[00:04] <onetwothee> hey everyone, I came here in search of the likelihood of ubuntu touch being compatible with the nexus 7 (2013, snapdragon, not cortext), can anyone feed me any insight?
[00:05] <popey> onetwothee: if someone ports it, sure
[00:05] <achiang> onetwothee: no one has ported it to new n7 yet
[00:06] <popey> nhaines: cyphermox https://bugs.launchpad.net/ubuntu-keyboard/+bug/1234985
[00:06] <onetwothee> popey, achiang, do you think that it is likely someone will port to it?
[00:06] <popey> as likely or not as any other android device, maybe slightly more likely as it's a nexus
[00:06] <onetwothee> i just am not too familiar with situations like these, thats why i came here, haha
[00:07] <popey> end_speculation
[00:07] <onetwothee> ok, for sure
[00:07] <nhaines> popey: thanks, I couldn't find an existing bug.
[00:09] <popey> np
[00:21] <sergiusens> doanac, hey, can you take a look at this? https://code.launchpad.net/~sergiusens/phablet-tools/click_testing_python_modules/+merge/189744
[01:07] <crowell> hi, I'm trying to get ubuntu touch to compile on my m7spr, and I get the following error currently
[01:07] <crowell> http://pastebin.kde.org/pv8xxdmae
[01:08] <crowell> where would the makefile rules for libaudioamp.so be?
[02:23] <lucenut> So I want to install ubuntu touch on my nexus 10. :-)
[02:24] <lucenut> Can I do it withou a computer?
[02:29] <iBotPeaches> lucenut: no
[02:34] <lucenut> I saw a video from february showing 2 zip files. With no computer.
[02:34] <lucenut> have things changed since then?
[03:31] <AskUbuntu> development for ubuntu tablet using python | http://askubuntu.com/q/355168
[06:01] <LLckfan> Hello
[06:01] <LLckfan> Does any1 know how to get a stale house smell out of a clean house without opening windows?
[07:05] <GoodPanda> With running apt-get update and dist-upgrade will that provide me the full updated verisons?
[07:11] <tsimpson> GoodPanda: dist-upgrade will update installed packages to their latest versions, as well as install any new dependencies for the updated packages
[07:12] <GoodPanda> tsimpson, so that would be the latest versions and all new packages the same as if I was to flash the latest daily?
[07:12] <tsimpson> no
[07:13] <LLckfan> Does any1 know how to get a stale house smell out of a clean house without opening windows?
[07:13] <tsimpson> GoodPanda: it'll be the latest version of all installed packages, plus any new packages those already installed packages depends on
[07:14] <GoodPanda> tsimpson, thanks!
[07:19] <dholbach> good morning
[07:22] <MacSlow> hey dholbach
[07:25] <dholbach> hi MacSlow
[07:38] <N0tTh30n3> Does anybody know if there's a chance that whosthere is going to be picked up again?
[07:38] <N0tTh30n3> I really can't use any phone without whatsapp
[07:38] <N0tTh30n3> It's my no 1 communications app...
[07:47] <nhaines> N0tTh30n3: no one knows if there's a chance because it's not up to any of us.
[07:48] <nhaines> Email Whatsapp and let them know.
[07:49] <N0tTh30n3> I will mail them but there was an whatapp client for ubuntu touch called whosthere. It is not commatible with the current version of the whatsapp protocol.
[07:50] <N0tTh30n3> I've seem a bunch of people on the net net stating the same problem I have, "no whatsapp client, no switch".
[07:51] <N0tTh30n3> I just thought that since the groundwork has been layed out, canonical might be interested in launching with suport for the most popular mobile messaging app in the world.
[07:52] <N0tTh30n3> For a lot of people whatsapp is more important then sms....
[07:56] <ogra_> N0tTh30n3, https://answers.launchpad.net/whosthere/+question/231055
[07:57] <N0tTh30n3> ogra, I know. I thought that canonical might be able to:
[07:57] <ogra_> canonical doesnt have any spare resources atm
[07:57] <N0tTh30n3> - check if there are legal concerns
[07:57] <N0tTh30n3> - help with protocol updates
[07:57] <ogra_> conversational stuff can surely be done, but there is still the issue that someone needs to do the work
[07:58] <N0tTh30n3> - help Matthias Gehre with this important app.
[07:59] <N0tTh30n3> ogra_, "canonical doesnt have any spare resources atm" | check
[07:59] <N0tTh30n3> :-)
[08:01] <ogra_> i suppose that wont change until 14.04 (or even 14.10, depends on when the convergence work is done)
[08:20] <N0tTh30n3> "Hello Whatsapp,
[08:20] <N0tTh30n3> First of all thank you for this great application connecting me to my friend on all different platforms! I really like your product and I've been using it for quite some time.
[08:20] <N0tTh30n3> I like android but I would like to switch/try Ubuntu Touch as it seems to be a very interesting mobile phone platform with some cool new concepts and features. At this moment the Ubuntu Touch OS seems to be (almost) ready for daily use.
[08:20] <N0tTh30n3> The only thing holding me back from using it on a daily basis is, you guessed it, lack of a ?"whats app"/"whatsapp app"?.
[08:20] <N0tTh30n3> So, via this email I would like to inform all of you at whatsapp that I would be interested in having (and alpha/beta -testing) a whatsapp client for Ubuntu Touch."
[08:25] <robjh> it'd be very easy if they didnt go mucking about with stardard jabber ¬.¬
[08:26] <checco> ciao a tutti, come si puo installare ubuntu touch su galaxy nexus s da pc ubuntu
[08:26] <checco> ?
[08:28] <checco> Hello everyone, as you can install ubuntu on galaxy nexus s touch pc from ubuntu?
[08:38] <N0tTh30n3> robjh, yeah but where would be in that?
[08:41] <asac> how do i recover from a busybox-only install?
[08:42] <ogra_> asac, whats a busybox-only install ?
[08:51] <nerochiaro> fginther: this MR passed all the tests when checking the commits, but then failed some during autolanding. how can I unblock it by re-running the tests ? there's no "trigger a rebuild" link anymore: https://code.launchpad.net/~amanzi-team/notes-app/notes-app-qmlscene-desktop-file/+merge/189563
[08:56] <checco> Hello everyone, as you can install ubuntu touch on galaxy nexus s  pc from ubuntu?
[08:56] <jibel> this new 'wifi password' dialog doesn't work, the UI components behind the dialog also receive the touch events
[08:57] <jibel> and it doesn't work at all on Mir. I cannot press the buttons
[09:04] <asac> ogra_: it doesnt boot, just ends up in busybox
[09:09] <ogra_> asac, boot into recovery and use -d <device> for phable-flash
[09:11] <JamesTait> Good morning all! Happy Face Your Fears Day! :-D
[09:12] <asac> ogra_: oh
[09:12] <asac> good
[09:12] <jibel> JamesTait, you're 1 week early, release is next week ;)
[09:12]  * asac flashes maguro for probably one of the last times :)
[09:12] <asac> :(
[09:12] <asac> ogra_: ^^
[09:12] <asac> seems to do its job. thanks
[09:12] <JamesTait> jibel, hah!
[09:15] <asac> ogra_: or not... guess it wants me to delete my  data :(
[09:15] <asac> but just a few holiday pictures lost, so guess its fine
[09:16] <OrokuSaki> Morning
[09:19] <ogra_> asac, oh sorry, i was assuming you use --no-backup
[09:19] <asac> ogra_: i prefer to not loose my photos
[09:19] <ogra_> right
[09:24]  * ogra_ notes that he seems to only have a panel clock every tenth boot or so nowadays
[09:25] <ogra_> oh, i love the new OTA updater !
[09:26] <pitti> seb128: bonjour Monsieur, ça va ?
[09:27] <pitti> seb128: can I talk you into doing a binNEW for an arch:all package with just a single upstart script? (ofono-phonesim-autostart)
[09:27] <pitti> seb128: we need that for dialer/messaging app testing in CI
[09:27] <pitti> seb128: FFE is bug 1236708
[09:27] <seb128> pitti, salut
[09:30] <asac> ogra_: btw, do we have a SIM-PIN unlock UI now?
[09:37]  * asac reboots
[09:37] <ogra_> asac, only on cmdline i think
[09:38] <RedPandaFox> Sorry to be one of those people who come into the channel just to ask questions but does Ubuntu Touch support desktop mode on a Galaxy Nexus, and if so how do I enable it?
[09:38] <ogra_> RedPandaFox, nope, 14.04 or 14.10 feature
[09:38] <ogra_> 13.10 will only be a phone OS (we need a stable base first before adding convergence)
[09:38] <RedPandaFox> ogra_, that makes me so sad :( but understandable
[09:40] <RedPandaFox> and there is no Ubuntu for Android build avalible at this time?
[09:40] <popey> RedPandaFox: correct
[09:41] <RedPandaFox> Oh well, I guess its still good. I do really like the way Ubuntu Touch is coming. I love to see how much progress is being made. Keep up the great work :D
[09:52] <seb128> pitti, sorry, I was in an hangout, looking in a bit
[09:54] <ogra_> Laney, seb128, is there a bug open about language settings not carrying currency and dateformat along ?
[09:54] <ogra_> i can set my system to german and have proper translations in most places, but neither currency nor the date format are set
[09:54] <Laney> https://bugs.launchpad.net/ubuntu/+source/ubuntu-system-settings?field.searchtext=language&search=Search&field.status:list=NEW&field.status:list=INCOMPLETE_WITH_RESPONSE&field.status:list=INCOMPLETE_WITHOUT_RESPONSE&field.status:list=CONFIRMED&field.status:list=TRIAGED&field.status:list=INPROGRESS&field.status:list=FIXCOMMITTED&field.assignee=&field.bug_reporter=&field.omit_dupes=on&field.has_patch=&field.has_no_package=
[09:55] <Laney> one day I'll learn to paste LP links propely
[09:55] <ogra_> well, doesnt seem among these
[09:56] <ogra_> i suspect the hardcoding in /etc7nevironment gets in your way here
[09:56] <ogra_> */etc/environment
[09:56] <seb128> ogra_, should be an easy fix, the item is "display language" so I guess we only call the method to set that
[09:56] <seb128> ogra_, can you open a bug?
[09:56] <Laney> File it, attach ~/.pam_environment
[09:57] <ogra_> aj, yeah, only LANG and LANGUAGE are right in that file
[09:57] <popey> mtp seems broken on #86. I don't see the device in nautilus.
[09:57]  * popey reboots phone
[09:59] <jibel> popey, works for me. A bit too well actually as I end up with tons of nautilus windows by the end of the day
[09:59] <popey> hmm, worked after a reboot
[10:01] <mhr3> eh, flashing failed for me, can get into recovery, but phablet-flash doesn't work from there, ideas?
[10:01] <davmor2_> Morning all
[10:01] <mhr3> ogra_, you always know ^ :)
[10:01] <ogra_> mhr3, use -d <devicename>
[10:01] <jibel> mhr3, add -d <device name> to phablet-flash
[10:02] <mhr3> ah, right
[10:02] <ogra_> Laney, seb128 bug 1236772
[10:04] <Laney> ok
[10:04] <seb128> ogra_, thanks
[10:06] <om26er> stgraber, Do you have suggestions to restore my phone when a flash went bad ? I did phablet-flash ubuntu-system something wrong happened during the process. Now I am dropped to busy shell when I try adb shell
[10:07] <om26er> I know I can download images from cdimage and get back to a working Ubuntu state. but I am looking for something to do in the Bootloader mode (fastboot) so that I don't have to download more stuff
[10:07] <ogra_> om26er, boot into recovery and use -d <devicename> with phablet-flash
[10:08] <ogra_> if you dont have recovery working, pull the recovery.img from cdimage and flash it with "fastboot flash recovery /path/to/recovery.img" while in bootloader/fastboot mode
[10:08] <om26er> ogra_, you are superman
[10:08] <mhr3> +1 ^
[10:09] <ogra_> heh, i just ran into that before :)
[10:10] <om26er> ogra_, I think its a new problem. I broke a maguro in the morning and now the same thing happened with my mako
[10:10] <ogra_> bot OTA upgrades ?
[10:10] <om26er> ogra_, rather both phablet-flash ...
[10:10] <ogra_> or both bad flashes ?
[10:11] <ogra_> do you make sure to update phablet-flash before using it ?
[10:11] <om26er> ogra_, didn't update it. I see there is an update pending for it
[10:12] <ogra_> k
[10:12] <sergiusens> popey, hey, so who can take a look at this? https://code.launchpad.net/~sergiusens/ubuntu-terminal-app/clicker/+merge/189578
[10:12] <ogra_> well, try to make sure to verify breakage in the latest version, it rarely happens that you need a new phablet-flash feature
[10:16] <sergiusens> ogra_, om26er new phablet-tools brings in nothing for flashing
[10:16] <ogra_> sergiusens, yeah, i just remembered
[10:16] <om26er> atleast my cable is not broken, since the issue happened with multiple cables. It could be my USB port
[10:17] <sergiusens> om26er, errors while flash?
[10:17] <ogra_> anything in dmesg on your host machine ?
[10:17] <sergiusens> om26er, or while recovery runs?
[10:17] <popey> sergiusens: i think mhall119 may be able to top approve that.
[10:17] <om26er> sergiusens, yes, it successfully pushes all the files to the phone, the failure happens when it tries to boot recovery again
[10:18] <om26er> some error code which i don't have right now. I will provide it if I see the issue again
[10:18] <om26er> sergiusens, is Mir default now ?
[10:19] <sergiusens> om26er, so it fails while on fastboot
[10:20] <om26er> sergiusens, the phone is stuck on Google logo when the issue happens, so the final reboot that it does before flashing is where the issue happens
[10:21] <ogra_> om26er, Sf or Mir ? and how long did you wait
[10:21] <om26er> ogra_, I always --no-backup so SF. But the issue is more like flashing goes bad because at that point adb shell show the busybox and not root@ubuntu-phablet
[10:21] <ogra_> if apport kicks in while the HUD starts (which usually takes 100% of all cores) the boot can take aeons
[10:22] <ogra_> hmm, yeah, that indicates it couldnt mount /
[10:24] <sergiusens> ogra_, I think it's the same issue you had when doing fastboot boot recovery.img
[10:25] <ogra_> sergiusens, well, that was clearly a host issue
[10:25] <ogra_> my USB went wild
[10:26] <ogra_> if thats the case for om26er, there should be something in dmesg
[10:35] <MacSlow> nic-doffay, can the OptionSelector also have no option selected initially? Just wondering...
[10:37] <nic-doffay> MacSlow, nope.
[10:41] <saxin> Is it possible to get Spotify to work in Ubuntu Touch ?
[10:41] <popey> saxin: we don't have a spotify app at the moment
[10:41] <popey> saxin: someone could write one though, they have a library and api available
[10:42] <pitti> seb128: merci
[10:51] <OrokuSaki> I have a feeling my firmware is not being loaded for my gpu?
[10:51] <OrokuSaki> hmmm
[10:53] <saxin> popey: Thanks for your answer.
[11:06] <saxin> When will the list with official supported phones and tablets be updated? I was thinking about buying a new phone soon, but don't want to buy an old phone just to be able to get the Ubuntu Touch flashed on it :-)
[11:07] <ogra_> updated ?
