/srv/irclogs.ubuntu.com/2013/12/09/#ubuntu-touch.txt

=== hakr is now known as h[a]kr
dakerBillyZane: the management needs to buy the phones then send them to developers, after that you can talk about porting :)00:02
BillyZanewho's the management?00:03
dakerBillyZane: Canonical, because needs the phones so they can test00:03
BillyZanedamn00:04
BillyZanei bought a nexus 5 thinking i would put ubuntu-touch on it00:04
BillyZanedo they pay the developers?00:04
BillyZanealso, and this is probably a dumb question00:06
dakerBillyZane: "developers" = Canonical employees00:06
BillyZanebut suppose i have ubuntu-touch on this phone...00:06
BillyZaneand i have it in "desktop mode"00:06
BillyZanethen i can install and run x86 apps?00:06
BillyZanefor example, eclipse IDE00:07
dakerBillyZane: i am not really sure00:07
BillyZanewell00:08
BillyZaneit says on the wikipedia page, supported platforms: ARM and x8600:08
dakerwhich wikipedia page ?00:08
BillyZane"Ubuntu Touch utilizes the same core technologies as the Ubuntu Desktop, so applications designed for the latter platform run on the former and vice versa. "00:09
BillyZanehttp://en.wikipedia.org/wiki/Ubuntu_Touch00:09
bkc_still needs to be compiled for whatever architecture you're running on00:09
bkc_see if qemu is build/ported/whatever for ubuntu-touch yet00:10
bkc_well, qemu-user *00:10
aquariusI believe the Ubuntu touch  keyboard is Maliit, yes? Can I customise the layout without recompiling? The base maliit keyboard comes with a bunch of plugins to actually do the hard work, some of which are QML (yay!) and some C++, and I don't know how to find out which Ubuntu is using...00:28
bkc_aquarius: slightly related, might give a hint on where to look: http://askubuntu.com/questions/307923/ubuntu-touch-cant-change-keyboard-input00:32
bkc_unless that's what you're looking for ^.^00:33
aquariusbkc_, hm, that's potentially somewhere to start looking at least! I shall have a glance at that...00:34
=== chriadam|away is now known as chriadam
aquariusaaaaaand... onscreen\active=libubuntu-keyboard-plugin.so:00:35
aquariuswhich looks a lot like the Ubuntu keyboard is a compiled C++ plugin rather than a QML thing :(00:35
aquariusnow to try and find the source on LP :)00:37
aquariuslp:ubuntu-keyboard looks relevant00:37
aquariusooh, fab, it might be QML after all00:38
aquariusI do believe it is. Excellent.00:39
aquariusand http://bazaar.launchpad.net/~phablet-team/ubuntu-keyboard/trunk/view/head:/qml/languages/en/Keyboard_en.qml seems to be the actual keyboard definition. Fantastico.00:41
aquariusalthough there is an xml file too, so clearly I need to ping bilf and ask some questions about how this works :)00:44
=== h[a]kr is now known as hakr
=== chihchun_afk is now known as chihchun
=== shuduo_afk is now known as shuduo
=== shuduo is now known as shuduo_afk
=== shuduo_afk is now known as shuduo
Elleoaquarius: in theory it should be able to load pure qml keyboards as well since it's maliit underneath, but that seems to be buggy in ubuntu touch, qml keyboards end up getting hidden by applications and only occasionally flicker through03:49
Elleoso you seem to have to do things through the ubuntu keyboard plugin at the moment03:50
=== Nothing_Much_ is now known as Nothing_Much
FuLgOrEgood morning06:14
FuLgOrEis anybody available at the moment?06:15
HunterHello06:32
=== Hunter is now known as Guest92298
=== Guest92298 is now known as Hunter2451
Hunter2451Anyone here?06:33
pittitvoss: thanks for your review! (clearly I've written C for far too long :) )06:53
tvosspitti, yw :) all minor stylistic things, and I thought I pointed out the things that could be done way easier with modern C++06:54
pittitvoss: indeed, std::chrono sounds just like what I want06:54
pittiand the ptr → ref changes are just habits I don't have yet for C++06:55
pitti(but I agree they should be done that way, and pointers be avoided)06:55
=== pipedrea1 is now known as pipedream
tvosspitti, yup, and wrap them in shared_ptr/unique_ptr if you really need pointers07:01
pittitvoss: I'll do some RTFM about these07:01
FuLgOrEdoes anybody know how to mail to the ubuntu touch launchpad email address? It doesn't work for me07:01
FuLgOrEnormally I guess all the emails should be collected on https://lists.launchpad.net/ubuntu-phone/07:02
FuLgOrEI can see several emails which were sent over the mailing list07:02
FuLgOrEbut for example nothing about Nexus 507:02
FuLgOrEin the past (3rd dec.) there were 2 emails about that topic sent via the mailing list07:03
FuLgOrEI cannot see them on launchpad07:03
FuLgOrEI hope launchpad is not "censored" for some critical topics ;)07:04
FuLgOrEI also tried to send a reply for that topic but I can not find that, too07:04
FuLgOrEstrange..07:04
FuLgOrEI used the thunderbird function "reply on list" (DE: "Liste antworten")07:05
FuLgOrEdoes anybody know more about the reason why the topic "nexus 5" is not recorded on launchpad?07:23
FuLgOrEand why I cannot see my reply to the mailing list07:23
FuLgOrEmaybe it's my fault07:23
FuLgOrEI don't know. It's the first time I use a mailing list07:24
=== chriadam is now known as chriadam|away
pittididrocks: can we land https://code.launchpad.net/~thomas-voss/process-cpp/fix_pkgconfig_setup/+merge/198025 soon? current package is obviously broken, and it's blocking my sensor tests08:22
