[02:23] Beldar, Note that install works but youll be loicked tot he old kernel and might have some issues updating it [02:23] *locked [02:29] ParkerR, Yeah, it's eol not worth it really. I have the multirom app installed just messing with options. [02:34] Beldar, I currently have that image with e17 on mine [02:34] Actually kinda nice [02:34] At this very moment Im attempting to upgrade to trusty. *ctosses fingers* [02:34] *crosses [02:35] 500MB of updates heh [02:35] Beldar, My wish is an updated desktop image with 3.10+ kernel [02:35] They did a lot in the more recent version for tegra chips [02:36] Including an open source graphics driver [02:39] that update would be nice [02:41] Beldar, but sadly this desktop image we have was just a testing ground when developing Touch [02:41] It wasn't really meant to be used that much :( [02:42] I for one LOVE the idea of desktop image on something i CAN EASILY CARRY WITH ME [02:43] lOT MORE VERSATILE THAT uBUNTU tOUCH IS [02:43] *Caps [02:43] *A lot more versatile than Ubuntu touch is === jhodapp is now known as jhodapp|afk [02:57] Beldar, :OO we may be in luck http://i.imgur.com/sSYAzG2.png [02:57] Look at the bottom of the terminal [02:58] There's some program that can flash the kernel and initramfs === chihchun_afk is now known as chihchun [03:55] Beldar, haha enlightenment crashed part way through and relaunched. this was after it got updated http://i.imgur.com/n0XJa9t.png [03:55] Is ubuntu phones contacts going to be pluginable to different clouds? [03:55] or just ubuntu one [03:55] (or nothing) [04:41] Beldar, you around? [04:43] http://i.imgur.com/QniIgSA.png :D it updated to trusty without a hitch [04:45] Cool, I see people on occasion asking the same grouper nexus general questions here. [04:46] Im going to see what I can do about the kernel now [04:55] Well that sure is some shit https://pastee.org/jcqjv [04:55] grouper doesn;t get the 3.4 kernel while the Nexus 4 and 10 do === Ursinha is now known as Ursinha-afk [04:59] :< gogoling linux-image grouper is just showing the results for 3.1. Im surprised nobody has done a newer build [04:59] *googling [04:59] I may try to build one on the device with localmodconfig... [04:59] I wonder if it would flash correctly [05:00] Tempting... === Ursinha-afk is now known as Ursinha [05:04] hey guys just wanted to say keep it up, just tried it out on the N4 and while it's not quite ready, i'm loving the gestures so far. [05:06] Who in the heck joins just to say one line and leave [05:06] :| [05:06] Frustrating [05:22] If I backup /dev/mmcblk0p2 can I reflash it with gastboot flash? [05:22] *fastboot flash [05:23] *fastboot flash boot === JoseeAntonioR is now known as jose [06:38] flash-kernel: installing version 3.13.6 [06:38] Flashing kernel and initramfs to /dev/mmcblk0p2... /dev/mmcblk0p2: updated is too big for the Boot Image (12195840 vs 8388608 bytes) [06:43] ogra_, I see yyou had the same issue haha [06:43] *you [06:43] How did you slim yours down exactly? [06:43] [13:25] even when droping everything from the initrd, setting MODULES=list and swithcing to xz compression i cant get the initrd small enough [06:43] [13:30] ah, shiny, weeding out more stuff manually makes it work (sadly right at the edge of teh size limit) === chriadam is now known as chriadam|away [07:20] Ok. This time around I enabled xz and also made sure to check DRM_TEGRA [07:20] Forgot that on the last build [07:27] hello [07:27] could someone help me with something [07:27] Just ask [07:27] If somebody is able to help hopefully they will repsond [07:28] trying to install ubuntu touch on my nexus 7 2013 [07:29] OK, so what's the issue? [07:29] didnt know which one to download and if i could install like i would a rom [07:30] MrSparkles, http://forum.xda-developers.com/showthread.php?t=2628398 [07:31] You can install MultiRom with the application in the playstore [07:31] does it actually dual boot? [07:31] where i can choose which one i wanna use [07:31] https://play.google.com/store/apps/details?id=com.tassadar.multirommgr [07:31] Yes [07:31] thats awsome [07:32] MrSparkles, looks something like this http://liliputing.com/wp-content/uploads/2012/12/multirom.jpg [07:33] good morning [07:33] Morning [07:33] good morning to you sir [07:36] parker does multirom use its own recovery? that would conflict with clockworkmod [07:36] It does. It uses TWRP [07:36] TWRP is fantastic though [07:37] okay so should i just restore my stock one or will it get rid of clockworkmod and install twrp [07:37] The app should handle it last I checked [07:37] It just needs root access [07:37] okay [07:39] gonna take me a hour or so to download one of those