[11:08] <ogra_> it didnt and doesnt change
[11:08] <ogra_> nexus4 and galaxy nexus are the supported ones
[11:12] <saxin> ogra_: No support for new phones? How is that gonna work?
[11:12] <ogra_> saxin, we only work on these two officially, everything else has to be done via community ports
[11:15] <OrokuSaki> "W/Adreno200-ES20( 1737): <qgl2DrvAPI_glReadPixels:190>: GL_INVALID_OPERATION"
[11:15] <OrokuSaki> Hmmm
[11:15] <ogra_> GL ?
[11:15] <OrokuSaki> yeah...
[11:15] <ogra_> weird
[11:15] <OrokuSaki> when trying to play a video in browser
[11:15] <ogra_> you should talk more to jhodapp|afk then :)
[11:16] <OrokuSaki> yeah... the firmware is from a 8690 board.. touchpad is msm8660.. but it works in android
[11:16] <OrokuSaki> maybe a hybris issue or.. uh... dunno
[11:16] <OrokuSaki> I get the bad context before...
[11:16] <OrokuSaki> so maybe it's saying my 3d is not available
[11:17] <OrokuSaki> but 2d is
[11:17] <OrokuSaki> ?
[11:17] <OrokuSaki> lets see what Mir does
[11:18] <OrokuSaki> touch .display-mir in phablet right?
[11:19] <ogra_> right
[11:19] <OrokuSaki> I was looking at the difference between ubuntu hybris and the actual libhybris... I guess that thing on the wishlist is in mir
[11:20] <OrokuSaki> looks like my GLES libraries keep reloading in logcat with Mir on
[11:20] <OrokuSaki> with logcat
[11:20] <OrokuSaki> flashy screen
[11:20] <OrokuSaki> I/ServiceManager( 1034): service 'display.hwcservice' died
[11:21] <ogra_> cking, yo ho ... i was just looking into bug 1234743 ... and was wondering if we couldn't  just remove omapfb_send_vsync_work from drivers/video/omap2/omapfb/omapfb-main.c
[11:21] <ogra_> i doubt anything uses the uevent
[11:22] <cking> ogra_, when I removed it unity8+Mir didn't work for some reason that I could not fathom
[11:22] <ogra_> did you ask the mir guys ?
[11:22] <ogra_> tvoss_, ^^^
[11:22] <OrokuSaki> "terminate called after throwing an instance of 'boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<std::runtime_error> >'   what():  display factory cannot create fb display"
[11:22] <cking> ogra_, not yet, I was fighting a bunch of other issues at the some time
[11:23] <ogra_> ah, seems it also calls that function on init of the driver
[11:23] <OrokuSaki> Can I increase my logging in init.rc?
[11:23] <ogra_> (it shouldnt after initing it though)
[11:23] <tvoss_> cking, agree with ogra_, omapfb does not use uevent as far as I know
[11:24] <ogra_> tvoss_, well, the function is used on first init of the driver ... i suspect you need it there but not later anymore
[11:24] <OrokuSaki> I changed it to loglevel 15 an it seems to be the same in logcat
[11:24] <ogra_> i cant seem to find any code beyond the iniut that uses it though ... thats weird
[11:25] <ogra_> ogra@anubis:~/Devel/packages/linux-maguro-3.0.0$ grep -R omapfb_send_vsync_work *
[11:25] <ogra_> drivers/video/omap2/omapfb/omapfb-main.c:static void omapfb_send_vsync_work(struct work_struct *work)
[11:25] <ogra_> drivers/video/omap2/omapfb/omapfb-main.c:	INIT_WORK(&fbdev->vsync_work, omapfb_send_vsync_work);
[11:25]  * ogra_ doesnt get why it still goes on spamming the dbus with uevents after init 
[11:25] <OrokuSaki> loglevel 8.. oh
[11:26] <cking> ogra_, http://paste.ubuntu.com/6208910/ is what I used
[11:27] <ogra_> cking, whats line 26 there ?
[11:27] <cking> oh heck, stupid me
[11:27] <ogra_> i assume you still want that
[11:28] <cking> that;s what happens when I rush stuff while doing 3 other things late on friday
[11:28] <ogra_> heh, yeah i know what you mean
[11:28]  * cking slaps himself
[11:29] <xnox> ogra_: may I upload: https://code.launchpad.net/~xnox/ubuntu/saucy/initramfs-tools-ubuntu-touch/missing-mkdir/+merge/189789 and https://code.launchpad.net/~xnox/ubuntu/saucy/lxc-android-config/add-generic-rules/+merge/189792
[11:29] <xnox> ?
[11:30] <cking> ogra_, actually, I recall now, that was my first idea, but then I reverted c2e591296568310c55ed79032aaf747a2d4d5587
[11:30] <ogra_> xnox, bothj look fine to me did you do a boot test with them added to a working image ?
[11:30] <cking> oops, I mean a4e449e8a08e727b8160cb4e4a0367c447148d2f.
[11:32] <xnox> ogra_: well they work with emulator, let me boot up grouper with those changes in.
[11:33] <ogra_> xnox, right, i just want to be sure we dont regress working images
[11:33] <ogra_> (i wouldnt see why from the code, but better test than be sorry)
[11:35] <cking> ogra_, actually, line 26 in the code just  calls omapfb_send_vsync_work, so that patch does look sane
[11:36] <ogra_> cking, well, we dont want to lose the vsync itself, just the uevent
[11:37] <cking> ogra_, that's all omapfb_send_vsync_work does - it sends the uevent doesn't it?
[11:38] <ogra_> cking, we want the even, but not the uevent on the dbus
[11:38] <ogra_> *event
[11:39] <cking> ogra_, ok, understand that, but not how to implement that at the moment. I seriously need to attend an appointment in -5 mins time, so I'll be back later
[11:39] <ogra_> cking, ok
[11:45] <jochenh> Hey! I have a Samsung Galaxy Tab 2 (3110) and want to run ubuntu touch on it. But nearly all of the ports are not up to date at the moment because of the flipped ports!?
[11:45] <AskUbuntu> Set more than one property (Pagestack) | http://askubuntu.com/q/355321
[11:47] <OrokuSaki> https://developer.qualcomm.com/mobile-development/mobile-technologies/gaming-graphics-optimization-adreno/tools-and-resources
[11:48] <OrokuSaki> I seem to be using the "early sample" firmware\binaries from qcom
[11:48] <davmor2> popey: ogra_: are web apps working for you on the current image?
[11:48] <popey> davmor2: define current?
[11:48] <ogra_> davmor2, half
[11:48] <popey> there was an issue yesterday
[11:49] <ogra_> i can start G+ but cant scroll
[11:49] <popey> I managed to flash the wrong phone ☹
[11:49] <ogra_> (though the UI elements have the right size for the first time)
[11:49] <davmor2> ogra_: bingo
[11:49] <ogra_> i can click links
[11:51] <jibel> on build 86 clicks apps are not listed on the app scope after boot. Is it known?
[11:51] <jibel> *click
[11:52] <davmor2> jibel: does it have a connection to the web?
[11:52] <davmor2> jibel: i.e. open a browser does it actually connect
[11:52] <jibel> davmor2, yes, and how would it affect preinstalled applications?
[11:53] <jibel> davmor2, I confirm it can connection to the web
[11:53] <jibel> *connect
[11:53] <davmor2> jibel: ah sorry I thought you meant the online click apps were not displaying
[11:53] <jibel> davmor2, no, preinstalled or even installed. After a reboot it is not listed
[11:54] <jibel> but they appear if I do a search
[11:54] <davmor2> jibel: and then do they appear in the correct section?
[11:54] <popey> jibel: which apps?
[11:54] <jibel> davmor2, yes
[11:55] <jibel> popey, stock ticker, calculator, ...
[11:55] <ogra_> jibel, search for "hello" in the app lens
[11:55] <jibel> lool, bug 1236807 seems to be a recent regression
[11:55] <popey> i see both here
[11:55] <ogra_> jibel, it will magically show up
[11:55] <ogra_> jibel, the population isnt fully automatic yet
[11:55] <jibel> ogra_, I know, but it used to list all installed applications click or not after boot
[11:55] <popey> jibel: http://popey.com/~alan/device-2013-10-08-125544.png
[11:56] <popey> current build number: 86
[11:56] <davmor2> jibel: confirmed some are showing for me but not all
[11:56] <davmor2> popey: that's on SF right?
[11:56] <popey> yes
[11:56] <davmor2> jibel: are you on mir?
[11:56] <popey> well, 86, whatever that is
[11:56] <popey> I am only using defaults
[11:57] <davmor2> popey: you got a screenshot so it's sf
[11:57] <jibel> sf or mir doesn't make a difference
[11:57] <jibel> popey, I get this immediately after boot. If I do a search then they appear on the list
[11:58] <popey> jibel: mine has been up an hour or two and been rebooted twice
[11:58] <popey> no searching done though
[11:58] <popey> can someone confirm bug 1236814
[11:58] <popey> its annoying ☻
[12:00] <Laney> https://code.launchpad.net/~laney/ubuntu-system-settings/up-from-panels/+merge/189669
[12:00] <lool> jibel: so after reboot you shoudl see them, yes
[12:01] <popey> yay
[12:01]  * popey marks dupe
[12:03] <davmor2> popey: works for me http://ubuntuone.com/4XrDcsaMyZkZ96FOEU4zfi  how were you getting into account though?
[12:04] <popey> davmor2: see Laney's comment
[12:04] <popey> davmor2: its on account creation, the bug has step by step ☻
[12:05] <lool> jibel: pinged thostr_1 on your bug
[12:05] <davmor2> popey: oh yes n7 has it but as I only logged in then closed the app I didn't see it on the phone :)
[12:10] <OrokuSaki> And I need ION to use the latest qcom drivers (3.0 kernel)... hmmmmmm
[12:11] <xnox> ogra_: does touch .display-mir at all works on grouper?
[12:12] <davmor2> xnox: the command does I don't think mir was though
[12:13] <davmor2> xnox: just trying now
[12:19] <davmor2> xnox: No it doesn't get past the Google logo for me
[12:20] <davmor2> xnox: however adb is working so you can plug it in and remove the line then it works fine on SF
[12:55] <alecu> jibel: ping
[12:56] <alecu> jibel: do you have any hints on how to setup my network to reproduce bug #1236807 ?
[12:57] <jibel> alecu, actually my device sometimes connects to an open network in my area which requires a web based authentication
[12:57] <asac> ogra_: we might need a governor override bandaid... where do we put such kernel config echo'ing?
[12:57] <jibel> alecu, so it acquires an IP but doesn't really have access to the net
[12:57] <jibel> alecu,I think that is what triggers the bug
[12:57] <ogra_> asac, cpufreq ? thats in android
[12:58] <ogra_> asac, ovveriding can indeed happen via sysfs
[12:58] <alecu> jibel: may I ask you for your click scope logs? they are in .cache/unity-scope-click.log
[12:58] <alecu> (in the phablet user)
[12:59] <jibel> alecu, I think you can reproduce with a FW rule that would deny access to the web to your device
[12:59] <jibel> alecu, I'll attach the log
[12:59] <ogra_> asac, whats the issue and what exactly do you want to do ?
[13:00] <ogra_> rsalveti, up already ?
[13:11] <xnox> davmor2: =(
[13:11] <davmor2> xnox: the n7 is tegra3 nvidia based iirc so not sure if that binary is available for mir yet
[13:13] <dobey> alecu, mhr3: i didn't see any flicker of the suggestsions when i tested the branch, nor any other indication of any infinite looping (high CPU usage, etc)
[13:13] <karni> sergiusens: Heya o/ I hope you're doing great. I was wondering if you could tell me just a bit about how we actually use the i18n.tr(string) calls. How do we provide translations for Ubuntu phone apps? Where are they located?
[13:14] <sergiusens> karni, hey; I'm not the subject matter expert for this
[13:15] <sergiusens> karni, someone for the sdk team can probably help you better
[13:15] <mhr3> dobey, well the code is clearly wrong, you don't invalidate the results from inside search(), that's just broken
[13:15] <karni> sergiusens: I see. Guys at my team mentioned you and dpm. I'll continue searching :)
[13:15] <sergiusens> karni, ah, yeah, dpm is a good candidate
[13:15]  * karni nods
[13:17] <tedg> lool, Thoughts on this branch?  https://code.launchpad.net/~ted/url-dispatcher/app-id-url/+merge/187913
[13:20] <lool> tedg: I think my needs fixing comment stillk applies?
[13:20] <tedg> lool, ?  I think we need that functionality now to move core packages to click.
[13:21] <tedg> lool, Otherwise we can't have URLs that resolve to them.
[13:25] <lool> tedg: where would we use these URLs again?
[13:26] <lool> tedg: is for url-dispatcher to open e.g. music in the right app?
[13:26] <tedg> lool, We need a way to resolve "calendar:///" to it's appid, so we need the search ability.
[13:26] <lool> tedg: right
[13:27] <tedg> lool, And so, for 13.10, we don't need the appid:/// URL, but that's a tiny bit of that merge.
[13:27] <lool> tedg: comment was: Should not poke at manifest from click directory directly, but use click list --manifest instead.
[13:27] <lool> tedg: that is, you should read the manifest from output of that command
[13:27] <tedg> lool, Yes, and that's in another MR, linked below
[13:27] <cjwatson> Or click info as I newly-provided and Ted has used
[13:28] <tedg> cjwatson, Not quite yet, that MR is dependent on the one we're discussing :-)
[13:28] <cjwatson> Yeah
[13:28] <tedg> It does clean things up a bunch though!
[13:28] <lool> tedg: click info change is nice
[13:28]  * tedg loves MRs that add features AND delete code
[13:28] <lool> tedg: you miss a dependency on new click perhaps
[13:29] <tedg> lool, I think it's there, line 8 of the diff.
[13:29] <tedg> We don't need the latest, just 0.4.9
[13:29] <lool> tedg: ah it's there
[13:30] <lool> right
[13:30] <dobey> mhr3: if it's wrong, make a branch to fix it? lack of documentation and being told to do the wrong things when i ask how to do them, isn't going to help me do it.
[13:30] <lool> tedg: so will you update app-id-url once that has landed?
[13:31] <tedg> lool, Once what has landed?  The click info branch is dependent on the app-id-url one.
[13:33] <asac> lool: so the unity-mir in the ppa has everything, right?
[13:33]  * asac installs
[13:34] <lool> yes
[13:36] <nerochiaro> bfiller: i have been trying to figure out what is going on with bug 1235956 and it looks like taps sometimes go through the OSK and trigger the note to collapse, as suspected
[13:37] <alecu> dobey, mhr3: let's work things out, and focus on what needs to be changed, how this should be working.
[13:37] <asac> lool: which binaries should i install?
[13:37] <ogra_> asac, upgrades to the ones you already have :P
[13:38] <bfiller> nerochiaro: does the same behavior happen with mir?
[13:39] <nerochiaro> bfiller: good question, trying
[13:40] <alecu> mhr3: would you mind opening a bug with the symptoms you've seen, and what we are doing wrong, and if possible, how we should be doing it right? We'll take it from there.
