=== beuno_ is now known as beuno [00:53] Later all === m_conley is now known as m_conley_away [07:33] so, mandel FILESYNC_ERROR is str and WARNING_MARKUP is unicode [09:46] rye, cute... I wonder why.. [09:47] mandel, are you using spanish locale? [09:47] where, in my ubuntu? no] [09:47] oh, and good morning! [09:48] Good morning all! [09:52] mandel, ¿Y U NO USE ESPAÑOL? [09:53] rye, cause is a stupid language for computing :P [09:53] mandel, but anything non-latin1 is good for catching unicode issues [09:55] rye, I've got enough with running a prerelease hehe [09:56] mandel, I am running precise AND spanish. I understand 30% of the labels :) [09:58] rye, sorry, that is too much for me [10:05] Morning guys o/ [10:11] * mandel restarting due to updates === mandel is now known as p === p is now known as Guest63379 [11:04] good morningg!!! [11:04] gatox, moarninggsss mwahaha! [11:04] gatox, https://bugs.launchpad.net/ubuntu/+source/ubuntuone-control-panel/+bug/937809 [11:04] rye, mmmmmm...... i didn't like that laugh jejeje [11:05] ralsina, more unicode? unbelievable! :P [11:07] gatox, are you using spanish locale on your development machine? [11:11] rye, it's an english version of win, with a chinese/spanish username [11:11] gatox, well, the bug is for ubuntu, gettext returns 'str' which i don't know whether is used in windows version, i guess it is [11:11] gatox, the windows version does not have translations [11:11] i suppose [11:12] rye, we are using gettext wrong, that is the provlem [11:12] problem* [11:12] rye, ahhh i didn't see that was for linux.... just saw unicode aand start hitting my head with the table jeje [11:12] should be easy to fix anyway [11:13] we should be using ugettext [11:13] Guest63379 is mandel btw [11:13] oh, sorry :P === Guest63379 is now known as mandel [11:13] much better :) [11:13] mandel, i could not figure out how to drop-in-replace gettext with ugettext [11:14] yet [11:14] mandel, ahhhhh you are the mistery guy! jeje [11:14] mandel, did you see this: https://plus.google.com/115212051037621986145/posts/6x1BrUNgTuA ??? [11:17] * mandel looking [11:17] rye, I think we have been using gettext instead of ugettext everywhere.. so is a pita [11:18] mandel, i wonder why it exploded only now [11:18] gatox, nice! [11:18] mainerror, yeah! :P [11:18] rye, a mistery.. [11:18] gatox, I'm here! [11:19] Huh? [11:19] mandel, ahhhhhh [11:19] hehe [11:19] mainerror, sorry.... code completion always confuse mandel with mainerror :P [11:19] mainerror, our nicknames are too close and people get confused with the autocmplete hehe [11:19] heh [11:19] code completion jeje [11:19] just completion [11:19] unless by removing gtk we changed something related... [11:20] gtk control panel [11:20] well, the net result is that control panel for qt does not work for any languages that use something outside latin1 [11:20] or ascii, to be precise [11:20] hello everyone! [11:20] rye, super cute! [11:20] rye, :P [11:20] nessita, morning! [11:20] rye: hola! where is that? [11:21] sweet! I got my ubuntu machine back [11:22] * mandel switches machine [11:22] nessita, in ubuntu, now, gettext returns 'str' but our language files for e.g. spanish are in utf-8 [11:22] nessita, https://bugs.launchpad.net/ubuntu/+source/ubuntuone-control-panel/+bug/937809 [11:22] nessita, https://bugs.launchpad.net/ubuntu/+source/ubuntuone-control-panel/+bug/937809 [11:22] :P [11:23] rye: awesome, will propose a fix. I just fixed that for sso last week :-) [11:23] nessita, oh, shiny! [11:23] rye: thanks for catching this! [11:24] nessita, you are very welcome [11:25] I'm back!!! hurray! [11:29] nessita, i was thinking about the UIFe for the bug about how to handle showing the errors inside the wizard..... but, we don't need a UIFe for that (i think), because in the wizard from sso we are just showing the errors in a new dialog, the inside display of the errors is only affecting the windows wizard [11:30] gatox: but we want to stop showing the dialogs in the sso screens, no? [11:30] and integrate the errors with the same wizard pages [11:31] nessita, ah yes, if that is the plan, so we need a UIFe that involve to move the errors inside the pages, handling the correct way to handle that is just consequencce..... am i right? [11:32] nessita, is that gettext issue the same we talked last time in your place? [11:32] nessita, hola, btw :) [11:32] gatox: yes [11:32] nessita, ok, i'll create that bug now! [11:33] facundobatista: hola! yes, though is not in sso (is fixed), but now is also in controlpanel, so will apply the same fix [11:33] gatox: wait! [11:33] nessita, what? [11:33] gatox: wasn't the bug already created? [11:33] nessita, i don't remember to have seen any bug related to move the errors inside the pages...... [11:34] nessita, ah, ok [11:34] gatox: you were supposed to file the UIFe for that along with the margin changes [11:35] nessita, i report the one with the padding [11:36] gatox: let me re-read our chat from last fri 'cause now I'm confused about what we agree [11:36] * gatox too [11:38] nessita, i file the uife bug about the padding..... only the one about the errors remain... but i found about that we are not using that in that way in linux..... now that you said that the idea is to move that into the