[00:00] ahayzen, so, I take advantage of your kindness: when I have to use a lot of and when I have to use a lots of? [00:01] 'a lot of' and 'a lots of' ? [00:01] ahayzen, yes, what's the difference? [00:01] i usually say... 'a lot of' or 'lots of' [00:02] eg 'a lot of something did something' or 'lots of something did something' [00:02] i don't think 'a lots of' makes sense [00:03] ah, ok, interesting. 'a lot of' and 'lots of' are interchangeable? [00:04] basically :) [00:05] ok, thanks :-) [00:05] Time to sleep, see you soon! [00:05] o/ rpadovani have fun :) [00:26] balloons, does jenkins use adt? or will it pick up my new depends from the deb control? or will we have to manually add them? [00:26] ahayzen, for now, jenkins uses deb [00:27] balloons, ok cool at least it 'should' work then :) [00:27] ...once it works locally :P [00:27] 'PASS' \o/ [00:28] balloons, would you put the url-dispatcher calls in a helper? i've just been doing the subprocess.check_call() directly in the test? [00:29] balloons, eg this is what one of the tests currently looks like http://pastebin.ubuntu.com/8294829/ [00:29] ahayzen, probaly a helper yea === _salem is now known as salem_ === salem_ is now known as _salem === daker_ is now known as daker === tsimpson_ is now known as tsimpson [06:23] good morning [08:10] morning all === ogra_` is now known as ogra_ [08:16] Good morning all; happy Wonderful Weirdos Day! :-D [08:16] morning popey, around? [08:17] ya [08:23] dpm, for the app dev training materials - would you suggest I put them under dev.u.c/community/ somewhere? [08:24] dholbach, I'd put them under http://developer.ubuntu.com/resources/ for now [08:24] dpm, ah yes, good thinking [08:25] popey, could you add the comment with the feedback mardy gave you on https://code.launchpad.net/~dpm/reminders-app/account-click/+merge/230768 ? I'd like this MP to be merged this week [08:26] ah yes. sorry I forgot... [08:26] will do now [08:26] np, thanks [08:27] done [08:28] thanks [08:50] Good morning :-) [08:51] zbenjamin: hey, could you quickly review https://code.launchpad.net/~nik90/ubuntu-calendar-app/remove-qmlproject-hacks/+merge/233831 [08:51] zbenjamin: I removed the qmlproject file and the desktop file hack. should take less than 15 secs [08:51] dpm: hey, could you review https://code.launchpad.net/~nik90/ubuntu-clock-app/make-alarmlabel-translatable/+merge/233828. related to pot file update [08:52] mihir: ping [08:54] nik90_: https://code.launchpad.net/~zeller-benjamin/ubuntu-calculator-app/sdkcompat/+merge/233693 [08:54] nik90_: i can not review Mps for that project [08:54] zbenjamin: as in you cannot top approve? [08:55] i can if zbenjamin says it's okay [08:55] well, if nik90_ does ㋛ [08:56] popey: let me test the calculator MP by zbenjamin now [08:56] thanks! [08:57] nik90_: popey: the MR from nik90_ looks good [08:58] nik90_: approved [08:58] zbenjamin: thnx [08:58] popey: can you top approve https://code.launchpad.net/~nik90/ubuntu-calendar-app/remove-qmlproject-hacks/+merge/233831 [09:01] done [09:02] popey: thnx [09:02] zbenjamin: btw if the name in the manifest is @PROJECT_NAME@, can I append .devel to it before runnning on device or emulator? [09:02] nik90_: yeps [09:03] zbenjamin: silly me kept removing it and replacing it with com.ubuntu.calculator.devel :P [09:03] nik90_: its just variable substitution [09:03] true [09:05] hey all [09:07] popey: top approve https://code.launchpad.net/~zeller-benjamin/ubuntu-calculator-app/sdkcompat/+merge/233693 as well pls. I just approved it [09:07] Akiva-Thinkpad: hey [09:07] o/ [09:07] done [09:08] popey: btw headsup, we cannot release a new version of clock before the next image promotion since ideally it requires the snooze backend that landed in the indicator-datetime package [09:09] nik90_: ok [09:17] has anyone tried to have the ubuntu emulator running in a virtualbox? [09:24] dholbach: i am trying now [09:24] assuming this is for that mail thread [09:24] popey, yes [09:24] yeah i said I'd look at it [09:24] popey, because I tried it on a 14.10 virtualbox and it didn't work - I was just trying to see if anyone else had tried it [09:25] i am testing on 14.04 [09:25] thanks [09:25] nik90_: type in wolverhampton in the clock app but stop at wolver, in other apps on the phone that would show a bunch of suggestions but in clock it say no