[13:41] <sergiusens> popey, can we get this? https://code.launchpad.net/~sergiusens/music-app/click_fixes/+merge/189851
[13:41] <mhr3> alecu, can that wait a bit? trying to debug nasty crasher in mediascanner
[13:41] <mhall119> popey: sergiusens : top-approved the terminal clickify MP
[13:42] <alecu> mhr3: sure. We are taking a look at the code, and trying to reproduce in the meantime.
[13:43] <sergiusens> mhall119, thanks, if you can look at that other MR would be great too
[13:43] <sergiusens> for music
[13:44] <nerochiaro> om26er: how do i try again to run tests and automerge this MR ? https://code.launchpad.net/~amanzi-team/notes-app/notes-app-qmlscene-desktop-file/+merge/189563
[13:44] <om26er> nerochiaro, isn't it merged already ?
[13:44] <nerochiaro> om26er: it was ok before, nothing changed, but at automerge jenkins refuse it with fake errors. i just want to re-run it until the fake erros are gone and it can be approved
[13:44] <dobey> alecu: well, if it's doing what he says it's doing, i guess you can re-open the infinite loop bug
[13:44] <dpm> hi karni, I'm at a conference atm, but if you send me an e-mail reminder I can reply to you on e-mail to describe how our i18n infrastructure works
[13:45] <karni> dpm: Thank you! I'm looking at ubuntu-ui-toolkit examples, that might suffice :) In case I'm in doubt, I'll e-mail you :)
[13:45] <om26er> nerochiaro, I can retrigger the CI if you wnt
[13:45] <dpm> karni, in the meantime, this can probably help you too: http://developer.ubuntu.com/api/devel/ubuntu-13.10/qml/ui-toolkit/qml-ubuntu-components0-i18n.html
[13:45] <dpm> ok, cool
[13:45] <karni> dpm: thanks!
[13:45] <nerochiaro> om26er: oh, nevermind, didn't notice it somehow got merged
[13:46] <jochenh> Hey! I get a error with phablet-dev-bootstrap: "ERROR:phablet-dev-bootstrap:Error while trying to sync repository" Can somebody help me?
[13:47] <cwayne> zsombi: ping
[13:48] <zsombi> cwayne: pong
[13:48] <nerochiaro> bfiller: good news, can't seem to be able to repro with mir
[13:48] <cwayne> zsombi: hey, we're seeing a lot of weird issues when we've tried to setup a custom theme, was wondering if you might have any ideas
[13:49] <bfiller> nerochiaro: that is good news
[13:49] <nerochiaro> bfiller: indeed
[13:49] <zsombi> cwayne: I'll have soon the branch to fix your theming stuff, will drop you a line to test it, hopefully will solve your issues too
[13:50] <cwayne> zsombi: does it fix Ambiance's MainViewStyle.qml?  That hardcodes the theme based on the background color
[13:50] <cwayne> (which i think is one of our issues)
[13:50] <zsombi> cwayne: one but at a time :)
[13:50] <zsombi> cwayne: you can completely override the logic from MainViewStyle, noone says you should use the same code :)
[13:51] <zsombi> cwayne: check this pls: https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/custom-theme-paths/+merge/185439
[13:52] <cwayne> zsombi: checking it out now :)
[13:52] <cwayne> zsombi: my concern was, I thought if you had ambiance as parent_theme, it would go back to that MainView
[13:53] <cwayne> so to get around that, you'd have to make a completely new theme, rather than just change a palette (or at least that's my partial understanding of it)
[13:54] <cwayne> ssweeny: ^ could you take a look at that branch?
[13:54] <zsombi> cwayne: if your theme is overloading the Ambiance MainViewStyle, the theme engine will load oit from yours, not the Ambiance one
[13:54] <cwayne> zsombi: ah, ok, that's right then :)
[13:55] <ssweeny> cwayne, on it
[13:58] <cwayne> zsombi: one issue that we've seen is our custom theme overwriting *some* bits in non-themed apps
[13:58] <cwayne> zsombi: like for example, the calendar app doesn't specify a theme
[13:58] <cwayne> and so when we make a new theme, it'll somehow get the new themes background, but not font colors
[14:01] <zsombi> cwayne: the app will load whatever theme is define din ~/.config/ubuntu-ui-toolkit/theme.ini file
[14:01] <alecu> sergiusens: do you know why some pre-installed click apps have not been uploaded to the click index webservice?
[14:02] <alecu> sergiusens: https://bugs.launchpad.net/ubuntu/+source/unity-scope-click/+bug/1227570
[14:03] <sergiusens> alecu, because it's single user
[14:03] <sergiusens> alecu, so needs coordination
[14:04] <alecu> sergiusens: so, should we add each app that gives a 404 to that bug?
[14:04] <alecu> jibel: ^
[14:05] <sergiusens> alecu, I'll create them now; but will I be able to allow other accounts to upload new versions in the future?
[14:06] <dobey> alecu: the problem is basically the same as the issue of there being some duplicate apps, i think
[14:06] <dobey> alecu: ie. the pre-installed app IDs are different from the ones on the server
[14:06] <sergiusens> alecu, the plan with lool was for each core app dev to upload them by taking the package from the build farm
[14:06] <dobey> don't know if that has been "fixed" yet or not
[14:06] <alecu> dobey: yes, sounds like that
[14:08] <alecu> so, it makes sense to add each individual app to that bug, so that only one copy is on the index, with the same package_name (not app_id!) that on the preinstalled click.
[14:08] <cwayne> zsombi: unless it explictly sets its theme right?  the dialer-app and contacts-app seem to explictly load surudark
[14:08] <zsombi> cwayne: yep
[14:10] <cwayne> zsombi: but what about apps that set some theming elements but not all of them
[14:10] <cwayne> for example calendar-app sets its background to some funky color
[14:10] <cwayne> should that be overwritten?
[14:11] <lucenut> OK, folks. I have a new Nexus 10 and would like to give it a rip.
[14:11] <lucenut> Is it possible to make it dual-boot?
[14:16] <zsombi> cwayne: well, if those apps define their styles in such a way that they use the theme palette, then the only thing you can change is the colors, not the shapes anymore
[14:19] <cwayne> zsombi: hm, ok.  i think all we want to change is the palettes anyway.  but when we tried that, it didn't quite work out, let me see if i can figure out why
[14:20] <cwayne> zsombi: so what should i have in the MainViewStyle.qml if I want to just change the  pallette?
[14:21] <sergiusens> alecu, I'll ping back later to see if you, beuno and me can talk
[14:21] <sergiusens> jdstrand, seems someone is eating up the last char somewhere http://paste.ubuntu.com/6209508/
[14:21] <MacSlow> Saviq, I elegantly picked work-in-progress branches for review... :)
[14:22] <MacSlow> Saviq, any MRs in particular you want to hand over?
[14:22] <Saviq> MacSlow, good choice ;)
[14:22] <Saviq> MacSlow, check this one out https://code.launchpad.net/~nick-dedekind/unity8/indicators.overflow/+merge/186395
[14:22] <zsombi> cwayne: check what you have in Ambiance, and omit all that does the theme switch
[14:23] <cwayne> zsombi: ah, so include the majority bu take out that last bit?
[14:26] <cwayne> stgraber: we're still waiting on that RT?
[14:26] <stgraber> cwayne: yes, IS said it was fixed but it's clearly not
[14:26] <stgraber> cwayne: https://rt.admin.canonical.com//Ticket/Display.html?id=65040 if you want to try and get them to fix it any faster...
[14:27] <jibel> alecu, I added all the preinstalled apps I could identify that returns a 404
[14:27] <zsombi> cwayne: you should define a MainViewStyle.qml, and copy all the code from Ambiance which doesn't alter the theme directly. This can be kept till we fix the other bug
[14:27] <cwayne> stgraber: i'll see what I can do :)
[14:27] <cwayne> zsombi: ah, ok, thanks
[14:27] <cwayne> i think that might have been what i was missing
[14:28] <alecu> jibel: great, thanks!
[14:28] <zsombi> stgraber: check, but essentially test it!
[14:28] <zsombi> stgrabber: ah, sorry :D wrong nick
[14:30] <zsombi> cwayne: ssweeny: pls have a test with that branch, kalikiana will check the MR
[14:30] <cwayne> zsombi: sure, im about to test it :) although I think we don't set XDG_CONFIG_DIRS, so we'll either need to get that fixed, or have it set to XDG_DATA_DIRS instead
[14:30] <cwayne> ssweeny: ^
[14:32] <zsombi> cwayne: ssweeny: I'd rather say it should be the config dir
[14:32] <jdstrand> sergiusens: yes, see bug #1235444
[14:33] <sergiusens> jdstrand, thanks
[14:33] <cwayne> zsombi: i'd tend to agree, but then we're going to need to put in an MR for ubuntu-touch-session package to add that dir
[14:35] <cwayne> which i'm fine doing, but perhaps we'd need to punt it til after 13.10
[14:35] <sergiusens> popey, can you go about checking this? https://code.launchpad.net/~sergiusens/ubuntu-calendar-app/click_fixes/+merge/189865
[14:36] <ssweeny> zsombi, out of curiosity why do you prefer CONFIG over DATA?
[14:36] <ssweeny> zsombi, i'd assumed that since the themes are in /usr/share it'd be DATA but i'm happy to be wrong :)
[14:38] <zsombi> ssweeny: I dunno, the DATA gives me other information that the CONFIG does, beside I've seen people saying that those two variables are having different values...
[14:40] <ssweeny> zsombi, right. one concern i have is that we already set XDG_DATA_DIRS and we'd have to submit an extra change to set XDG_CONFIG_DIRS as well
[14:40] <zsombi> ssweeny: ok, what if I add both?
[14:40] <ssweeny> zsombi, what different info do you get from the two?
[14:41] <ssweeny> zsombi, that would work i suppose
[14:41] <lool> stgraber: which package do you think we should add upgrade hooks to?
[14:42] <lool> stgraber: I'd like to add one to switch to mir on upgrades
[14:42] <zsombi> ssweeny: for instance there were talks on the web about $XDG_CONFIG_DIRS=/etc/xdg and XDG_DATA_DIRS=/usr/local/share:/usr/share
[14:43] <stgraber> lool: since they may end up being pretty similar to maintainer scripts, I think it's best to put them into the affected source package. In this case, Mir.
[14:43] <lool> stgraber: do you have a link to a template handy?
[14:43] <ssweeny> zsombi, right. that was my understanding. we currently add /custom/xdg/data to XDG_DATA_DIRS. so since upstream themes are in /usr/share we could have our custom stuff in /custom/xdg/data. if you look through XDG_DATA_DIRS you'd get both
[14:43] <jibel> alecu, adding timestamps to unity-scope-click.log would help to determine interesting events. I'll attach a new log file when I reproduce
[14:45] <stgraber> lool: "start on boot-hooks WHEN=new-version and starting lightdm" "pre-script [ check if already done ] && stop && exit 0" "script do-whatever-you-need-there"
[14:46] <lool> tedg: approved https://code.launchpad.net/~ted/url-dispatcher/app-id-url/+merge/187913 but did you test binaries of this?
[14:47] <alecu> jibel: indeed, it does help a lot in every other log we have. I don't know why we don't include timestamps in it. I'll open a bug to add them.
[14:47] <lool> stgraber: can I tell which image I'm upgrading from?
[14:47] <lool> stgraber: rather than "already done"
[14:47] <tedg> lool, I tested those on the device a while back.  I'll check again when we have a trunk that is final.
[14:47] <lool> stgraber: I'd like to touch .display-mir, but not touch it again if people rm it
[14:47] <lool> could keep a flag file, but seems superfluous if I can test imge versions
[14:48] <stgraber> lool: you can but it won't do you much good
[14:48] <stgraber> lool: since the actual version will vary based on the channel
[14:51] <lucenut> OK, I have been reading all last night and this morning. Can't decide where to start with my nexus 10.
[14:51] <lucenut> Do I have to "root" it?
[14:53] <cwayne> zsombi: so what's the plan? to include both DATA and CONFIG dirs?
[14:54] <vincentbosch> I would like to know if there's any possibility that we will see Ubuntu Touch on the Galaxy Note 2014 edition (SM-P600) soon? Exynos 5 version
[14:55] <zsombi> cwayne: yes, updating the MP
[14:55] <cwayne> zsombi: awesome, thanks, will test as soon as its up :)
[14:55] <zsombi> cwayne: then go for it :)
[14:56] <cwayne> zsombi: pulling it now.  shall i just pbuild it? or is there some other preferred way of building
[14:57] <zsombi> cwayne: yes, pbuilding i sfine
[14:57] <cwayne> zsombi: perfect, doing so now
[15:01] <zsombi> cwayne: I'll not be around for long, so if you have something that doesn't seem to work well, and you sense it is because of the MP, drop a comment on the MP
[15:01] <cwayne> zsombi: sure thing, i'll be sure to update the MP if it works too :)
[15:04] <lool> stgraber: aha
[15:04] <lool> stgraber: so I'll create a flag then
[15:09] <stgraber> hmm, something is really wrong with the system-image client or the download service. I have two devices both with the same update path, both hanging at the exact same spot of the download
[15:14] <mterry> boiko, so I'm testing your uri_handler branch for telephony-service...
[15:15] <boiko> mterry: ah yes, it needs to be tested together with the branches for address-book-app, dialer-app and messaging-app
[15:15] <mterry> boiko, OK, I was wondering
[15:15] <boiko> mterry: let me get the URLs for you
[15:15] <mterry> boiko, thanks
[15:16] <popey> mhall119: can you happrove? https://code.launchpad.net/~sergiusens/music-app/click_fixes/+merge/189851
[15:16] <boiko> mterry: https://code.launchpad.net/~renatofilho/address-book-app/fix-1230404/+merge/188473
[15:16] <boiko> mterry: https://code.launchpad.net/~boiko/dialer-app/uri_handler/+merge/189423
[15:16] <boiko> mterry: https://code.launchpad.net/~boiko/messaging-app/uri_handler/+merge/189428
[15:17] <mterry> phew, ok
[15:17] <boiko> mterry: I was about to get some bug reports, right? :)
[15:17] <mterry> boiko, we'll see :)
[15:18] <boiko> mterry: the activation of already running apps I didn't test yet, I'm going to be testing that in a couple hours
[15:18] <mterry> boiko, ok
[15:20] <mterry> boiko, are these url schemes documented somewhere, or just in code?
[15:21] <boiko> mterry: it was somewhere on wiki, let me find it
[15:21] <boiko> mterry: hmm, it only shows the available schemes, but the addressbook specific ones are not listed there
[15:21] <boiko> mterry: https://wiki.ubuntu.com/URLDispatcher
[15:22] <lucenut> Hey, can someone help me get started loading ubuntu-touch on nexus 10?
[15:22] <lucenut> Do I first have to root it?
[15:22] <nebajoth> https://wiki.ubuntu.com/Touch/Install
[15:22] <lucenut> The only ubuntu PC I have to work with is a Xubuntu notebook.
[15:23] <lucenut> Does the developer preview upgrade into the real think when it's released?
[15:24] <nebajoth> Ubuntu Touch has the ability to perform OTA updates
[15:24] <nebajoth> I'm going to go ahead and assume it will allow you to upgrade directly
[15:24] <nebajoth> but I'm no official spokesperson
[15:25] <lucenut> Cool.