files but also restoring tablet [07:40] MrSparkles, also after it reboots and installs [07:40] Boot into andoird, and reopen multirom manager [07:40] ok === duflu_ is now known as duflu [07:40] It has an option to install Ubuntu Touch (it may or may not have the option. Not sure about the flo) [07:41] does it download the files required in the app? [07:41] So if it doesnt have the ption inside the application just use the XDA thread [07:41] Yep [07:41] *option [07:41] that'd be great === chihchun is now known as chihchun_afk [08:05] you still here parker? [08:07] ParkerR multirom manager is asking me to install and update do i need to install the kernal on it? [08:07] Yes [08:07] It needs some kernel features that arent in stock [08:07] It should be fine [08:08] ok [08:09] which one though? [08:09] has stock 4.3 4.4 cm 10.2 and cm.11 [08:14] MrSparkles, what android version [08:14] Match it up [08:14] ok [08:14] Settings > About [08:15] yea running 4.4.2 [08:15] Alright so choose that kernel [08:15] it says on playstore that it has option to install ubuntu touch but said for the 4g model === chihchun_afk is now known as chihchun [08:15] Well either installation method this part is required :P [08:15] dunno if would work for mine [08:16] yea figured that [08:18] nvm on that last msg [08:24] ok parker do i just install those 2 zip files in recovery? [08:27] MrSparkles, the recovery it installed should do it automatically [08:27] Oh the XDA post? [08:27] yea [08:27] Recovery > Advanced > Multirom [08:28] reboot to Recovery > Advanced > Multirom [08:28] do i have to install them seperatly? [08:28] k [08:28] Just go there and do what the post says [08:28] hmm, what's the easiest way to get my c [08:28] oops [08:28] The multirom section has its own little menu [08:28] hmm, what's the easiest way to get my contacts from my android phone to ubuntu touch? [08:28] without using a google account ;D [08:29] ok parker im in multirom in recovery [08:30] i was thinking about using the somewhat standardized (iirc) vcf format as intermediate step [08:31] google crashed === didrocks1 is now known as didrocks [08:34] i see mentions of syncevolution on e.g. askubuntu [08:35] but it seems the only way to install it is by using the apt tools directly [08:35] which breaks OS updates iirc? [08:35] Yeah you have to make system writable [08:35] Which turns off OTA updates [08:36] Hii. [08:36] ParkerR: "OTA" i an abbreviation for what? [08:36] its possible to see Android GUI on phablet (ubuntu touch) [08:36] ParkerR: in that case, can i still keep the system up to date manually? [08:36] over the air [08:36] fps, Over the air [08:37] anil, You are in Ubuntu touch. How would oyu see android UI? [08:37] *you [08:37] surface Flinger is running [08:38] can we disable lightdm and see android graphics [08:38] ... I dont think it works like that [08:38] hmm ..how I can achieve that [08:38] It's just a compositioner [08:39] That Ubuntu Touch just so happens to borrow from android [08:39] ok.. [08:39] can I disable lxc and boot the system [08:39] ok im installing ubuntu touch now [08:40] because I want to do some experiment with lxc container [08:40] I want to start it manaully once device is up [08:41] i dont think it installed properly parker [08:41] i chose the 2 zip files and it went to black screen within a minute [08:41] I created /etc/init/lxc-android-config.override contain manual but it did not work [08:41] guess ill try in multirom manager [08:48] Hello [08:49] Hello [09:20] ogra_, I really need you ;_; [09:21] parker you still here? [09:21] Aye [09:22] i installed ubuntu touch with the multirom manager and it gets an error whenever it starts [09:22] What does it say? [09:22] umm where would i look its showing me a bunch of lines from klog [09:23] Hmm [09:23] I really dont know :| [09:23] dang [09:26] ParkerR, you are operating under wrong assumptions ;) the kernel team in #ubuntu-kernel maintains the kernels :) [09:27] did i install wrong kernal? [09:28] ogra_, Hmm? Im trying to compile a new kernel and am having ha hard time getting it dowen to 8mb [09:28] As some IRC logs pointed out you have too :D [09:29] turn more stuff into modules then [09:29] Hmm ok will try [09:32] but i can update manually if i make the system fs writable? [09:38] parker [09:38] i got it installed finally [09:38] did the xda method [09:39] god i love this its so clean looking and smooth [09:39] I thought thats what you already did... [09:39] i did [09:39] fps, you can but packages may break, we dont officially support apt upgrade/dist-upgrade on these images [09:40] first time it went to black screen and never did anything [09:40] MrSparkles_, meh