened to one machine on lan out of 3 [14:44] rye: sigh [14:44] * dobey realizes his spanish isn't that bad, it's that he can only understand the swearing. [14:45] the only idea I have is a very long race condition but that couldn't be this consistent [14:45] joshuahoover: btw, https://bugs.launchpad.net/ubuntuone-client/stable-2-0/+bug/692597/comments/8 [14:45] * ralsina slaps mandel with a wet employee handbook. Again. [14:45] Ubuntu bug 692597 in Ubuntu One Client stable-1-6 "If the user has the date/time/timezone wrong, connection fails (Expired timestamp error)" [High,In progress] [14:45] dobey, i already told you..... you know "night-spanish" [14:45] dobey: thx [14:46] ralsina: you'd be surprised how consistent race conditions can be, on other peoples' machines [14:46] dobey: sigh [14:46] * ralsina goes take an antiacid [14:46] ralsina: i recall a race bug i had to deal with recently in the music store :) [14:47] ralsina, that was not even rude at all! [14:47] ralsina, I say it nearly everyday.. [14:48] heh [14:49] mandel: https://code.launchpad.net/~dobey/libubuntuone/alertbar-align-3-0/+merge/103116 [14:49] mandel: yes you do. [14:49] mandel: easy review of backport from trunk [14:49] ralsina, so it lost meaning :) [14:50] dobey, on it, I've seen that code before [14:50] aquarius: one more question - right now i have a bit of files that happen to be MakeFile'd but no content was added. They can be seen in the children list but cannot be queried directly. Seems wrong [14:50] rye, um [14:50] rye, that's not supposed to happen. sounds like a bug, that; webm0nk3y may have thoughts on it [14:51] hmm [14:51] karni: may remember if we return them or not [14:51] mandel: yeah, i don't like the code, but me cago en el puto gtk+ de mierda :P [14:51] mandel: see? you are a bad example [14:52] oops... rye, karni may know if we return the file place holders [14:52] webm0nk3y: i know we do return, i did not know they are inconsistent [14:52] dobey, yes, is weird that the vbox have to be added, plus the warning in the compilation, we should remove those if we can.. [14:52] mmc_: is this you? http://michael-mccracken.net/ ? [14:52] rye: are you familiar with the process to see if we can get some server-side info on those SSL errors? [14:53] urbanape, yes that's me [14:53] rye: because I never tried that before [14:53] mandel: we can't remove the compiler warnings about vbox. the "fix" doesn't work with gtk2 and i'd rather not have a bunch of #ifdef crap in there :-/ [14:53] rye: what do you mean by inconsistent? [14:54] rye, regarding the ssl bug: is the one ubuntu user with problems using proxies? [14:54] dobey, certainly, I prefer warnings to ifdef [14:54] rye: inconsistent? [14:54] future will hopefully be better though. would much rather just inject js/html into the DOM [14:55] rye: webm0nk3y: I believe we return them when the dir is asked with_children=true (or include_children, was it). Dunno what is returned if you query directly. [14:55] karni: I don't remember if we do it by design or by coincidence (aka accident) [14:56] rye, one more ? I think I have it running but do I need to tell it to sync [14:56] karni: ok, nevermind, i will test it more and will file a bug [14:57] rye: please let me know of your findings :) [14:59] Captain_Proton: ubuntuone-syncdaemon syncs automatically [14:59] cd - [15:00] rye, nevermind, I just read in the logs that proxies were disabled for the ubuntu user. [15:00] alecu: uh-huh :( [15:01] alecu, FYI this https://code.launchpad.net/~alecu/ubuntuone-client/timestamp-autofix-1-4/+merge/82755 and this are not landing https://code.launchpad.net/~alecu/ubuntuone-client/timestamp-autofix-1-6/+merge/82754 [15:02] mmc_, I've added you to the osx port google doc [15:02] thanks alecu [15:02] are we doing standup? [15:02] me [15:02] me [15:02] hehehe [15:03] me [15:03] skip it! :) [15:04] @skipIf('dobey') [15:04] me [15:04] me [15:04] mandel, lols [15:04] thisfred, ralsina, urbanape, mmc_: standup? [15:04] me [15:05] me [15:05] me [15:05] DONE: u1fb [15:05] mmc_: you are exent for today ;-) [15:05] excepted. Or something like that. [15:05] mandel, go [15:05] mandel, go [15:05] DONE: Reviews. Wome work regarding squid on windows tests (I have added several bug