didrockspitti: trunk as other changes that I'm not confortable to push until we can promote an image08:23
didrockspitti: which I hope will happen today08:23
pittididrocks: ah, ok08:23
pittididrocks: merci08:24
didrocksde rien :)08:24
oSoMoNMirv, hey, care to approve https://code.launchpad.net/~osomon/webbrowser-app/fix-ftbfs-qt52/+merge/197947 ?08:31
=== dholbach_ is now known as dholbach
MirvoSoMoN: ok, looking and approving if it builds for me too08:47
dholbachgood morning08:56
aquariusElleo, the ubuntu keyboard plugin seems to itself be QML, though? Or at least its keyboard definition is. I'd like to just... tweak a few things ;)09:08
infinito84Hi09:16
infinito84anybody?09:17
popey!ask | infinito8409:17
ubot5infinito84: Please don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) See also !patience09:17
aquariusElleo, the thing I'm not sure about is *what* I need to tweak to change the keyboard: specifically, do I just change the QML? The XML files as well? How do I get the keyboard to load my changes? Stuff that someone who works on this will know instantly :)09:20
* popey points aquarius at tmoenicke09:22
infinito84ok, ubot5 sorry is my firts entry (and sorry for the english), im Android programmer and like sockets application for remote control, Is anyone working on a similar project in Ubuntu-Phone?, i made this app http://www.taringa.net/posts/linux/16605472/Software-Propio---Controla-tu-PC-mediante-telefono-Android.html09:23
aquariusooh! tmoenicke, ping about the Ubuntu keyboard :)09:26
tmoenickeaquarius: just qml, ignore the xml09:27
aquariustmoenicke, excellent! If I just edit the QML directly (to see if it works) how do I get the keyboard to reload it?09:28
tmoenickeaquarius: you can use initctl stop|start maliit-server09:29
aquariusbeautiful.09:29
tmoenickeaquarius: for easier debugging, stop the server, and do this on your command line on the device: export MALIIT_DEBUG=enabled09:30
tmoenickethen just run maliit-server09:31
aquariustmoenicke, so to, for example, make the "q" key on the English keyboard provide a "1" when long-pressed, I can just edit languages/en/Keyboard_en.qml to say CharKey { label: "q"; shifted: "Q"; extended: ["1"] } and restart the service, and that's *it*?09:32
tmoenickeaquarius: you would have to add your long-press event. but yes, thats it09:33
aquariustmoenicke, how do I add the long-press event? (That has to be defined separately? It's not automatically created if a CharKey defines a "extended" property?)09:34
tmoenickeaquarius: i see what you mean. yes you are right09:35
aquariushm, I'm not allowed to write to it.09:38
aquariusdarnit. root filesystem is mounted ro.09:38
aquariusif I remount it read-write will I break things? (Will I break image updates, specifically? I do not mind if my changes get overwritten by a new update.)09:39
popeyyes09:43
popeyif you switch to r/w mode (which you dont do by just remounting it) you break OTA updates09:44
popeyso will have to reflash later via phablet-flash to make that work again aquarius09:44
aquariuspopey, that's what I was worried about. So... if I want to make a trivial alteration to one of the system files, even if I change it back afterwards, I have to step away from OTA updates? :(09:45
popeyyes09:45
* aquarius looks depressed.09:45
popeyhowever09:45
popeyyou can re-flash and not lose your data09:45
popeyso it's not _that_ bad09:45
popeyi think you can even re-flash on the device09:46
aquariuspopey, OK, cheers. Are there simple instructions somewhere for going rw, and hten going back?09:46
aquariusI'm poking around https://wiki.ubuntu.com/Touch/Install09:46
popeyyes09:46
popeyIf you have a device which is already flashed and you've made it read-write, and want to update it and keep your data and retain read-write mode09:46
popeyadb shell system-image-cli --build 009:46
popeyfrom that page09:46
popeyhmm.. that isn't quite accurate, needs an update09:47
* popey edits09:47
popey    adb shell rm /userdata/.writable_image09:48
popey    adb shell system-image-cli --build 009:48
popeydo that09:48
aquariuspopey, I'm not sure I understand how to make it rw, or how to turn off rw mode and go back to OTA updates once I've finished poking around?09:49
popeyto go rw: adb shell touch /userdata/.writable_image  && adb reboot09:49
aquariusI do not want to be in "developer mode" and have to do updates forever from the command line :)09:49
popeyto go back: adb shell rm /userdata/.writable_image && adb shell system-image-cli --build 009:49
aquariusaah, that's cool09:51
aquariusso I can go rw, then change whatever I want and play with it, then go back again and that will put me back on OTA updates? (And, presumably, remove my changes.)09:51
popeyyes09:51
popeythe system-image-cli is effectively "download a whole read-only userland, and forget whatever was there before"09:52
aquariusI shall edit that "go rw" command into the wiki page09:52
popeyhmm...09:53
popeythat's more a developery option09:53
popeywe don't really want everyone doing that09:53
popeyit's like the "killall pulseaudio" of "fixing things"09:53
Hourdhaha09:53
=== shuduo is now known as shuduo_afk
aquariusagreed, but at the moment there's no way of knowing that it's even possible without asking on IRC, which is dead frustrating. My thought was to have a line explaining what it *does* and why you don't want to do it.09:54