I like the desktop image better :) http://imgur.com/a/HSJOw#QniIgSA [09:40] figured id try again but this time i tapped on the screen when it would dim and it worked [09:40] * http://i.imgur.com/QniIgSA.png [09:40] =O [09:41] you make a desktop version? [09:41] I updated the old raring image to trusty [09:41] dont understand that really [09:41] How so? [09:41] when it boots it has my default android and 2 ubuntus [09:42] im not really great with linux used it few times but i mainly play games so use windows [09:42] One of the Ubuntus might be the original one you thought failed [09:42] it didnt show up there [09:43] it only showed my original android [09:43] hmm under recent apps browsers bit buggy [09:51] ogra_: ok [09:52] how can i install apps? [09:58] MrSparkles_, from the application page [10:00] MrSparkles_, it pulls the list of available apps from the network, so make sure you are online in some way [10:00] (WLAN, 3G etc) [10:05] sometimes keyboard doesnt open =\ [10:11] hey oSoMoN! how are you? [10:12] hey didrocks [10:13] didrocks, recovering from a big bad flu, I’ve known better days… [10:13] oSoMoN: urgh, you want to be reminded that we are still in winter? (even if the weather doesn't say so :p) [10:14] didrocks, yeah, it looks more like spring here :) [10:14] heh, here as well :) [10:14] ha my weathers sucking to [10:14] oSoMoN: IIRC, the calendar-app and messaging-app are done by your team, right? It seems that they both have flaky test results if you look at the past 2 images on mako [10:14] oSoMoN: mind giving that a look? (are you as well upstream for music-app? I don't think so…) [10:15] didrocks, calendar-app is a community-developed app, messaging-app is developed by my team indeed [10:15] oSoMoN: ok, mind looking at the last one then? [10:15] music-app is community-driven as well [10:15] sure [10:15] I’m on it [10:15] popey: are you in touch with music-app and calendar-app upstream? [10:15] oSoMoN: thanks :) [10:16] well thanks for helping me ParkerR and ogra_ [10:16] ima mess with this a little thanks [10:16] MrSparkles_: also make sure you have dash searches enabled to also search online ;D [10:16] oSoMoN: btw, direct link is http://ci.ubuntu.com/smokeng/trusty/touch/mako/238:20140314.1:20140304/7158/messaging_app/897403/ [10:17] didn't fail on latest image even if it seems there is no reason for it to be fixed [10:21] didrocks: yup! on it. [10:21] thanks! [10:28] " ParkerR, you are operating under wrong assumptions ;) the kernel team in #ubuntu-kernel maintains the kernels :)" what was that a reply to exactly? [10:32] ParkerR, to the unexpected PM you sent me [10:33] ogra_, Oh haha sorry about that. I started there but then realized the channel might be a better place === ondrat is now known as ondra [10:56] Morning all [10:58] Morning === vrruiz_ is now known as rvr [11:38] Hii [11:38] How to enable /dev/fb0 logging on console === jhodapp|afk is now known as jhodapp [11:58] beuno: hey, FYI, the new frameworks names are now added to the touch metapackage, you can add them to the store as well (but keep -dev1 until we have this gallery-app in please) [12:03] :D http://i.imgur.com/f0ZiZuH.png [12:22] ololosha: just a little followup on my previous question: is there a way to make a click package out of a deb and install it on the phone without breaking updates OTA? [12:23] ololosha: sorry, wrong nick completion [12:23] ogra_: just a little followup on my previous question: is there a way to make a click package out of a deb and install it on the phone without breaking updates OTA? [12:24] i'd really like to try syncevolution so i can get my contacts on the ubuntu touch phone :( [12:25] fps, syncevolution is already installed [12:26] sergiusens: aaah ok :D [12:26] i was confused by askubuntu answers... [12:28] sergiusens: hey, did you get any news from the store guys (if there is any progress) to get gallery-app in? [12:28] didrocks, yeah, needs to be fixed in the store admin interface [12:28] ok, before breaking everything horribly this is the plan then: connect to the device using adb shell [12:28] sergiusens: ok, let's wait for them to be awake I guess to be ready to rock it :) [12:28] didrocks, so I think we just need to wait for beuno to be awake :-) [12:29] yep :) [12:29] I am awake! [12:29] 1] change user to phabet 2] manually run syncevolution to get contacts from my owncloud [12:29] Heh [12:29] is that plan sound? [12:29] beuno: ah great! did you see as well my request about the other framework support? === MacSlow is now known as MacSlow|lunch [12:30] Hi ogra_ [12:31] I have compiled kernel with CONFIG_FRAMEBUFFER_CONSOLE=y [12:31] to enable console