[15:26] <mhall119> sergiusens: popey: does the music app need the unconfined template?
[15:26] <sergiusens> mhall119, yes, it's using grilo
[15:27] <mhall119> ok
[15:27] <mhall119> top approved
[15:27] <mhall119> in the future I assume we'll have more restrictive templates/policies for it
[15:31] <popey> thanks mhall119
[15:34] <sergiusens> mhall119, not sure what the deal with grilo is going to be; jdstrand ?
[15:38] <jdstrand> sergiusens: I am lacking context. what about grilo?
[15:39] <mhall119> grilo is the API for querying the mediascanner database
[15:39] <sergiusens> jdstrand, are we going to be able to run confined apps with grilo
[15:40] <mhall119> jdstrand: https://wiki.gnome.org/Grilo
[15:40] <mhall119> runs over DBus, IIRC
[15:42] <jdstrand> currently, there is no policy for it so no for 13.10. it is possible to add policy, but that might also require changes to grilo to integrate with trust-store depending on what its api is like
[15:42] <jdstrand> someone should create a UDS session for it
[15:45] <Ikagara> Is there an x86 version of the ubuntu touch, does the ubuntu dekstop have the apropreate features to support bring run on an x86 tablet?
[15:46] <asac> om26er: hy
[15:46] <asac> om26er: https://bugs.launchpad.net/ubuntu/+source/mir/+bug/1235190
[15:46] <asac> om26er: is that 100% reproducible? sure its not caused by crashes etc.?
[15:47] <mhall119> Ikagara: it depends on what you include as "Ubuntu Touch"
[15:47] <om26er> asac, definitely not caused by any crash. It happens all the time.
[15:47] <om26er> 100% reproducible right after boot
[15:47] <asac> om26er: can you do what ogra says?
[15:47] <asac> int he bug?
[15:47] <mhall119> Ikagara: the shell (Unity 8) runs fine on x86, as do the apps, and most of the platform services
[15:47] <asac> om26er: he asks for a longer top run while this is happening etc.
[15:47] <om26er> top really does not show anything from the RAM/CPU perspective
[15:47] <ogra_> asac, the bug is just heavily discussed in #ubuntu-mir
[15:47] <asac> om26er: also note that unity might crash all the imte as well :)
[15:48] <asac> (but guess you checked /var/crash?
[15:48] <asac> )
[15:48] <asac> ah ic
[15:48] <asac> om26er: lets discuss there then #unity-mir
[15:48] <om26er> asac, yes there are no crashes, that's for sure
[15:48] <mhall119> the phone images contain some Android bits, because they're running on Android phones, and those bits won't be available on x86, though chances are you don't need them because something that does the job already exists on x86
[15:48] <om26er> ok
[15:49] <Ikagara> mhall119: i am playing around with a viewsonic 10 tablet, and wanted to try out ubuntu on it. would the desktop version of ubuntu be enough to support a touch screen type input (onscreen keyboard and stuff)
[15:49] <mhall119> Ikagara: technically yes, you can run the desktop on a touch device
[15:50] <mhall119> but the Unity 8 experience will be better, in Ubuntu 13.10 there should be a way for you to try Unity 8 on x86
[15:50] <mhall119> kgunn: ^^ can you confirm that statement?
[15:50] <Ikagara> how would i go about making this happen?
[15:51] <kgunn> mhall119: for the moment...touch is really based on arm config (due to our use of some pre-compiled arm libs)
[15:52] <kgunn> mhall119: this isn't to say that intel ones don't exist...but it would take some cobbling together i would imagine.
[15:52] <kgunn> mhall119: not famiiliar with viewsonic tablet...even if its intel, you would need to follow the touch/porting wiki
[15:53] <Ikagara> that would be a bit intence for me kgunn
[15:54] <mhr3> charles, ping?
[15:55] <mhr3> charles, i'm trying to repro #1236705 got it to crash on both sf and mir
[15:55] <mhr3> now trying to get stacktrace from the mir crash
[15:55] <mhr3> gdb is generating core file as i type this
[15:55] <mhr3> just don't have debug symbols
[15:55] <mhall119> kgunn: I was more asking about whether there will be a "Unity 8" session option in 13.10
[15:55] <mterry> boiko, I can't save a new contact?  probably not related to your branch, but are you seeing that too?
[15:55] <davmor2> what package is the user metrics under?
[15:56] <mterry> davmor2, libusermetrics
[15:56] <davmor2> mterry: thanks
[15:56] <kgunn> mhall119: sorry...yes, even today you can run unity8 as a large overlay window to the desktop
[15:56] <mhall119> kgunn: can you run SDK apps inside of it yet?
[15:56] <mhr3> charles, although even the nonsymbolic one is pointing fingers - http://paste.ubuntu.com/6209880/
[15:56] <mterry> boiko, oh, it just let me...  odd
[15:57] <davmor2> popey: you have 2 phone correct?
[15:57] <mhall119> kgunn: and will there be an option to run Mir+Unity8 as the user session on a 13.10 desktop, rather than inside a windows under Xorg
[15:57] <boiko> mterry: hmm, I need to check, right now my phone has different versions of all the apps (in a hangout with designers doing visual tweaks)
[15:57] <cwayne> zsombi: seems the CI failed
[15:57] <popey> davmor2: i have two ubuntu touch devices, yes
[15:57] <boiko> mterry: I'll check that afterwards
[15:57] <cwayne> also my pbuild failed as well
[15:58] <davmor2> popey: do they have photos on either?
[15:58] <popey> davmor2: yes
[15:59] <davmor2> popey: have you teken them with the camera on the phone and does the usermetrics show that in the circle on the welcome screen?
[16:00] <popey> davmor2: yes
[16:00] <davmor2> popey: thanks
[16:00] <davmor2> now to reflash and see if I can reproduce
[16:02] <jodh> xnox, stgraber: any thoughts on https://bugs.launchpad.net/ubuntu/+source/upstart/+bug/1235649/comments/35 ?
[16:03] <mterry> boiko, yeah, launching already opened apps doesn't seem to work greate
[16:04] <xnox> jodh: i'm not a dbus-foo guru, i wonder if dbus and/or us are expecting a reply/ack for each event we emit on dbus =?
[16:05] <jodh> xnox: ditto. possibly. maybe slangasek/tedg could comment?
[16:05] <boiko> mterry: I guess promoting the apps to the front has not landed yet
[16:05]  * mterry looks at watch
[16:05] <tedg> xnox, signals, no, there are no replies there
[16:06] <tedg> xnox, If you're getting a call, yes, you should respond.
[16:07] <xnox> tedg: and if client is listening on a signal, should it free the received msg / signal?
[16:07] <xnox> jodh: i thought memory usage was on receiving side, not sending side.
[16:07] <tedg> xnox, Well, there's no shared memory, so that doesn't effect the emitter.  But, yes, it should or it'll leak itself.
[16:07] <jodh> xnox: we're freeing the signal correctly fwics on our end.
[16:07] <zsombi> cwayne: CI fails on everything
[16:07] <xnox> jodh: horum.
[16:07] <xnox> =/
[16:08] <rsalveti> ogra_: I'm up now :P
[16:08] <rsalveti> (checking backlog)
[16:08] <ogra_> lol
[16:08] <ogra_> rsalveti, well, ignore :)
[16:08] <ogra_> go back to bed !
[16:08] <rsalveti> ogra_: ;-)
[16:08] <ogra_> :)
[16:08] <cwayne> zsombi: my pbuild failed cus of failed tests, i wonder if my environment wasnt setup right?
[16:09] <rsalveti> yeah, might be a better idea
[16:09] <slangasek> xnox, jodh: init/control.c:control_emit_event_emitted() doesn't exist, this seems to be nih-autogenerated code; could one of you pastebin that autogenerated code if you have it handy?
[16:10] <zsombi> cwayne: hmm...
[16:10] <stefan__> hi, when porting ubuntu touch to an unofficial cyanogen device? how to delete all stuff that is not needed from that source?
[16:11] <mterry> boiko, makes it a little hard to test these branches without that.  Is there a branch I can install for the bringing-app-to-front bits?
[16:11] <xnox> stefan__: if you remove a repository from the manifest, and do repo -u, then repository will be gone. if there is no repository, then all the extra pieces are not built.
[16:11] <xnox> stefan__: alternatively change your product manifest to remove apps, and big stuff
[16:11] <zsombi> cwayne: actually it failed on the theme engine unit tests, which are passing on my laptop... need to check it further, and will let you know when it passes
[16:12] <xnox> stefan__: in build/product/ something.
[16:12] <jodh> slangasek: yes - just run make, or: http://paste.ubuntu.com/6209950/
[16:12] <mhr3> charles, https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1236249/comments/3
[16:13] <slangasek> jodh: right, that looks correct
[16:13] <slangasek> jodh: have you checked whether adding dbus_connection_close() doesn't fix the problem?
[16:13] <zsombi> cwayne: actually that CI failure is from the first push...
[16:13] <jodh> slangasek: ack. I cannot see anything wrong with that code at all wrt memory mgt.
[16:14] <jodh> slangasek: yes - see #37.
[16:14] <slangasek> jodh: that comment doesn't tell me you've tested it, you say "could do with" :)
[16:15] <slangasek> jodh: was valgrind any use here?
[16:16] <jodh> slangasek: updated with #38. valgrind -> not that I could divine, no.
[16:18] <slangasek> jodh: I think there are missing calls to dbus_connection_unref()
[16:18] <slangasek> jodh: we unref() when called via control_bus_close(), but there are other paths into control_disconnected() which bypass this step
[16:18] <jodh> slangasek: that's what I thought, but note that nih_dbus_connection_disconnected() calls that for you (in a somewhat convoluted way :)
[16:19] <slangasek> jodh: where does nih_dbus_connection_disconnected() get called?
[16:19] <jodh> slangasek: let me update the bug with the call sequence...
[16:25] <slangasek> jodh: ah, found it; nih_dbus_connection_disconnected() wraps our own disconnect handler, and calls dbus_connection_unref() in the right place
[16:26] <jodh> slangasek: right.
[16:26] <slangasek> jodh: could you share a valgrind trace?
[16:26] <jodh> slangasek: all on the bug :)
[16:26] <slangasek> ah, so it is
[16:34] <slangasek> jodh: valgrind suggests there might be a memory leak in environ_remove(), which has no allocation tests in init/tests/test_environ.c; should probably add some?
[16:35] <mterry> tedg, in the url-dispatcher world, are there changes in how apps get focused?
[16:36] <mterry> (with mir especially?)
[16:36] <slangasek> jodh: of course, valgrind also says < 100k has been lost, so that's probably not really our culprit :/
[16:39] <jodh> slangasek: yes, we should indeed write some tests for that, but I'm struggling to see a leak just eye-balling it :)
[16:41] <plars> renato: anything new on https://bugs.launchpad.net/mediaplayer-app/+bug/1231418
[16:56] <lool> pete-woods, tedg: Hmm https://code.launchpad.net/~ted/hud/fix-hud-activation/+merge/189869 is fishy; why do we need to do this?  I thought we fixed dbus job to set this?  shouldn't we just add a "start on started dbus" to whatever is triggering hud?
[16:58] <pete-woods> lool: it's not that simple unfortunately
[16:58] <pete-woods> lool: I have the up to date dbus/upstart packages with those supposed fixes in, and they don't help
[16:59] <lool> pete-woods: do we understand why not?
[16:59] <lool> pete-woods: it seems like we're working around a state we're in that we don't understand
[16:59] <nerochiaro> oSoMoN: when you have some time, https://code.launchpad.net/~amanzi-team/notes-app/notes-app-save-on-deactivate/+merge/189919
[17:00] <pete-woods> lool: I agree, but we need it fixed now, this workaround also has the benefit of launching apport when the error occurs, for helping us track it down
[17:00] <pete-woods> lool: from my investigation it looks like a bug in upstart, but I don't think anyone is taking me seriously
[17:00] <lool> pete-woods: oh gosh saw /usr/lib/x86_64-linux-gnu/hud/dbus-activation-hack.sh
[17:01] <pete-woods> lool: that's a red herring, yes it's evil, but it's not the cause
[17:01] <lool> pete-woods: why doesn't hud have start on started dbus?
[17:01] <pete-woods> lool: we don't want that, we want lazy init
[17:01] <lool> so why dont we just have dbus activation?
[17:02] <slangasek> jodh: yah... it would be good to know if nih agreed with valgrind about there being a leak, though :)
[17:02] <ogra_> heh
[17:03] <pete-woods> lool: that is a good question, maybe we should, we should be able to use upstart, though, right?
[17:03] <lool> pete-woods: supporting both upstart and dbus activation doens't seem to have value and adds moving parts
[17:03] <ogra_> pete-woods, it would be good to pick either :)
[17:04] <pete-woods> lool, ogra_: my real concern is that any processes managed by upstart are susceptible to this
[17:04] <lool> pete-woods: think about all the code we could remove!
[17:04] <pete-woods> for all I care yes, we could switch to dbus activation
[17:04] <lool> pete-woods: well I think it's because of the dbus activation thing that we're seeing this
[17:04] <pete-woods> lool: it's definitely not
[17:04]  * ogra_ thinks it is the mixup 
[17:04] <lool> pete-woods: In fact, I wonder... is it not because we call start from within start?
[17:05] <pete-woods> you can discount the dbus activation
[17:05] <pete-woods> you can reproduce this by simply doing "start hud"
[17:05] <pete-woods> and, indeed when upstart gets into this staet
[17:05] <pete-woods> you can add new jobs (I added a foo,conf job)
[17:06] <pete-woods> that just runs a bash script that dumps its env
[17:06] <pete-woods> and that suffers the same problem
[17:06] <lool> pete-woods: but what made sure dbus was running?
[17:06] <pete-woods> lool: the fact that I was logged into my desktop
[17:06] <pete-woods> this is a running desktop session, with dbus
[17:07] <pete-woods> lool: I can do initctl list-env, and it shows the DBUS_ADDR environment variant
[17:07] <pete-woods> but it does not appear inside any upstart job's environment
[17:07] <pete-woods> variant->variable
[17:07] <lool> pete-woods: I did a start hud
[17:07] <lool> pete-woods: it had DBUS_SESSION_BUS_ADDRESS=unix:abstract=/tmp/dbus-yyOXNA71cT
[17:07] <om26er> pete-woods, hey! you work on user matrices right ?
[17:07] <oSoMoN> nerochiaro: looking
[17:08] <pete-woods> lool: I know
[17:08] <lool> pete-woods: but not on touch?
[17:08] <pete-woods> lool: it doesn't happen every time you log in
[17:08] <lool> pete-woods: isn't it because it gets dbus activated and that's racy?
[17:08] <nerochiaro> oSoMoN: having trouble running tests ATM, but the change shouldn't mess up anything that was working before
[17:08] <pete-woods> lool: no
[17:08] <pete-woods> lool: I can create a _new_ job_ no dbus activation
[17:09] <pete-woods> and that suffers the exact same problem
[17:09] <lool> pete-woods: I just created a job, and I got DBUS_SESSION_BUS_ADDRESS=unix:abstract=/tmp/dbus-yyOXNA71cT
[17:09] <lool> pete-woods: you're saying it's random?
