=== chriadam|away is now known as chriadam === salem_ is now known as _salem === _salem is now known as salem_ === salem_ is now known as _salem === JanC_ is now known as JanC === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === mardy_ is now known as mardy [08:22] zsombi: hey, from your comment on bug #1485222, do I understand you are looking into it? [08:22] bug 1485222 in content-hub (Ubuntu) "Using ContentStore disables silently any StateSaver capability" [Undecided,New] https://launchpad.net/bugs/1485222 [08:23] didrocks: no, I don't... [08:24] zsombi: who should I talk to to get it on some UI toolkit list? I think this is part of the bugs that would be nice to get fixed for our developers [08:24] or then, we can't have StateSaver [08:24] and I guess I quite spent some time to localized where the issue is with a simple reproducer [08:24] (like tracking down to ContentStore already took quite some time in my large application when things started to not work anymore) [08:25] didrocks: I know, what I don't get how ContentStore can disable StateSaver... [08:25] didrocks: if a simple ContentStore disables StateSaver, then perhabs that should be checked fiorst... [08:26] didrocks: all StateSaver does is to hook to SIGINT and SIGTERM signals to get notified when the app is killed/stopped, not more [08:26] didrocks: and then, if there're states saved, it will restore them one by one as they come alive [08:28] didrocks: there is a piece of code to which kenvandine might refer, which arises when an app is launched thru ContentHub, which delays state restoration till the URI is handled, but that piece of code is not in the toolkit, so these two components have nothing in common really [08:41] zsombi: sorry, at debconf and the network dropped. Last sentences I got were: [08:42] 10:25:01 zsombi | didrocks: I know, what I don't get how ContentStore can disable StateSaver... │ [08:42] 10:25:28 didrocks | I don't know either, didn't write that code, however, I do know it's easy to reproduce │ [08:42] [11:25:35] didrocks: if a simple ContentStore disables StateSaver, then perhabs that should be checked fiorst... [08:42] [11:26:19] didrocks: all StateSaver does is to hook to SIGINT and SIGTERM signals to get notified when the app is killed/stopped, not more [08:42] [11:26:53] didrocks: and then, if there're states saved, it will restore them one by one as they come alive [08:42] [11:28:30] didrocks: there is a piece of code to which kenvandine might refer, which arises when an app is launched thru ContentHub, which delays state restoration till the URI is handled, but that piece of code is not in the toolkit, so these two components have nothing in common really [08:43] zsombi: ok, so i guess it's back on ken's plate? [08:44] didrocks: so far, yes [08:44] ok, commenting and playing the ping pong ball then… [08:44] didrocks: and we will only deal with it if it is OTA bug, convergence is on our sholder [08:45] zsombi: I kind of get those, just sad that there are already way too many ways to break our toolkits, so before getting more features, would be nice to have a working SDK for devs [08:45] but I understand, priorities and so on, so having to leave with the bunch of workarounds I've done [08:45] (however, this one is the only one without any workaround, that's why I worry about it a little bit more) [08:48] didrocks: I know, but itás not my call... if it would be me, I'd rather spend 3-4 months to fix everything and only then to introduce new features... [08:51] zsombi: yep, I'll raise this anyway once I'll start publishing my apps to management, we'll see if that will enable some people to have some time to fix those issues, changing prioties, hopefully :) [08:52] didrocks: what comes to StateSaver, it's me :( [09:21] ralsina: if I register the second user for push, I'm at the dead end -- when the second user is logged out, I 1) can't call unregister passing the token, because that would unregister both users 2) can't not call unregister, because after logging out user #2, the phone would still receive notifications for both. [09:22] shoot, wrong chan, he's not here === greyback_ is now known as greyback|shops === greyback|shops is now known as greyback [11:38] I'm keen to package a node.js app (https://github.com/magne4000/quassel-webserver) for the phone. But I don't know much about node, and I'm not sure how best to package it. I think it needs a newer version of node than the one in the repos. Can anyone give me any pointers as to what to do? [11:40] mcphail: i use node from a ppa... one mo [11:41] mcphail, i have only created snappy nodejs packages yet ... and all of them were using node as a webserver [11:41] https://launchpad.net/~chris-lea/+archive/ubuntu/node.js [11:41] ogra_: yes, this would be a webserver too [11:41] popey: cheers! [11:42] to just get the nodejs binary and npm modules installed you can use my node-snapper script [11:42] doesnt look like he's updated it recently [11:42] it creates two tarballs with all the binary bits needed ... one per arch for armhf and i386 [11:42] ogra_: I tried your script, but it just hung when building the modules. Don't know if it is a wily problem? [11:42] could be [11:43] i only use it on trusty .. but have seen people use it on vivid too [11:43] ogra_: also wondered if C++ binaried built on wily would work in vivid due to the c++ ABI changes [11:43] most likely not [11:43] ogra_: your script builds in a wily chroot [11:43] oh, that would need some fixes then [11:44] namely you would need to convince it to download a different ubuntu-core tarball for the chroot [11:45] (just hardcode it in the wget command in the script) [11:45] ogra_: I'll give that a try === chihchun is now known as chihchun_afk === _salem is now known as salem_ [12:17] dholbach, do you remember if we had, on the old API website, a QML api for thumbnails === chihchun_afk is now known as chihchun [13:13] liuxg, hi [13:13] renatu, hi [13:14] liuxg, nik90 help to create some extra components that can be easily shared btw apps [13:14] renatu, if you have any resources on how to make use of the contact APIs, it is fine to me. [13:15] renatu, nik90, that sounds great, I think it is normal to have the component to import contacts from other apps. a developer does not need to re-invent the wheel every time if there is a component there. [13:18] nik90, ping [13:18] liuxg, http://bazaar.launchpad.net/~phablet-team/address-book-app/trunk/files/head:/src/imports/Ubuntu/Contacts/ [13:18] this are the components that messaging and dialer app uses [13:19] renatu, it is really a lot of stuff. does it mean every app needs to do the same thing? [13:21] renatu, there should be some images, right? is it possible to have a component for it so that a developer to make use of it directly without involving too much into it? [13:24] linuxg, I am in a meeting right, I will be back to you as soon as I finish it [13:24] sorry [13:25] renatu, it is OK. Many thanks for your answering! [13:32] davidcalle, no idea === chihchun is now known as chihchun_afk [14:03] ahayzen, what's happening? :-) [14:04] balloons, see in -ci-eng, a calendar job has been stuck for 5 days http://91.189.93.70:8080/ [14:04] balloons, preventing any further landings/jobs :-/ [14:04] ahayzen, yikes! [14:04] yeah :-/ [14:05] ahayzen, I killed it [14:05] how on earth did that happen. [14:05] #blameballoons [14:05] thanks, let see what happens now [14:05] btw the owner of blamepopey.com updated the kernel version [14:06] ahayzen, I see the issue [14:06] ooh yes they did...kernel 4.1 [14:06] yikes, I'm running kernel 4.1! [14:06] * balloons closes page [14:06] you got popey_exploit-0.3-alpha? :-) [14:07] hah [14:07] when do we expect popey_exploit-beta or v1.0? [14:07] i want a stable exploit! [14:07] exactly! We dont run alpha s/software/exploits [14:08] true.. I'm safe, it's only a broken alpgha [14:08] ahayzen: I will run your latest changes to the content export MP tonight and comment again. Mostly it looks good. I suppose once victor does the code review it can go in. [14:09] ahayzen: correct me if I am wrong, but you guys are waiting on the listitem fixes to land in OTA-6 before you push out a new music app update? [14:09] nik90, yeah we're waiting mainly for review from victor, but it would be nice to have the clock side MP'd as well so we can test it end-to-end before landing both [14:09] ahayzen: clock app MP is mostly rdy