such place and repeats that until you finish typing in the whole name [09:26] davmor2, the clock app only shows real places [09:26] davmor2, what's the 'wolverhampton' you're inventing now? [09:26] davmor2: bug 1366283 [09:26] bug 1366283 in Ubuntu Geonames "Possibility for search locations by name starting with given characters" [Undecided,New] https://launchpad.net/bugs/1366283 [09:27] davmor2: sry but that doesn't qualify for a clock bug ^^ [09:27] :P [09:27] nik90_: meh I'll just hit it harder :P I thought I was doing really well finding one that quick [09:28] popey: ta [09:28] davmor2: sure ;) [09:38] nik90_, done [09:38] cool thx [09:40] dpm: I just read your comment. I was asked to remove the space in "%1 h" since in the UI it should appear as "3h 4m" [09:41] nik90_, by design? [09:41] dpm: yes === nik90_ is now known as nik90 [09:45] nik90, in any case, they will need to be fixed so that they work with translations [09:45] nik90, was it giorgio who suggested that? [09:45] dpm: yes [09:46] dpm: he is online atm on irc, we can discuss this [09:48] dpm: btw in catalyn is the time supposed to be shown as h.mm ? [09:49] clock does this but the indicator-datetime doesn't [09:50] nik90: pong [09:50] nik90, yes, it should be . [09:50] the indicator used to use . at some point [09:50] but at some other point it changed [09:50] actually, it should be shown as H.mm [09:50] mihir: hey, zbenjamin made some changes to the calendar cmake files which now lets you run it on your device directly from qtc :D [09:51] mihir: you will have to open the cmakelist.txt file instead of the .qmlproject file from now on [09:51] nik90: yeah I saw that in emails, but didnt get time to try on my machine, will do that tonight :D [09:51] nik90: awesome :D [09:51] mihir: btw what are we waiting on your new-event branch? [09:52] mihir: I want to get started on another big calendar branch, but first need your branch to go in first [09:52] nik90: nothing I am done my side, popey was suppose to take look [09:52] mihir: oh ok [09:52] nik90: what is that about ? [09:53] mihir: massive reorganization of the files in the source tree. atm all the files are in the root of your calendar app. I am going to move them into separate folders [09:53] mihir: similar to the clock app structure [09:53] nik90: yeah , i thought of same.. [09:53] mihir: I will need your help. So same drill as before [09:53] mihir: although this time no AP changes required [09:53] most likely [09:54] nik90: also, wanted to move some of the functions to saperate files [09:54] nik90: yup it shouldn't required. [09:54] nik90: what i will do , i will try to start strcutre tonight , or tomorrow ,and will push that [09:54] mihir: I am already on it. [09:55] mihir: I will push to a common branch [09:55] nik90: okay then no issues :) just push that we can start together [10:02] dpm, http://developer.ubuntu.com/resources/app-dev-training/ - good to go? :-) [10:03] ogra_: does one need developer-mode just for adb shell? [10:03] nik90, yes [10:03] ogra_: will I be able to use phablet-shell withouth developer mode? [10:05] popey: ping [10:06] nik90, nope, it uses adb [10:06] mihir: pong [10:06] popey: if you get some time to review new -event design we can get it merge :) [10:06] kk [10:06] mihir: can you quickly do a trunk merge [10:06] nik90: i can't in office right now :| [10:07] mihir: alrite let me do it [10:08] dholbach, looks great, not sure if the picture is going to put people off ;) [10:08] haha, I don't think so [10:19] popey, dholbach, could one of you follow up with jdstrand on bug 1357211? [10:19] bug 1357211 in Canonical Click Reviewers tools "Validation errors when adding an Online Account to a click package" [Undecided,New] https://launchpad.net/bugs/1357211 [10:20] dpm, sure [10:21] awesome, thanks [10:21] Hi [10:25] dpm, do you think you could upload a click package for the bug you mentioned? [10:25] dpm, is this generally working or is c-r-t just complaining? [10:25] dholbach, there is one in the merge proposal linked to, but I'll add a direct link [10:25] IIRC it's just the scripts complaining [10:25] ah ok, no worries [10:27] done [10:29] thanks [10:38] !ping [10:38] pong! [10:58] nik90: thanks :) [10:59] np [10:59] nik90: if you can have final approve, lets merge that to the trunk :D [11:02] * mihir hugs nik90 for helping to