[17:09] <pete-woods> lool: only at login time
[17:09] <pete-woods> i.e. if you've logged in and it works, it stays working
[17:10] <pete-woods> if you've logged in and it's broken, it stays broken
[17:10] <ogra_> that cant happen on dbus activation
[17:11] <pete-woods> ogra_: I really think this has nothing to do with dbus activation, as I said above you can create a new job without any crazy dbus-hack.sh and that suffers the same as HUD
[17:12] <ogra_> pete-woods, well, dbus activation means you have been triggered by a dbu message, that definitely makes sure you have the proper env
[17:12] <pete-woods> ogra_: that's a very good point
[17:12] <lool> pete-woods: I just feel it's the only weird thing about this service
[17:13] <lool> pete-woods: but you say you reproduce with other jobs
[17:13] <lool> pete-woods: do we have an upstart or dbus bug for that failure?
[17:13] <ogra_> only hud iirc
[17:13] <pete-woods> lool: the bug got logged against url-dispatcher
[17:13] <ogra_> i think there was one from ted for the dbus env though
[17:13] <pete-woods> https://bugs.launchpad.net/url-dispatcher/+bug/1234731
[17:14] <ogra_> oh, tight i'm mixing that up with "hud is slow to start"
[17:14] <ogra_> *right even
[17:14] <oSoMoN> nerochiaro: approved, feel free to top-approve once CI has run
[17:15] <nerochiaro> oSoMoN: thanks
[17:17] <charles> larsu: could you take a look at https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1236249/comments/3
[17:20] <ogra_> pete-woods, lool .... LOL ! i know what the issue is
[17:20] <pete-woods> ogra_: ????
[17:21] <larsu> charles: interesting. I'll try to reproduce it tomorrow
[17:21] <ogra_> look at yoour upstart job ... dbus/session is created by a profile.d snippet on login
[17:21] <ogra_> .cache/upstart/dbus-session that is
[17:22] <ogra_> pete-woods, hud starts before that i think
[17:22] <charles> larsu: does that trace look like a unity menumodel issue to you?
[17:22] <pete-woods> ogra_: this is on the desktop, I don't think that applies there, also this happens when you manually start and stop jobs
[17:23] <larsu> charles: very likely
[17:23] <larsu> charles: I mean, yes, it looks like that to me ;)
[17:23] <ogra_> pete-woods, ah, ok ... red herring then
[17:23]  * pete-woods hopes dashed :'(
[17:23] <charles> larsu: want me to assign it to you, then?
[17:24] <larsu> charles: yes please :)
[17:27] <charles> larsu, cool
[17:29] <slangasek> jodh: so the fact that valgrind does *not* report this as a leak implies that we do have references to it all somewhere, and that on exit upstart actually manages to clean it all up
[17:29] <slangasek> so I wonder where that is
[17:30] <nerochiaro> om26er: any idea why this would happen ? https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-mako/2180/console
[17:30] <om26er> nerochiaro, that's jenkins problem
[17:31] <slangasek> pete-woods, ogra_: where is the actual upstart job that reproduces this problem? the /usr/share/upstart/sessions/hud.conf I currently have on my phone is dbus-activated
[17:31] <nerochiaro> om26er: should i just retrigger ?
[17:31] <om26er> nerochiaro, does not look good. probably
[17:31] <ogra_> slangasek, http://paste.ubuntu.com/6187560/
[17:31] <om26er> nerochiaro, retrigger may help. its a bug in jenkins IMO
[17:31] <ogra_> slangasek, its the hud session upstart job
[17:31] <pete-woods> slangasek: this is on the desktop, see the e-mail chain with me and ted for more details
[17:31] <nerochiaro> om26er: shouldn't fginther know about it ?
[17:31] <nerochiaro> om26er: i retriggered
[17:31] <pete-woods> slangasek: but basically, you can create a new job /usr/share/upstart/sesions/foo.conf
[17:32] <slangasek> ogra_: but where is it, not what :)
[17:32] <om26er> yo fginther that ^ looks familiar ?
[17:32] <jappie> hello
[17:32] <pete-woods> slangasek: /usr/share/upstart/sessions/hud.conf is the one
[17:32] <fginther> nerochiaro, om26er, something looks bad
[17:33] <slangasek> pete-woods: that shows as dbus-activated here
[17:33] <pete-woods> slangasek: right, but you can manually start it for testing purposes (i.e. simply start hud)
[17:33] <slangasek> pete-woods: anyway, you're saying that if the setup goes wrong, it goes wrong for everything?  So e.g., 'initctl list-env | grep DBUS' would be empty post-login?
[17:33] <pete-woods> slangasek: no, list-env shows the right values
[17:34] <slangasek> ok
[17:34] <pete-woods> slangasek: but the jobs have the DBUS one missing
[17:34] <slangasek> so I should be able to reproduce this by creating a job and starting it?
[17:34] <pete-woods> slangasek: in that e-mail you're CC'ed in, there is all this detail (and more :) )
[17:34]  * slangasek goes to re-read
[17:34] <nerochiaro> fginther: om26er: have to EOD, I'll leave you guys to it
[17:35] <pete-woods> slangasek: yes, I created a new job pointing to a script that just does env|sort>/tmp.foo.txt
[17:35] <jappie> it is possible to put ubuntu on my htc wildfire s?
[17:35] <fginther> nerochiaro, ack, looks like that node is broken, will offline it
[17:35] <om26er> nerochiaro, bye, have fun.
[17:35] <slangasek> pete-woods: ok, reproduced
[17:36] <pete-woods> slangasek: :D
[17:36] <pete-woods> well, :D / :(
[17:36] <ogra_> slangasek, that hud activation has a very werid startup chain btw
[17:37] <mterry> pitti, I love your dial-number script, thanks :)
[17:37] <mterry> pitti, particularly for faking an incoming call
[17:37] <slangasek> ogra_: compared to what? :)
[17:37] <ogra_> slangasek, look at /usr/share/dbus-1/services/com.canonical.hud.service
[17:38] <ogra_> slangasek, then look at /usr/lib/arm-linux-gnueabihf/hud/dbus-activation-hack
[17:38] <slangasek> ok, but that's unrelated to this issue
[17:38] <ogra_> the upstart job is the last bit in that chain
[17:39] <ogra_> slangasek, is it ? we go through several different environments here
[17:39] <slangasek> yes, because it's reproducible with an unrelated job.
[17:40] <ogra_> ok
[17:40] <slangasek> and also, because upstart doesn't give a fig for any environment but its own. :)
[17:40] <ogra_> well, i was wondering if something got overwritten on the way ... but yeah, if thats the case
[17:46] <slangasek> pete-woods: ah; note the difference between 'initctl list-env' and 'initctl list-env -g'
[17:46] <slangasek> I would say that initctl set-env --global is not working as intended
[17:46] <slangasek> xnox: ^^ does that sound right to you?
[17:47] <jibel> alecu, added log file and screenshot to 1236807, from a freshly flashed device.
[17:48] <jibel> alecu, there are several "Error calling webservice: Web request failed: HTTP 4 Cannot connect to destination (search.apps.ubuntu.com)" at the end of the log
[17:49] <AskUbuntu> Porting to Nook HD | http://askubuntu.com/q/355450
[17:49] <AskUbuntu> it is possible to put ubuntu on my htc wildfire s? and how? | http://askubuntu.com/q/355453
[17:49] <pete-woods> slangasek: that's a good observation, didn't try the global switch myself
[17:56] <slangasek> pete-woods: ok, and it works if I run the 'initctl set-env -g' command post-login... curiouser and curiouser
[17:58] <pete-woods> slangasek: well I don't know the upstart codebase at all, so it could be anything
[17:58] <pete-woods> I have no idea about it's modes of operation of anything
[17:59] <pete-woods> I just know how to track a bug down
[18:00] <alecu> jibel: great, thanks! I'll take a look now.
[18:14] <dobey> what is "unity8-fake-env" ?
[18:15] <alecu> jibel: I managed to reproduce the bug, thanks for all the logs.
[18:15] <dobey> alecu: really? how?
[18:15] <racarr> ricmm: Saviq: So what's the
[18:16] <racarr> verdict on monitor channels and such
[18:16] <alecu> dobey: I connected the device to the wifi of a spare router, and unplugged the ethernet on it
[18:16] <davmor2> pmcgowan: I have an issue with wired headset in that it plays music through the headset but if I make a call that happens on the phone and I have to remove the headset to make the call.  Any ideas what I would file that against?  I'm guessing at pulse or ofono being the culprit
[18:17] <racarr> alf_: Thanks for digging deeper
[18:17] <dobey> alecu: weird. i was just looking at the code, and don't see how that would affect the installed apps search at all
[18:17] <racarr> on hold-surface-alive...
[18:17] <racarr> I think this will fix a lot of crashes :)
[18:17] <pmcgowan> davmor2, good question, make tasks for both and the guys can sort it
[18:18] <alecu> dobey: and now, I can't reproduce it any longer. Not great.
[18:18] <davmor2> pmcgowan: will do
[18:18] <dobey> heh
[18:19] <dobey> hrmm, and the unity8 in the latest image doesn't have that fix either
[18:22] <Saviq> ricmm, monitor channels? ;)
[18:22] <Saviq> racarr, ↑
[18:22] <Saviq> racarr, we seem to be good
[18:22] <pmcgowan> mterry, re bug 1234903 how do we did we used to do it on SF?
[18:22] <racarr> Saviq: for HUD and keyboard focus?
[18:22] <racarr> seems like it should be fine for both I guess
[18:23] <racarr> (still need the qtubuntu keymapping bit)
[18:23] <Saviq> racarr, yes, all works
[18:23] <Saviq> racarr, last related issue standing ricmm was fighting was was the keyboards InputArea
[18:23] <Saviq> racarr, that we didn't rotate on orientation changes
[18:23] <racarr> mm
[18:23] <sergiusens> mhall119, according to jdstrand apps shouldn't be using dconf http://paste.ubuntu.com/6210460/ (from https://jenkins.qa.ubuntu.com/job/saucy-touch_ro-maguro-smoke-ubuntu-clock-app-autopilot/115/artifact/clientlogs/application-click-com.ubuntu.clock_clock_0.7.log/*view*/)
[18:23] <mterry> pmcgowan, looks like there is a deprecated API we used
[18:24] <racarr> Saviq: Apparently we had this discussin before lol
[18:24] <racarr> and it just got lost...
[18:24] <racarr> https://code.launchpad.net/~robertcarr/mir/support-monitor-input-channels/+merge/176775
[18:24] <racarr> "This it to enable hooking the current Unity8 gesture recognizer up to Mir"
[18:24] <ricmm> :p
[18:25] <ricmm> so much for following up
[18:25] <ricmm> we should be generally good now... other than the keys propagating to input fields in apps
[18:25] <ricmm> but I'm not too sure about landing an event filter in unity-mir right now
[18:25] <ricmm> not really critical
[18:26] <racarr> which keys propagating to what?
[18:26] <racarr> when?
[18:26] <pmcgowan> mterry, ah, can we de-deprecate it?
[18:27] <mterry> pmcgowan, I think I have a solution
[18:29] <mhall119> sergiusens: jdstrand: why not?  isn't dconf the standard was for apps to store config data?
[18:29] <davmor2> pmcgowan: https://bugs.launchpad.net/pulseaudio/+bug/1236989
[18:30] <mhall119> sergiusens: though I don't see where the clock ap is actually using dconf
[18:30] <jdstrand> mhall119: not appstore apps. see https://wiki.ubuntu.com/SecurityTeam/Specifications/ApplicationConfinement#GSettings.2BAC8-dconf-1
[18:31] <jdstrand> mhall119: you can also see https://wiki.ubuntu.com/SecurityTeam/Specifications/ApplicationConfinement#GSettings.2BAC8-dconf for why
[18:48] <davmor2> pmcgowan: how the hell did you find that pulseaudio I looked though 8 pages and it wasn't listed :)
[18:49] <pmcgowan> davmor2, I am lp savant ;)
[18:50]  * davmor2 bows to pmcgowan 's lp-fu
[18:50] <pmcgowan> hah
[18:51] <alecu> jibel: what device are you using to test touch?
[18:55] <slangasek> pete-woods: well, interestingly, I can't reproduce the problem in a guest session at all.  I'll restart my main session (now that I've added a bit of debugging to upstart) and see what that gives me...
[18:56] <pete-woods> slangasek: I find I have to reboot a few times until it happens
[18:56] <pete-woods> yay for VMs
[18:56] <slangasek> pete-woods: oh, so it's not 100% consistent?
[18:56] <alecu> jibel: doh, it says so in the log: mako
[18:56] <pete-woods> slangasek: once it's happened, it's happened
[18:56] <slangasek> VMs> what magic configuration do you have that doesn't make compiz unbearably slow in a VM?
[18:57] <pete-woods> slangasek: er, I run Parallels on OSX :$
[18:57] <pete-woods> I think that's the setup for quite a few devs here
[18:57] <pete-woods> (Parallels has an excellent pass-through GLX driver)
[18:57] <slangasek> heh
[19:02] <slangasek> pete-woods: about what percentage of the time do you hit the error condition?
[19:02] <pete-woods> slangasek: I'm not really sure, to be honest
[19:05] <mfisch> mhall119: who is working on the video player or knows the state of it?
[19:05] <popey> mfisch: that would be jhodapp
[19:06] <jhodapp> mfisch, what part, the backend or the actual app?
[19:06] <mfisch> jhodapp: as part of customization we
[19:06] <mfisch> jhodapp: as part of customization we're dropping a movie trailer on the device (a free one), but right now I have no video when I play it. And no icon in the scope.
[19:06] <mfisch> jhodapp: so I was wondering if there's a codec I should/should not use
[19:07] <jhodapp> mfisch, what codec are you using?
[19:07] <jhodapp> mfisch, h264 is the most tested
[19:07] <jhodapp> mfisch, h264 for video, aac for audio, mp4 container
[19:07] <mfisch> jhodapp: I've got ogg and ogv now, but I'm not up to speed on containers/codec relationships
[19:08] <slangasek> pete-woods: I've tried > 10x now with a guest session, and can't reproduce it.  Should I have expected to hit the bug by now?
[19:08] <jhodapp> mfisch, ogv isn't going to be supported for hardware accelerated playback
[19:08] <pete-woods> slangasek: er, hmm, that's usually been enough for me
[19:08] <jhodapp> mfisch, because most of the hardware decoders don't support decoding ogv
[19:08] <mfisch> jhodapp: okay, let me try mp4
[19:09] <pete-woods> slangasek: I would say (very roughly) I'm getting it maybe 1 in 3 reboots
[19:09] <jhodapp> mfisch, the codecs are the formats the audio/video are encoded in...the container is the wrapper of those two streams of data along with metadata
[19:10] <slangasek> pete-woods: ok, I'll try a real user session; since there's clearly a timing thing here, maybe something about the guest session itself throws it off
[19:11] <mfisch> jhodapp: in general playing videos should work?
[19:11] <jhodapp> mfisch, absolutely
[19:11] <mfisch> jhodapp: or should I wait a week or so
[19:11] <mfisch> jhodapp: okay and what about icons in the unity scope?