aquariusbesides, it won't let me log into the wiki anyway ;)09:54
aquariuspopey, suggested text (I'm logged in now): You can make your device read-write (a "developer mode") to allow editing system files directly. Doing this will stop your device from being updated, and is not recommended unless you're developing Ubuntu itself). adb shell touch /userdata/.writable_image  && adb reboot09:57
aquariuspopey, do you think that'd make it clearer?09:57
popeythats fine09:57
aquariusedited.10:00
aquariushm. I *still* can't save the keyboard file from nano.10:00
aquariusvi works, though. party like it's 1979 :)10:02
popeyare you using adb?10:02
popeymaybe ssh into it for an overall better experience10:02
aquariusyeah, I just thought of that :)10:03
aquariustmoenicke, I don't seem to have a "maliit-server" service?10:03
popeydo it as the phablet user10:04
popeyit's a user daemon10:04
aquariusah, I bet it's a user service, isn't it?10:04
aquariusha!10:04
aquariusvictory! numbers on the top row!10:05
aquariusman, now I can make apostrophe easier to get to, too10:05
aquariusquestion: has the keyboard been officially designed yet by the design team, or is it just basically stock(ish) maliit?10:06
aquariusthat is: should I just propose a bzr branch, or start a discussion on the design list?10:06
=== iahmad is now known as iahmad|afk
aquariuspopey, I can't ssh in, for some reason. (I've tried the adb forwarding approach, and get "ssh_exchange_identification: read: Connection reset by peer"; sshing to the phone's IP complains there's no route to host.)10:34
aquariuspuzzled. I'm sure this worked before :(10:34
popeyaquarius: start the ssh daemon10:36
popeyits not started by defaultr10:37
popey-r10:37
aquariusdoh.10:37
aquariusyes, that works10:37
aquariusthank you :)10:37
aquariusmuch easier for editing! yay10:42
aquariusalso, nano works.10:42
aquariustmoenicke, if I propose a branch which adds numbers to the extended property on the top row of the English keyboard, will you reject it because I should be making the same change in all language keyboards? :)10:45
=== Namidairo`bnc is now known as Namidairo
mandellool, I'm going to try and get roberto to do the spreadsheet so that we have a new version of u-d-m in the image asap11:17
mandellool, they have been busy in a sprint11:17
=== chihchun is now known as chihchun_afk
=== _salem is now known as salem_
loolmandel: ack11:33
tmoenickeaquarius: it should be consistent ;)11:39
davmor2Morning all11:44
nerochiarozsombi: when i use the StateSaver, when is a saved property restored exactly ? and what happens to the binding that might already be declared for that property ? is it permanently overriden ?11:44
zsombinerochiaro: state is restored next time the component is re-created. Bindings will be overwritten, yes11:45
nerochiarozsombi: so basically if i have a binding to the state of some other component, i can't really use StateSaver as it will restore the prev state but then forever prevent the binding to work :(11:46
zsombinerochiaro: unfortunately yes... what we could do is to show a warning in case the property in subject has a binding...11:47
nerochiarozsombi: i guess what i can do is to have a separate property which i state-save, then the binding depends on the property if the property is != "". and when the component gets loaded i reset the property to ""11:49
nerochiarobut it seems really convoluted11:49
nerochiarozsombi: for doing things like saving the current view and position in a list i'm not really sure StateSaver makes a lot of sense11:50
nerochiarozsombi: i'll play around with it some more though11:50
zsombinerochiaro: yes, that is the way unfortunately. if we would try to set the property before the binding is evaluated, then we wouldn't be able to do it reliable, and we would end up like we were with teh previous theming...11:50
nerochiarozsombi: can't StateSaver do a one-off assignement instead of overwriting the binding ?11:51
zsombinerochiaro: any kind of property write breaks previous bindings... :(11:51
nerochiarozsombi: i.e. the binding is evaluated, then StateSaver changes the value, then gets out of the way11:51
nerochiarozsombi: ok11:52
nerochiarozsombi: another Q, when does StateSaver save its state ?11:52
zsombinerochiaro: but... perhaps if we would try to save the binding then restore after the write finished...11:52
zsombinerochiaro: when the component is destroyed11:53
zsombinerochiaro: also when the app is deactivated11:53
zsombinerochiaro; app deactivate leads to have all the states saved at once11:53
nerochiarozsombi: ok. let's see if i can do what i want with it, but i fear it's not really usable for my task11:54
zsombinerochiaro: when Notes app looses focus (goes to bg ex) all states will be saved11:55
nerochiarozsombi: i'm trying to save the state of the gallery's current screen so i can go back there when it's restarted11:55
zsombinerochiaro: ah, ok11:56
zsombinerochiaro: is that in a PageStack?11:56
zsombior tabs... hmm...11:56
nerochiarozsombi: in gallery no, i will have to do it with a Loader.source11:56
nerochiarozsombi: there might be something to do with PageStack or tabs too once a screen is loaded11:57
zsombinerochiaro: aham... hmm... curious to hear the result!11:57
nerochiarozsombi: ok11:57
nerochiarozsombi: it seems to work fine with selectedTabIndex, since it's normally not bound to anything12:06