revamp new event and merge in trunk :D [11:03] mihir: yw our work is not yet done :D [11:03] nik90: i know :) === MacSlow is now known as MacSlow|lunch [11:23] mzanetti: ping [11:24] nerochiaro: hi [11:25] mzanetti: is the tagger app a project you manage ? [11:26] nerochiaro: yes [11:26] mzanetti: i can't seem to be able to build it in qtcreator. it fails to run qmake [11:28] nerochiaro: works fine for me as long as only building for desktop [11:28] nerochiaro: our SDK for phone doesn't support qmake [11:28] mzanetti: oh, and i think it's only supporting in-tree builds and qtcreator by default does shadow builds [11:28] mzanetti: so how do i build it for the phone ? [11:28] no... definitely supporting shadow builds [11:28] nerochiaro: there's a run_on_ubuntu_touch.sh script [11:29] zbenjamin: is there a reason installing ubuntu-sdk doesn't pull in ubuntu-emulator on 14.04? (this means the wizard doesn't work because it offers installing emulator device, which can't possibly work without the emulator package) [11:29] mzanetti: ok, i'll try that [11:29] nerochiaro: what do you want to do with it? [11:29] mzanetti: i'm fixing a bug [11:29] mzanetti: it should let go of the camera when it goes to background [11:30] nerochiaro: uh oh... does the app need to take care of that? [11:30] mzanetti: at the moment yes [11:30] IMO that should be handled by the QCamera backend [11:30] mzanetti: i agree but it's not the way it works now, and apparently tagger will be preinstalled on OEM images, so we need it to play nice with camera [11:31] popey: that sounds like a bug [11:32] nerochiaro: can't we fix it in the backend now? shouldn't be too hard I guess [11:33] there are a bunch of other apps too that use the camera [11:34] mzanetti: the problem i think is that once you get back to focus you need to reconfigure the sensor to match what your app needs. either the backend keeps state for apps and restores it when the app comes back to being active, or I don't see any other way to do it automatically [11:34] mzanetti: like video versus photo, flash mode, etc [11:35] mzanetti: that requires some significant work if you want the backend to do it [11:35] mzanetti: but i guess we can have the backend disable the camera on unfocus, at least [11:36] mzanetti: assuming there's a way to get to the application object from the QCamera library [11:36] hmm... [11:36] not sure if you need the application object [11:36] or rather, from our QCamera backend [11:36] mzanetti: how do you know you've been unfocused by the WM otherwise [11:36] ? [11:37] ok, true... but qApp() should give you that [11:37] zbenjamin: ok. where do i file it? [11:38] nerochiaro: well... just saying... if its too much just work it around. but it'll leave a bunch of other apps broken and it still doesn't sound too complex at a first glance to save those 4 properties [11:39] but I haven't ever digged into the camera backend stuff [11:39] so I might be wrong [11:40] mzanetti: there's a bunch of properties that camera exposes, and more importantly you need to save them per-app, and then know when an app closes so you can clean up its state to avoid leaks. it's not easy I think [11:40] mzanetti: the easy part is stopping the camera [11:40] mzanetti: i guess that one we can do [11:40] nerochiaro: the backend should be in process anyways, no? [11:41] popey: file it against qtcreator-plugin-ubuntu [11:41] so if you have 3 apps fighting for the camera, the backend code would run 3 times and you'd just need to save the current state on unfocus and restore it on focus [11:41] assuming the ApplicationManager plays nice and calls the unfocusing on one app before the focusing on the other [11:42] mzanetti: need to figure out who's responsible for the backend at the moment and have a chat, but yeah, you have good points [11:43] zbenjamin: bug 1367246 - this affects a partner who wants to port their app to ubuntu which is how i discovered it [11:43] bug 1367246 in qtcreator-plugin-ubuntu "ubuntu-sdk doesn't pull in emulator on 14.04" [Undecided,New] https://launchpad.net/bugs/1367246 [11:44] popey: thanks [11:47] I'm trying to find how to use the new header features instead of the deprecated toolbar, but can't find any documentation about it, how can I find it? [11:47] The documentation, I mean [11:49] popey, could you test