[19:11] <mfisch> I have a blank icon with a 3d outline
[19:11] <jhodapp> mfisch, known bug, working on a fix atm
[19:11] <mfisch> jhodapp: perfect
[19:11] <mfisch> jhodapp: bug#?
[19:12]  * mfisch has mp4 playback
[19:12] <mfisch> the phone just locked itself during my movie :(
[19:13] <jhodapp> mfisch, bug #1236874
[19:13] <jhodapp> mfisch, yep, also a known bug
[19:13] <mfisch> jhodapp: thanks
[19:14] <tedg> mterry, Did you get your questions answered?
[19:14] <mterry> tedg, I think I'm OK now
[19:24] <dobey> olli: hey. when you hit this back button deleting the u1 account bug, what page were you looking at exactly when you hit the back button?
[19:26] <slangasek> pete-woods: so... can you actually still reproduce this bug with a completely up-to-date saucy system?
[19:27] <slangasek> pete-woods: because my already-running desktop session was affected by the bug... but I'm sure not having any luck reproducing it now
[19:27] <pete-woods> slangasek: yes, it was happening until I rebooted just now
[19:27] <pete-woods> slangasek: it is a complete pain to reproduce, which is why no-one until you has believed me it existed
[19:27] <slangasek> pete-woods: and you're certain there were no updates applied before that reboot?
[19:27] <pete-woods> slangasek: yes
[19:28] <slangasek> ok
[19:28] <slangasek> yeah, this is pretty painful
[19:32] <slangasek> pete-woods: was it always reproducible for you immediately after login?
[19:33] <pete-woods> slangasek: I'm fairly confident it was
[19:33] <slangasek> I wonder if my debugging patch is fiddling the timing just enough to reduce the chances of hitting the bug... seems unlikely, since upstart itself is single-threaded
[19:33] <slangasek> pete-woods: oh right, because it was affecting other dependent jobs immediately after startup. hmm.
[19:34] <pete-woods> slangasek: the easiest check on login is does hud work?
[19:36] <slangasek> pete-woods: hrm.  so this would affect the hud but not the dash?
[19:36] <slangasek> I don't remember whether I brought the hud up at all during my last session
[19:38] <olli> dobey, the accounts/settings overview, that lists the newly created u1 account
[19:38] <olli> dobey, this is on #87
[19:38] <dobey> mhr3: is the behavior of unity8 with respect to results_invalidated() documented well in API docs somewhere?
[19:38] <dobey> olli: from the main page, with the list of accounts?
[19:43] <ascii00> is the build for motorola razr dead?
[19:44] <olli> dobey, reproducing, 1 sec
[19:44] <olli> dobey, opening settings/account, seeing 1 U1 account
[19:45] <olli> tapping on the account
[19:45] <dobey> ah ok
[19:45] <olli> when in there, using back in anticipation of going to the accounts overview
[19:45] <olli> ending up there, previously listeds acount is gone
[19:46] <dobey> ok, thanks. mmcc will have a branch to fix that soon
[19:46] <dobey> just wanted to verify there wasn't another similar bug
[19:48] <SonikkuAmerica> Because the 2013 Nexus 7 isn't (yet) supported, what would happen if I tried to flash the current Ubuntu Touch image onto it?
[19:49] <dobey> it will fail
[19:49] <dobey> so outside of having rooted the device; basically nothing
[19:50] <dobey> at least, using phablet-flash. if you take a more manual approach, then i don't know.
[19:52] <SonikkuAmerica> Thanks. Someone should get on it though. Not me, however, I'm not a coder
[19:53] <mterry> boiko, do you know why QDesktopServices::openUrl caused crashes in your telephony-service branch?
[19:54] <boiko> mterry: I didn't dig too much into the details on that, but my assumption is that it is because QDesktopServices belongs to QtGui, and thus assumes the application is a QGuiApplication
[19:54] <mterry> boiko, huh, that's bogus
[19:54] <boiko> mterry: which is not the case for those daemons, they are all QCoreApplication instances
[19:54] <boiko> mterry: either that, or our implementation assumes that
[19:55] <stefan__> I get an error while doing phablet-dev-bootstrap : error: could not verify the tag 'v1.12.4', gpg public key not found
[19:55] <stefan__> why that?
[19:55] <boiko> mterry: I ended up using the platform-api directly
[19:55] <mterry> boiko, but your other branches just use QDesktopServices::openUrl
[19:55] <mterry> er
[19:55] <mterry> Qt.openUrlExternally
[19:55] <mterry> rather
[19:55] <mterry> boiko, could telepathy-services use that?
[19:55] <boiko> mterry: yes, those are UI applications running on top of a QGuiApplication
[19:56] <mterry> boiko, ok, so Qt.openUrlExternally ends up using QDesktopServices::openUrl?  bummer
[19:56] <boiko> mterry: I would assume it does
[19:57] <boiko> mterry: maybe the platform plugin is not loaded on QCoreApplications? who knows?
[19:57] <boiko> mterry: I plan to investigate that, but not right now :/
[19:58] <mterry> boiko, fair
[19:58] <mterry> boiko, I fixed the greeter side of that bug, I think
[19:58] <boiko> mterry: oh. nice to know
[19:59] <mterry> boiko, you mentioned some problems with bringing apps to focus?
[19:59] <mterry> boiko, is there a branch that fixes that somewhere?
[20:00] <boiko> bfiller: ^
[20:00] <boiko> bfiller: do you know which branch fixes that?
[20:01] <bfiller> mterry: upstart-app-launch and unity-mir from daily build ppa, tedg and ricmm would know if others needed as well
[20:02] <mterry> bfiller, so that's just trunk then, no special branch?
[20:02] <bfiller> mterry: yes
[20:02] <mterry> bfiller, awesome, thanks!
[20:03] <bfiller> no special branch that I'm aware of
[20:03] <mhr3> dobey, calling it emits a signal that gets picked up by the dash and if the dash page of that (sub)scope is active new search() request is sent to the scope
[20:05] <robert_ancell> mhall119, hey, not sure if cjwatson contacted you but I was just attempting to clickify my apps - the main question I have is what do I put in the .desktop file for the Exec line (given I don't know exactly where the package files will be installed to). Also, how does click associate the icon in the .desktop file with the icon file?
[20:06] <mhall119> robert_ancell: are the C++ or QML apps?
[20:06] <robert_ancell> mhall119, straight QML
[20:06] <mhall119> then you should be able to use Exec=qmlscene foo.qml
[20:06] <robert_ancell> mhall119, so the Exec line is run from inside the install directory?
[20:06] <dobey> mhr3: and if it's not active? does it queue the query to wait until it is?
[20:06] <mhall119> IIRC, click will add a Path= line to the .desktop pointing to the install location
[20:07] <robert_ancell> ah
[20:07] <mhr3> dobey, right
[20:07] <mhall119> robert_ancell: and that should work for both Exec and Icon
[20:07] <dobey> mhr3: so i can avoid calling the internal search method in the timeout, and just rely on the search being called again when i call results_invalidated()?
[20:08] <robert_ancell> mhall119, I installed on my desktop box - is that a supported case? It didn't seem to modify the .desktop file or install it anywhere
[20:08] <mhall119> robert_ancell: you'd have to ask cjwatson if and how it works on desktop
[20:08] <robert_ancell> mhall119, ok, thanks
[20:08] <mhall119> robert_ancell: just make sure you have the desktop hook listed in your click manifest
[20:08] <robert_ancell> mhall119, also, is there a way for click not to install everything in the project directory?
[20:09] <robert_ancell> mhall119, yep, I've got that
[20:09] <mhall119> I don't think so, no
[20:09] <mhall119> with the exception of the click package hooks
[20:09] <mhr3> dobey, why is it even called inside search() itself?
[20:09] <cjwatson> robert_ancell: it's not expected that any of this will work on desktop yet
[20:10] <cjwatson> robert_ancell: I officially Do Not Care for 13.10
[20:10] <robert_ancell> cjwatson, k
[20:10] <robert_ancell> :)
[20:10] <cjwatson> I expect I will care for future releases
[20:10] <cjwatson> But it's not in scope right now
[20:11] <pete-woods> slangasek: just checking, but is it possible that the "fix" that's being applied the hud.conf, etc is masking the problem from re-occuring for you? (assuming you have the modified hud.conf)
[20:11] <dobey> mhr3: because we need to wait to do the query, if the network is not accessible
[20:12] <cjwatson> robert_ancell: (it may be possible to cobble it together, and feel free to try and take notes, but don't expect me to have cycles for it until at least November)
[20:12] <robert_ancell> cjwatson, no problem, I assumed it wouldn't be supported, was just trying to test my packages locally
[20:13] <mhr3> dobey, then the query failed and what you should do is run a timer which will emit the invalidation later
[20:13] <mhr3> dobey, but even that shouldn't be necessary once dash issues requests when network state changes
[20:15] <dobey> mhr3: well, for things that only do network stuff, waiting for the dash to tell it to do something might be useful
[20:15] <dobey> but the click scope has to deal with network searches and local searches both
[20:16] <dobey> although, that could also be separated a bit better now, given the other scope thing you landed yesterday, which is local only
[20:17] <mhr3> dobey, doesn't change the fact that it needs to be requeried when you get network
[20:18] <dobey> mhr3: the network part does, yes. the local part doesn't
[20:18] <mhr3> dobey, sure, but it's just one scope...
[20:18] <dobey> when it gets network it should only the do the work it needs to do, not all the work it is capable of doing
[20:18] <mhr3> then it should have been two scopes
[20:19] <dobey> yes, it's just one. imagine 100 scopes all doing stuff at the same time when network comes up
[20:19] <mhr3> but that doesn't really matter
[20:19] <dobey> whatever, i don't wnat to argue about the architecture right now
[20:19] <mhr3> dobey, you're forgetting that only the visible ones will be requeried ;)
[20:20] <dobey> i'm not going to rewrite the whole scope to resolve an issue i can fix with changing 6 lines
[20:20] <mhr3> dobey, i'm not asking you to, i'm describing what should ideally happen
[20:21] <mhr3> but we all know ideal work never meets deadlines :)
[20:26] <mterry> boiko, what about my comment in https://code.launchpad.net/~boiko/messaging-app/uri_handler/+merge/189428 ?
[20:26] <mterry> about removing functionality
[20:27] <boiko> mterry: ah yes, sorry, I didn't see that comment, I am working on something else right now
[20:27] <mterry> boiko, no worires, I was just going to approve the branches, but just wanted to confirm that the removal was OK
[20:27] <boiko> mterry: those options are indeed not used anymore. They are leftovers from previous versions of phone-app
[20:27] <mterry> cool
[20:28] <boiko> mterry: ok, don't approve them yet
[20:28] <boiko> mterry: we want to give them some more testing before having them merged
[20:29] <boiko> mterry: (as the current implementation is working well on surfaceflinger we don't want to introduce regressions)
[20:29] <boiko> mterry: in any case, good to know everything is working :D
[20:48] <tedg> sergiusens, Do you have all these click packages installed?  Can you get me the app names?
[20:51] <robert_ancell> jibel, did you want to review https://code.launchpad.net/~robert-ancell/lightdm/handle-no-xrdb/+merge/189971
[20:52] <slangasek> pete-woods: I haven't modified hud.conf, no
[20:52] <pete-woods> slangasek: never mind then :/ (I am seriously worried about this bug, btw)
[20:53] <slangasek> pete-woods: and I'm checking this at the source ('initctl list-env -g'), so it wouldn't have been masked anyway
[20:53] <slangasek> I'm rolling back to the pristine upstart package now to see if I can finally catch it
[20:53] <pete-woods> slangasek: I mean, masked in the sense that the hud job (or something else) was changed in a way to successfully inject the environment variable
[20:54] <sergiusens> tedg, which ones are 'these'?
[20:54] <slangasek> pete-woods: right, nothing else is injecting the vars into the global env
[20:54] <tedg> sergiusens, The ones that we need to add the the url dispatcher.  Sorry, "these" from the list in your e-mail :-)
[20:54] <sergiusens> tedg, one sec, I'm flashing fresh
[20:55] <pete-woods> slangasek: ah well, there's not much I can suggest for this, perhaps a repeatedly run autopilot test or something?
[20:55] <slangasek> pete-woods: I have yet to successfully run an autopilot test; I don't suppose you could orient me?
[20:56] <tedg> thomi, slangasek needs schooling ^
[20:56] <tedg> :-)
[20:56] <sergiusens> tedg, full app_ids or just pkgname?
[20:56] <pete-woods> slangasek: I've only written one myself, I figured you might be the expert
[20:56] <slangasek> pete-woods: not my department :-)
[20:56] <pete-woods> slangasek: basically if requires you to be in the right dir (it doesn't search your tree or anything)
[20:56] <thomi> hmmm?
[20:56] <tedg> sergiusens, pkg and app name.  If you've got full app_ids I can subtract the version.
[20:56] <slangasek> I'm happy to learn, but last time I gave this a try on mako I was told "it should just work from adb" and that was double-plus untrue
[20:56] <pete-woods> I've only run them on desktop too
[20:57] <thomi> slangasek: the runner works just like other python test runners... you can even use testtools.run or whatever if you like
[20:57] <thomi> the difficulty at the moment is that running them on top of unity8+mir is a little delicate
[20:58] <slangasek> thomi: sorry, but that tells me approximately nothing.  The last time I attempted this, I tried to figure out the commands to run based on looking at the jenkins job, and that was full of fail.  Is there documentation that describes the one true way to launch autopilot tests?
[20:58] <slangasek> (because there should be - one true way, and documentation of it)
[20:59] <thomi> I agree. What I'm saying is that there's more than one concern there. If you've already got the test suites locally, it's easy: autopilot run [-v] my.test.id.here
[20:59] <slangasek> thomi: pretty sure that's the command I ran; pretty sure it didn't work when I tried it over adb
[21:00] <slangasek> maybe something's been fixed since
[21:00] <sergiusens> tedg, adb shell ls /home/phablet/.local/share/applications | sed -e 's/.desktop//' |pastebinit -> http://paste.ubuntu.com/6211066/
[21:00] <thomi> Right - sthe other part is the environment setup, which can be tricky, especially if you're running mir. For example, you need to be the phablet user, unity8 needs to be running, and you need to have the screen unlocked
[21:01] <sergiusens> tedg, need to install gallery and camera, but names are obvious com.ubuntu.camera_camera and com.ubuntu.gallery_gallery
[21:01] <tedg> sergiusens, Cool, nice that we've got those standardized now.
[21:01] <tedg> sergiusens, Thanks!
[21:03] <sergiusens> np
[21:04] <mhall119> sergiusens: did you find where/how the clock app is trying to use dconf?
[21:05] <sergiusens> mhall119, still need to strace it, but my guess is timezone config perhaps?
[21:08] <tedg> sergiusens, So addressbook is staying deb?
[21:09] <mhall119> sergiusens: hmm, maybe, he had a C++ plugin I think to read tzdata
[21:09] <sergiusens> tedg, yeah, the tres amigos are staying as deb
[21:09] <mhall119> which would explain why I didn't see any dconf stuff in the qml project branch
[21:09] <tedg> sergiusens, Heh, the code names are getting spicy!