pages, what we really need to file is the bug to move the errors inside the pages..... that's what i understand [11:38] gatox: as per my logs, you were supposed to create that bug asking for the UIFe on Monday :-P [11:39] gatox, alecu, nessita FYI I have control panel working through a proxy atm \o/ [11:40] mandel: yey! ah, btw [11:40] mandel: saw my comment on the bug you filled yesterday re: strings for sso dialog? [11:40] nessita, no, let me check [11:41] nessita, yeah, I though about the doc being private.. I'll provide more details regarding the strings to be used [11:41] mandel: ok, and consider we need a UIFe for that, since we're making releases today [11:42] nessita, yes, there are several branches of mine for review that will require that, I'll coordinate with josh about this [11:42] mandel: thanks! [11:42] no problem :) [11:56] gatox: yes, more unicode :-( [11:57] good morning everyone! [11:57] ralsina, hi! :P [12:03] hola ralsina! [12:04] hola nessita! I did very early reviews of the stable branches, just for you! [12:05] ralsina: thanks! already merged, so I'm moving forward with the releases :-) [12:05] nessita: cool [12:05] ralsina: so, I will have to leave for an hour approx after lunch :-/ (I'm filled with hate :-D) [12:06] nessita: ack [12:15] rye: would you be able to review this? https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/unicode-gettext/+merge/94136 [12:17] nessita, interesting [12:22] nessita, muy bien! it works for spanish/russian for sure [12:22] rye: YEY (can you please try both, -gtk and -qt?) [12:23] nessita, well, yes, both -qt and -gtk work equally good, gtk one has more strings translated though [12:23] nessita, also, qt one does not want to get killed by Ctrl+C [12:23] but that is not relevant [12:24] rye: right to the first one, 'cuz -qt just got into the ubuntu repo last week, so translator are still getting the new strings. About the second... yes, we know but have no idea why (and have not debugged yet) [12:26] nessita, well, here qt one seems to be killed KeyboardInterrupt but event loop continues, in case of u1sdtool this is sigkill only [12:28] nessita, ralsina that branch will brake if you receive a http://developer.qt.nokia.com/doc/qt-4.8/qevent.html#Type-enum of type LanguageChange because the retranslateUi methopd provided by qt designer will be using QtGui.QApplication.translate instead of gettext [12:29] that happens when the QTranslate is changed.. I don't know when that happens [12:30] mandel: what branch, the gettext-unicode? if so, how the code that was before there was any different? [12:30] nessita, the compiler modification you changed was the one that ensured that the correct method was used, do a python setup.py build and compare to _ui files, one from trunk and one from that branch [12:31] this guy: _wrapuic [12:31] mandel: doing it ATM, but then why is working without that code? [12:32] nessita, because we are manually setting the labels in the widgets using setText(GETTEXT_TEXT) but that is done during the __init__ of the widgets, if we get the LanguageChange event, which is when the QTranslater changes the retranslateUi method won't use gettext [12:33] mandel: so this only affects when someone changes the language while the app is opened? [12:33] nessita, yes, only then AFAIK [12:33] mandel: so, we can workaround by not setting any string from the designer files, no? [12:34] nessita, the it will set everything to '' [12:34] mandel: hum? I mean, settings the button's string using code [12:36] nessita, read the following: http://developer.qt.nokia.com/doc/qt-4.8/qcoreapplication.html#installTranslator [12:36] mandel: the thing is that we need to do something that works without having to import the _ [12:36] mandel: since we need to install it with gettext.install so we can force to have unicodes from it [12:37] nessita, I'd say, remove it, branch that code and then add a bug saying that we do not support dynamic-translation and we won't [12:37] nessita, I mean, not branch, land [12:38] mandel: but I see another issue now, using what you showed me: there are strings that will never get translated, for example: [12:38] self.edit_profile_button.setText(QtGui.QApplication.translate("Form", "Edit personal details online", None, QtGui.QApplication.UnicodeUTF8)) [12:38] mandel: we used to have _("Edit personal details online") and that will make the string show up for translation [12:38] now I guess it won't? [12:38] nessita, indeed [12:39] nessita, but is nicer, since it would be confusing for translators.. since we have the common ones to translate and then the ones from the QDesigner, so duplication everywhere [12:40] mandel: the problem is that some strings are only set once in designer, so now I need to move those out and set them by code [12:40] nessita, best thing I can think of, use your code and allow us to use strings in designer, so that we can see what we are doing [12:40] then state, dynamic translation is not supported [12:40] if you change the lang, close and open the app [12:41] mandel: does gettext support dynamic translation? [12:41] nessita, very very good question.. and I have no idea, maybe dobey knows it [12:42] buen día, gente [12:42] nessita, maybe Chipaca knows :) [12:42] Chipaca, buenos