[07:09] mandel: ping [07:17] ralsina, pong [07:17] morning all!!! [07:17] hola manuel! [07:17] ralsina, que leches haces despierto! [07:17] loco! [07:17] :) [07:17] mandel, I just replied to your mail, please check it to see what you can be doing this morning [07:18] I had to piss. You know, almost 40, that kind of thing. [07:19] ralsina, I'v just read it, I'll be doing keyring? its hosted in bitbucket, so I'll start doing pull requests and will contact the maintainer when he wakes up [07:19] he is in the states [07:19] Cool, we can switch to your fork for release in the meantime if needed [07:20] do you have a XP box to test? I think it's only one of the windows backends that's broken, on XP it works [07:20] oops, on 7 it works [07:20] unless I have a patched keyring, of course :-) [07:21] ralsina, yes, the issue is that on XP it does not delete from the registry [07:21] doesn't sound hard to implement [07:21] ralsina, nope :) [07:21] cool [07:21] ralsina, I'll do the branches and will let you know, also I'll like to move txsecrets there [07:22] ralsina, maybe alecu and I could become maintainers of pykeyring and hav control over it? buahahahahaha [07:22] Also, we will eventually have to document how to delete credentials manually in all platforms, but that's for llaaaaaaater [07:22] mandel: looks like it would be useful. And I will take over pyxdg :-) [07:22] Since I already reimplemented it for windows. [07:22] ralsina, I think we could create a super tiny script for that [07:23] mandel: actually, implementing the linux equivalente of the "suggested UDFs list" is a lot of work. [07:23] ralsina, just a small .exe that gets ran, remove the creds by using standard qt dialogs, [07:23] yeah [07:23] ralsina, I'm talking about the removal of the creds :) [07:23] but in that case, it can be done via control panel [07:23] control panel -> remove local device [07:24] it should always do the right thing [07:24] ralsina, cool, I'll start working on it right now [07:24] great, I will go sleep anouther 5 hours :-) [07:24] ralsina, I might appear and disappear due to errands, new house you know? :) [07:24] you already in madrid? [07:25] ralsina, yes, I'm that good :) [07:25] he [07:25] ralsina, you might not even notices that I moved? but there are some things I have to get done [07:25] cool, how's iron taking it? [07:25] ralsina, bank, internet at home (I'm at my parents office) etc... [07:26] ralsina, he's like WTF but with no problems? I think the ex is taking it worse hehehe [07:26] ha [07:26] ok, have fun! [07:27] ralsina, duerme! === aquarius_ is now known as aquarius [08:27] Happy Monday, everyone! :) [08:29] morning JamesTait [08:54] fagan, hey! what are you doing here, I thugh you had a job ;) [09:04] mandel: ha I just opened irc [09:04] doesnt really distract from the work :) [09:18] sure... [09:24] mandel: so do you miss having someone in your timezone? [09:26] fagan, well, ralsina was here like an hour ago :) [09:26] and diego starts early, but it is indeed quite lonely [09:30] mandel: well at least you wont have any interns to inflict your loneliness on :D [09:37] mandel, hey! i'm here jeje [09:38] Hi everyone! [09:39] I'm here as well ;d [09:39] =D [09:41] karni, :P hi [09:42] gatox, buenas! [09:43] gatox: hello :) [09:45] karni, mornings, a quick android question - have you ever seen an application that only had BroadcastReceiver and nothing else? No service, nothing. I wonder how can I make it be called since my tests show that android does not call the onReceive() at all even though i put the correct intent filters [09:47] rye: hello! which broadcast is it? for instance, apps installed on external storage, do not receive some system broadcasts. [09:47] rye: I don't think I have, but I wouldn't say it's anything bad. a single BroadcastReceiver can be useful as well :) [09:48] karni, i am trying to capture the event that media scanner has finished scanning file. Then i tried to get at least anything, obtained READ_PHONE_STATE permission, started listening to at least anything and still nothing [09:48] rye: And you're sure you've declared the receiver properly? [09:48] it looks like I don't understand the concept, is Android really initiating the apps based on their intent-filter actions? [09:49] rye: Yes it is. [09:49] rye: Show me your manifest :) [09:51] rye: That's our network "listener" http://paste.ubuntu.com/687508/ [09:51] rye: well, it's not really interesting. note that you have to provide full action, including the package. [09:52] rye: and we start with android:name=".receiver..." because manifest tag declares package="com.ubuntuone.android.files" [09:53] karni, uh, huh. Well, I will boot the netbook and grab the code later today and will poke you once again. [09:53] rye: Sure :) === zyga is now known as zyga-afk [10:47] sorry, internet issues :) === zyga-afk is now known as zyga [11:13] good morning everyone! [11:14] nessita, buenos dias! [11:14] hola mandel [11:27] Good morning! [11:29] hi ralsina [11:30] mandel: you in madrid already or still barcelona? [11:31] hello nessita [11:42] nessita, madrid :) [11:43] nessita, seems like my twitter client on the phone does not work very well :( [11:43] mandel: why? [11:43] nessita, its posting things like hours late :P [11:44] nessita, I though you asked because I new pict was posted of my old house :) [11:44] ralsina, nessita I'm of to walk the dog, lets hope he likes the new place, bbl [11:44] mandel: ack [11:46] Hello Androids! If you have an older device / low-end device, I would mostly appreciate for givint this Ubuntu One Files 1.0.3.1 a spin http://goo.gl/RlQRk.qr [11:46] Most notably, I'm interested if auto-uploading multiple pictures causes any visible load on the system, thanks! [11:47] karni: I have a galaxy 3, that counts as low end? [11:48] ralsina: That's a 667 MHz, I would appreciate if you could give it a spin, please! [11:48] karni: sure! [11:49] ralsina: Please: open the app > Menu > Settings > Configure auto-upload > enable it, if it's off, and try taking few pictures (5-10 would be great) [11:49] ok, let me go find it [11:49] ralsina: Please let me know if the phone feels visibly slower [11:49] ralsina, when you have a minute (already has nessita approval): https://code.launchpad.net/~diegosarmentero/ubuntuone-control-panel/over-quota/+merge/74837 [11:49] ralsina: Much thanks! [11:50] hi gatox! [11:50] nessita, hi [11:50] * karni turns to support tracker to request feedback from these users [11:51] * karni left his HTC Hero in Warsaw [11:56] gatox: I just assigned bug #847232 to you, so you and ralsina can evaluate implementing it when doing the build [11:56] nessita, ok [12:01] gatox: got it! [12:01] ralsina_, thanks [12:06] nessita, gatox: doing what bug #847232 suggests is really not trivial. [12:06] I added a comment in the bug [12:06] ralsina_, yes, i assume its not [12:06] oh, we are botless again :-( [12:06] http://msdn.microsoft.com/en-us/magazine/cc300434.aspx [12:07] For example, you can't have more than 5 entries, and you can replace the list but not add to it! :-/ [12:07] ralsina_, mmmmmmm....... that sounds like a "perno" :P [12:08] gatox: NO KIDDING [12:08] ralsina_: I think the bug refers to adding UBuntu One to the favorites while using explorer, not using a file chooser... [12:08] sorry, hit capslock by accident :-) [12:09] Hmmm could be [12:09] ralsina_: this http://www.ghacks.net/2010/07/19/how-to-add-folders-to-windows-explorer-favorites-in-windows-7/ [12:09] but done programatically [12:10] i'm finishing with some bugs now... let me close this and i'll take a look at that one [12:11] nessita: yes. Not that I find that with a quick gogle :-) [12:11] ralsina_: that == doing it programatically? [12:11] yes [12:11] right, we may need to dive in the msdn site [12:12] Here they are: HKEY_CLASSES_ROOT\CLSID\{323CA680-C24D-4099-B94D-446DD2D7249E}\ShellFolder [12:12] gatox: Whoa, this looks like a complicated system. We can talk this later though. [12:13] nessita: I talked with mandel earlier and he's starting with the keyring problems first [12:13] ralsina_, ok! [12:13] ralsina_: sounds good.... which reminds me, I downloaded a third iso last friday, I will start installing that [12:14] nessita: ok, cool [12:15] bu. the image is not "bootable" [12:16] ralsina_: managed to find the galaxy 3 perhaps? :) [12:16] karni: got sidetracked, will check in 5' [12:17] ralsina_: sure! [12:18] ralsina_: is it "sane" to hace cleanupUp page calling initializePage? I mean, can we trigger an endless loop somhow? [12:18] nessita: only if going back takes you to the same page [12:19] ralsina_: are we 110% that does not happen? :-) [12:19] nessita: well, we are in these two cases :-) [12:21] nessita: in fact, not even then would it cause