about it for u1-dev-tools). SSL errors in 3.0 [15:05] TODO: SSL errors in 3.0, tcp-activation mac, squid on windows [15:05] BLOCKED: no, well, the ssl debugging :) [15:05] gatox, tira! [15:05] ralsina: exempt [15:05] DONE: [15:05] Review, run tests and propose a branch against the mac port branch. Lot of reviews on Friday. Working on an env setup script for mac. [15:05] TODO: [15:05] Finish with the env setup script, review/fix bugs in my queue [15:05] BLOCKED: [15:06] No [15:06] dobey, go [15:06] mandel, you skipped briancurtin [15:06] λ DONE: reviews, bug #962860 (stable-3-0) [15:06] λ TODO: backports, SRUs, u1db packaging/buildsys [15:06] λ BLCK: none. [15:06] Launchpad bug 962860 in libubuntuone (Ubuntu Precise) "The prompt for installing MP3 playback support dialogue box does not fit in a 1024x600 display" [High,Triaged] https://launchpad.net/bugs/962860 [15:06] i'll go at the end [15:06] urbanape [15:06] briancurtin, ouch, perdon! [15:06] DONE: Fixed last little bits of the branch for final acceptance. TODO: Back on iOS, acting as reference for mmc_. BLOCK: None. NEXT: thisfred [15:06] DONE: u1db example app done status TODO: hook example app up to real u1db database BLOCKED: no NEXT: briancurtin [15:06] DONE: i can't really remember what i did outside of the branches i proposed...it was a long weekend [15:06] TODO: get windows installer into shape [15:06] BLOCKED: none [15:06] NEXT: alecu [15:07] ralsina: i think you can go now :P [15:07] DONE: rendering of clothes, pulling of hair, etc. over darned ssl issue. Effective results = none. Reviews, administrivia. TODO: hope and dream. Tech leads call. BLOCKED: by own inability to reproduce. [15:07] DONE: conference day: Mozilla MDN Hack day in Buenos Aires [15:07] TODO: try to find the cause for "SSL handshake failed" bug. Write conference report [15:07] BLOCKED: no [15:07] NEXT: [15:07] thisfred, you go [15:07] ralsina: he already did [15:07] ralsina, thisfred already went. and returned. [15:08] * dobey wonders if The Disorderlies is on Amazon Prime Instant [15:08] mmc_: so, the idea of standup is that we go around the whle team, and everyone does "DONE/TODO/BLOCKED/(possibly something else if needed)" and then passes the baton to the next guy [15:08] mmc_, so, that's how the standup should work. But we should be respecting the turns, you see :-) [15:08] ok, sounds easy enough [15:08] mmc_: other than that, there is a weekly 1-1 call with me, and on thursdays, instead of standup we have a 1-hour team call [15:09] then, after the round ends, I say "Comments?" [15:09] ralsina, how often is standup? daily? [15:09] Comments? [15:09] mmc_, "BLOCKED" is something that's blocking you from getting your job advancing. [15:09] mmc_: daily [15:09] Disorderlies is not on amazon streaming. so sad :( [15:10] if there are no comments, EOM [15:11] EOM it is [15:11] back to your usual squabbling [15:12] in that case, dobey: IMO, having the fat boys on demand, whenever you want them, would be unhealthy. [15:12] mandel, so, what shall we do about this ssl bug? any further idea? [15:13] mandel, I've tried installing 3.0.0 in clean vms with 7 and xp and I can't reproduce it. [15:13] alecu, did we confirm that the user in ubuntu had the issue? [15:13] mandel, http://pastebin.ca/2139113 [15:13] mandel, that log says "he had it". [15:14] mandel, also, his proxy was disabled. [15:14] mandel, (according to the log) [15:14] not the cleanest standup i've seen from you guys :) [15:14] Chipaca: indeed [15:14] alecu, I'm being evil and testsing what happens if I play with the security settings of the internet connection of the system [15:14] mmc_: did you get the ubuntu sso account yet? [15:14] alecu, first one, disable the use a v3 ssl certs, that should give the error [15:15] mmc_: and, with that, access to the canonical wiki? [15:15] ralsina, no - I haven't heard anything after the initial request for info from the sysadmin [15:15] mmc_: ok, let me ping the guys [15:16] ralsina, ok thanks [15:16] mandel, if you have a moment (inn your review day :D), please review this branch: https://code.launchpad.net/~urbanape/ubuntu-sso-client/initial-darwin-port/+merge/101112 (i already approve it) [15:16] mandel: i got live.com broken after i tested the