btw..https://code.launchpad.net/~nik90/ubuntu-clock-app/custom-alarm-sound/+merge/268499 [14:10] ahayzen: I was about to ask for testing help during clock app's meeting today [14:10] nik90, we are waiting for the ota6 landing before we land the MP for the listitems... we will push trunk to the store when we feel it is worth it (which the content-hub on its own could be enough :-) ) [14:11] ahayzen: sweet. Ideally it would be nice if we can release both apps together. Pushing clock app without music update makes the custom-sound feature useless [14:11] ;) [14:11] exactly [14:11] ahayzen: I'll keep you updated on that front. [14:12] thanks [14:13] balloons, if jenkins has run out of space i assume all the other jobs are gonna be stuck as well? And will need to be restarted once there is space? [14:13] I just killed the other jobs as well [14:13] thanks [14:16] it's pretty toasted. We'll need someone to free some space [14:16] hehe :-) [14:19] Can someone tell me if there is a way in debian packaging to install a suggests via a separate PPA? for example, if I have package foo, and it suggests package bar, but package bar is only available via the bar-dev/bar-tools ppa, can I specify in the packaing for foo to use that PPA or do I still need to manually (outside the package) add the PPA first [14:22] bladernr_, indeed apt only knows about what's in it's sources list [14:26] balloons: ok... damn I was hoping to be all superpowered and stuff... heh. thanks! (I really didn't think it would work, but the intricacies of debian packaging are like magic to me) [14:30] bladernr_, yea it's fun because the package gives a dependency name, but doesn't require you to install it from any specific location :-) [16:20] ogra_: was it you looking at kodi in a snap>? [16:21] popey, yeah, a few times already ... trying different approaches [16:21] be nice if we had a few extra brain cells on that [16:21] but it is more work than fun to get all the deps built first :/ [16:22] before you can even start on kodi itself [16:22] can you not just yoink them from the repo? [16:22] or do they all need special rebuilds for new sdl2 etc? [16:23] which repo ? the PPA doesnt have any arm builds [16:23] vivid does [16:23] you ean in the archive ? [16:23] *mean [16:23] yes [16:23] ugh [16:23] i always tries the upstream PPA [16:23] *tried [16:23] * ogra_ slaps forehead [16:23] i didnt even know it was in the repo [16:24] well, I'm on about deps [16:24] not kodi itself [16:24] (i think it was ripped out in wily though ... ) [16:24] ah [16:24] there are many packages not in the repo i think [16:25] root@localhost:/# apt-cache policy kodi |grep armhf [16:25] 500 http://ports.ubuntu.com/ubuntu-ports/ wily/universe armhf Packages [16:25] hmm [16:25] i'll probably play with it again on the weekend [16:25] for snappy i'd actually like it to be built with direct GLES support ... [16:26] then you neither need Xorg nor Mir [16:36] ogra_: what about ripping the build from the raspberry pi? [16:36] ogra_: it was raspbmc, but is called something else now [16:42] morphis, not sure, is there a non-raspbian build ? [16:43] err [16:43] mcphail, [17:42] ogra_: the raspbmc one was based on raspian, but there is also openelec [22:11] nik90, appear to have busted clock on my device, how do you normally build it? I tried click-buddy but that tried to build an amd64 version, then I tried doing "Install application on device" via QtC but that seems to have broken it :-/ [22:12] ahayzen: I don't usually do it via console..can I instead send you a click package ;) ? [22:12] nik90, yeah if you want :-) but how do you usually make the click ? [22:13] ahayzen: I usually set the device to phone in qtc, and then through the menus tell it create a click package [22:13] i'll try that.. [22:13] usually when I run clock through qtc on device, it creates a armhf click in the builddir [22:13] but that should be no different to install app on device via QtC surely :-/ [22:13] indeed [22:14] it shouldn't be [22:14] wait which version did you try installing? [22:14] * ahayzen tries run on device first [22:14] lp:~nik90/ubuntu-clock-app/custom-alarm-sound -> rc-proposed mako [22:14] I broke the clock app in the MP I gave you few minutes