http://people.canonical.com/~dpm/click/com.ubuntu.reminders_0.5.221_armhf.click ? I included the fix mardy suggested - you'll need to uninstall account-plugin-evernote too [11:50] jdstrand, do you think you can let me know once you have a bit of time to chat about 1357211? [11:50] dpm: sure [11:50] jdstrand, I'm happy to work on a fix, if you can guide me a little bit, or at least confirm that the .click provided by dpm is the way it should be :) [11:52] popey: who works on weather app? [11:52] davmor2: m-b-o [11:53] popey davmor2 o/ [11:54] thanks popey [11:54] m-b-o: I have a small layout issue with the new version, you have edit and settings in the same menu now and it isn't particularly clear which is which, would + Add Cities maybe be better for edit? [11:55] davmor2 agreed [11:55] s/Cities/Location/ [11:56] davmor2 I guess problem is the width of the overlay, which is a problem when the label gets too long [11:56] davmor2 popey: like you add a location with a long name [11:56] yeah [11:56] m-b-o: how about just Locations then? === MacSlow|lunch is now known as MacSlow [11:57] davmor2 yeah, agreed :) [11:57] popey: ^ [11:58] Could you all help a bit in promoting http://developer.ubuntu.com/resources/app-dev-training/? Thanks a lot in advance. [11:58] I will as far as I can === seb128_ is now known as seb128 [12:17] dholbach: I need someone to guide me on what it is supposed to look like :) [12:17] jdstrand, hum hum... mardy maybe? :) [12:17] mardy, do you think you can help us with bug 1357211? [12:17] dholbach: I guess my questions is why is "evernote-account-plugin" not in "reminders"? [12:17] bug 1357211 in Canonical Click Reviewers tools "Validation errors when adding an Online Account to a click package" [Undecided,New] https://launchpad.net/bugs/1357211 [12:17] jdstrand, I think they were two separate binary packages earlier on [12:18] dholbach: the tools currently assume that will be the case [12:18] dpm: http://popey.mooo.com/screenshots/device-2014-09-09-131535.png [12:18] dpm: i dont get a login screen [12:19] bummer [12:19] right, so mardy can tell us if !account-application should be in its own entry in the hooks db. if so, need to change things up [12:19] how long until the on-air thing happens? [12:20] Akiva-Thinkpad: 2 hours 40 [12:20] popey, could you have a chat with mardy to debug this? [12:20] jdstrand: I'm not sure I understand the problem; AFAICT, those hooks could certainly go inside the "reminders" block [12:20] popey davmor2 https://code.launchpad.net/~martin-borho/ubuntu-weather-app/edit-label/+merge/233912 [12:20] mardy: dpm has a click http://people.canonical.com/~dpm/click/com.ubuntu.reminders_0.5.221_armhf.click - and it doesn't quite work [12:21] mardy: the problem is the tools fail if they are not in the reminders block [12:21] mardy: and I'm wondering if that is correct behavior? (we can maybe add better debugging) [12:21] mardy: would appreciate help debugging [12:21] popey: we are talking about that now [12:21] popey, dpm: then let's move them inside "reminders", I don't see a big issue with that [12:21] oh ☻ [12:22] jdstrand: the only difference will be noticed when we support running account plugins under confinement [12:22] jdstrand: if the hooks are under "reminders", then the account plugin will run confined as the "reminders" app [12:23] mardy: what kind of confinement are we talking about? [12:23] jdstrand: while it might be better to use its own confinment rules [12:23] jdstrand: aa-exec [12:23] mardy: like, we define a new policy group or something? [12:24] or template [12:24] jdstrand: I'm not sure that that will be necessary, your call :-) [12:24] jdstrand: what we will need is just to run the account plugin code in a confined profile [12:25] ok, that suggests a different entry that will need its own apparmor entry [12:25] mardy: so, that only affects account-qml-plugin? [12:26] jdstrand: and account-provider [12:26] mardy: let me rephrase [12:26] jdstrand: I guess the question is whether we want to allow click packages which ship just an account plugin (and no app), right? [12:26] m-b-o: could you please take a look at https://code.launchpad.net/~zeller-benjamin/ubuntu-weather-app/sdkcompat/+merge/233752 [12:27] mardy: that isn't the question in my mind :) [12:27] jdstrand: ok :-) [12:27] mardy: that is a different question [12:27] mardy: the