zsombinerochiaro: kewl!!12:07
nerochiarozsombi: that was easy, let's see how it goes when i try to save a position in a list. i predict pain ;)12:07
zsombi:/12:08
nerochiarozsombi: also i don't want to think at when i'll need to write unit tests for this stuff12:08
nerochiaro:(12:08
zsombi:)12:08
nerochiarozsombi: where is the state actually saved ?12:09
zsombinerochiaro: in an conf file under ~/.config/<appname>12:09
=== MacSlow is now known as MacSlow|lunch
nerochiarozsombi: so in unit tests i can go and zap it away before running the test, so i'm sure it's all clean12:12
aquariustmoenicke, http://bazaar.launchpad.net/~sil/ubuntu-keyboard/numbers-on-top-row/revision/120 -- I haven't proposed for merging, because I've only poked the English keyboard. Most importantly, do you think that the refactoring to make each row a separate thing is a good idea? (Those rows are the same in each keyboard; it seems silly  to duplicate the code.)12:15
zsombinerochiaro: you can do that yes, perhaps you could also check how it is done in the state saver unit tests.12:16
nerochiarozsombi: ok12:16
nerochiarozsombi: seems to be doing stuff from c++ with StateSaverBackend which i can't do in AP tests12:19
nerochiaromzanetti: i'd ask omer but he's not around, but maybe you know: is it possible to use unity8 in autopilot to ask it to terminate an app ? or should i get the PID and kill it ?12:20
zsombinerochiaro: yep... unfortunately that's true12:21
nerochiarozsombi: no big deal, i can zap files easily12:21
nerochiarozsombi: makes for a cleaner test anyway, so i know all config is gone for that app12:21
mzanettigreyback: hmm... what would be the right way to close applications? scripted that is12:21
zsombinerochiaro: also, your unit tests may have different app name, so it won't interfere with the gallery app states12:22
mzanettigreyback: just killing it would crash unity in certain states still, right?12:22
nerochiarozsombi: still need to remove the state between test12:22
zsombiyep12:22
greybackmzanetti: well unity should never crash. If app stops, upstart should notify unity12:22
nerochiarogreyback: mzanetti: really ? is it that fragile ?12:22
mzanettigreyback: sure... I gree it shouldn't :P12:23
mzanettinerochiaro: well. I guess you can just kill it.12:23
cwaynemardy: heya, got a minute for an MR? :) https://code.launchpad.net/~cwayne18/ubuntu-system-settings-online-accounts/account-plugin-click-hooks/+merge/19812812:24
mzanettinerochiaro: if everything goes well, fine, otherwise let us know and we have to fix it12:24
greybacknerochiaro: kill the app, it should be fine. If not, I'll fix it12:24
=== boiko_ is now known as boiko
=== tvoss is now known as tvoss|lunch
nerochiarogreyback: mzanetti: thanks12:25
zsombiboiko: StateSaver is what U R looking for ;)12:25
aquariustmoenicke, also...I don't suppose it's possible to drop my own QML files into /home/phablet and then point at them somehow with .config/maliit.org/server.conf? :)12:25
boikozsombi: nice, let me take it a look12:27
nerochiaroboiko: if you're trying to save the current tab, you just do StateSaver.properties: "selectedTabIndex" in your MainView.Tabs12:29
davmor2greyback: you don't have a maguro then do you, if you think unity should never crash ;)12:29
boikonerochiaro: nice! and if I need to save extra data, I can just declare it as a property and it is going to be saved, right?12:29
mardycwayne: I added a comment; your MR looks good, but since I approached the same topic with another solution, I'd prefer you to rebase your branch12:29
nerochiaroboiko: looks like12:29
greybackdavmor2: I have one I can test with. But I do focus more on the nexus4 I'll admit12:30
zsombiboiko: http://developer.ubuntu.com/api/devel/ubuntu-13.10/qml/ui-toolkit/statesaving.html12:30
boikozsombi: yep, just found it12:31
cwaynemardy: hm, they're not apparmor-hooks though, they're click-hooks12:31
nerochiaroaaaargh !! autopilot-qt and autopilot-python STILL conflict on the desktop. is anyone else ever running tests on desktop ???12:31
zsombiboiko: also remember that if the serialized property has binding, that may be broken when th eproperty is restored12:31
zsombinerochiaro: I used to... haven't noticed yet though...12:31
mandellool, FYI we have a lot of memleaks fixes in udm and we are missing reviews, we are going to get eh reviews today and will propose the new u-d-m landing for tuesday, is that ok with you?12:32
boikozsombi: how about more complex stuff like pagestack?12:32
davmor2greyback: I challenge you to use the maguro as a daily driver for week, I give you 1-3 days before you want your n4 back :)12:32
mardycwayne: uh, what was I smoking? :-) I'll rename that now :-)12:32
mandellool, all the fixes are mem related, nothing else, after that we start with the code for the client lib (I have done a lot of work and gotten no reviews :-/ )12:32
zsombiboiko: that's also a no-go, if you want to save the pages stacked...12:32
boikozsombi: I can probably workaround it12:33
zsombiboiko: we were also planning to arm UITK components with state saving, however have not had time to even think about those12:33
nerochiarozsombi: try it now, you won't be able to run them anymore. it's been like that for weeks12:33
zsombiboiko: sure you can, on the PageStack if you work with URL pages, you can. Also, array saving is not solved either12:34