back...which I just fixed now [22:14] try pulling again now [22:14] oh! [22:14] the last revision should rev 357 [22:14] sry [22:15] wrong format in the .json file [22:15] #blamenik [22:15] ahayzen: no no you're not starting any meme similar to #blamepopey or #blameballoons :P [22:15] you don't have a website yet no ;-) [22:17] ahayzen: if everything goes well, we will have this feature merged tomorrow. And then wait for translations for another few days. [22:18] During the meeting, we were planning for an update hopefully next week sometime [22:18] hehe i'm just waiting on victor :-) as you have an MP then i should be able to poke him with something to test end-to-end [22:18] and our meeting is tomorrow so can raise it then [22:18] ok [22:19] I wish the file manager was also a valid music source [22:19] nik90, it appeared as one for me from the hub-importer? [22:19] it shows up in the peer picker page but doesn't actually work properly. [22:19] ah [22:19] as in I cannot see any music files to pick from [22:20] file manager is strange though as IIRC it still tries to open music files under ~/Music via content-hub instead of urlhandler [22:20] oh [22:20] which causes duplicate tracks :-) [22:21] * nik90 checks if filemanager has any code contributions [22:21] https://bugs.launchpad.net/ubuntu-filemanager-app/+bug/1413821 [22:21] Ubuntu bug 1413821 in Ubuntu File Manager App "Music files are imported into the Music app, rather than simply opened" [Undecided,New] [22:22] as i said in comment #7 anything in /home/*/Music or /media/*/*/Music can just be opened via urlhandler [22:23] hmm strange I see 7 MP top-approved to land..but no jenkins activity in them. strange [22:23] * nik90 looks at the bug report now [22:23] nik90, maybe due to the jenkins chaos earlier [22:24] (it ran out of disk space) [22:24] ahayzen: no its blocked since beginning of august [22:24] oh [22:25] nik90, i did add custom sound..selected a track...then it went back to the list but suru arpeggio is still checked and i can't see my track? [22:25] ahayzen: the custom sounds are added right at the top, just below the "Add custom sound" button [22:26] if it didnt add it, you might have to reboot [22:26] "Alarm Clock" is the frist [22:26] i see this in the log as well.. qml: [LOG] Final Custom Alarm Sound URL: file:///home/phablet/.local/share/com.ubuntu.clock/07 - Part II.flac [22:26] but i'll try restarting just to be sure :-) [22:26] no dont [22:26] hah [22:26] it only shows .mp3 and .ogg files [22:26] ugh [22:26] y u no flac? [22:26] I wasn't sure of how to include all music formats [22:27] can you not import them to their own folder and then just list all of them? as if they came via content-hub they at least should have been music [22:28] or you could just attempt to play/pre-roll each track ? [22:28] or you could lookup what formats the phone supports... [22:28] I was thinking of doing that, but not sure if contenthub move() function created a folder on the fly if it doesnt exist [22:28] * nik90 tries that now [22:29] i think it does [22:30] as we dump it in ~/Music/Imported/yyyy/MM/dd/hhmmss when we import [22:31] oh nice...I think I just need to change in one place and then test [22:31] :-) [22:31] have like a custom-tones folder or something :-) [22:31] yup exactly [22:32] * ahayzen wonders how file manager determines what is 'music' [22:32] ahayzen: btw why was "suru arpeggio" checked for you by default? We changed the default alarm tone to "Alarm clock" [22:32] a couple of commits ago in trunk [22:32] probably because i have suru arpeggio normally? idk [22:32] oh [22:32] i edited an existing alarm :-) [22:32] ah ok [22:33] nik90, http://bazaar.launchpad.net/~ubuntu-filemanager-dev/ubuntu-filemanager-app/trunk/view/head:/src/app/qml/content-hub/contenttyperesolver.js#L33 file managers 'solution' [22:33] oh my [22:34] btw in your testing, does the music-app stay in the unity8 spread despite choosing the sound? [22:34] content-hub or unity8 bug? [22:34] nik90, sometimes yes, but that is a shell bug i've been told [22:34] ah ok [22:34] nik90, if you try to go to it you can see it has actually be quit'd [22:35] just the surface is still being shown in the spread for some reason [22:35] Ken seemed to know about it already [22:35] yeah its blurred out and stuff [22:35] yeah [22:37] nik90, i like the changes to the app against whats in the store btw :-) [22:37] :-) [22:37] Hopefully we continue that trend [22:38] are you going to allow swiping left/right on the start page to switch between the sections ? [22:39] ahayzen: yes, we had to disable it in the very last minute since it is conflicting with the list item swipe-delete gesture and we couldnt fix it properly [22:39] oh for the laps ? [22:39] and the world city listitems [22:39] yes [22:39] do you think the bottom edge on the stopwatch page should show the laps ? === salem_ is now known as _salem [22:40] I tried a couple of ideas by mzanetti which included placing a mouse area below , but that only works partially [22:40] as its difficult to read the laps in such a small view when you have loads [22:40] well we were thinking of using the bottom edge to allow users to save laps permanently...for use cases like comparing lap times later at any point of time with your past records [22:40] that'd be cool [22:41] this is the first implementation of the stopwatch feature. Expect some more design refreshes coming up for clock app. [22:41] coolio :-) [22:42] nik90, let me know when the custom sound one is ready to rock for testing again :-) [22:42] ahayzen: will do. I just noticed the logs saying that it transfered the sound to com.ubuntu.clock/CustomSounds/...*.mp3 etc..but I dont see it in the UI..investigating now [22:43] will ping you [22:43] thanks === _salem is now known as salem_ === salem_ is now known as _salem [22:56] ahayzen: go ahead pull. I also fixed music names not eliding as well. [22:56] and all formats supported [22:56] nik90, thanks :-) [22:57] well thnks for raising the issue. Would hate to hear users report it after the store update [22:58] yeah hehe [22:59] nik90, i have "clock-app@30" and "ubuntu-clock-app" now in my list? [22:59] eh? [22:59] in the alarm sound list? [22:59] yup at the top [23:00] wat..that's crazy..it should show the contents of .com.ubuntu.clock/CustomSounds [23:00] which is empty [23:00] what happens if that folder doesn't exist as in my case? [23:00] does it then drop back to "." [23:00] checking now [23:00] which would be where the app is running? [23:00] hence the icon appearing [23:01] god damn QML FolderListModel if that be true [23:01] * ahayzen shakes fist [23:02] * nik90 wonders how to fix this.. [23:03] I could add a custom c++ function which checks the existence of the directory first [23:03] could just ensure its created on startup i guess? [23:03] is there no error from the FolderListModel? [23:03] i can't see anything in the docs [23:03] well the folder property is by default a invalid url [23:04] file:///opt/click.ubuntu.com/com.ubuntu.clock/3.5./share/qml/alarm/AlarmSound.qml:92:17: Unable to assign [undefined] to QObject* [23:04] * ahayzen checks what that is [23:04] that is the contenthub activeTransfer [23:04] line92 [23:04] ah yes have the code infront of me now :-) [23:05] ok I can fix it easily. GIve me another 5 mins [23:05] surely if the folder doesn't exist it should really error or just return nothing? [23:05] exactly, that's what I assumed as well [23:05] crazy QT :P [23:05] i would have as well [23:06] it is under Qt.labs i suppose ;-) [23:06] haha yes I suppose [23:07] stop making sense ahayzen and just complain alongside pls :P [23:07] we should raise something upstream :-) [23:19] ahayzen: fixed and pushed. [23:19] \o/ [23:19] * ahayzen pulls [23:20] folder will be created on app startup if it does not exist [23:20] and if a dev deletes the folder with the app running its their own fault for being silly? ;-) [23:20] lol yes... [23:21] I mean I could have added that check in the EditAlarmPage, but was afraid of a race issue between the folderlistmodel and the folder creation [23:21] yeah that'd be the other option, unless you could refresh it, but then you could get a flicker [23:21] this is probably simplest for now [23:21] yeah [23:22] so they have disappeared, now time to import :-) [23:22] * nik90 crosses his fingers, hoping for the best [23:23] nik90, ok they import, but how do you stop them from playing? [23:23] you