question is what are the different account-* types that need their own policy? [12:28] popey yes [12:28] dpm popey: weather app has a problem with localizations, abbreviations for max/min in particular. [12:28] dpm: https://bugs.launchpad.net/ubuntu-weather-app/+bug/1366714 don't knwo how to handle this correctly [12:28] Ubuntu bug 1366714 in Ubuntu Weather App "max and min temperatures are overlayed (text corruption)" [High,Confirmed] [12:29] jdstrand: I don't understand the question, but if I answer "an account plugin is defined by the data from the account-qml-plugin and account-provider hooks", does that help? :-) [12:30] jdstrand: that is, if it were a deb package, I would expect it to contain those files [12:31] m-b-o, a stopgap measure (but not a solution) would be to add a translator comment to the strings, telling translators to use 1-character translations only [12:31] mardy: yes. let me paraphrase: anything specifying account-application should have an accompanying "desktop" and "apparmor" entry. "account-qml-plugin" and "account-provider" need to specified together, and need a separate "apparmor" entry [12:31] dpm: in the pot file, right? and request later a review later on in the tranlsation tool? [12:31] jdstrand: correct [12:32] s/a separate/an accompanying/ [12:32] m-b-o, just stack them vertically if they don't fit side to side? [12:32] mardy: ok, what about account-service-- should it be with one, the other, or on its own? [12:32] jdstrand: account-service goes with the app [12:33] so, account-application? [12:33] m-b-o, no, on the code. You can add a comment like this to the line above where the text appears. E.g. "// TRANSLATORS: use 1 character for the translation at maximum". Then when you generate the .pot file, these comments will be added there for you [12:33] seb128 that would require to adapt the sizes of the other shapes [12:33] yes, we use those comments in other apps. [12:33] m-b-o, other solution is to not have those labels, just min/max with colors? [12:33] jdstrand: yes, account-service and account-application should be defined by the app [12:34] mardy: ok. is either account-application or account-service optional are are both required? [12:34] jdstrand: it's also possible for an app to not specify an account-service entry, if the service is already in the root fs [12:34] mardy, so you're saying that if I do this the app should then work (as opposed to not showing any accounts for login)? -> http://pastebin.ubuntu.com/8299521/ [12:35] mardy: so account-application could be alone. can account-service be alone? [12:35] jdstrand: I tend to think that if an application uses the "accounts" app armor policy, it should have at least the "account-application" hook -- but let me think more :-) [12:36] jdstrand: you mean a click app having an account-service but no account-application? That sounds weird, I cannot think of any case where that would be useful [12:36] seb128 yes, or small icons.... but a redesign is on the way, and it can be solved by fixing the translations. So keeping the current solution would be ok for me... [12:36] mardy: ok [12:37] m-b-o, k, so just add the translator comment others suggested? [12:37] seb128 yes, will make a MP [12:37] mardy: can account-application be specified with a "scope" as opposed to only with a "dekstop" entry? [12:37] dpm: yes, that should be ok :-) [12:38] dpm: wait, hold on :) [12:38] I'm going to make a recommendation in a moment [12:38] after I'm done grilling mardy [12:38] jdstrand: you mean *by* a scope? [12:38] :-) [12:38] mardy: I mean: [12:38] "foo": { [12:39] jdstrand, I'm not too worried about the click reviewers tools validation atm, but rather on getting the app to work at all first [12:39] "scope": ..., [12:39] "account-applicaton": ... [12:39] "apparmor" [12:39] } [12:39] popey, could you give this new click a go? -> http://people.canonical.com/~dpm/click/com.ubuntu.reminders_0.5.222_armhf.click [12:40] sure [12:40] thx [12:40] dpm: sure, but the tools enforce what is expected. other things may need to change and at a mimimum, you don't want to upload a new manifest that has to be changed immediately [12:41] mardy: see my "foo" example, above. is that legal? (well, it should be "apparmor": ..., but you get the idea) [12:41] jdstrand, sure, but I'm not uploading yet, I'm trying to get the