dias! [12:42] hola Chipaca! [12:42] tapping into the proxies experience we have, if I wanted a glib app to support proxies, should i make it use libsoup? (this is for the remote videos scope) [12:44] Chipaca, yes, although you have to write code dealing with the 407 and the 401 coming from the proxy, I'm playing with the libsoup code atm so I should be able to point you to our webclient for that [12:44] what does dynamic translation mean? gettext is rather crufty, but you can call it twice in a row with a little setup in between and it'll switch languages [12:44] Chipaca: is my understanding that using libsoup is the best option, though you still need to do some not trivial tasks to proxy actually being used. Among those are retrieving proxy credentials, validating ssl cert and handling any errors, etc [12:45] plain ol' http, here [12:45] it's a lovely place where responses are fast and uncomplicated (?) [12:45] mandel: i'd appreciate those pointers :) [12:46] Chipaca, so, it depends how libsoup was compiled, if it was compiled with the --gnome-support or something like that flag, it will pick the proxy settings, otherwhise you have to manually set it, which is a pita [12:46] Chipaca, I should be done with the code this afternoon, can you wait that long? [12:46] Chipaca: there is an email from alecu in reply to "Preferred C HTTP Library", that has a little more info... you saw that? [12:46] this is python :) [12:46] Chipaca: ah! [12:47] Chipaca: then the lens may use the ussoc's webclient (new) library? [12:47] url! url! :) [12:47] indeed, if you re using python you can use our code [12:47] which has integration tests with squid etc.. [12:48] * Chipaca pulls ussoc trunk [12:48] Chipaca: http://bazaar.launchpad.net/~ubuntuone-control-tower/ubuntu-sso-client/trunk/files/head:/ubuntu_sso/utils/webclient/ [12:48] yep, bzr and grep got me that far :) [12:49] Chipaca: the choosing of which webclient is "automagic", so you just import the factory and start enjoying [12:49] davidcalle: ping, about lens, dependencies, and proxies [12:49] Chipaca, well.. no in trunk there is not proxy support yet [12:49] nessita,mandel: about the translation stuff, I really really hate .ui files that have no text in them, so unless we actually know how to crash it, please don't remove them :-/ [12:49] Chipaca, ? [12:50] Chipaca, lp:~mandel/ubuntu-sso-client/webclient-use-dialog [12:50] davidcalle: hi. The remote video lens does not support proxies, and uses urllib2 to get things [12:50] Chipaca, no 100% ready yet.. [12:50] davidcalle: would you be ok with depending on an external library (part of the ubuntu sso client) for that? [12:50] Chipaca, yes [12:50] ralsina: the problem is that those strings are not getting translated... I guess we can leave some dummy strings in place, as long as we always override them with code (and thus set a proper string instead) [12:51] ralsina, +1 on not removing the strings, lets use them but not translate them using gettext, so I thin nessita code is 100% correct, we simply say that if you change the lang of the system you will have to close and open [12:51] Chipaca, if it doesn't affect the speed. [12:51] davidcalle, it will affect the speed if you don't have the proxy settings ;) [12:51] davidcalle: the other option is to use libsoup, but there are some corner cases with proxies that it alone doesn't cover [12:53] Chipaca, davidcalle the API is asyn using deferreds, so it should not block, and the overhead should be in the interaction with the user to get the creds, either from him or from the keyring [12:53] davidcalle: would you rather the lens interrupted the user to ask for the proxy credentials if they aren't set up, or would you rather it just bailed? [12:53] that might be a design question [12:54] mandel: since you already looked at the branch, would you please review (I filed bug #938626 to fix the other things we discussed)? https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/unicode-gettext/+merge/94136 [12:54] mandel,nessita: +1 on that [12:54] nessita, yes :) [12:54] Chipaca, I didn't knew that urllib2 was an issue with proxies. Anyway, interrupting the lens for credentials is pretty bad. I might be missing something, why the lens with what we currently have, would need credentials? [12:55] davidcalle: authenticated proxies [12:55] davidcalle: proxies can require credentials [12:55] Ok [12:55] davidcalle, my nightmare.. :) [12:56] davidcalle: I agree interrupting from the lens is bad. I think we should just let it bail. [12:57] We have the possibility to display an error message in the lens if no results have been fetched : "No credentials found for your proxy " or something, and let everything being handled by the keyring. [12:58] Hum, I'm not saying that 'No credentials found for your proxy" is a godd user facing string, but you get the idea :) [12:59] nessita, +1 [12:59] nessita, I'll try to research at some point in my life how to deal with this in the future.. only thing I can think of is a generic QDesigner extension, but would be a PITA to work with.. and maybe not worth the effort [13:01] nessita, did you see the ubuntudev twit? :D [13:02] ok, does anyone understand this: [13:02] [13:02] [13:02] libsoup-CRITICAL **: soup_session_pause_message: assertion `item != NULL' failed [13:02] gatox: no! link? [13:02] nessita, i retweeted [13:02] Chipaca, I just need to investigate libsoup then. I'm used to urllib2 for calls and lxml if I need to parse something (faster than libsoup). [13:02] gatox: :-D [13:03] nessita, grosa! :D [13:03] heheheh [13:03] lol [13:05] davidcalle, Chipaca so, the deal is, if you don't have the creds you can tell the user to go to the system settings and use user:password@domain, gsetting will lie to you because it will say no auth required, then if you use that for the proxy url in the libsoup session you will be fine [13:05] davidcalle, Chipaca if you get a auth error and use auth.is_for_proxy() you will know if it was a 401 from the proxy or from the source [13:06] where source I mean which ever server thing you are using [13:06] Ok [13:14] mandel: you lost me there a bit [13:14] Chipaca, ok, open your system settings and go to network, you will see that the ui does not provide a single thing for auth creds [13:14] Chipaca, to work around that you set the url of the proxy to be user:password@domain [13:15] Chipaca, for example mande:proxy@daredevil.marvel [13:15] mandel: but if the user has done that, things just work via libsoup? [13:15] Chipaca, yes [13:15] Chipaca, well, if libsoup was compiled with the gnome flag [13:17] Chipaca, you could get a libsoup that uses nothing from gnome :) [13:17] mandel: yah, yah. Which it is in ubuntu if anybody is sane, yes? :) [13:17] Chipaca, yes [13:17] mandel: you say "compiled with the gnome flag", but you still have to add the SoupGNOMEProxyWhatever thing to the session, yes? [13:19] Chipaca, automatic proxy detection should be there: http://mail.gnome.org/archives/libsoup-list/2009-May/msg00008.html [13:21] http://developer.gnome.org/libsoup/2.37/libsoup-client-howto.html [13:21] mandel: ^ adding the gnome proxy thing (in C) [13:22] Chipaca, which is exactly the same that we do via pygobject [13:22] mandel: yep [13:23] mandel: so you're saying it's automatic once you've manually set it up :) [13:24] Chipaca, well, yes, but you just have to set the proxy-uri thing [13:33] ok, I'm off to have lunch [13:37] Chipaca, about "Recently Viewed", have you discussed it with John Lea? [13:37] davidcalle: yes [13:37] davidcalle: only filed the bug after getting his go ahead [13:37] Chipaca, cool, merging then. [13:37] davidcalle: ditto with the others [13:38] bah, i haven't checked the "refresh" one with him, but only because i can't imagine a world where we *don't* want to refresh :) [13:38] davidcalle: I'll be fixing most of these today [13:38] davidcalle: unless you want to work on any of them? [13:39] (the "refresh" one is one that can wait, because it's clearly not a feature nor ui change :) [13:42] Chipaca, I'm currently writing a lens tuto for dev.ubuntu. I don't mind if you want to work on them. About the refreshing, when do you want to do it? On each search? [13:42] davidcalle: nah, once an hour tops [13:42] Chipaca, great :) [13:43] Chipaca, I'm merging the i18n branch, and adding your string change with it. === yofel_ is now known as yofel [14:10] nessita: what do you mean by "dynamic translations" ? [14:10] hola dobey! [14:10] hola [14:10] dobey: it was a term that mandel was using, I think he means changing the language while having an app opened (and thus expect that the strings change to teh new language) [14:11] it's certainly possible, but not easy, and not worth the trouble of implementing, i don't think [14:11] dobey: right [14:11] hi team [14:11] hola briancurtin! [14:12] briancurtin, hi [14:12] hi briancurtin! So, you didn't get selected? [14:13] ralsina: ha. i was in the list of 40 people to be questioned for a trial, but i know the lead attorney [14:13] briancurtin: lucky break! [14:14] i'm still on the hook through friday, but they said once you've been called in it's unlikely they call you back in during that period [14:28] nessita: we have our 1-1 but I am feeling kinda crappy, so I'd rather we had it tomorrow [14:28] ralsina: is ok, let's have it tomorrow [14:28] * ralsina doesn't want to puke on mumble (yes, too much information) [14:52] nessita, dobey open and close the app is good enough, is not that people keep switching from one lang to another all the time [14:59] argh, standup! [14:59] me [14:59] me [14:59] 16 seconds to go! [14:59] me [15:00] me [15:00] me [15:01] alecu, dobey? [15:01] meh [15:01] me [15:01] ralsina: go! [15:01] DONE: calls, reviews, interviews, canonicaladmin, misc, reading about OSX development, installed a hackintosh, minor hacking TODO: feeling ill, so maybe go to bed BLOCKED: by feeling like crap [15:01] gatox, please [15:02] DONE: [15:02] Relaxing a lot during holiday (monday-tuesday), File some UIFe issues, Propose a branch for UI styling in SSO. Start working in SSO padding problem (almost done!). Happy about tihs: https://plus.google.com/115212051037621986145/posts/6x1BrUNgTuA (yes, i needed to say it) [15:02] TODO: [15:02] Finish with the padding issue, work in the other UIFe issue. [15:02] BLOCKED: [15:02] No [15:02] briancurtin, go [15:02] DONE: holiday, federal jury duty [15:02] TODO: 851810, figure out the very last level of exposing the notification since i have the lower levels figured out [15:02] BLOCKED: none [15:02] NEXT: mandel [15:02] DONE: Finished and proposed branches with the ssl dialog. One review for nessita. Debugged why libsoup webclient is giving issues. Fixed my P machine that went to hell. [15:02] TODO: Talk with josh about exceptions. More proxy. Mumble with alecu, urbanape and ralsina [15:02] BLOCKED: no [15:02] nessita, please [15:02] DONE: lots of holidays [15:02] TODO: ubuntu releases, fix bug #938626 and bug #937809 and build patches for packages [15:02] BLOCKED: nopes [15:02] NEXT: dobey [15:02] λ DONE: releases, pyqt pokery [15:02] λ TODO: finish releases, deal with some twisted upstream stuff [15:02] λ BLCK: none. [15:02] alecu [15:02] DONE: got a working tunnel server, started with tunnel client [15:02] TODO: mumble with mandel & urbanape, work on tunnel client (bug #929207) [15:02] BLOCKED: no [15:02] EOM? [15:02] comments anyone? [15:03] nop [15:03] I was wondering if is everyone familiar with the freeze exception process... JIC [15:03] ralsina: speaking of OSX development, i looked around and can't find the install CD so i'll need to get a new one [15:03] nessita, I'm a little, but I've never had to deal with it [15:03] briancurtin: ok, let me check what we can do about that [15:03] nessita: it wouldn't seem to affect me for now, but for the future i should probably know [15:03] nessita, any wiki I can get the infor from? [15:04] briancurtin, mandel: https://wiki.ubuntu.com/FreezeExceptionProcess, any question about that, please ask. Yes, is mostly FYI, what I care the most is that we don't land branches that violates some of the freezes that are happening these weeks [15:04] ok [15:05] dobey: is our precise tarmac blocked? [15:06] i don't think so [15:07] dobey: hum, this is not landing https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/unicode-gettext/+merge/94136 [15:07] (approved 57 minutes ago and has commit message) [15:08] nessita: i can't get to it to verify anything anyway [15:09] Chipaca, dead bip? [15:10] aquarius: yes. just sent you a message. [15:11] ralsina, should be briancurtin in the mac os x chat? [15:13] briancurtin, FYI mac os x chat in 17 mins aprox [15:13] mandel: mumble or somewhere in IRC? [15:13] briancurtin, mumble so that you can hear or horrible accents :) [15:14] mandel: briancurtin: yes [15:15] mandel: i'll have to make up an accent, or get one of those voice encoders like movie criminals [15:16] briancurtin, If I were you I'd fake a mexican accent to blend in the team hehehe [15:17] nessita, mandel, gatox: btw, you have earned scorn for https://code.launchpad.net/~diegosarmentero/ubuntuone-client/network-detect/+merge/88545 [15:17] Chipaca, I have no message? [15:18] dobey, why? [15:19] dobey, scorn or corn? :P [15:19] gatox: you commented out two constant imports, and use arbitrary numbers instead of the constants, where they were being used before [15:19] aquarius: whatsapp [15:19] gatox: and mandel and nessita approved it! [15:20] dobey: why? [15:20] Chipaca, ah. blew away my phone and haven't reinstalled it ;) [15:20] dobey, bloody true: 10 +# NM_STATE_CONNECTED_GLOBAL, [15:20] 11 +# NM_STATE_DISCONNECTED, [15:20] aquarius: ah. Well. "bip server down" was pretty much the extent [15:20] mandel: yes, that [15:20] and the related changes to that [15:21] Chipaca, heh, ok :) [15:21] dobey: besides the commented lines, that should be removed, what else? [15:21] dobey, that's why we were removing that things.... but couldn't remove everything yet [15:21] dobey, I'm up to emend that error.. [15:21] dobey: the replacemente of that is not a fetaure adding, but using an existent feature [15:21] nessita: they shouldn't be removed. [15:21] dobey: yes they should (along with the fake network manager, there is a bug for that) [15:22] dobey: bug #935389 [15:22] oh no bot [15:22] nessita: then they should have been removed in that, rather than simply replacing them with magic numbers in this branch [15:22] https://bugs.launchpad.net/ubuntuone-client/+bug/935389 [15:23] dobey: yes, agreed that the magic number replacing is bad. But is in test_dbus.py, so I filed the bug to fix in another branch [15:23] nessita: branches that replace well defined and descriptive constants, with magic numbers, should not be approved [15:24] nessita, I agree with dobey we should have removed it all at once, but since its going away, is not a huge issue [15:24] it is an issue when i propose a backport branch for making the release, and it includes that change, and a reviewer asks me why it is doing that [15:24] dobey: I agree, and we should not do that again. Though for this particular case, since is a test file, is not that bad. I know gatox was trying to remove the fakednetworkmanager in that branch, but since it got hairy, we left that for another branch [15:25] dobey: ok, that's bad, yes [15:25] dobey: want gatox to propose a revert for that? [15:25] no. i want to get the release done. and for you all to know to never do that again :) [15:27] dobey: duly noted. gatox, got that? ^ [15:27] yes [15:27] dobey: thanks! [15:55] alecu, dobey: you are both confirmed for UDS [15:55] ralsina, cool [15:56] oh right, i've been meaning to ping about that anyway [15:56] dobey: so, preemptively ponged [15:57] ralsina: and no sprint or some-hands or whatever the