zsombinerochiaro: ok, will try12:34
boikozsombi: how bout dictionaries?12:34
=== Ursinha-afk is now known as Ursinha
boikohow about12:34
zsombiboiko: meaning?12:34
nerochiaroboiko: check the unit tests for that feature, all supported types are listed in one of them12:34
boikozsombi: javascript dictionaries as properties, can those be saved?12:34
nerochiaroboiko: the unit tests for ubuntu-ui-toolkit and statesaver i mean12:35
zsombiboiko: ah, ok. not really... whatever can be saved is listed in the tutorial I gave U12:35
zsombiboiko: basically QML base types12:36
cwaynemardy: great! i'm happy to rebase mine then, shall i just wait til yours is merged?12:36
boikohmm, that complicates things a bit I think12:36
nerochiarozsombi: basically the problem is: ubuntu-ui-toolkit-autopilot requires libautopilot-qt, but libautopilot-qt conflicts with python-autopilot > 1.4, which is required to run AP tests.12:36
zsombiboiko: file a wishitem for SDK, so we can investigate the possibilities12:36
nerochiarozsombi: that's on saucy12:36
boikozsombi: ok, thanks12:36
mardycwayne: no, you don't have to wait. Just wait till I ping you in a short while when I'm done with the renaming12:36
nerochiarozsombi: i tried moving to trusty but the upgrade crashes half way12:37
nerochiarozsombi: maybe today i'll have more luck12:37
zsombinerochiaro: not sure is the same on trusty... 'cause I'm on that, and I was able to run it last week np12:37
nerochiarozsombi: trusty should be ok12:37
nerochiarozsombi: if only i could update to it without having to wipe my system :(12:37
zsombinerochiaro: I'm a trusty oldtimer :D12:37
mardycwayne: renaming done and pushed, you are free to go :-)12:38
nerochiarozsombi: yeah, i would be too if I could actually do the damn upgrade12:38
cwaynemardy: thanks!12:38
nerochiarozsombi: i'm gonna try again now, see if i'm more lucky12:39
zsombinerochiaro: comment out extras from sourcelist, then it should work12:39
nerochiarozsombi: extras ?12:39
zsombiubuntu extras12:39
nerochiarozsombi: doesn't it disable automatically all PPAs when upgrading distro ?12:39
zsombinerochiaro: that was failing 2 me 212:39
zsombinerochiaro: that's a channel nota ppa12:39
nerochiarozsombi: oh, ok12:40
nerochiarozsombi: is that universe or multiverse ?12:40
zsombiphhhfffff... don't remember12:41
nerochiarozsombi: i'll try multiverse first12:41
zsombinerochiaro: I took deb http://extras.ubuntu.com/ubuntu..... lines away12:42
zsombinerochiaro from /etc/apt/sources.list12:42
nerochiarozsombi: ok, not sure why i was doing it with the GUI anyway12:42
janimo`ogra_, hi, can the touch build be now performed locally?12:43
=== LarrySteeze is now known as LarrySteeze|Away
cwaynemardy: like this? https://code.launchpad.net/~cwayne18/ubuntu-system-settings-online-accounts/click-hook-rebase/+merge/19824912:50
mardycwayne: can you set the app-access branch as dependency? In that way, only your changes will be shown in the diff12:51
cwaynemardy: i'm not quite sure how to do that tbh12:52
mardycwayne: when you submit the MR, you can set a prerequisite branch12:56
pittididrocks: sorry for the late followup: so we can't top-approve/land process-cpp in trunk because that's a package which would get auto-uploaded?12:56
didrockspitti: no no, you can top-approve12:56
pittididrocks: I can't (not in phablet-team)12:57
pittididrocks: so I can find someone to review/top-approve, it'll just not land yet; thanks!12:58
popeyi can add you to that team if needed pitti ..12:58
pittipopey: might make sense, is there a particular initiation rite or something new members should be aware of?12:59
* pitti gets ready for the pain sticks12:59
popeyoh, you don't know?12:59
popeyThe first rule of Phablet Team...13:00
pitti"Don't talk about the Phablet Team!"13:00
popey /kick pitti13:00
popeydone13:00
pitti/block popey13:00
popey\o/13:00
pittipopey: thanks13:00
popeyyw13:00
cwaynemardy: ah, im not sure you can do that from bzr lp-propose-merge, but the UI has it, done :)13:01
mardycwayne: thanks!13:01
=== alan_g is now known as alan_g|lunch
=== iahmad|afk is now known as iahmad
loolmandel: I'm not on a deadline to get this or that for my personal usage; it's just that I noticed that we hadn't done any update in trusty, and were in this awful situation were the stable release has some bug fixes that the devel series lack (which should be a no-no)13:22
mandellool, AFAIK all fixes in S are in trunk, I made sure of that. If you are referring to the logs issue. I'll make sure that all fixes are landed in the img asap I though we would be back to automatic updates, mea culpa for not keeping track of that, sorry13:23
loolmandel: trunk is badly out of date with the image though13:23
loolmandel: being in trunk only serves the developers  :-/13:24
mandellool, correct, I have to push that, ralsina already updated the doc, waiting for reviews right now13:24
loolmandel: it's ok, I'm not looking for apologies or anything; I raised this to you some time ago and now you're taking care of it, all good :-)13:24
mandellool, no worries, I dont take it badly :)13:25
mandellool, will make sure it does not happen again13:25
loolmandel: I think we are in a slow shift towards making all developers care for the end-result that users are getting more than for individual components13:26
loolmandel: and it takes time for our processes and for people to adjust13:26