a loop. It would have to involve page 1 moving to page2 from initialize, and page2 going back from initialize. [12:21] nessita: so, a loop is really really really unlikely because it would mean page2 "bounces" you all the time, and things would be broken anyway [12:21] ack [12:24] ralsina_: branch looks good! IRL testing now [12:25] nessita: cool! [12:39] karni: trying to install that file locked up my phone [12:39] ralsina_: o_O [12:39] ralsina_: the install, or post-login? [12:40] karni: the install itself, I have the red arrow thing on notification and can't even open that :-) [12:40] karni: so it's likely not your fault [12:41] ralsina_: Oh, that's strange.. Anyway, thanks for trying! blajk justtested on HTC Legend, and it looks fine [12:41] ok! [12:47] mandel, nessita, alecu, dobey, gatox: standup in 10 or so! [12:47] ralsina_, ack [12:48] ack [12:49] gatox, ralsina_: I just found that we have some regressions in the setup account page regarding button enablement, see bug #847796 (screencast added) [12:49] nessita: looking! [12:49] gatox: I marked that as High, please note is Higher that other Highs :-) (not not that much to be critical) [12:50] nessita, jeje ok... i'm finishing with 2 bugs, and then i will take that one [12:50] ralsina_: I found that while testing your branch, but I confirmed the bug is in trunk as well [12:50] ack [12:51] it's probably there since the button got moved to the wizard [12:52] ralsina_: ah, perhaps. The tricky part is that the "password too weak" error appears but the loading will not go away [12:52] gatox: ^ [12:53] ack [12:56] ack [13:00] me [13:00] me [13:00] me [13:00] me [13:01] dobey, alecu, standup [13:01] alecu, dobey? [13:01] let's! :-) [13:01] DONE: Tons of testing with clean ISOs, RELEASE WOOHOO [13:01] TODO: more testing, bug triage, grab some controlpanel UI bug (probably bug #800705). [13:01] BLOCKED: nopes [13:01] NEXT: gatox [13:01] DONE: [13:01] Bug #829358, Bug #828983, Bugs #845083 and #845735 almost done. [13:01] TODO: [13:01] Keep fixing UI bugs. Try to survive the flu. [13:01] BLOCKED: [13:01] No [13:01] mandel go [13:02] hello! [13:02] alecu: hello! happy release! [13:02] alecu, hi! [13:02] mandel: go? [13:02] hi nessita! :-) [13:02] DONE: Looks at python keyring and modify part of the tests running system. I'm talking with benji (mantainer and canonical colleage) about the changes and how to continue so that I can push the changes there. [13:02] TODO: more pykering and bitbucket [13:02] BLOCKED: no [13:02] COMMENTS: I might be out for the last 2 ours of my day for errands :P [13:02] next ralsina_ [13:02] DONE: release! fixed my pending branches, reported bugs, did reviews TODO: merge a couple of branches, assign tasks, fix bugs (not chosen yet), tech leads call BLOCKED: no [13:02] alecu? [13:03] * alecu is writing notes [13:03] ralsina_: shall I attend that teach leads call? [13:03] nessita: it's called tech leads, but it's more of a "tech management call before the management call" I think [13:03] ralsina_: ah :-) [13:03] so we clean up what we present without getting bogged in technical details [13:04] Then again, I am not *sure* you shouldn't be in it :-) [13:04] makes sense [13:04] ralsina_, do we know anything good/bad about the beta testers? [13:05] we have 588 of them [13:05] mandel: not many bug report were opened (yet) [13:05] DONE: took a conference day off, but due to sick child was able to attend CISL only for a few hours [13:05] TODO: find and fill pending tickets from expenses, mis bugfixing [13:05] BLOCKED: no [13:05] they reported very few bugs, but there is this on askubuntu: http://askubuntu.com/q/60754/711 [13:05] * mandel is a karma whore => goes to ask.ubuntu :P [13:05] alecu: how's Amelia? [13:06] ralsina_, she's great today. I'm supposed to take her to kinder in a few minutes [13:06] alecu: you have kinders today? In province they are closed :-/ [13:06] * "misc bugfixing" [13:06] they sent use the note on friday [13:07] ralsina_, we realized that, so we called yesterday one of the other mothers to see if today the kinder was open [13:07] ralsina_, (because Amelia has skipped thu and fri). [13:08] right [13:08] ralsina_, alecu, nessita, gatox are we having mumble? I will probably have to miss it :( [13:08] ok everyone, go fix a bug or two ;-) [13:08] I think we can skip mumble for today [13:08] ralsina_: will you be handling