various settings but U1 kept working [15:16] rye: that should make us happy, actually. A bug this hard to reproduce can not affect all high a percentage of users. [15:17] rye, even the v3 ssl stuff? [15:17] urbanape: what OS X version is everyone working with? I'm using 10.6 here, for no particular reason [15:17] I recently upgraded to Lion [15:18] but I don't think there's any reason the stuff in that doc wouldn't work on Snow Leopard. [15:18] thisfred: apparently allhands forgot my evaluation of you [15:19] thisfred: so I am going to redo it and try to straighten things out with HR [15:19] ralsina: ok, as long as it doesn't mean I'm fired :) [15:19] thisfred: not in my hands anymore! ;-) [15:19] ha [15:20] urbanape: ok, I guess I'll check that assumption for you :) [15:20] SCIENCE! [15:23] mandel, I think we should hook into the "sslErrors" signal, and log with more detail the ssl errors that are detected. [15:24] alecu, that is for sure, at least will give more info than that little 'meh, I'dont work' message [15:25] alecu, I think an exception only with 'because' would have been less useful [15:27] rye, joshuahoover, ralsina, alecu, removing support for ssl v3 will return an error from sso, is not the one we have, but just so you know you will get a download error in the ui [15:27] mmc_: do you have your account details? you are supposed to have access to wiki.canonical.com now [15:27] mandel: ack [15:27] mandel, why do you propose removing support for ssl v3? [15:28] briancurtin, what is the version of Qt that is used for the windows 3.0.0 package? [15:28] mandel: want to re-weigh in on my branch? [15:29] ralsina, I got an email saying that I was added as a member of Canonical, but still no wiki. Here's what I'm doing: [15:29] cd - [15:29] mmc_: if you are now a member of canonical, the wiki.canonical.com should just work. May need you to re-login though [15:29] alecu: it was Qt 4.8.0 - SIP 4.13.1 - PyQt 4.9.0 [15:29] ralsina: 1 - logged into launchpad, 2 - tried to access https://wiki.canonical.com/PeopleAndCulture/Induction/NewStaffTasks. 3 - got error. 4- read your message. will try to relogin... [15:30] oh come one, irssi. cd - means going to the previous window... Y U NO SHELL? [15:31] mmc_: try login.ubuntu.com [15:31] urbanape, yep, I'm running tests on another one and will move to yours [15:31] alecu, I'm not proposing, but if a user did, we brake [15:31] alecu, for the next time we see anything like that [15:32] alecu: i just installed the PyQt 4.9.1 but it did not include Qt 4.8.1 like i was hoping, it's still 4.8.0 so i'm going to try to upgrade Qt by itself [15:32] mandel, "if a user did" what? [15:33] alecu, go to internet options, advance, security, un-check ssl v3 [15:33] alecu, why would they do it, because people are stupid [15:33] briancurtin, right, 4.8.1 is shipped on ubuntu [15:33] * alecu thought we were still using 4.7.x [15:35] alecu, already asked that question in case we had the qt bug, not the case :) [15:36] doh. [15:36] * gatox lunch! === gatox is now known as gatox_lunch [15:38] ralsina - now it works. thanks! [15:43] mmc_: ok, you probably have enough to entertain you for today there. ping me here for any questions, or things you get stucked on [15:44] ralsina, yep looks like there's lots to do. thanks [15:52] ok need to get lunch. bbiab [15:53] Quantal Quetza meh I would have call it Quantal Quesadilla which sound funnier :) === mmc_ is now known as mmcc === mmcc is now known as mmc_ [16:02] FYI: It looks like 'mmc' is in use, so I'm switching to 'mmcc' for my nick, to avoid the underscore… [16:03] mmcc: ack [16:03] mmcc: you can register it in freenode [16:04] ralsina, yeah, that's how I found out that 'mmc' was in use [16:04] mmc: ack [16:04] mmcc, you can ask for it if the guy did not use it in ages [16:05] mandel: last seen = now :\ [16:05] mmcc, ouch.. well mmcc is not that bad :) [16:06] mandel, yeah I'm not complaining. [16:06] mmcc: plus, it's a roman year [16:06] mmcc: now you are not a multimedia card. Says a grass... [16:06] mmcc: and welcome :) [16:06] rye: you are actually a cereal! [16:07] ralsina: wikipedia says I am a grass [16:07] rye: yes, all cereals are grasses [16:07] rye, thanks for the welcome. And I assumed rye = liquor [16:07] of course, liquor from the grass/cereal [16:07] * ralsina wants a picture of a tuna on roman [16:07] ralsina, do you know where s test_qtwisted.py? [16:07] mmcc: hm, never tried that [16:08] mandel: that was moved around by urbanape's branch I think [16:08] mandel: used to be test_windows.py or something [16:08] aye [16:08] and I added it to the exclusion list in run-tests [16:08] ok [16:09] urbanape, so the issue of the segfault was mixing qt and glib, cool :) [16:09] rye: "Rye is a cereal grain and should not be confused with ryegrass, which is used for lawns, pasture, and hay for livestock." [16:09] right [16:09] rye is also a really good whiskey [16:10] ah, catching up [16:10] herp [16:10] urbanape, so, this might be an issue: +PLATFORM_QSS = ":/qtwisted.qss" since the qss will have to be platform specific and not reactor specific.. but since you are moved to ios I'm not going to be a PITA and will add it as a bug for gatox [16:11] Ah, then that should be easy enough to punt back up to the relevant modules. [16:12] urbanape, can you do that, I just realize the following the old qss is there (data/qt/windows.qss) so that line will be an issue since on windows we wont load the style [16:12] Syre thing [16:12] sure even [16:12] urbanape, so either revert, so less work for you and we sort it out later, or add it in a platform specific module [16:13] nah, I got it. [16:13] gatox_lunch, bad boy ^ [16:14] should it be darwin.qss for darwin, do you know? [16:15] urbanape, yes, you can copy the data/qt/windows.qss for darwin [16:16] urbanape, once done that the resources should add it and we qt will be able to find the css (the : means look in the resources path) [16:17] urbanape, and even in this case I would do a simple if sys.platform, no need to do anything more complicated [16:20] done and pushed in rev 964 [16:20] * mandel looks [16:25] urbanape, since you are just using source to set the qss, why not setting it directly in the __init__py? you can use the very nice idea of the qss map for all the platforms (I don't know when sys.platform will return linux3) [16:25] because of that [16:26] but if we're okay with that just breaking at some future python point [16:27] in which case, I'd rather just rename the resource win32.qss and do PLATFORM_QSS = ':/%s.qss' % sys.platform [16:31] urbanape: +0 let's try not to make up things to do ;-) [16:33] which is why I only went so far [16:35] urbanape, ralsina, what about something like: http://paste.ubuntu.com/942755/ [16:35] mandel: nice [16:36] ralsina, urbanape, linux will be there for linux2 and linux3, also centralizes the way to add new qss per platforms, amiga? [16:36] urbanape, but is not a blocker for the branch perse, I'm just being an ass (sorry 'after hand') but I want to avoid the problems we had moving things around too much [16:38] we're supporting amiga? [16:38] :) [16:38] briancurtin, at this rate I would not be surprised hehehe [16:39] since we're starting to support Mac, let's go a step further back and support NeXTSTEP [16:40] lol === gatox_lunch is now known as gatox [16:42] mandel, what happend [16:42] ? [16:43] gatox, read the backlog, you missed an error jejeje [16:44] ok, lunch is starting to look like a good idea === salgado is now known as salgado-lunch [16:50] afk, quick lunch and A/C repairman [16:59] ralsina, urbanape: i'd rather just delete the .qss ;) [16:59] dobey: you can use -stylesheet whatever.qss ;-) [17:00] doesn't help other users though :) [17:11] * ralsina is failing at lunchbreaking === salgado-lunch is now known as salgado [17:51] alecu, ralsina, please review the following https://code.launchpad.net/~mandel/ubuntu-sso-client/fix-ssl/+merge/103147 that will add much better logging for the qtwebclient (atm will not add it for libsoup) [17:51] mandel, alecu: are we using qt webclient on ubuntu? [17:52] mandel: looking [17:52] ralsina, we should be using it since we have a qt control panel [17:52] mandel: he [17:52] ralsina, it was logged :) [17:52] mandel: because we may have a guinea pig in facundobatista [17:53] ralsina, I have been told so, but he has a problem in sd not control panel, right? [17:53] mandel: right, in protobuffers [17:53] alecu: ping. do we need to do anything