[21:10] <mhall119> nik90: ping
[21:10] <sergiusens> mhall119, if you have the deb just run a dbus monitor against it
[21:10] <sergiusens> tedg, long days have that effect :-)
[21:10] <mhall119> sergiusens: it'll be faster for me to ping the dev :)
[21:11] <tedg> sergiusens, mediaplayer ?
[21:11] <sergiusens> tedg, right, didn't build that one yet, but will be com.ubuntu.mediaplayer_mediaplayer
[21:11] <mhall119> why the double name?
[21:12] <tedg> sergiusens, ack
[21:12] <sergiusens> mhall119, it's not a double name ;-)
[21:12]  * mhall119 needs to have his vision checked then
[21:12] <tedg> mhall119, it's not a double name ;-)
[21:12] <sergiusens> mhall119, pkgname and appname; it's a mere coincidence
[21:12] <mhall119> now I'm seeing double replies
[21:12] <sergiusens> lol
[21:13] <mhall119> y'all make my brain hurt
[21:17] <slangasek> pete-woods: have you ever reproduced the bug with a logout/login?  Or only ever in the first login after a reboot?  (trying to rule out some sort of race condition related to the system init)
[21:17] <pete-woods> slangasek: I've always rebooted my VM
[21:18] <pete-woods> slangasek: to be honest I'd kinda given up trying to reproduce it, then it happened again
[21:18] <pete-woods> so I copied down all that debug info you have in the e-mail
[21:20] <slangasek> pete-woods: ah.  "given up trying to reproduce it" sounds a little less than 1 in 3 frequency, maybe?
[21:21] <pete-woods> slangasek: it seemed to just go away for a while, then it was happening every login
[21:21] <slangasek> ok
[21:21] <marrabld> Hello, is it possible to get cellular data on the Nexus 4 with Ubuntu touch?  I tried this link but it was vague and didn't seem to  work.  http://theravingrick.blogspot.co.uk/2013/06/dog-fooding-success-sort-of.html
[21:22] <lool> stgraber: Hey mind helping me debug why my job is blocking startup?
[21:23] <lool> stgraber: the boot-hook ran
[21:23] <slangasek> marrabld: sure, that works out of the box with current images
[21:23] <lool> stgraber: but then lightdm never came up, it's start/starting
[21:23] <lool> stgraber: this is the final version
[21:23] <lool> http://paste.ubuntu.com/6211148/
[21:23] <marrabld> mmm. I am pretty sure I have the latest image but I cna't get it to work
[21:23] <slangasek> (in fact, it works so well I accidentally blew my monthly data cap with mine while testing image-based updates over 4G ;)
[21:24] <stgraber> lool: initctl list | grep <your job name here>
[21:24] <lool> stgraber: boot-hooks/set-display-mir stop/waiting
[21:24] <lool> stgraber: the files were touch-ed as expected
[21:25] <slangasek> marrabld: check System Settings -> Cellular?
[21:25] <stgraber> lool: initctl list | grep boot-hooks
[21:25] <lool> stgraber: boot-hooks-emit stop/waiting
[21:26] <lool> and: lxc-android-config start/running, process 634
[21:26] <Chocanto> mhall119: Hey ! :) Do you know if I can set arguments by using "Run application on Device" on QtCreator ? I'm finishing zoom feature
[21:26] <stgraber> lool: (unrelated but note that you're supposed to place your hook in /etc/init/boot-hooks)
[21:26] <lool> stgraber: I did
[21:27] <stgraber> lool: oh yeah, sorry, you were clever filtering the output of the commands I gave you :)
[21:27] <marrabld> I don't have System Settings -> Cellular.  But I do have Network-> ceullular-> Cellular data.  And it is ticked, and so it Data roaming .  But, I can't get any data on my browser
[21:27] <lool> stgraber: I greped for set-display-mir
[21:28] <slangasek> marrabld: and it lists your carrier?
[21:28] <marrabld> I get network error.
[21:28] <stgraber> lool: I think I know what's wrong, let me go read the doc to confirm
[21:28] <marrabld> Yes, I am with Virgin in the UK . Does that matter?
[21:28] <marrabld> and it lists it
[21:28] <marrabld> I can make calls text etc
[21:28] <slangasek> marrabld: right, just asking to make sure you have *any* cell service. :
[21:28] <slangasek> :-)
[21:28] <marrabld> yep.  no probs
[21:29] <slangasek> I don't know why that wouldn't be working with the current images
[21:29] <marrabld> Any ideas how to debug.  is there a /var/log eqivilent ?
[21:29] <stgraber> lool: try removing the "task" stanza
[21:30] <Chocanto> Hey, does anyone here know if we can send arguments to the command when we test our app on a device using QtCreator ?
[21:30] <lool> stgraber: rebooting
[21:30] <slangasek> marrabld: there are a series of scripts in /usr/share/ofono/scripts (IIRC - typing this blind) that you can use to probe the current state of the cell data stack
[21:30] <lool> with this change
[21:30] <Chocanto> Maybe someone from the music-app team know it ?
[21:30] <tedg> sergiusens, Not sure when this should land, but it's here when things align.  https://code.launchpad.net/~ted/url-dispatcher/click-core-apps/+merge/189976
[21:31] <stgraber> lool: my guess here is that the job first started because it got boot-hooks + starting lxc-android-config
[21:31] <stgraber> lool: then the state was reset and starting lightdm got emitted
[21:31] <sergiusens> tedg, thanks, from what I hear, after mir
[21:31] <stgraber> lool: at that point the job is now waiting for another boot-hooks event to be emitted to run a second time and unblock lightdm
[21:31] <mhall119> Chocanto: I don't know, bzoltan might be able to tell you though
[21:31] <Chocanto> mhall119: Ok, thank you :)
[21:31] <lool> stgraber: ah forgot to reset build-id
[21:32] <lool> stgraber: now lxc-android-config start/starting
[21:32] <stgraber> lool: dropping the task statement may fix that, if it doesn't the easiest way is still to keep the task statement out of it and just merge everything into the pre-start, that way the job will be considered as running once it's execute and won't attempt to run a second time
[21:32] <lool> stgraber: I think it's stuck even earlier now
[21:32] <lool> boot-hooks/set-display-mir stop/waiting
[21:32] <lool> boot-hooks-emit stop/waiting
[21:32] <stgraber> lool: try moving the content of your script section to the pre-start
[21:33] <lool> hmmmm
[21:33] <mterry> boiko, still there?  I have a fix for your addressbook branch
[21:33] <boiko> mterry: that's actually renato's branch, but I think he is still here :)
[21:34] <lool> stgraber: I've reset build to 0, I'm without task and I've moved everything to pre-satrt
[21:34] <lool> rebooting
[21:34] <mterry> oh so it is
[21:34] <boiko> mterry: the other two apps I fixed, but this one renato had already the changes made
[21:34] <mterry> boiko, I don't see him on.  I'll comment in MR
[21:34] <lool> stgraber: boot-hooks/set-display-mir start/running
[21:34] <lool> boot-hooks-emit stop/waiting
[21:34] <lool> stgraber: it did run
[21:35] <lool> I thnk it's still booting
[21:36] <boiko> mterry: ok., thanks
[21:36] <lool> stgraber: I think it's stuck again
[21:36] <lool> lightdm start/starting
[21:37] <stgraber> annoying
[21:37] <stgraber> slangasek: suggestions? ^
[21:39] <lool> stgraber: Hmm there should be a dep between lxc-android-config and lightdm
[21:39] <stgraber> lool: can you confirm you don't have any errors in /var/log/upstart?
[21:40] <lool> stgraber: no error
[21:40] <lool> stgraber: only left over messages from my earlier debugging  in /var/log/upstart/boot-hooks_set-display-mir.log
[21:40] <sergiusens> jdstrand, hey regarding the click-reviewer-tools, it's saying icon is not a supported entry
[21:40] <marrabld> slangasek: Thanks looking now.
[21:41] <jdstrand> sergiusens: can you show me the desktop file and the error?
[21:41] <sergiusens> jdstrand, it's not desktop file, it's manifest
[21:41] <jdstrand> oh
[21:41] <jdstrand> we have a new hook or something?
[21:41] <sergiusens> jdstrand, http://paste.ubuntu.com/6211240/
[21:41] <sergiusens> jdstrand, http://click.readthedocs.org/en/latest/file-format.html
[21:42] <lool> stgraber: oddly, I see no dep between lxc-android-config.conf and lightdm.conf
[21:42] <sergiusens> jdstrand, package icon (instead of app icon which can be n+1)
[21:42] <sergiusens> jdstrand, icon: icon to display in interfaces listing click packages; if the name refers to an existing file when resolved relative to the base directory of the package, the given file will be used; if not, the algorithm described in the Icon Theme Specification will be used to locate the icon
[21:43] <jdstrand> sergiusens: ok, I'll update the tests
[21:43] <slangasek> pete-woods: oho, here's something interesting.  after a suspend and resume, it appears that the envvars have disappeared from my session.
[21:43] <sergiusens> jdstrand, sounds good
[21:43] <slangasek> pete-woods: no evidence that init itself has restarted, however
[21:43] <sergiusens> beuno, ^^
[21:43] <stgraber> lool: so I'm pretty sure the problem is with waiting on (starting A or starting B) since upstart will block the start of A and B and only unblock those if the start condition matches
[21:44] <stgraber> lool: I'm not sure if we ever had something else try to do that
[21:44] <lool> stgraber: Yeha, that's also what I think but I can't think of a way to do differently
[21:44] <beuno> sergiusens, cool, so pull from that when jdstrand lands it, and ping me when you get a pass from it and I'll approve
[21:45] <slangasek> stgraber: lightdm is start/starting, with no pid?  is something 'start on starting lightdm' and blocking it?
[21:45] <lool> stgraber: I could have two jobs
[21:45] <lool> slangasek: likely that hook job
[21:45] <lool> slangasek: http://paste.ubuntu.com/6211148/
[21:45] <slangasek> lool: which job exactly?
[21:45] <stgraber> slangasek: so lool needs a hook that's "start on boot-hooks WHEN=new-version and (starting lightdm or starting lxc-android-config)"
[21:45] <lool> slangasek: we tried without task
[21:45] <lool> slangasek: and we tried with everything in pre-start
[21:46] <stgraber> slangasek: currently he gets the hook executed as expected and one of lightdm or lxc-android-config is allowed to start, the other is stuck in starting
[21:46] <stgraber> I think that makes sense, I'm just not sure what's the right way to achieve lool's goal
[21:46] <lool> stgraber: I think two jobs would be safer
[21:47] <slangasek> so taking a slightly different tack, why is it lightdm or lxc-android-config?  Doesn't lightdm start strictly after lxc-android-config?
[21:47] <lool> stgraber: I could also try (A and B) or (A and C)
[21:47] <slangasek> two jobs also doesn't solve it
[21:47] <lool> slangasek: well that's whatI just thought too but I didn't see any deps between the two
[21:47] <stgraber> two jobs would solve it by having the code simply trigger twice
[21:48] <slangasek> because in the case that boot-hooks WHEN=new-version is /not/ emitted at boot, this will still deadlock
[21:48] <slangasek> stgraber: nope
[21:48] <stgraber> slangasek: ah true
[21:48] <lool> right, I didn't get how it was supposed to work on second boot
[21:49] <slangasek> so the best idea I have so far seems rather hackish, but
[21:49] <lool> I can have a job starting on lightdm, another starting on lxc-android-config, both checking whether they need to do their thing by checking a flag set by a boot-hook job
[21:49] <lool> but that's fairly convoluted
[21:49] <stgraber> slangasek: that's actually making it a bit problematic for any hook that needs to start before another service... I guess they then have to do something like "start on boot-hooks and starting service" and then check $WHEN in pre-start...
[21:49] <slangasek> stgraber: yes, that may be simplest
[21:50] <marrabld> slangasek: There are quite a few scripts in /usr/share/ofono/scripts .  None of them obviously useful to me.  Any ideas on what might be useful? Cheers.
[21:50] <slangasek> stgraber: and then the job needs to be left in state 'started' afterwards, to not block any restarts of service
[21:50] <slangasek> (which means: not task, do the work in pre-start)
[21:51] <stgraber> slangasek: right, that's the first thing I told lool to do, drop task, move everything to pre-start.
[21:51] <slangasek> you do also need it split into two jobs
[21:51] <slangasek> because otherwise you'll only delay startup of *one* of the services
[21:52] <stgraber> slangasek: so I guess in this case, there's two ways of doing it, either have two mostly identical jobs with "start on boot-hooks and starting <service>" which each check $WHEN in pre-start for new-version and DTRT based on that and stamp file. Or have a single job and make lightdm somehow depend on android.
[21:52] <slangasek> stgraber: I think the former is clearly preferable, architecturally
[21:53] <slangasek> unless lightdm does technically have an undeclared dependency on android anyway
[21:53] <lool> slangasek: yeah, I think that's the case
[21:53] <slangasek> marrabld: I'm trying to pull up the package now to see, since I'm not in a position to look at my phone right now
[21:53] <stgraber> well, lightdm spawns mir which needs the binary drivers that android has
[21:53] <lool> ah actually only on SF I guess
[21:53] <slangasek> lool: which case?  a dependency that's undeclared?
[21:53] <lool> since with Mir, we can actually display without SF
[21:53] <lool> slangasek: yeah
[21:54] <slangasek> lool: even in mir we need the drivers though, and the driver init is done by the android container
[21:54] <stgraber> lool: but can we run Mir without access to the Android bits and without ueventd?
[21:54] <lool> ok
[21:54] <lool> what's the best way to fix this missing dep?
[21:54] <lool> lxc-android-config start on starting lightdm?
[21:55] <stgraber> so it may be better to make lightdm on touch depend on android and then go with a single job that's "start on boot-hooks and starting lightdm" + check of WHEN in pre-start
[21:55] <marrabld> slangasek: oh, I appreciate your help, if it's its a pain in the  .. don't stress.    Its just the data thing is the only thing stopping me using it a my every day phone.
[21:55] <marrabld> that and no spotify app yet. :)
[21:55] <stgraber> lool: nope you don't want that, otherwise "restart lightdm" will try to restart lxc
[21:55] <slangasek> lool: no, better to put the dependency in the other direction and change lightdm's start condition with an override
[21:55] <lool> stgraber: another question, is it ok to "stop" the job in pre-start?
[21:55] <stgraber> lool: I think we want lightdm "start on android"
[21:55] <marrabld> slangaske: its pretty cool I can use these scripts to make phone calls via adb on my computer.  :-)
[21:55] <slangasek> lool: no, you need to definitely *not* stop the job in prestart
[21:56] <lool> slangasek: right, that's what I understood
[21:56] <lool> so I need to guard the whole pre-start
[21:56] <lool> so many pitfalls
[21:57] <itmannen> Testing the connection from Sweden
[21:57] <lool> stgraber, slangasek: Something like http://paste.ubuntu.com/6211296/
[21:58] <lool> slangasek: oh there's an override file for lightdm I hadn't seen
[21:58] <lool> but it doesn't wait on android
[21:58] <lool> start on ((filesystem
[21:58] <lool>            and runlevel [!06]
[21:58] <lool>            and started dbus)
[21:58] <lool>           or runlevel PREVLEVEL=S)
[21:58] <stgraber> slangasek: does stopping a job that's "start on starting a" really cause "a" to fail to start? I thought you were supposed to do "stop $SOMEVARIABLE" for that (or exit 1)?