week prior, right? uds only? :) [15:59] dobey: AFAIK just UDS [16:00] ok, good [16:00] alecu, suertudo! [16:01] i wonder if i can get a flight that's not too insane. [16:02] dobey: flying there won't be too bad. flying back is another story :/ [16:03] briancurtin: ? probably wouldn't be as bad for you. it seems i'd have at least 2 stops on the way [16:04] dobey: ok i change my answer then [16:09] nessita: care to do https://code.launchpad.net/~dobey/ubuntuone-client/update-from-trunk/+merge/94053 ? [16:09] dobey: not at all [16:23] * gatox lunch! [16:34] lunch, bbiab [16:35] * mandel wants to kill himself.. [16:36] mandel: first! called it! ;-) [16:36] mandel: why? [16:37] Damn, lost the opportunity to call shotgun instead [16:38] ralsina, stupid proxy raising Unauthorised exceptions that are cough by the backend of control panel which then cleans the creds.. then everything explotes! [16:39] mandel: libsoup or qt? [16:39] ralsina, atm qt [16:40] ralsina, I'm fixing it atm on the sso code.. but it has made me waste some time.. [16:42] dobey: approved [16:42] good morning all.. I'm having issues syncing tomboy across several ubuntu/kubuntu machines. When I manually sync the system asks if I want to make the copy on my machine marked as (old) and get the server version. Now all my machines have duplicate notes half of which are labeled (old). How do I remedy this sync process? [16:42] mandel, ralsina: can I haz a review please? https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/all-ui-from-code/+merge/94220 [16:43] nessita: of course [16:43] thanks! [16:43] nessita, sure [16:44] ugh, i forgot to submit per diem from the sprint. quick question if anyone knows: 1) is it too late, 2) what is the max value? [16:46] briancurtin: it's not too late, and there is a table for what you claim, it's a fixed amount per day [16:47] briancurtin: um ... maybe in private with your manager :) [16:47] Chipaca: ah yes i'll find the table, and oops [16:49] nessita, this looks funny: if getattr(self.ui, 'label', None) is not None: why is that? [16:50] briancurtin, I know that 1 trillion dollars is not acceptable as per diem.. :( [16:51] mandel: that looks like The Right Way to do hasattr(self.ui, 'label') [16:51] mandel: well, assuming self.ui.label cannot be None. Otherwise, the not-so-good-way :) [16:51] mandel: test blowing up in my face 'cause we're faking the object that setupUI returns :-/ [16:51] Chipaca: why not so good way? [16:52] nessita: if self.ui.label can be None, then getattr(self.ui, 'label', None) can't tell "self.ui has no attribute 'label'" from "self.ui.label is None" [16:53] Chipaca: ah, yes. Anyways, if it's None I can't call what's next, so the check is what I want (tm) :-) [16:53] nessita: if self.ui.label can be None, you'd do: marker=object(); if getattr(self.ui, 'label', marker) is not marker [16:53] nessita, looks good, I would have been A LOT lazier than you and I would have done the following, created a dict with the name of the label/button/blah as key and the string as value [16:53] nessita: yes, of course :) [16:53] mandel: only if you had dinner in pluto [16:53] nessita, then I would have iteritems and would have done a setText in each [16:53] mandel: half a trillion for lunch in pluto, though [16:54] nessita: not-so-good only in the "right way to do hasattr", not necessarily bad when you can opportunistically do less to do more [16:54] mandel: but sometimes is setText, sometimes is other method [16:54] Chipaca: :-) [16:55] nessita, yes, setTitle, but that is 6 methods vs the rest, nevertheless is because I'm lazy, now that it has been typed, who cares :) [16:55] +1 on the branch [16:55] mandel: thanks! [16:58] nessita: isn't it more pythonic in lines ~688 to just do the remove and catch KeyError? [16:58] ralsina: looking [16:58] sorry, I meant "do the setText" [16:59] nessita: also, why would it not have a button? [16:59] ralsina: sorry, got lost. What would be more pythonic than what? [16:59] nessita: lines 688-689 of the diff [17:00] ralsina: the button may not be the 'cause in tests self.ui is a extremely generic fake [17:00] ralsina: what would you use instead of the getattr? [17:00] nessita: I don't understand why that may happen. And if it may happen, why do a getattr instead of a try/except KeyError [17:00] ralsina: exception are expensive [17:00] nessita: having the code do strange things because of a test seems backwards to me [17:01] ralsina: yes, though I need to put this in ASAP before I build the packages (will be a vendor patch) [17:01] nessita: but hey, I am nitpicking [17:01] ok [17:01] ralsina: I agree, but fixing the test suite will take me some time that I don't have (today) [17:01] can you add a FIXME, maybe? It you say no, it's fine too [17:01] ralsina: I will create a bug report, is that ok? [17:01] or a FIXME< as you prefer [17:02] nessita: wonderfultastic [17:02] nessita: +1 [17:02] thanks [17:04] mandel, ralsina: disclaimer comment added in each case [17:06] I'm not sure exceptions are expensive in python.. how do you get out of iterators? [17:07] raise StopIteration [17:10] mandel, ralsina: equivalent branch for ussoc: https://code.launchpad.net/~nataliabidart/ubuntu-sso-client/all-ui-from-code/+merge/94228 [17:11] nessita: on