cant, unless you change the selection or go back to the previous page [23:23] or have i now got to listen to 6mins of paramore? ;-) [23:23] ah [23:24] that wasn't a design issue up until now since default tones were small [23:24] so the settings has a "stop playing" button that is always visible [23:24] I will check with design [23:24] or why not just clicking on the same listitem to toggle play/pause ? [23:24] (like we do in the queue) [23:24] hmm true that's actually a good idea [23:26] nik90, ok so the first time i couldn't save the alarm for some reason, but then i went back and set the alarm again for 1min time and it worked \o/ [23:26] logs indicate anything as to why you couldnt save? [23:27] don't think so [23:27] I hate that the sdk warnings about some background color of ubuntu shape are filling my logs uselessly. [23:27] hopefully when we switch to 1.3, they will all go away.. [23:27] file:///opt/click.ubuntu.com/com.ubuntu.clock/3.5./share/qml/alarm/EditAlarmPage.qml:103: TypeError: Type error [23:27] lots of those possible around the time [23:28] hmm that's about the alarm label assignment [23:28] its a harmless simple assignment [23:28] not sure [23:28] yeah can't see anything obvious [23:28] was that a one-time issue or does it happen on every new alarm creation? [23:28] could have just been media-hub locking up or somethign with me playing many many tracks :-) [23:29] i was editing an existing alarm.. [23:29] hmm ok [23:30] it looked more platformy as the button didn't even animate onpress [23:30] maybe the header was thinking the view was scrolling or something as i've noticed it then dims the header actions [23:30] not been able to reproduce as of yet [23:30] oh we don't touch the header button state at all in the main edit alarm page. We only do it in the child pages. [23:30] ok [23:31] anyway this is looking sweet :-) [23:31] * ahayzen wonders when system-settings can add the same support :-) [23:32] I am testing your idea atm..was only a few additonal lines of code [23:32] * nik90 hopes as well [23:32] nik90, does the swipe delete work? or like incredibly hard to hit on mako ?! [23:32] ah got it! [23:33] ahayzen: thank you! Finally someone is able to confirm that issue. [23:33] ahayzen: apparently it is fine in the BQ images [23:33] http://pad.lv/1486008 [23:33] Launchpad bug 1486008 in Ubuntu UX "leading actions width is too small making it very difficult to press" [Medium,Incomplete] [23:33] hahaha hmmm are they using the new clock/{app} with the new listitems those? [23:33] *though [23:33] they are using the new 15.04 SDK list items [23:34] as music is easy to hit, but thats because it is using the old listitems until we land it [23:34] the old renatu's listitems have larger widths, it could be because of that [23:34] i mean the folks testing on the bq are using the new listitems right ? [23:35] ahayzen: well they are...but the sdk folks fixed a bug I reported about the width only a few images ago in rc-proposed :P [23:35] that's when the problem started for me on mako [23:35] haha ugh so more reasons not to land the new listitems :'( [23:36] i'll check our listitem branch and if that is affected confirm the bug [23:36] well alan confirmed that it isn't an issue on his bq running rc-proposed [23:36] so yeah [23:36] ..and add as a blocker :-/ [23:36] its too late for clock app to revert since we did so much changes after merging that to trunk :/ [23:36] :/ [23:39] nik90, yup affects music as well FFS :'( [23:39] :/ [23:41] * ahayzen wonders how a screenshot would help [23:42] maybe they think visually it is small to press I suppose...no idea [23:42] I am going to add the screenshot and add a comment. [23:42] just confirm it on your end [23:44] i have and added to the #blocked list on our MP [23:46] nik90, were you going to make that other change to toggle the playing state by tapping in the same MP or another? As i was going to pass it for the functional part of the review [23:46] I was thinking of adding it in this MP, but my mind is too tired atm..will do it tomorrow morning. [23:47] okies, i'll approve for now so i don't forget [23:47] cool, thnx [23:48] I'm going 2 sleep now. Gnite \o === _salem is now known as salem_