changes in the MP to run the app at all. Right now nothing is working, let alone the validation [12:41] jdstrand: so, at the moment, scopes and apps are treated by OA in exactly the same way: they need an account-application entry and a desktop file; however, I've been told that having a desktop file might cause problems for them === _salem is now known as salem_ [12:41] jdstrand: however, I haven't heard any news about that, so the current situation is that the "desktop" entry is still needed for scopes as well [12:42] mardy: "scope" and "desktop" may not be specified in the same entry in the hooks db [12:42] dholbach: I tried to download the App Dev training material referenced in your e-mail. Google Drive gives me a "Server error - We are unable to process your request at this time, please try again later." at the moment. Too much traffic maybe? [12:42] because they must use different apparmor templates [12:43] jdstrand: that's fine, I don't mind how it is specified, but then I'd expect that the hook processing the "scope" would generate a .desktop file for the scope [12:43] dpm: still blank [12:43] jdstrand: OA needs a name and an icon, and it gets them from the .desktop file [12:43] mardy: afaik, it does not currently. you should talk to pete-woods about that [12:43] dpm: http://paste.ubuntu.com/8299574/ [12:44] jdstrand: OK [12:44] * jdstrand is writing a summary of the conversation. gimme a sec [12:45] sturmflut-work, no idea - let me take a look [12:45] jdstrand, pastebin of the conversation is here: http://paste.ubuntu.com/8299580/ [12:45] I'll add it to the bug [12:45] mardy, it seems after the changes the system is still not finding any accounts. See http://paste.ubuntu.com/8299574/ - any ideas? [12:46] the "(process:3631): accounts-glib-WARNING **: _ag_manager_exec_query: runtime error while executing "INSERT INTO Services (name, display, provider, type) VALUES ('com.ubuntu.reminders_reminders', 'Evernote', 'evernote', 'sharing');": attempt to write a readonly database" seems to be a hint [12:48] dpm: I'm afraid it's again the same issue we found some time ago: try deleting ~/.config/libaccounts-glib/accounts.db* (you'll lose all of your accounts), then reinstall the click package [12:49] mardy: is this accurate: if "account-application" is specified, it must have a corresponding apparmor entry and one of "desktop" or "scope". "account-service" may be specified with "account-application" but must not appear without it. "account-qml-plugin" and "account-provider" must always be specified together and without "account-application" and "account-service". For now, "account-qml-plugin" and "account-provider" triggers a manual re [12:50] dpm seb128: https://code.launchpad.net/~martin-borho/ubuntu-weather-app/fix-1366714-min-max-overlayed/+merge/233916 [12:50] mardy, so even if that works, does that mean we cannot ship an updated Reminders app until there is a way to write to the database? [12:51] jdstrand, I think you were cut off towards the end? [12:51] mardy: http://paste.ubuntu.com/8299618/ [12:51] meh [12:52] mardy: http://paste.ubuntu.com/8299628/ [12:52] Hey, anyone know what has caused this jenkins failure? http://91.189.93.70:8080/job/music-app-ci/1113/ It seems to be stating something todo with GPG keys? 'gpgkeys: key BB82C3A8CA549E74B3C79789322391EC382C744F can't be retrieved gpg: no valid OpenPGP data found.' [12:53] ahayzen_: that seems to be related to the click package signing landing. I don't have more details, but I think they are/have discussed it in #ubuntu-ci-eng [12:53] mardy: i tried deleting those, installing the click, reboot and still see blank in the accounts dialog - http://paste.ubuntu.com/8299654/ [12:54] jdstrand: yes, though I don't see a big issue if we break this constraint "and without "account-application" and "account-service"" [12:54] jdstrand, ah :/ ... i thought things were done via debs within jenkins though? [12:54] mardy: well, we answered that before. if the qml plugin is going to use a different template, it needs to have a different apparmor entry [12:55] jdstrand: OK, fine by me [12:55] mardy: desktop has one set of templates. scopes another and the qml plugin another [12:56] popey: I don't understand what's happening there; I just see that there are not accounts. What is happening, and what is expected to happen? [12:56] mardy: which leads us back to your question> can a click ship a qml plugin without a corresponding appliction? [12:56] jdstrand: cool .