for bug #692597 in u1cp stable-2-0 and stable-1-0? [17:53] Launchpad bug 692597 in Ubuntu One Client stable-1-6 "If the user has the date/time/timezone wrong, connection fails (Expired timestamp error)" [High,In progress] https://launchpad.net/bugs/692597 [17:53] * facundobatista has just a pig inside [17:54] facundobatista: hmmmm pig [17:54] facundobatista: just in case: you don't get the SSL thing with u1cp? [17:54] ralsina, the MP is there, adds a lot of details to the logs, including each cert for each error its cert details and its error number and string [17:55] ralsina, what? [17:55] mandel: launchpad is taking a bit to scan the branch [17:55] facundobatista: the SSL handshake errors, you are not getting them in u1cp, only in syncdaemon? [17:56] ralsina, ah! yes, you're right [17:56] ralsina, you do know the trick of bzr diff --old lp:ubuntu-sso-client --using meld or kdiff for you? [17:56] mandel: yes, I know it, but I would rather just see the diff in the page first [17:56] mandel: since I usually have other things to fill the minute before it appears ;) [17:56] ralsina, ok, is a simple one and has tests. I need to walk the dog, might be back later [17:57] mandel: and because the diff line numbers are useful to discuss it [17:57] mandel: ack [17:58] ralsina, Iets approve that branch and get mre info, but I think there is a problem in the other side and not ours, version 2 is not as secure [17:58] ok, laters! [18:00] ralsina, ejemplo de un par de errores: http://paste.ubuntu.com/942880/ [18:00] ahora si, adios [18:00] ups, sorry, nothing to important :) [18:02] bye mandel! [18:03] alecu, briancurtin: after we do some basic testing of mandel's branch, we should do an installer and have joshuahoover pass it to some experimental^W user [18:03] ralsina: yes, let's do that :) [18:03] joshuahoover: it's just for diagnostics, will not fix it [18:04] ralsina: that's fine...anything that might help us track it down further [18:04] ralsina: which branch is this? i'm working on getting pyqt with qt 4.8.1...the PyQt 4.9.1 only includes Qt 4.8.0 :/ [18:04] oh, the fix-ssl one [18:05] briancurtin: yikes, that's going to be tricky [18:06] ralsina: yep, it's already been a pain in the ass, but im getting Qt 4.8.1 on here to try and rebuild my own copy [18:06] briancurtin: ack [18:13] hmm [18:13] do i spend time chasing people, or try to do other SRUs instead [18:14] ralsina: btw, when should we talk about release scheduling? [18:14] dobey: hmmm... early tomorrow morning ok with you? [18:15] ralsina: there is some roadmap/release call i am supposed to be on tomorrow at 11:30 my time [18:16] dobey: ok, after that one [18:16] dobey: remember to ping me [18:16] ok [18:38] alecu: I +1d https://code.launchpad.net/~mandel/ubuntu-sso-client/fix-ssl/+merge/103147 it needs another review [18:38] ralsina, got back from lunch; looking [18:38] and mandel's review day is already over [18:38] alecu: cool, thanks [18:39] ralsina: well i guess mandel shouldn't review his own branch :P [18:53] ralsina, for asking IS about new account info, is it actually the #is channel, not the #/is channel? The wiki says #/is but that seems to be wrong… [18:53] facundobatista: ping? [18:53] mmcc: it's #is in the private irc server [18:53] ralsina: ok, thanks [18:53] mmcc: you are there already? If yes, please join #u1-internal [18:54] mmcc: which is our secret channel [18:54] ralsina, done. [18:56] alecu: did you see my question from earlier? [18:57] beuno: do you know if facundobatista is around today? [18:58] dobey, looking [18:58] achiang, he was around earlier today. [18:58] right [18:58] I'd expect to see him around again soon [18:59] alecu: ok, thanks. not urgent, just revisiting the "sync massive amounts of data" issue again, since i'd like to have my stuff safely in the cloud before travelling at the end of this week [18:59] (travel = bring my external backup hd on a plane) [18:59] achiang, leavcing already? when! [18:59] beuno: fly out on saturday [18:59] back to the States [19:00] :-/ [19:00] too bad, i just discovered a great lunch cafe near me too [19:00] achiang, you can always come back ;) [19:01] beuno: yes, i'll come back when i don't need to ship hardware anymore. :-/ [19:01] * achiang curses argentine