logging [12:31] didrocks, yes, can you give me exact strings? :) [12:31] and added cmdline = console=tty0 loglevel=8 in bootimg.cfg [12:31] is that ok [12:32] beuno: ubuntu-sdk-14.04-html-dev1, ubuntu-sdk-14.04-papi-dev1, ubuntu-sdk-14.04-qml-dev1 (but please keep the old one for now until we have everything moved back) [12:32] anil, yes, but it will break the user session [12:33] didrocks, ack [12:33] thx! [12:33] keep us posted on the gallery-app issue so that we can push it [12:33] anil, (the driver on the N10 does not work when fbcon is enabled) [12:35] sorry I could not get ..which driver will not work [12:36] anil, the graphics one [12:37] you will have fbcon and can use the framebuffer ... but the accelerated graphics driver will fall over [12:37] sergiusens, your app should be fixed [12:37] (as i told you last time we talked) [12:38] hmm [12:39] didrocks, and it's ubuntu-sdk-14.04-html-dev1, not ubuntu-sdk-14.04-html5-dev1, [12:39] ogra_ : I dont want accelerated graphics .. normal boot up should happen [12:39] ? [12:39] with graphics [12:39] anil, right, if you dont want to use any graphical stuff it will work [12:40] mterry, shy today ? [12:40] ogra_, :) [12:41] mterry, you got some comments on the usc spinner branch [12:41] :) [12:41] ogra_, oh good, will check [12:41] beuno: that's what colin used (and it seems that's what was agreed on https://wiki.ubuntu.com/Click/Frameworks) [12:41] seems you actually forgot to add the option :) [12:41] ok .. I am compiling kernel ...let see how it goes [12:41] sergiusens: keep us posted! :) [12:41] didrocks, ack. All done. [12:41] beuno: excellent! thanks a lot [12:42] didrocks, sergiusens's app should stop erroring (sorry about that) [12:42] yesterday, as suggested , to disable lxc container, I created /etc/init/lxc-android-config.override containing manual [12:42] no worry, let's hope everything will be fine, then, we "just" have some minor issues [12:42] it did not work [12:42] ogra_, were you expecting the spinner to work out of the box, or did you manually pass --spinner=... ? [12:43] by your explanation i was expecting to just have to add --spinner ... (without = ) [12:43] which i added to usc-wrapper for testing [12:47] beuno, didrocks just for the FAQ, if this happens again, do we just contact you? [12:47] ogra_, oh sorry. I let you specify the executable (for customization reasons or whatever) [12:48] ogra_, so you need to pass the path [12:48] ogra_, I'll make that clearer in the desc [12:48] ogra_, also for your other question, the design is not coordinated yet. I asked design to look the spinner over, but they haven't gotten back to me [12:48] Method__, so --spinner=/usr/bin/unity-system-compositor-spinner ? [12:48] sergiusens, yes. We're working on making sure it doesn't happen again :) [12:48] err [12:48] mterry, ^^ [12:48] ogra_, yes [12:48] * ogra_ tries [12:49] ogra_, the look of it now is far from perfect. But it's just a placeholder [12:49] beuno, oh, not just this one ;-) we are all troublemakers here :-P [12:49] mterry, yeah, its enough until we get proper design [12:49] ogra_, also, if you are testing without all the other split branches and mir 0.1.7, then spinner will come up, then black screen, then unity8. That's expected [12:49] sergiusens, so anything related to the store, always me [12:49] mterry, ok [12:50] yay ! [12:50] mterry, works fine with the path [12:50] ogra_, awesome [12:56] orgra_ : if we disable lightdm then unity is also disabled [12:56] or not [13:03] yes [13:04] thanks === alan_g is now known as alan_g|lunch === lool- is now known as lool [13:13] popey: are you around? [13:13] m-b-o: i am [13:14] popey: o/ [13:14] I've removed and added icons to the weather app at the weekend, now jenkins refuses to work [13:15] an I haven't found any place to change icon paths [13:15] popey: http://91.189.93.70:8080/job/generic-mediumtests-trusty-weather-app/71/console [13:23] m-b-o: oh [13:25] popey, m-b-o bzr bd should be able to reproduce it; bump the upstream version and try again === w00t_ is now known as w00t [13:29] serguisens popey: hmm, bzr bd seems not to have a problem with the icons [13:29] serguisens: what do you mean with bumping the upstream version? === MacSlow|lunch is now known as MacSlow [14:08] If I want to install Ubuntu Touch on a a 2012 Nexus 7 (grouper), which channel should I install from? === alan_g|lunch is now known as alan_g [14:11] Rev_DE: it's no longer supported [14:12] Officially, or "It won't run" not supported? [14:13] There are quite a few ports out there [14:17] I also see an image of Trusty for the Grouper here: http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/ === jono is now known as Guest78266 === psivaa_ is now known as psivaa [14:31] Rev_DE, https://wiki.ubuntu.com/Touch/Install Claims to be deprecated but the install insctructions are there [14:32] The flashing tool still works with it [14:33] Rev_DE, currently I'm working on an updated desktop image for grouper (unofficially). I have it up to trusty and am working ona new kernel now === dandrader is now known as dandrader|afk [14:48] stgraber, hey ho ... do you remember why we do the dbus session job like we do it atm ? if you look at people.canonical.com/~ogra/touch-bootcharts/ubuntu-phablet-trusty-239.png we seem to have certain processes starting tehir own dbuses which indicates the session bus address isnt properly exported to them [14:48] (there are 3 started by other processes ... all of them die at some point) [14:50] (or do i read the bootchart wrongly ?) [14:54] ogra_: it's pretty hard to tell when we don't have the arguments... I know that on the desktop at least we can get up to 3 dbus per session as things like accessibility use their own separate bus [14:54] well, i would expect things to hook into the same session bus inside the same session [14:55] ogra_: that'd make sense but that's not the case for accessibility [14:55] stgraber 2759 0.0 0.0 39464 2096 ? S Mar08 0:01 /bin/dbus-daemon --config-file=/etc/at-spi2/accessibility.conf --nofork --print-address 3 [14:56] stgraber, right, but these arent a11y services [14:56] one is dconf ... which according to desrt only fires up its own daemon if it cant attach to the session bus [14:56] ogra_: now it's not impossible that some of the touch upstart jobs are wrong too, I don't know. dbus starts on "starting xsession-init" which is as early as technically possible. If another job was set on say "starting dbus" or "starting xsession-init", they'd indeed lack the required environment variables [14:56] one is zeitgeist [14:57] and one seems to be indicator-messages [14:57] jibel: hey.. is it easy to run britney locally? [14:57] jibel: lets say i have a package dsc/changes with sources and binaries and want to see if that thing would nicely go in [14:57] jibel: can i dry run that locally? [14:57] ogra_: I really know nothing about the unity8 upstart jobs so I can't help much... I just checked on the desktop and things do appear to be reasonably ordered there [14:57] stgraber, right, though i wonder if the exporting is somewhat racy [14:57] (leaving autopkg tests alone for now) [14:58] ogra_: considering upstart is single threaded, races are not technically possible, unless the start conditions are wrong [14:58] well, we do the exporting from pre and post scripts [14:58] which has race potential [14:59] asac, you can do it but I wouldn't say it is easy because you'd have to modify index files manually [14:59] anyway, if you really suspect a race in upstart rather than broken jobs on your side, I'd suggest talking to jodh [14:59] index = source and binary package lists [14:59] stgraber, well, zeitgeist for example doesnt even have a job === vying is now known as Guest16803 [15:02] ogra_: right, but on the desktop gnome-session itself won't ever start until dbus is fully started, so zeitgeist which is spawned by something that's spawned by gnome-session will have the right environment === jasoncwarner___ is now known as jasoncwarner === bfiller is now known as bfiller_afk [15:03] stgraber, k [15:04] i'll try to track down the chain here [15:11] what's the LP project for ubuntu-location-serviced? [15:14] root@ubuntu-phablet:/# dpkg -S /usr/bin/ubuntu-location-serviced [15:14] ubuntu-location-service-bin: /usr/bin/ubuntu-location-serviced [15:14] location-service is the source package [15:14] https://launchpad.net/location-service [15:15] thanks popey === awafaa_ is now known as awafaa === mandel` is now known as mandel [15:41] * ogra_ wonders if he should have a word with cjwatson to not upload click when it has a full blatter ... according to http://people.canonical.com/~ogra/touch-bootcharts/ubuntu-phablet-trusty-239.png it spawns "content-hub-pee" on boot [15:43] ogra_: :-P [15:43] :) [15:44] niemeyer: ping -- was trying to package my go+qml app to include in a ppa, but it complained about not being able to find the qml package in my pbuilder... shouldn't it automatically 'go get' it? http://paste.ubuntu.com/7108793/ [15:44] sergiusens: ^ you did some deb packaging with go, right? any ideas? [15:44] cwayne, yes [15:45] cwayne, you don't go get when packaging; you need to first package niemeyer's package and then build depend on it and any other golang package [15:46] ah, i figured i might have to do that, was trying to avoid it though