loolit means that you should arrange to verify the quality of your component *as used* in the image (by end-users and end-developers) and raise issues around its usage (e.g. if the network connection doesn't behave correctly, you should raise this even if it's not your component)13:27
loolbut anyway, I think this is in hand now13:27
mandellool, I had no idea we would not have jenkins and left it all in trunk, but knowing this I'll take a closer look and AFAIK I'm getting a nexus 4 for proper phone testing (right now I just can test on the nexus 7)13:28
loolmandel: we do have jenkins13:28
=== JamesTait is now known as Guest4828
mandellool, for the image landing I mean13:29
loolyeah, it's still there too  :-)13:30
loolit's just that the default isn't to copy PPA into archive automatically13:30
loolbut things land up to the PPA automatically all the time13:30
loolmandel: http://people.canonical.com/~platform/cu2d/results13:30
mandellool, oh! super cool!13:31
mandelI had no idea13:31
loolmandel: it kind of implies we're still not very good at communicating how the whole process works  :-/13:32
loolmandel: but now you know and you can tell your peers  :-)13:32
mandellool, will do in our next team meeting, thx for the info!13:32
loolmandel: FYI, we're generally moving towards even stricter handling of trunk13:33
cwaynemardy: updated MR based on your comments :)13:34
loolmandel: today as you know you can autoland things in trunk once they pass peer review and automated tests13:34
loolmandel: but as much as possible, we'd like to keep trunk and image close13:34
loolmandel: I mean close to each other13:34
loolmandel: that means landing only known good things in trunk13:34
mandellool, ok, so are we going to have a devel branch or something of the kind?13:35
loolmandel: you'll have feature branches like today13:35
loolmandel: but only land them once they are good to ship13:35
looldevel branches are bad IMO, but might be necessary in some cases13:35
mandellool, well, as a rule of thumb I do not land anything in trunk unless it brings and improvement and know that it works, I never ever land something not tested in the device in trunk13:36
mandelis that what you mean?13:36
loolmandel: I mean trunk should also be shippable and shipped to users too13:37
mandellool, ok, in my case is shippable but it has not been shipped, I need to work on that then :)13:37
=== alan_g|lunch is now known as alan_g
mardycwayne: thanks!13:53
=== MacSlow|lunch is now known as MacSlow
rickspencer3no update today? :(14:24
=== tvoss|lunch is now known as tvoss
=== Namidairo is now known as Namidairo`bnc
=== gatox_ultra is now known as gatox
=== JamesTait is now known as Guest86827
mhall119how do I make system-image installs writable again?15:06
mhall119I need to install strace to get some info for a bug15:06
blarochemhall119: adb_shell touch /userdata/.writable_image15:09
mhall119nvm, found it15:09
mhall119blaroche: thanks though :)15:09
blarocheno worries :)15:09
=== greyback is now known as greyback|london
mhall119pmcgowan: do we have plans for data syncing?  Will we have one place to manage contacts/calendar/etc syncing setup, or will each appropriate app need to add that?15:29
bkc_UbuntuOne?15:30
pmcgowanmhall119, we will definitely manage contacts sync, not sure yet on cal15:30
bkc_perhaps15:30
bkc_gCal, gContacts *coughing*15:30
mhall119pmcgowan: given that they both use e-d-s, it would make sense to have them together15:31
pmcgowanprobably right15:31
pmcgowanbfiller, would know more15:31
mhall119bfiller: ^^ ?15:32
mhall119bfiller: I managed to get calendar sync quasi-working with syncevolution15:32
mhall119but that's not a user-friendly way of doing it15:33
mhall119bfiller: so really what I'm asking is whether I need the calendar app devs to build screens to configure syncevolution, or if we have plans for adding that in one central place15:33
bfillermhall119: unclear15:33
bfillermhall119: in discussions wiht design about that15:34
mhall119ok, is somebody currently thinking about that?15:34
mhall119ok15:34
mhall119bfiller: in that case, if you could just keep me updated when a decision is made15:34
bfillermhall119: will do15:35
bfillermhall119: current thinking is it's apps specific I believe15:35
bfillermhall119: the plan would be to use syncevolution for all of our platform syncing, used by contacts, calendar and possible other apps15:36
bfillermhall119: there is work in process right now to integrate syncevo with ubuntu online accounts to perform the auth15:37
bfillermhall119: mardy is working on that and has proposed changes upstream to support it in syncevo15:37
=== alan_g is now known as alan_g|tea
mhall119bfiller: ok, so at a minimum we're going to need a QML plugin for syncevolution15:47
mhall119so calendar-app can read and write configurations to it15:47
mhall119which I assume will need a new apparmor policy to allow15:47
bfillermhall119: not necessarily15:47
mhall119jdstrand: ^^ is that right?15:47
mhall119bfiller: to which?15:48
bfillermhall119: qml plugin to syncevo, might be more of a qml api to some sync service that backends to syncevo15:48
mhall119true, is there a Qt API abstraction for syncing?15:49
bfillermhall119: not sure actually, but will check15:51
=== alan_g|tea is now known as alan_g
jdstrandmhall119: apparmor policy would presumably need to be updated, yes16:00