aduana again [19:02] heh [19:02] dobey, don't recall the details of bug #692597, but probably we need some fixes for old stable versions too. [19:02] Launchpad bug 692597 in Ubuntu One Client stable-1-6 "If the user has the date/time/timezone wrong, connection fails (Expired timestamp error)" [High,In progress] https://launchpad.net/bugs/692597 [19:03] alecu: right, i'm asking, because there were no branches merged to stable-2-0 or stable-1-0 of cp, and i was trying to get at least the stable-2-0 (oneiric) SRUs done [19:06] achiang, indeed. [19:10] if i was to put Windows release steps/tips/requirements somewhere, where would that be? wiki.canonical? wiki.ubuntu? something else? [19:12] dobey, I need to run to the kinder. I'll take a look later today to see what's missing for stable-2-0. [19:20] alecu: ok, thanks [19:22] briancurtin: wiki.canonical.com [19:23] ralsina: cool, starting a page there to act as a checklist (ie. don't forget to sign autoupdate) and also because i need to document how i got Qt 4.8.1 in here [19:23] briancurtin: awesome [19:29] achiang, I'm here [19:31] facundobatista: hi, i've been trying to sync a large (75GB) amount of data to U1... using an external drive, bind mount into home directory, then selecting it for sync to U1. the daemon consumes a lot of CPU, i hear disk activity, but then i get dbus errors [19:31] is there a size limit or type that can not be uploaded to ubuntu one? I setup a sync folder and if I put a txt file or something small in there it works fine if I put a bzip2 file in there that 250mb it does not upload [19:31] facundobatista: watching my network interface, i don't see much traffic cross the wire. also, u1sdtool --waiting just times out [19:34] achiang, the dbus errors are because of the bubbles, or something similar, alecu would know better [19:35] achiang, the --waiting should not timeout unless the daemon is working too hard, sometimes that happens, though, but it should work later [19:36] achiang, and it's normal to use a lot of CPU and disk, when you throw a lot of data... however, if you're connected, it should use the network [19:36] achiang, if you have logs in DEBUG mode we can tell exactly what's going on [19:37] ralsina: heard back from one user impacted by the auth issue in 3.0.0 and she mentioned that norton gave her no shortage of warnings when installing and setting up the application...i don't know that any of us test with norton or similar software [19:37] facundobatista: how do i start it in debug mode? [19:38] achiang, you should do the following: [19:38] 1. stop the syncdaemon client ("u1sdtool --quit") and be sure it's fully stopped ("ps -eaf | grep ubuntuone-client" should give you nothing). [19:38] 2. put a file named syncdaemon.conf in your $HOME/.config/ubuntuone directory with the following information: [19:38] [logging] [19:38] level = DEBUG [19:38] 3. restart the client. [19:38] ok [19:38] Captain_Proton: the size limit for individual files is several GB. 250MB file should upload fine. i've uploaded many of them many times [19:39] achiang: you're getting dbus errors? what version of ubuntuone-client package? [19:39] dobey: whatever is latest in precise [19:40] * achiang sets up the bind mount again [19:41] then dbus errors probably aren't from the notification bubbles. [19:43] facundobatista: u1sdtool --quit doesn't actually force the quit. i'll just kill -9 them [19:48] facundobatista: ok, restarted everything in debug mode [19:48] facundobatista: where do the logs go? [19:49] joshuahoover: not with norton. When we had warnings from antivirus in the past, I contacted the companies [19:49] joshuahoover: probably norton also gives firewall warnings [19:49] achiang, ~/.cache/ubuntuone/logs [19:50] achiang, you can leave it for a while, if you don't see activity, grab the logs, and send me and alecu by mail (or use a pastebin if they are small... they shouldn't) [19:52] ralsina: yeah...i'm going to give it a shot just for fun [19:52] facundobatista: http://people.canonical.com/~achiang/u1/ [19:52] facundobatista: i get that error pop-up about 1x every minute [19:53] achiang: does controlpanel.log have anything to say about it? [19:55] dobey: http://pastebin.ubuntu.com/943075/ [19:55] ugh [19:55] https://bugs.launchpad.net/ubuntu/+source/ubuntuone-control-panel/+bug/982754 [19:55] Ubuntu bug 982754 in ubuntuone-control-panel (Ubuntu) "ubuntu one, crashes on start up and the details say it's an IPC error." [Undecided,Confirmed] [19:59] achiang, ok, that's not the syncdaemon itself, dobey and alecu will be able to help you more than me [19:59] facundobatista: i guess i just need to be more patient. i think there are like 20K files for syncdaemon to iterate over and do hash checking. :( [20:01] achiang: do you have a ~/.config/ubuntuone/syncdaemon.conf file? [20:02] achiang: is syncdaemon actually running? [20:02] dobey: yes, the syncdaemon.conf file is getting updated [20:03] i guess so. if it's doing local rescan from first startup, then probably the dbus interface just hasn't come up yet :-/ [20:03] 2012-04-23 17:02:51,471 - ubuntuone.SyncDaemon.EQ - DEBUG - push_event: FS_FILE_OPEN, kwargs: {'path': '/home/achiang/Pictures/2007/0922 - Tour de Fat/IMG_4054.CR2'} [20:03] unfortunately, i have 5 more years of photos for it to go until it reaches 2012. :( [20:03] does syncdaemon always do a local rescan? [20:05] when it starts, yes [20:05] oh, that makes me sad. [20:06] seems like there should be a way to cache that data somehow [20:07] it does cache. it won't do the same work every time [20:07] but if you have changes to files or new files, it will have to look at them [20:07] interesting. perhaps it's because i've never let it complete successfully? (i got bored after 2 hours of local scan) [20:08] it's because you killed it, and restarted, before it rescanned them all previoiusly, yes [20:08] https://code.launchpad.net/~thisfred/u1db/indexing-deleted-docs/+merge/103165 [20:08] so it still had a lot of work to do when restarted [20:08] ok, will let it sit longer then. thanks dobey [20:09] achiang: 20K files that are ~8MB each is going to take a long time for it to chomp through, unfortunately :( [20:09] dobey: yeah, and i upgraded my camera in 2011 to something that creates ~20MB raw files. :-/ [20:10] ok, have to take a break and get some sugar [20:10] will be back in 15' [20:10] oh yeah, 20K files of 20MB will be a while. :-/ [20:11] i wonder why hashing is so slow though. it seems to be one of our big bottlenecks [20:13] dobey: hashing 20K files of 20MB means reading aproximately a hugabyte of data [20:14] * achiang is happy to donate a huge dataset if folks need stuff to test on [20:14] i'll have my external hdd at UDS if someone just wants to slurp it up [20:15] i don't think finding a huge dataset to test with is a problem :) [20:15] well, you need unique files, right? [20:15] it's why you can't just cp the same file over and over again to generate the dataset... [20:16] but ok, if you have what you need, then i'm happy too. :) [20:16] achiang: well random device is hopefully not the same data over and over again :) [20:16] if it is, we also find a security bug :P [20:16] ha [20:16] can you find enough entropy though? :) [20:17] yey!!!! i can run the tests in mac just with ./run-mac-tests :D \o/ [20:18] actually, i don't think uniqueness matters anyway, for this specific problem [20:18] gatox: COOL [20:20] gatox: why a separate run-mac-tests script? [20:20] more path insanity? [20:21] dobey, because at this moment we are ignoring some tests in mac... and because we are setting the environment in a different way than linux... [20:21] dobey, you have to point to a specific python that execute u1trial that is not in the path, etc....... usiing the buildout [20:22] i create a env-mac script.... that add the necessary paths from the buildout to the sysenv..... so you can run the run-mac-tests script [20:22] right [20:22] the "path insanity" :) [20:22] dobey, :P [20:34] alecu: not to be annoying, but any progress in reviewing mandel's fix-ssl branch? [20:34] * ralsina doesn't want to be *more* annoying [20:36] briancurtin, joshuahoover: I am suspecting we will not get the test-ssl bundle done before tomorrow unless alecu finishes that branch review. briancurtin, maybe you could start an installer build with that branch merged manually? [20:36] ralsina: sure, i can do that [20:36] briancurtin: cool, please do it [20:43] EOD here!! see you tomorrow people! (i'll propose