because lazy :) [15:47] cwayne, if it's a temporary thing, import the package's sources locally [15:47] temporary == lazy :-P [15:47] ha [15:48] well im thinking go-qml should definitely be packaged (and probably put into universe) since as there's a competiton for it, people are likely going to want their apps up in ppa's somewhere === dandrader|afk is now known as dandrader === gatox is now known as gatox_lunch === bfiller_afk is now known as bfiller [15:57] sergiusens: bah, exactly what i was afraid of, now need to package 'pkg-config' as well i think [15:59] cwayne, well go devs don't necessarily use PPAs or debs [15:59] cwayne, unless it is for packaging [16:00] yeah i know.. just figured itd make sense to have this one packaged.. [16:03] cwayne, for app developers doing click; not really [16:03] if it's something the system would depend on, then yes [16:03] fair enough re: clicks [16:07] tedg, what are your opinions about my last comment in https://code.launchpad.net/~mterry/indicator-sound/greeter-volume/+merge/209159 ? [16:15] ParkerR: Awesome. I'm flashing it now [16:24] kenvandine: ping [16:24] kenvandine: don't reply it's a trap! [16:24] it's *always* a trap [16:25] * davmor2 screenshots mhall119 last statement and emails it too all the developers :D [16:26] they already know [16:26] gah, I lost 2-finger scrolling after updating [16:27] * kenvandine runs [16:27] what's up mhall119? [16:28] mhall119: check the settings incase they got flipped somehow [16:28] nope [16:28] tedg: can you flip https://code.launchpad.net/~cjwatson/upstart-app-launch/porting/+merge/210385 to rejected and drop it from any landings? we shouldn't need it any more [16:28] kenvandine: hey, regarding a click package for wallpapers, what meta-data would you need? [16:29] tedg: xnox dealt with this a different (and better) way [16:29] sizes could be nice [16:29] but not sure how'd we'd use that info right now [16:29] Rev_DE, it it boots and you just see a black screen touch it a few times [16:29] cjwatson, Cool [16:29] It'll wake up [16:29] *if it [16:29] davmor2: removing and re-adding the psmouse driver fixed it [16:29] mhall119, basically none needed, just where the wallpapers are [16:30] mterry, Yes, I guess I disagree, but it's not worth holding up the MR. [16:30] mterry, I think that seb128 wants me to write an FFe for all the AS stuff. [16:30] mterry, So I'll put that in there as well. [16:30] tedg, ah OK [16:30] cwayne: Hmm [16:31] tedg, I'm bummed that vala doesn't let you have async getters/setters [16:31] cwayne: I would just bundle it in your app rather than going over the trouble of maintaining a package for it [16:31] mterry, That is odd, no clue why they did that. Probably a TODO on their side. [16:31] cwayne: This would mean putting a $GOPATH/src within your package, instead of just the app itself [16:32] cwayne: You can either do that yourself, by just creating a GOPATH and running go get.. or you can use godep (github.com/kr/godep) to do the local dependency management for you [16:33] niemeyer: ah, thanks. i'll try that out [16:33] yeah, id already given up on trying to deb package go-qml, so maybe this will work :) [16:42] Could somebody mark https://code.launchpad.net/~cjwatson/unity-scopes-api/lttng-optional/+merge/211357 as rejected? [16:42] mhr3, ^ [16:43] Hey guys, I'm trying to port Ubuntu Touch. I compiled the files and I put the contents of ubuntu-root (in the out directory) on the root partition. It mounts the root partition correctly, but there's a kennel panic because it cant find /init, but it is there. Also, the init in ubuntu-root is a script and not a binary (not sure if this makes a difference). Any thoughts? [16:43] Oh, wait, I can do it. [16:43] mhr3, Ignore that :-) [16:43] kenvandine: ok, here's a simple wallpapers click package: http://people.ubuntu.com/~mhall119/com.ubuntu.developer.mhall119.trusty-wallpapers_0.1_all.click [16:44] kenvandine: I called the hook 'backgrounds' and pointed it to a folder in the package [16:44] mhall119, excellent [16:44] i'll hack something together in a bit === guest826 is now known as hedz09 [16:44] so the hook will just need to put those under a namespaced folder and symlink to wherever the settings app looks [16:44] thanks kenvandine === gatox_lunch is now known as gatox === jhodapp is now known as jhodapp|doc [16:47] jhodapp|doc: has the media-hub landed in device images? [16:47] ogra_: mterry: just saw the video of bootsplash -- nice! [16:48] :) [16:48] Hey guys. I'm trying to port Ubuntu Touch. I compiled the files successfully. If I understand correctly, I put the contents of ubuntu-root on my rootfs partition. When I boot, the rootfs