it! [17:11] mandel: just FYI, whena adding constants to the ubuntu_sso/utils/ui.py, we try to add it in alphabetical order (sometimes we fail just because we're human). I corrected the PROXY_* constants (just FYI) [17:11] nessita, they are corrected in one of my proposed brancjes [17:12] branches* [17:12] mandel: ok, then you will need to merge trunk in :-/ [17:12] nessita, they are in the same place, I'd be surprised if we have issues [17:12] ack [17:13] nessita, why the new param in 4169 [17:13] looking [17:14] mandel: every wizard page needs to know which app_name is being displayed for, so we can build proper messages like 'In order to connect to %(app_name)s please have an internet connection foo bar' [17:14] mandel: see lines 4216 and 4217 [17:14] nessita, so doing something like page.wizard().appname is not a good idea? [17:15] mandel: not at all, we're trying very hard to remove the dependency on the parent wizard in each page [17:15] mandel: we get tons of AttributeError because .wizard() is None [17:16] mandel: so our code turns into a complex workflow where we can only do stuff when the page has been added to a wizard [17:16] which, from my POV, is not good, since we want the pages to be "self-sufficient" [17:16] so we can re-use them from other places, such as embed those in other UIs [17:18] nessita, ok, I don't know how you guys are dealing with the ui its you code :D [17:18] but maybe doing a widget instead of a wizard page is a better idea.. seems odd to say you have a wizard page with no wizard.. [17:18] what does a wizard page add to a widget? [17:19] mandel: that's the next step, but we couldn't do all the desired refactoring in time [17:19] ok [17:21] mandel: a wizard page has a wizard() method and has a method where it can do things like setting up the wizard's button layout [17:22] mandel: but of course we can split that into a widget and a wizardpage with the widget, or have the wizard setup its own buttons on next() and other ways [17:36] nessita, forgotten_password.ui changed sizes, is that an issue? [17:37] 446 => 230 [17:37] mandel: no, I just adjusted the form size with "adjust size" (it was big for no reason) [17:37] ok [17:39] nessita, I see you added a spacer for setup_account.ui, we had resizing issues? [17:39] mandel: hum, that was not intentionally... let me check [17:40] nessita, the object name is verticalSpacer_4 [17:40] nessita, and verticalSpacer_5 [17:42] mandel: what diff line? [17:43] nessita, in lp is 3449 [17:44] mandel: how are you seeing the diff? [17:44] nessita, meld [17:45] mandel: in LP the whole file is diffed since I changed the EOLs, and in meld I don't get that difference. Is your base branch (trunk) up to date? [17:45] nessita, double checking [17:47] nessita, sorry false alarm something landed from a few hours, sorry [17:47] mandel: ack [17:49] nessita, +1 [17:50] mandel: thanks! [17:50] dobey: hey, I have a non trivial question re-packaging... [17:51] dobey: both u1cp-gtk and u1cp-qt have some shared dependencies on some data under usr_prefix/ubuntuone-control-panel/*.png... how shall we deal with this at packaging level? [17:52] dobey: ATM, the -gtk binary is installing those... but if someone installs only the -qt package the images will be missing [17:55] nessita: make a new binary package called ubuntuone-control-panel-common, with those files, and have both depend on it [17:56] dobey: ack, thanks [17:58] brb! === gatox is now known as gatox_brb [18:05] dobey: another question, shall I remove the -gtk package? (I don't think so since we have not resolved the cd-yadda issue yet) [18:05] dobey: but perhaps you have a special plan in your head :-) [18:05] nessita: i don't see any reason not to remove it [18:05] nessita: other than the fact that we need a ffe to do so i guess [18:05] dobey: wouldn't desktops "break"? the launcher icon, the messaging icon [18:05] nessita: no [18:06] dobey: you already released to ubuntu ubuntuone-installer? if so, I need to bump the depends on controlpanel [18:07] nessita: ubuntuone-installer has been installed by default for probably 6 or 7 months already :) [18:07] dobey: yeah, I know that. But I was asking about the latest change you made, where it checks for -qt instead of -gtk [18:07] nessita: i haven't released it yet. am about to do that [18:08] dobey: ack, what version will that be? [18:08] nessita: same as everything else. 2.99.5 [18:08] great, thanks. So I will not remove the -gtk controlpanel yet, until we get the FFe [18:12] nessita: we need to get ubuntuone-control-panel-gtk removed from Depends of ubuntu-desktop package as well (which is main reason for ffe) [18:12] dobey: right [18:13] ok, I gotta go to run an unpleasant errand... will be back in one hour I hope [18:13] ralsina: ^ === nessita is now known as nessita-away [18:14] nessita-away: ack [18:15] this is why i hate parking my car on the street in the city [18:17] ok, EOD for me, laters! === JanC_ is now known as JanC === gatox_brb is now known as gatox [18:44] nessita-away: on bug #851810, i'm kind of at a dead-end. i'm not totally sure how to move forward exposing this signal. i have it doen and tested in VM and interaction_interfaces, but where above that does it need to be exposed? [18:44] (whenever you get back, of