jdstrandthat gets tricky though depending on how the calendar app is supposed to use synevolution16:01
vesarbfiller, hi Bill. Do you know if there is a way to get some dummy content (images, music, contact etc.) on the phone? We would need that for user testing.16:25
cyphermoxawe_: so, uploaded urfkill with your fixes. I'm going to look into flight mode next16:30
awe_k16:31
cyphermoxawe_: I looked though and I see it was last touched a year ago... I couldn't reach the maintainers16:31
cyphermoxso I'm not sure what will happen with the project itself, whether we'd kind of have to fork it or not16:31
awe_cyphermox, yes... that was a concern of mine16:31
cyphermoxI'm going to keep an eye open for the maintainer and send an email with our patches16:31
=== jevon is now known as Guest14702
awe_I'd read/been told that it was supposed to be a replacement for the deprecated rfkill_input module in the kernel16:32
=== Guest14702 is now known as jevon_
awe_however maybe that's not happing anymore16:32
awe_cyphermox, maybe we could check with sforshee, or possible johannes?16:33
bfillervesar: yes I believe there is, sergiusens do you recall best way to get demo content on the phone? can't remember if it's through phablet-flash or installing a deb16:35
sergiusensbfiller, phablet-config16:35
cyphermoxawe_: will do later16:35
sergiusensbfiller, phablet-demo-setup I mean16:36
cyphermoxawe_: in any case, the concept is still useful, it's good to have something central to take control of the killswitches of all kinds, to avoid applications fighting each other and keeping inconsistent states16:36
sergiusensvesar, phablet-demo-setup16:37
bfillersergiusens: thanks16:37
dakerdoes anyone know how can i add a hidden wifi network on touch ?16:38
dakerother than editing NM conf files via terminal16:39
vesarbfiller, sergiusens: great! seems to work.16:39
awe_cyphermox, ack16:44
=== ricardodrosales is now known as ricdros
cyphermoxawe_: so, I emailed the maintainers, we'll see what's up tomorrow (timezone will get in the way, I think)17:15
awe_cyphermox, ok thanks!17:18
=== gatox is now known as gatox_lunch
cwayne_cjwatson, is the Exec line of a click hook run after the symlink is created?17:25
cjwatsoncwayne_: That's what the documentation says.17:27
cjwatson     If the ``Exec`` key is present, its value is executed as if passed to17:27
cjwatson     the shell after the above symlink is modified.17:27
cwayne_cjwatson, doh, sorry i missed that in the doc, thanks17:28
cjwatson[5~1/wg 2517:28
cjwatsongah17:28
cwayne_mardy, hey, so an account-plugin looks for the qml-plugin dir based on what's in <provider> in the .service file, right?17:33
=== bschaefer_ is now known as bschaefer
=== nerochiaro is now known as nerochiaro|afk
=== ricardodrosales is now known as rcidros
=== bfiller is now known as bfiller_afk
=== jevon is now known as Guest41377
=== Guest41377 is now known as jevon_
=== mhall119 is now known as mhall119|afk
=== gatox_lunch is now known as gatox
=== mhall119|afk is now known as mhall119
davmor2kgunn, pmcgowan: hey guys this might be a maguro issue but there is some serious lag on the lastest build19:21
pmcgowandavmor2, when do you see it?19:22
kgunndavmor2: lag to what event exactly? touch to change on screen? app launch ?...?19:22
davmor2kgunn, pmcgowan: in general it doesn't feel as snappy as it did on Friday and Saturday I didn't use my phone much on Sunday19:25
pmcgowandavmor2, make sure you dont have something grabbing cpu, like a crash?19:26
pmcgowanmako is snappy19:26
davmor2yeah I'm about to check that19:26
davmor2pmcgowan: hmmm might be memory and swap that is causing the issue looking at free, I'll paste it19:31
pmcgowandavmor2, how many apps do you have running?19:32
davmor2pmcgowan: 1 web app19:32
pmcgowanok19:32
pmcgowanhmm19:32
davmor2pmcgowan: paste.ubuntu.com/654739019:33
davmor2pmcgowan: it's slowly going back down now 46M free19:35
davmor2and it's feeling snappier again19:35
davmor2top isn't showing anything hog memory in general or cpu at all19:35
pmcgowanhmm19:36
pmcgowanit was definitely close to out of mem19:37
pmcgowanbe nice to see whats got it19:37
davmor2pmcgowan: ah there it is, unity8 keeps leaping to the top of top using 25.7% memory and 23% cpu19:39
pmcgowandavmor2, it tends to, did you add a lot of content or apps?19:40
davmor2pmcgowan: the g+ app is all I added to test app installs.  the only other things I've done is calls texts added music/videos/backdrop19:41
davmor2oh and connected to wifi19:42
pmcgowando you have a lot more dash content?19:42
davmor2pmcgowan: I see the music and videos I've added other than that nothing else shows in the dash19:43
pmcgowandavmor2, I am not sure how well memory is managed when there are lots of content entries, lots being 50+19:43
pmcgowanbut that could be using up the memory, showing hte thumbnails for music and video19:44
davmor2pmcgowan: is 4GB a lot of music?19:44
pmcgowandavmor2, its more the number of files that show up as album art in the dash19:44
davmor2pmcgowan: most aren't19:44
davmor2pmcgowan: just the default music symbol19:45
pmcgowanhow many music files?19:45
davmor27 video iirc and apparently 632 items but that would include folder etc19:46
davmor2pmcgowan: but in saying that it's the same music that was on, on friday19:47
pmcgowandavmor2, ok, was just guessing19:47