the branches for run-mac-tests tomorrow as soon as i fix u1lint in mac) [20:45] EOD for me as well. briancurtin please upload/send joshuahoover the installer once you have it [20:45] have a good night ralsina [20:47] ralsina: if you're bored: [20:47] https://code.launchpad.net/~thisfred/u1db/indexing-deleted-docs/+merge/103165 [20:47] https://code.launchpad.net/~thisfred/u1db/u1todo-3/+merge/103170 [20:47] ralsina, joshuahoover: ack, building now [20:47] a [20:47] EOD, never mind :) [20:49] thisfred: that branch makes it look like a real application! ;-) [20:49] thisfred: I will review before your start tomorrow [20:49] no hurry! [20:49] thisfred: I review your example app for fun because I want to use u1db myself ;-) [20:50] thisfred: I consider it on-the-job training [20:50] I'm not gonna stop you ;) [20:50] * dobey wants to use u1db too [20:50] dobey: PYTHON! ;-) [20:50] using it is very simple [20:50] ralsina: on android, iphone, and webos. [20:51] me too [20:51] ralsina: on ubuntu, i just need to finish my vala impl [20:51] dobey: he [20:51] I am gonna build the recipe app [20:51] with shopping list integration [20:51] dobey: and to become a mathematician, I just need to take two exams [20:51] i am gonna build something much more insidious [20:51] dobey: but yes ;-) [20:52] ralsina: well, hopefully writing a u1db impl in vala is easier than taking those two exams. ;) [20:52] ralsina: or is this a certificate from BrainBench? [20:52] dobey: differential calculus and nonlinear optimization [20:53] dobey: not the toughest of subjects even [20:53] dobey: just the ones I never bothered with [20:53] so pretty close to brainbench [20:53] dobey: for the past... 20 years [20:53] heh [20:53] i'm sure there's a youtube video to teach you [20:54] dobey: most certainly there are plenty. [20:54] * ralsina could do his thesis in "hand->little bunny's head homeomorphism" [20:55] and defend it as chinese shadows [20:55] I have an MIT courseware channel on my roku box. So far it's not beating netflix ;) [20:55] * ralsina is sure that's not the right term in english [20:55] thisfred: yeah, I imagine "Deuce Bigelow 2" packs more entertainment wattage [20:56] that'd be a tough choice [20:56] * ralsina would take an animal husbandry course before that [20:56] yeah [20:56] as long as it involved no lab work [20:57] thisfred: as long as the cow doesn't sit down, it's not a problem [20:57] haha [20:57] * ralsina was a country boy [20:58] both my grandfathers were farm hands, so I'm not too far removed from wearing wooden shoes. [21:01] thisfred: animal husbandry AND wooden shoes is too much. You have to have standards. [21:01] in the Netherlands you can't have one without the other [21:01] OTOH, if my dad had used wooden shoes, I would have been a much nicer boy. [21:01] hehe === m_conley is now known as m_conley_away [21:21] brew install qt [21:21] …wait… [21:21] (ouch) [21:21] er, that last line was supposed to say: blah blah 2746 files, 200M, built in 165.2 minutes [21:22] I guess make -j is just not a thing in QT? [21:22] Qt uses cmake doesn't it? [21:23] dobey: no [21:24] dobey: Qt uses qmake but it builds it as part of the process [21:24] oh right [21:24] although I have not built Qt in ages [21:24] i had totally removed the horror of qmake from my brain [21:25] I wonder if it'd die if I used MAKE='make -j' $cmd [21:25] mmcc: probably will not die [21:26] ralsina: I'm almost, but not quite, curious enough to try. Hoping I won't have to. [21:30] mmcc leave it overnight :-) [21:31] dobey: you are taking bug #978903 ? really? [21:31] Launchpad bug 978903 in ubuntuone-client (Ubuntu Precise) "[precise] Client is stuck due to Upload executing before MakeFile" [High,Triaged] https://launchpad.net/bugs/978903 [21:33] dobey: oh, merging to stable. Ok. [21:33] ralsina: i assigned the packaging side to me, since i will presumably be doing it :P [21:44] dobey: ack [21:53] later all! have a good evening [22:19] yay, my local scan finally finished [22:19] u1sdtool is finally responsive again [22:25] joshuahoover: http://u1.to/~brian.curtin/g/3.0.0-windows-installer - also emailed to you [22:25] and with that, i'm out === m_conley_away is now known as m_conley [23:13] * mmcc is leaving for the night…