course) [18:57] https://code.launchpad.net/~dobey/ubuntuone-installer/update-from-trunk/+merge/94246 [19:05] can have a review please? [19:07] dobey: i looked and it seems reasonable to me [19:09] dobey: +2 with mine [19:39] sigh, gtk-doc is evil === nessita-away is now known as nessita [19:45] hey briancurtin [19:46] nessita: hio [19:46] briancurtin: so, you need to also add the signal to linux/dbus_interface.py and windows/ipc.py and ipc_client.py [19:46] briancurtin: that may be the only bit left for you [19:46] hm, so i have it in there. i guess i just need to test those pieces [19:47] briancurtin: you have that there and is not working? [19:47] nessita: well i have it in there because i had been following other similar types of notifications, but i havent written the tests yet to ensure that they work (i did it slightly backwards) [19:48] briancurtin: OH NO.... how dare you? :-) [19:48] the world didnt end yet, but i'll be on the lookout [19:48] :) [19:48] ;-) [20:00] 1 hr left. [20:03] dobey: what for? [20:03] oh no [20:04] that'd be 25 hrs wouldn't [20:04] am thinking it is thursday already :P [20:04] nessita: ui freeze [20:04] dobey: right, 25 hours, stop scaring me :-) [20:14] dobey: I'm adding the new u1cp-common bin package... I see that debian/tmp/usr/share/indicators/messages/applications/ubuntuone-control-panel should also be installed by that package, am I getting that right? (the content of that file is the path to ubuntuone-installer.desktop) [20:14] * gatox wants to kill a qlabel [20:14] nessita: yes; and you can make that package depend on ubuntuone-installer [20:16] dobey: right, and remove that dep for the -gtk one [20:16] right; well -gtk and -qt should depend on the -common [20:16] so will get it that way [20:17] dobey: should we consider having the com.ubuntuone.controlpanel.gui.service provided by also the -qt UI? (will require further fixings, but now that I see that, I wonder...) [20:18] rye, the file you published at the end of your vala+dbus post was very funny! [20:18] alecu, i suppose this should trigger a natural reflex of opening in every geek [20:18] lol [20:18] PLAINTEXT PASSWORD ME WANT! [20:20] alecu, even i went forward to open it, because it looked so interesting :) [20:20] nessita: well i guess not. are we actually using it anywhere? [20:20] dobey: I think the we're using that to integrate with the launcher... [20:24] the indicator. we open to the folders tab [20:24] we'll have to make it run the qt panel directly i guess [20:24] dobey: and we also set the urgency on the control panel by using that from the launcher, I think [20:25] nessita: no [20:25] nessita: the urgency is set using the unity launcher api [20:26] though i am not sure it works at the moment (though it should) [20:26] but we also make the controlpanel do something when the --alert is passed [20:26] nah, it was just checking if the launcher hint was set, and then unsetting it, when it gets launched [20:26] of course, it can't do that now, becuase of the gi issues [20:27] ah. Boo. [20:27] the qt panel could; or the installer could [20:27] we should probably have the installer do that [20:28] since it should be getting run when you click on the launcher, or on a message; and it then opens the control panel [20:29] ok..... i'm off for today..... EOD.... see you tomorrow! [20:32] dobey: right [21:05] hey folks... [21:06] I am on a machine I want to connect log into my ubuntu one account. But I don't want my ~/Music to try and download (its big). [21:06] will I gte the chance to not sync that folder? [21:10] dobey: is this correct for the new u1cp-common package? http://pastebin.ubuntu.com/853236/ my question is mainly regarding the Depends: ${misc:Depends} which I'm not 100% sure what's that for [21:11] j0nr, if that folder is a UDS yes, no problem what so ever [21:11] nessita: i think in general, people like the long description to be a bit longer. [21:12] dobey: have a suggestion, to understand what other info may be relevant in that desc? [21:12] nessita: and you need appropriate Breaks/Replaces; and you might also want to Recommends: indicator-messages or whatever the package is for it [21:12] nessita: i don't. i don't really like the needlessly long descriptions [21:13] dobey: breaks/replaces for ubuntuone-control-panel-gtk, right? [21:13] j0nr: if it's outside of ~/Ubuntu One/ then it isn't synced by default [21:13] nessita: yes [21:16] oh am tired now :-/ [21:16] dobey: so if I log in to the U1 client, by default it will only try to sync ~/Ubuntu One ? [21:17] dobey: why the Recommends on indicator-messages? so the icon in on the envelop and can display the shares/udfs/etc messages? [21:21] j0nr: yes; you'll have to subscribe the other folders you want synced [21:21] nessita: because the package provides a file that integrates with it [21:22] dobey: right, but then why is not a Depends? [21:22] nessita: because it's ok if it's not there [21:22] ack [21:51] ok, I'm gone to pilates, will get back in one hour and will finish the uploads to ubuntu (still waiting on some nightlies build to confirm proper packaging) [21:51] brb [22:14] later all === gatox is now known as gatox_away === lifeless_ is now known as lifeless [23:24] * nessita is back === m_conley_away is now known as m_conley