is mounted, but there's a kernel panic because it can't find /init. Also, I noticed init is not a binary but a readable script. Not sure if this makes a difference. Any thoughts? [16:48] mterry: you know Im gonna ask how to customize it though :) [16:48] cwayne, pass a different --spinner=/path value to it [16:48] cwayne, to USC [16:48] mterry: ah, so we'd have to an upstart job override then probably? [16:48] we'll make it configurable for you :) [16:49] cwayne, which you should be able to do with a lightdm.conf.d config value [16:49] ah [16:49] cwayne, we have a unity-system-compositor-command= field [16:49] mterry: as long as it can be read from /custom, i'm sold :) [16:49] cwayne, I actually am not sure which paths it checks. I can confirm [16:49] could even be read from a property :) [16:49] ooooh [16:49] do hat [16:49] ogra_, property? you mean env var? [16:49] then we can just set it in custom.prop [16:50] mterry, nope an android property [16:50] cwayne: do they scarf at the same time as hatting? [16:50] ogra_, how does that work? Which field can be set like that? [16:50] mterry, just run getprop on a phone :) === dandrader is now known as dandrader|lunch [16:50] mterry, we can ship made up properties or set them with setprop etc [16:51] mterry, for desktop i would suggest an env var or some such, for phone customization we should allow it to be overridden from a property [16:51] ogra_, but USC doesn't support using a property for --spinner= automatically, right? [16:52] not yet ... thats minor upstart job hackery [16:52] or usc-wrapper [16:52] should be doable in three lines or so [16:53] ogra_, fair... [16:53] first lets land it ... then lets think about themeing ;) [16:53] sorry, just wanted to bring it up now so it's not a pain later if we didnt have this in mind :) [16:53] cwayne, FYI, lightdm does *not* look in XDG_DATA_DIRS for config files [16:53] its a minor change to make a property override it [16:53] cwayne, it probably should [16:56] ParkerR: Okay, so what would I do to enable root access and to get down to something more closely resembling a desktop? [16:56] I'm trying to use either LXDE, XFCE, or KDE [16:58] I've got SSH working, and I think the system image is writable, but what's the equivalent of the GDM screen to select a DE === guest826 is now known as hedz09 === plars_ is now known as plars [17:00] Hey guys. I'm trying to port Ubuntu Touch. I compiled the files successfully. If I understand correctly, I put the contents of ubuntu-root on my rootfs partition. When I boot, the rootfs is mounted, but there's a kernel panic because it can't find /init. Also, I noticed init is not a binary but a readable script. Not sure if this makes a difference. Any thoughts? thanks in advance. [17:01] Rev_DE, there is no X11 on the phone and only unity8 can run with Mir yet [17:09] ogra_: Ah, dang. Is that just in the 14.04 builds, or is X11 not a part of the 13.XX images? [17:10] Supposedly Arch Linux ARM has native support for Tegra on X11 [17:10] X11 isnt a part of any of the Ubuntu system-image builds === om26er_ is now known as om26er === ondra is now known as ondra|away === ondra|away is now known as ondra [17:30] barry: I've noticed something a couple of times now, if I've opened settings just before an update is available it doesn't show the update when it is available is there a file cached somewhere? [17:31] davmor2: um, not sure what you mean. you mean "just before an update is available" on the server? [17:32] Hey guys. I'm trying to port Ubuntu Touch to the Beaglebone Black development board. I've already ported CM 10.1. I compiled the Ubuntu Touch files successfully. If I understand correctly, I put the contents of ubuntu-root on my rootfs partition. When I boot, the rootfs is mounted, but there's a kernel panic because it can't find /init, even though it's there. Also, I noticed init is not a binary but a readable script (not sure i [17:33] barry: so an image update is available at say 19:00, I opened settings at 18:55, I open settings again at 19:00 still shows no image but system-image-cli -n says there is [17:33] hedz09, repeating yourself all the time wont get you answers quicker [17:34] hedz09, if you have something called ubuntu-root and you have /init beeing a shellscript, thats definitely not a recent rootfs [17:34] barry: if I reboot and open setting it then says there is an update [17:34] this setup was only used in the very first (and very broken) iteration === ondra is now known as ondra|away === ondra_ is now known as ondra__ [17:36] davmor2: do you have manual or automatic downloads set? [17:37] barry: default so Auto Download on wi-fi === ondra|away