the askubuntu question? [13:08] or shall I? [13:08] nessita: mandel was volunteering ;-) [13:08] great! [13:09] mandel: please do not drop that package ;-) (we want all the feedback we can have from our users) [13:11] nessita, sire :) [13:11] I mean, sure hehehe [13:12] nessita, I'm answering to send more info since it seems that changes in the fs where not notified or the file was not uploaded. [13:13] mandel: request logs, always [13:13] I nearly failed the captcha test hehehe [13:13] nessita, FYI: http://askubuntu.com/questions/60754/latest-windows-ubuntuone-client-not-working-correctly/61075#61075 [13:13] mandel: and the name of the file that did not upload [13:13] nessita, oh, goo point hehehe [13:14] mandel: FYI: this is not the correct path for logs C:\Users\$username$\AppData\Local\xdg\ubuntuone [13:14] mandel: the proper path is: [13:14] C:\Users\$username$\AppData\Local\xdg\cache [13:14] nessita, ups, sorry [13:15] can you please clarify that? (ask to zip all the folder and attach that) [13:15] nessita, sorted [13:16] using quicksort, I hope [13:16] :-D [13:16] bubble, always bubble hehehe [13:17] meh [13:17] nessita, alecu, ralsina_ I have good news related to pykeyring, it does not longer use C but c types which means that it is longer not a pain to build on windows :) [13:17] mandel: that is good [13:18] mandel: great news! [13:18] mandel: what about the delete method? [13:18] λ DONE: bug #788532 [13:18] λ TODO: bug #840072, bug #838778, get stuff off CD [13:18] λ BLCK: None. [13:19] nessita, working on it atm, only problem is that for it to land I have to implement it on all platforms, is not hard, but a pain to test :( [13:19] mandel: what all platforms? [13:19] nessita, atm I'm workingon improving how tests are done so that is easier to work with it, benji is the poin of contact for this [13:20] nessita, it would be linux (gnome/kwallet) windows and mac os x [13:20] mandel: this is working in linux ATM, no? [13:20] I already have the fix for windows XP so we could just create a patch with that for ralsina_ to use [13:20] nessita, delete? or pykeyring? [13:20] mandel: delete [13:20] mandel: if you have a bitbucket fork, that's good enough for our releases, yes [13:21] nessita, delete was indeed implemented, but you do not want to be using pykeyring just yet because it does not use txsecrets [13:22] nessita, I wanted to talk with benji about implementing a txsecrets backend for pykeyring so that we do not longer have it in sso [13:22] but that is in the very long run [13:22] mandel: exactly, so let's try to stick to what we need and can do in the short term [13:23] mandel: we still have a tight schedule, so if you can do the delete for xp fix, that's enough :-) [13:23] mandel, that means "wait till this friday for other platforms" :-) [13:23] we don't need txsecrets support for another month I think [13:23] exactly [13:23] ralsina_, I'm not sure it's a good idea to make pykeyring use txsecrets [13:24] sure, xp only is not a problem [13:24] ralsina_, afaik pykeyring is synchronous [13:24] mandel: so, let me see if I get this correct: we do have the patch for win7, but that does not work for XP? [13:24] alecu: yes, it is sync [13:25] nessita, yes, the issue is that on windows the creds_management is not present on XP which means that the remove_credentials does not work, the patch ensure that when the sso creds is added to the regstry on xp we can indeed delete it [13:25] alecu, ralsina_ that is why I have to talk with benji about tx_secrets, we could add an async api besides the sync one [13:25] nessita, does it make sense? [13:26] mandel, well, that sounds a bit more reasonable. [13:26] mandel: no... sorry. Let's start again: I know you built a patch a while ago to support credentials deletion in windows. In which windows version did that patch work? [13:26] guys and gal: I have to go to kinder for 30 mins aprox [13:26] nessita, windows vista and 7 [13:26] * alecu bbl [13:26] ciao! [13:26] nessita, xp stores the data in the registry and I did not patch that [13:26] mandel: that was the Vault backend, right? [13:27] ralsina_, exactly [13:27] mandel: ah, I see. Ok, keep us posted. [13:27] nessita, will do, I'm re doing the vista and patch too so that they work in the latests pykeyring version [13:28] ralsina_, we should jot this down for the pycon talk, it is an interesting thing to talk about [13:28] mandel: yeah [13:29] mandel: all we need for that talk is a list of things and rant. We can organize it over two, three beers [13:29] dobey: we are botless again, can you re-do what you did? [13:29] ralsina_, yes, and I have a good idea of the slides? :) [13:29] ralsina_: I'm approving back-is-wrong but I found some buggy behaviors when moving between those folders, so I reported bug #847847 (added screencast so you can see the issues) [13:30] nessita: ack! [13:30] ralsina_: those folders == those folder pages [13:31] ralsina_, nessita, gatox, dobey, alecu I need to go away to do some serious errands (like moving the account from barcelona to madrid) etc.. I'll ba back asap, if its to late I'll send you an email with the progress in pykeyring but I believe it should be ready for tom with no problems, if benji likes me even in trunk :P [13:31] cool [13:31] mandel: add good tests! :-) [13:32] nessita: ok, will work on that bug today [13:32] nessita, yeah, that is one of the big changes to be added to pykeyring, tests are kinda ugly (it tests all backends in all platforms?) [13:33] mandel: ack, though if you're asking me something I'm not what that question is :-) [13:34] nessita, me perdi :P [13:34] ok, I need to go before bank closes? lazy little bastards! [13:34] mandel: bye [13:39] nessita: i don't think mandel is very good at using proper punctuation. :) [13:39] dobey: looks like it :-) [13:40] dobey, nessita hey, I'm still here!!! what pronuntiation? [13:40] PUNCTuation [13:40] ok, what punctuation :P [13:41] mandel: "tests are kinda ugly (it tests all backends in all platforms?)" [13:41] I did not add a ? [13:41] nessita, tests are kinda ugly (it tests all backends in all platforms?) [13:42] do you see a ? [13:42] yeah [13:42] * mandel se va a cagar en x-chat [13:42] dobey: do you? [13:42] yes [13:42] * mandel se caga en la puta de oros!! [13:42] also "before bank closes?" [13:42] nessita, dobey and here .. [13:42] ?? [13:42] mandel: no [13:43] mandel: don't you need to go to the bank? :P [13:43] he, so it looks like three dots are converted to a ? [13:43] 0x2026 is … [13:43] dobey, yes, but I prefer to miss the back than to be told I dont know how to use ? ;) [13:43] mandel: do your errands instead of argue! :-P [13:44] mandel: that's what she said! [13:44] nessita, dobey so, that was three dots that got transformed to a ? so, me cago en irc y sus muertos! [13:44] ok, now I'm away [13:48] nessita, ralsina_ when you have a minute: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/err-dict/+merge/75001 [13:49] xchat is probably trying to convert to an ellipsis ... and failing ;-) [13:51] gatox: looking! [13:53] gatox: where did you take the email idea from? [13:54] hello! [13:55] nessita, i see that in some cases for RegistrationError the dict only contain errtype and email, and i saw the tests, and applied the same way of checking that key as it does for message and __all__ [13:55] gatox: the goal of the bug is to generically handle any err dict [13:55] gatox: in a way that we never show the string repr of a dict to the user [13:56] gatox: so we need to improve the fix so any dict is properly handled [13:56] nessita, so, the only missing is the "password" key...... should i check for that one too...... and avoid the result that just print the dict? [13:57] gatox: no, we need to support any key [13:57] gatox: let me grab the GTK handling algortithm [13:57] nessita, its the same [13:58] gatox: what do you mean is the same? [14:00] nessita, i don't know if you are referring to something else...... but the "_build_general_error_message" method in gtk, does the same as the one in qt (except for the new thing about 'email' key)..... [14:01] nessita: I think we had seen that the root bug was that not the same errordict was being passed on linux and windows [14:01] nessita: the last time I touched that function because of a similar error? [14:02] gatox: right, because that method is not used for specific entries. If you go an see on_user_registration_error, we deal with error separately [14:02] so, we first check if 'email' is there and add an warning to the entry itseld [14:02] itself* [14:02] gatox: same for password [14:02] and then, we build a general message [14:03] gatox: so, besides doing specific handling for fields, we should do a generic handling [14:04] gatox: where the dict received as