pmcgowanI know that been an issue19:47
pmcgowandavmor2, are you testing on manta?19:47
pmcgowanever19:47
davmor2pmcgowan: I don't have a manta, maguro and grouper is all I have19:48
pmcgowanok19:48
=== bfiller_afk is now known as bfiller
=== rcidros is now known as ricdros
mardycwayne_: not exactly: the online-accounts-ui searches for the account plugin in qml-plugin based on the name of the .provider file (eventually we'll use the <plugin> element in the .provider file, if present)20:27
cwaynemardy: ah, that's even easier then :D cus with the click hooks, it's gonna be <click-id>.provider and qml-plugins/<click-id>/ anyway20:28
cwayneso really at this point all we'd need to do really is make u-s-s-o-a actually read from .local/share/accounts/qml-plugins, which i believe ssweeny is doing20:28
ssweenyyep20:30
cwayneso we dont even have to modify the <provider> tag.. awesome!20:32
kgunnricmm: yo20:43
kgunnricmm: was just triaging a bit...noticed https://bugs.launchpad.net/unity-mir/+bug/125397920:44
ubot5Ubuntu bug 1253979 in unity-mir (Ubuntu) "Application state isn't always saved" [Undecided,New]20:44
kgunnricmm: isn't this more of a feature to be developed ? (e.g. designed expressed way to warn/save out app state)20:45
kgunnaltho...not enforceable except for core apps20:45
pmcgowankgunn, are the guys still working on performance on the nexus10?20:49
kgunnpmcgowan: performance never stops20:50
kgunnpmcgowan: but in all seriousness20:50
kgunnit shouldn't be far off from mako (save maybe pixel overhead delta)20:50
kgunnpmcgowan: we do have a list of stuff to churn thru...but isn't n10 specific20:51
pmcgowankgunn, webpages and long apps scope results are quite glitch, but I assume thats known20:51
kgunnpmcgowan: https://blueprints.launchpad.net/ubuntu/+spec/client-1404-mir-performance20:51
kgunnpmcgowan: mmm, well...not sure if its something specific to mir then...20:51
pmcgowankgunn, jumpy as compared to SF20:52
kgunnpmcgowan: is there a bug outlining those 2 specific apps ?20:52
pmcgowankgunn, dunno, thats kinda why I was asking20:52
=== salem_ is now known as _salem
pmcgowanif its just a know thing I wont bother them20:52
kgunnpmcgowan: i wouldn't say this is known20:53
kgunnnews to me anyway20:53
kgunn(...and potentially app specific)20:53
pmcgowankgunn, I will log it then20:54
kgunnpmcgowan: video would be  nice20:55
kgunnjust to ensure consistency etc20:55
pmcgowanwhat am I QA?20:55
pmcgowan;)20:56
pmcgowanI see some similar bugs logged, but mostly old and some addressed20:56
kgunn:)20:56
=== hakr is now known as h[a]kr
lopsgood morning folks. is there any way to read the console.logs from my tablet running my app?21:02
dakerogra_: help! http://i.imgur.com/nYZOub4.png21:05
randomcpplops, cat ~/.cache/upstart/application-click-<appid>.log21:05
randomcpp;)21:05
randomcppdaker, I had the same problem today21:07
randomcppreboot your phone and it should go away21:07
dobeywhat should i file a bug against for the system update bit of system-settings? is that in lp:system-settings or a different module?21:11
dakerdobey: https://bugs.launchpad.net/ubuntu-system-image21:12
dobeydoes that include the panel that is in system-settings?21:13
pmcgowankgunn, actually, I just updated and dash is way better, browser still a bit wonky21:13
kgunngood to hear21:13
dakerdobey: yep see the title : Ubuntu system image (server/client/updater)21:13
dobeydaker: i see that, but i am looking at the code and i don't see any qml. only python21:14
pmcgowandobey, ubuntu-system-settings I think21:14
dobeyindeed21:15
dakerdobey: here https://bazaar.launchpad.net/~system-settings-touch/ubuntu-system-settings/trunk/files/head:/plugins/system-update/21:17
dakerrandomcpp: system-image-cli -i worked fine, upgraded to 5421:18
randomcppdaker, great :)21:19
rsalvetiboiko: the only crash we got in r55 is from dialer app: http://ci.ubuntu.com/smokeng/trusty/touch/mako/55:20131209.1:20131203/5356/dialer-app-autopilot/21:19
rsalvetimako only it seems21:19
rsalvetidon't know if this is a known issue, was just checking the test results21:19
boikorsalveti: not known to me at least, I will investigate this tomorrow21:20
rsalvetiboiko: great, thanks!21:20
rsalvetibfiller: ^21:20
lopseverything goes to that log file huh? thx randomcpp. and hi :P21:21
randomcpplops, yes all the stdout stderr stream go there21:22
randomcpplops, how's you app going? :)21:24
randomcppyour*21:25
=== timp is now known as t1mp
lopsuh, hopefully it's going well. i must have it finished on wednesday. it's basically crap and I won't submit it to the store or anything. I might do another one from scratch later though21:27
lopsrandomcpp, I don't have a nexus though, so it's hard to really test stuff.21:31
randomcppif you want I can do some tests for you..21:31
lopsi have a nexus 7 for this assignment in particular.21:32
lopsmaybe i'll spend some time trying to port it to my own phone21:32
lopsrandomcpp, my app is not generating any cache logs. altough i have pleanty of stuff being written.22:41
=== mfisch is now known as Guest82920
PoltoSHow can I download latest binary version for my Nexus? Are they available on launchapd? I don't want to compile it myself. Or how to easily compile only ofone. I'm searching for bugfix made only few days ago: 05 Dec23:56

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!