[21:58] <lool> just removes the plymouth stuff
[21:58] <lool> sorry forgot WHEN in the paste
[21:59] <lool> http://paste.ubuntu.com/6211299/
[21:59] <stgraber> slangasek: http://upstart.ubuntu.com/cookbook/#stop-a-job-that-is-about-to-start
[21:59] <itmannen> exit
[22:03] <stgraber> slangasek, lool: I've unfortunately got to run now, be back later tonight.
[22:03] <jdstrand> sergiusens: ok, pushed in r108
[22:04] <lool> slangasek: (Cc: stgraber) https://code.launchpad.net/~lool/session-manager-touch/lightdm-depends-on-lxc-android-config/+merge/189981
[22:04] <lool> slangasek: just boot tested this on Mir
[22:04] <lool> err SF
[22:04] <lool> since I had removed .display-mir for testing
[22:09] <sergiusens> jdstrand, hey, so why is Path blacklisted? the app hook replaces this correctly
[22:10] <jdstrand> sergiusens: can you show me the error?
[22:10] <sergiusens> jdstrand, http://paste.ubuntu.com/6211325/
[22:11] <lool> slangasek: boot tested on Mir as well
[22:12] <jdstrand> sergiusens: I suppose I could turn that into a warning or info. why would an app specify it?
[22:12] <jdstrand> sergiusens: in its source
[22:12] <sergiusens> jdstrand, because it also builds as a deb (it's originally hardcoded, let me just template it)
[22:13] <jdstrand> sergiusens: sure, I understand your reasoning, I was just thinking about normal developers
[22:13] <jdstrand> sergiusens: who aren't trying to support two things in one
[22:14] <sergiusens> jdstrand, yeah, normal yeah, are good :-)
[22:14] <thomi> who should I be talking to if I want to better understand the ARM calling convention?
[22:14] <jdstrand> sergiusens: so, are you fixing it?
[22:15] <sergiusens> jdstrand, easy fix is to leave an empty line, going to see if the tool complains about that
[22:17] <xnox> slangasek: pete-woods: i was trying to reproduce "dbus env not set" with system-image / ro, on grouper (which does not have mir). BTW pete-woods, you can create .override file and/or .conf file in ~/.config/upstart/. E.g. "echo pre-start exec env" > ~/.config/hud.conf to go back to "hud without pre-start" & env debugging.
[22:19] <pete-woods> xnox: thanks for the tip
[22:21] <nik90> mhall119: pong
[22:21] <nik90> mhall119: sry really late
[22:26] <lool> stgraber: I think there's a bug in the boot-hooks; WHEN is always new-version because of the && var=value thing being in a subshell
[22:27] <racarr> Is there a way to stop the nexus 4 wireless from getting so slow when the screen turns off?
[22:27] <slangasek> stgraber: 'start on starting a' does not cause 'a' to fail to start if 'a' stops.  The problem here is that the job isn't 'start on starting a', it's 'start on starting a and foo', where foo is a one-time boot event, so if you stop this job, it will cause a /restart/ of a to hang
[22:29] <slangasek> lool: pastebin> this still needs to not be 'start on boot-hooks WHEN=new-version [...]', it needs to be 'start on boot-hooks' with a check for the value of WHEN that skips making the change when WHEN != new-version
[22:30] <lool> slangasek: yeah, I pasted a second one minutes later: http://paste.ubuntu.com/6211299/
[22:30] <slangasek> lool: the lxc-android-config change looks sane to me
[22:30] <slangasek> lool: ah, looking
[22:31] <marrabld> I can't get data over network to work so I am trying to restore my android backup.  I have tried the old 'adb restore ./backup.ab'.   But I get the message  "Now unlock your device and confirm the restore operation." on the terminal.  but nothing happens on my phone
[22:31] <marrabld> Any help would be appreciated.
[22:31] <marrabld> not sure what to try
[22:31] <slangasek> lool: yeah, that version looks sane to e
[22:32] <slangasek> marrabld: I believe the link in the channel topic includes instructions on how to restore android
[22:33] <slangasek> xnox: so the fact that my initctl list-env -g is different on my laptop after a suspend/resume is really amazing
[22:34] <sergiusens> jdstrand, what about "lint_maintainer_domain": "(MANUAL REVIEW) email domain too long 'ubuntu-devel-discuss@lists.ubuntu.com' for app name 'com.ubuntu.notes'" ?
[22:34] <pete-woods> ted: I'm EOD'ing now
[22:34] <pete-woods> ted: stuff you need to know - we need this MR in unity-mir (https://code.launchpad.net/~pete-woods/unity-mir/window-stack-get-property/+merge/189984)
[22:34] <marrabld> slangasek: okay, please excuse my ignorance but I am not sure which one.  Ubuntu Touch Support & Discussion | Home: http://bit.ly/YEqEfo | Porting guide (advanced) https://wiki.ubuntu.com/Touch/Porting | File bugs at https://bugs.launchpad.net/ubuntu/+filebug | Under active development! | Nexus Status: http://bit.ly/18kIrhM | Dashboard: http://bit.ly/12AQV53 | Changelog: http://bit.ly/18xvTAA
[22:35] <marrabld> I followed the ones on https://wiki.ubuntu.com/Touch/Install wihtout success
[22:35] <slangasek> marrabld: right, that's exactly the page I meant.  I don't know why adb talking to Ubuntu Touch would prompt you for unlocking
[22:35] <pete-woods> ted: I think the hud watchdog stuff was nonsense (I'd stopped the widow bridge service)
[22:36] <xnox> slangasek: =/ that is scary. is it because our memory was not restored, or jobs stop & restart. Note that e.g. dbus job clobbers environment each time it starts.
[22:36] <slangasek> marrabld: you may need to do a factory reset first, and then restore
[22:36] <marrabld> slangskek.  Okay, thanks Ill try that
[22:37] <slangasek> xnox: "memory was not restored" - after a suspend/resume?  I wouldn't expect such a specific problem, I would expect general crashiness :-)
[22:39] <lool> slangasek: ah
[22:39] <lool> slangasek: I think my job is not run when the second boot-hook is emitted
[22:39] <lool> slangasek: because it is already running once the every-boot ones come in
[22:39] <jdstrand> sergiusens: so, that is correct, but we could special case that email address, in theory
[22:40] <jdstrand> sergiusens: since the appname is com.ubuntu.notes, it is expecting <foo>@ubuntu.com
[22:40] <jdstrand> sergiusens: s/appname/pkgname/
[22:41] <jdstrand> sergiusens: I'd like to think about that
[22:41] <sergiusens> jdstrand, ack, can we let it in for now as is and change later? beuno ?
[22:42] <jdstrand> sergiusens: sure-- that is why it says MANUAL REVIEW. a human just needs to make the call
[22:42] <jdstrand> beuno: ^
[22:43] <sergiusens> jdstrand, good :-)
[22:43] <jdstrand> sergiusens: is that really accurate though for the email?
[22:43] <jdstrand> sergiusens: I mean, a coredev or ubuntu member isn't able to perform the upload
[22:43] <xnox> slangasek: catching up on bug mail. So in the upstart manpages there are vague notes that "set-env -g" might be lost by jobs transitioning from pre-start to spawned.
[22:43]  * beuno nods
[22:43] <sergiusens> jdstrand, it's the original debian/control email
[22:43] <jdstrand> sergiusens: right, and that works for the archive
[22:43] <xnox> slangasek: and maybe making xsession-init start on started dbus, might not be such a bad idea for 13.10.
[22:43] <sergiusens> jdstrand, I would hope that when the multiple users thing comes in we would allow that
[22:44] <jdstrand> indeed
[22:44] <beuno> jdstrand, only one person can claim com.ubuntu, so the email issue will be mostly moot in this case
[22:44] <jdstrand> it isn't in place now though, so I don't know how to adjust the script
[22:44] <beuno> of course, we'll get the message each time for now
[22:44] <beuno> until we land autmated checks and approved permanent exceptions
[22:44] <jdstrand> com.ubuntu is already considered a special domain
[22:45] <jdstrand> honestly, the email/domain handling is quite difficult to automate
[22:45] <beuno> it is
[22:45] <beuno> I have an item on my ToDo to revise it
[22:45] <beuno> maybe we don't allow custom domains
[22:45] <jdstrand> so it is trying a few things, with the understanding that a human will have to get involved in certain circumstances
[22:45] <beuno> I mean, we don't really get anything from it
[22:46] <sergiusens> jdstrand, fwiw, these are the fixes https://code.launchpad.net/~sergiusens/notes-app/click-fixes/+merge/189980 if you want to ack it
[22:46] <sergiusens> beuno, custom as in not com.ubuntu.developer. ?
[22:47] <beuno> sergiusens, yes
[22:47] <beuno> proving you own a domain in an automated way is a pita, really
[22:48] <jdstrand> man, I don't think I understand cmake at all
[22:48] <jdstrand> if(NOT CLICK_MODE)
[22:48] <jdstrand> else(NOT CLICK_MODE)
[22:48] <jdstrand> endif(NOT CLICK_MODE)
[22:48] <jdstrand> sergiusens: is that correct? ^
[22:48] <xnox> jdstrand: looks wonderful.
[22:48] <sergiusens> jdstrand, the else with params is just to match the original if
[22:48] <jdstrand> neat
[22:48] <jdstrand> that isn't confusing at all :P
[22:49] <xnox> jdstrand: you are now 5m closer to the dark side.
[22:49] <sergiusens> jdstrand, works without it, but give warnings
[22:49] <sergiusens> jdstrand, once you get used to it, cmake is awesome
[22:49] <sergiusens> although people can build really aweful configs too
[22:49] <jdstrand> it is on my todo to get into it more
[22:49] <jdstrand> I have a rather long todo list though
[22:49] <xnox> sergiusens: i think end developers ruin all build-systems.
[22:49] <xnox> =)))))
[22:49] <sergiusens> xnox, LOL
[22:50] <xnox> no matter how nice it was intended to be.
[22:50] <mhall119> nik90: hey, sergiusens found that the clock app is accessing dconf/gsettings somewhere, and that's not allowed for click apps
[22:50] <jdstrand> sergiusens: do you want me to approve that or an upload?
[22:50] <sergiusens> jdstrand, just ack it in the votes section
[22:50] <nik90> mhall119: hmm...I dont use gsettings since the security team said that it was not secure and shouldn't be used
[22:50] <mhall119> sergiusens: ^^
[22:51] <jdstrand> sergiusens: actually, while I have you...
[22:51]  * jdstrand finds bug comment
[22:51] <nik90> mhall119: on reading your comments above, I do not have a c++ plugin yet..all timezone is gathered from online APIs
[22:51] <nik90> mhall119: so no dconf usage etc.
[22:52] <sergiusens> jdstrand, sure, what about?
[22:53]  * jdstrand is looking
[22:53] <jdstrand> I thought I had it
[22:53] <jdstrand> sergiusens: https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/1234544/comments/9
[22:53] <jdstrand> sergiusens: fyi only
[22:54] <sergiusens> jdstrand, looks good
[22:54] <sergiusens> jdstrand, I'm wondering how this will hit us with multiple apps though
[22:54] <sergiusens> jdstrand, as application name defines the package name today
[22:55] <nik90> mhall119: do you happen to anyone who worked on the phone app? I just want to know where the ringtones in the phone are located at for playing a tone when the timer is completed.
[22:55] <jdstrand> sergiusens: what do you mean, 'multiple apps'? clicks that ship multiple desktop files?
[22:55] <nik90> mhall119: happen to know*
[22:55] <marrabld> slangasek: After factory reset my phone wont boot from recovery.  Do you know I can push an image (any image) while in recovery.  None of my old cyanogen backups seem to be available from the recovery screen anymore, like they used to be.
[22:57] <sergiusens> jdstrand, yup
[22:57] <jdstrand> sergiusens: right, so the apparmor policy was designed specifically for that
[22:57] <jdstrand> eg: if APP_ID=com.example.foo_bar_0.1
[22:57] <sergiusens> jdstrand, but if I have to applications, is applicationName supposed to be the same in both?
[22:57] <jdstrand> then that app has write access to ~/.local/share/com.example.foo
[22:58] <jdstrand> then later the app ships a second desktop file
[22:58] <jdstrand> APP_ID=com.example.foo_baz_0.1
[22:58] <jdstrand> it also has write access to ~/.local/share/com.example.foo
[22:58] <sergiusens> jdstrand, that means applicationName has to be the same on both though
[22:58] <jdstrand> this way different applications in the same click package can share data
[22:59] <jdstrand> in the qml, yes
[22:59] <jdstrand> does that not work?
[22:59] <sergiusens> jdstrand, for c++ we can do organizationName and problem solved though
[22:59] <sergiusens> jdstrand, well the translations also come out of applicationName
[22:59] <jdstrand> the SDK could do the same
[22:59] <sergiusens> jdstrand, I think the sdk should support organizationName
[22:59] <jdstrand> sergiusens: you're right in that the SDK folks need to figure all that out
[23:00] <jdstrand> they decided to use applicationName in this manner. multiple applications were mentioned as a future requirement
[23:01] <jdstrand> I need to step away for a bit
[23:02] <jdstrand> feel free to contact me on irc if needed (I read backscroll)
[23:18] <marrabld> Can anyone help me flash an image (Ubunutu or otherwise) while stuck on recovery.  I followed the instruction on restoring Android here -> https://wiki.ubuntu.com/Touch/Install#Restoring_Android . It didn't work.  and now I am stuck.  I can reboot in to the recovery image but non of my backups that I did from Cyanogenmod can be seen from the recovery backup.
[23:19] <marrabld> resotore ./backup.ad doesn't work eitehr
[23:24] <marrabld> with my old phone I used to be able to put images on to the sd card and load them from there.  I can't seem to do that with the Nexus 4
[23:29] <marrabld> Ah Ubuntu, it doesn't seem to matter how much I was
[23:29] <marrabld> x
[23:29] <marrabld> lyrical about you.  You always seem to hurt me
[23:41] <racarr> lool: How comes the image?
[23:42] <lool> racarr: I have a super serious issue while looking at upstart-app-launch update
[23:42] <lool> racarr: can't open OS updates
[23:42] <lool> racarr: double-checking the issue
[23:43] <lool> racarr: I finisihed the task of the upstart job to enable mir, after fixing a prereq bug
[23:43] <lool> racarr: so this works on Mir
[23:44] <lool> racarr: I was trying unity8 testsuite from SF
[23:44] <lool> so that's pretty bad for grouper
[23:44] <lool> or community
[23:44] <lool> I guess we need to keep it in -proposed while we fix this
[23:45] <lool> racarr: nevermind, it worked upon reboot
[23:56] <lool> racarr: but it's still making progress, just more slowly than I'd like  :-)
[23:57] <racarr> lool: Sounds ilike an adventure as always :D
[23:57] <racarr> ok let me know if I can be useful
[23:59] <lool> racarr: I'd love diving into the unity8 test failures with you after I'm done with uploads