error, is *never* shown as a dict to the end user. In the worst case, we should iterate over the keys, values and show that in a user-friendly way [14:04] gatox: this is, for example, the GTK controlpanel generic error handling http://pastebin.ubuntu.com/687633/ [14:05] ralsina_: is related, but not the same. My goal is that we never show a dict to the user, despite we may have different error keys in linux and windows [14:05] ack [14:05] nessita, when you say "check if 'email' is there and add an warning to the entry", do you mean add that message to the email_assistant label besides the email line edit? [14:05] gatox: that would be ideal, but not sure what design spec'ed for that [14:06] duanedesign: joshuahoover: Can you guys see version 1.0.3.1 on the right side of https://market.android.com/details?id=com.ubuntuone.android.files ? [14:06] gatox: in any case, in the generic callback handler, we never should query for 'email' in particular, and we never should default to "result = "Error: %r" % errordict" [14:06] gatox: so, when registering, we may *also* handle possible specific errors such as email and password [14:07] gatox: email and password errors will only show up as registration results [14:07] gatox: the same happens for email-token validation, the errdict will add a new pay key called email_token [14:08] (see sso, gtk code, gui.py:1060) [14:08] karni: i don't see it there [14:08] joshuahoover: thanks. probably not populated yet. [14:08] gatox: not sure if I can explain myself about this: summary is that we should deal separately with form-specific errors vs general errors [14:10] nessita, so...... let me see if i understand...... the method that build the messages, if it doesn't has the __all__ and message key, it should return None.... and avoid showing that..... and for the methods that catch the errors for Registration and so, we should parse the additional data as password and email and show it somewhere else [14:10] gatox: I would recommend the other way around [14:10] nessita, sorry..... i'm kind of lost.... [14:10] gatox: each error handler may do some specific handling, and then, if there is something left to handle in the error dict, build a general message and show that [14:10] gatox: we can mumble if you want [14:11] nessita, ahhhh ok, i think that i understand now [14:11] nessita, first check local errors, then the rest [14:12] gatox: right. So, for example: when handling registration error, you can have specific_error_1 and specific_error_2. You pop out those from the errdict and then, build a general error that may be the empty string [14:12] gatox: same for email token result, reset password result, login result, etc [14:12] nessita, okkkkkkkk....... now i think that is clear [14:13] gatox: is more robust, and will generate a cleaner error message handling and visualization [14:13] gatox: for a guide, you can follow sso/gtk/gui.py from lines 993 to 1114 (the end) [14:13] gatox: ack? [14:14] nessita, ok..... so, my only doubt now.... is where to show the specificic errors.... can i put that in the assistant labels besides the line edits? or should i show all the error messages together one below the other at the bottom of the form? [14:15] gatox: do we have any spec for that? [14:15] nessita, nop [14:15] nessita, but [14:15] gatox: I think there is one... trying to find it [14:15] nessita, i think that we can show that error message besides the line edit, because it's kind of the same message that you receive when you are completing the form [14:16] gatox: yes, that's the idea [14:16] nessita, ok [14:16] nessita: indeed you have now gained a weekly call! [14:17] ralsina_: :-( [14:17] ralsina_: I hate myself for asking [14:17] :-P [14:26] nessita: for you, it's opt-in [14:26] nessita: it's mandatory for managers [14:27] Chipaca: I was kidding ;-). What day, what time? :-) [14:28] nessita: adding you to the invite [14:28] Chipaca: thanks [14:28] nessita: done [14:28] got it! [14:51] karni: i see it know [14:54] surprisingly, it seems no windows tester has find actual unknown bugs except maybe the askubuntu guy [14:57] duanedesign: thanks :) [15:05] lunch time, bbiab [15:06] ralsina_: so, I've been doing some follow up in this bug report https://bugs.launchpad.net/ubuntuone-windows-installer/+bug/846373 [15:07] ralsina_: and the wizard is crashing and is not handling the error [15:07]