-) [12:56] mardy: i go to add account, and i get a blank account screen, not the web page I expect where I'd login to evernote [12:57] jdstrand: since you are creating a template for the account plugins, we might want to answer "yes" [12:57] mardy: if it is useful, I don't see any reason why we wouldn't allow it [12:57] jdstrand: or at least, you re making it theoretically possible [12:57] mardy: http://popey.mooo.com/screenshots/device-2014-09-09-131535.png there should be a login page there [12:57] jdstrand: yep [12:58] popey: ah, ok; and what does "find ~/.local/share/accounts/" say? [12:58] mardy: right. for now, it would still trigger a manual review, but in the future once they can run under confinement, the template is written and everything is hooked up to do it, it could be an automatic check [12:58] jdstrand: indeed [12:58] cool [12:59] dpm: ok, please have the manifest follow http://paste.ubuntu.com/8299628/ [12:59] dholbach: I'll do the patch to the tools [12:59] mardy: http://paste.ubuntu.com/8299692/ [12:59] thanks a lot jdstrand! [12:59] popey, dholbach: if it comes in before I am done with the tools, please just make sure it conforms to http://paste.ubuntu.com/8299628/ [13:00] added all info to the bug [13:01] thanks jdstrand [13:02] dholbach, dpm: np [13:02] popey: that paste looks terribly familiar :-) [13:03] popey: the qml-plugins directory does not contain Main.qml, it looks like the account-qml-plugin hook is given the wrong data [13:04] popey: that @ACCOUNT_QML_PLUGIN_DIR@, if you remember [13:07] mardy, that was fixed in http://bazaar.launchpad.net/~dpm/reminders-app/account-click/revision/221, unless I did something wrong [13:09] dpm: and you also don't have any qml files under ~/.local/share/accounts/ ? [13:12] popey, ^ [13:23] i pastebinned the output dpm mardy [13:24] http://paste.ubuntu.com/8299692/ [13:25] zbenjamin: i'm still having problems with ssh. what i'm doing is on the desktop ssh-keygen -t dsa, then adb push ~/.ssh/id_dsa.pub /tmp, then go in as phablet via adb and do cat /tmp/id_dsa.pub > ~/.ssh/authorized_keys2, then try to ssh into the phone (ssh -v -p 10000 phablet@127.0.0.1) and it tells me that it's "Offering DSA public key: /home/nerochiaro/.ssh/id_dsa" but then "Permission denied (publickey)" [13:25] zbenjamin: i don't undersand why it doesn't accept my key [13:26] nerochiaro: i have connection problems myself, seems the phone goes into a reconnect loop when qtc is started [13:26] nerochiaro: but its different from your problem [13:26] zbenjamin: yes, this is outside qtc [13:30] zbenjamin: ok i got it. the problem was that the permissions on .ssh were wrong. the dir needs to have chmod 700 and the files inside it chmod 600, or sshd will be angry [13:30] nerochiaro: arrgh yeah of course [13:32] zbenjamin: but the biggest problem is: qtc will create the .ssh directory as owned by root:root [13:32] zbenjamin: so of course it won't work [13:34] zbenjamin: so the bug is: .ssh is created as root, and the chmod bits are wrong on both the dir and the authorized_keys2 file. Want me to file a bug ? [13:34] nerochiaro: huh, do you have the latest image? [13:34] zbenjamin: it's qtc that creates it [13:34] zbenjamin: and yes, latest image, flashed this morning [13:34] nerochiaro: what user are you if you do adb shell? [13:35] zbenjamin: root [13:35] nerochiaro: thats wrong, you should be phablet [13:35] balloons, FYI i just merged this with trunk so you'll need to rerun that trial run (as it had conflicts before) https://code.launchpad.net/~andrew-hayzen/ubuntu-calculator-app/autopilot-move-py3/+merge/233799 [13:35] nerochiaro: which channel? [13:35] zbenjamin: rtm proposed [13:36] zbenjamin: ubuntu-touch/ubuntu-rtm/14.09-proposed [13:38] salut davidcalle - ça va? [13:40] dholbach: Any luck with the training material download link? [13:41] zbenjamin: another thing. in the manifest, what is exactly the key i need to change to make it possible to install the app via qtc ? "name" ? [13:41] Salut dholbach :) [13:41] zbenjamin: i changed it from com.ubuntu.gallery to com.ubuntu.gallerydev and it still asks me to change the name [13:41] nerochiaro: check out click list if the app is already installed [13:41]