is now known as ondra [17:38] * ondra is back (gone 00:00:27) [17:38] davmor2: could be a ui issue. what happens if you explicitly click on Updates icon? [17:38] ondra, ! [17:39] ogra_: what? :) [17:39] (switch off auto away in public channels please) [17:39] ogra_: OK [17:39] thanks :) [17:40] barry: same thing, hence asking if there was a cached file somewhere. Rebooting is the only thing that seems to make it visible in the ui, however if I update today I can update tomorrow with no issues. Just seems to be if you open the setting app just before the release is available [17:41] ogra_: How would I get a recent rootfs? [17:41] anyway teatime === bfiller is now known as bfiller_afk [17:42] hedz09, http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/ has the original tarballs (before they are turned into system-image images) [17:43] davmor2: there's no cached status file. the s-i-dbus process would still be hanging around, but it should still respond to new CheckForUpdate dbus methods. the thing to do is to tail -f /var/log/system-image/client.log while you're reproducing this and see if the CFU method gets called (you'd see a bunch of activity in the log file). bumping up to verbose logging might also provide more information. it's unclear whether the problem [17:43] is in the ui or in s-i-dbus [17:43] ogra_, lool, asac: FYI, I'm planning on doing the -> ubuntu-touch/ rename on prod system-image later today. As I said on the mailing-list, aliases will be in place for quite a while so people won't notice anything and both barry and sergiusens checked that the flashing tools and updater will cope properly. [17:43] so I'm not expecting anyone will notice anything, but that's just a quick heads up anyway. [17:44] hedz09, you might want to take a look at the installer script of rootstock-ng (bzr branch lp:rootstock-ng) to get an impression what has to happen with your android system.img file [17:44] stgraber: ok, do you have a checklist to validate after the switch? and a way to backout in case that - against all odds - things are broken? [17:44] stgraber: time for me to go on vacation. :) [17:44] stgraber: important that someone checks that ci.ubuntu.com still works [17:44] sergiusens: ^^ [17:44] stgraber, just make sure the tools get updated before removing the aliases :) [17:44] sergiusens: have you checked that whatever outdated tool they use supports this? [17:44] sergiusens: remember that they are trapped in raring for some reason [17:44] doanac`: ^^ [17:45] ogra_: one change for you though (and whoever else does that) is that when promoting image, you'll need to put the new channel name so "ubuntu-touch/trusty-proposed" and "ubuntu-touch/trusty". Using the old names will tell you that "redirects" aren't allowed as arguments. [17:45] doanac`: seems folks want to change how we do channels. oplease check with stgraber and sergiusens so we are sure that the lab doesnt explode [17:45] stgraber, ok [17:45] doanac`: and also tell them how they can actually validate right after the switch that stuff works (e.g. like sending a test job?) [17:45] stgraber, so leave me a ping if thats done and i'll update my README [17:46] asac: QA and all the other tools will simply follow the aliases so there won't be any breakage, we'll just need to make sure everyone has moved to the new names by the time we drop the aliases (my plan is to do that when saucy goes EOL, so still quite a while) [17:46] asac, it was emailed two weeks ago; I tested ubuntu-device-flash from two weeks ago and added unit tests covering this specific case to make sure [17:46] stgraber: right. guess do the switch early and ensure that a new image gets sent through the pipe to know for sure [17:46] asac, wrt to raring and phablet-flash we will need to handle it; but as stgraber said; there are going to be links in [17:47] stgraber: switch early during your day and at best work with robru/doanac etc. that we get a new image through the infra so we dont wake up with mysery :) [17:47] ogra_, asac: as far as post-change testing, I have two out of date devices here which I'll upgrade after the change to make sure that the prod server works identically to the test server I've been using so far [17:47] right [17:47] ogra_: can you help stgraber geting an image through ci right after? [17:47] so we see it green on white that its all going well? [17:47] what should i help with exactly ? its all automatic [17:47] stgraber, right, channels.ini needs to be updated; I didn't test that part [17:48] asac, i can help if the automation breaks ;) [17:48] definitely need green today :) === dandrader|lunch is now known as dandrader [17:48] ogra_: you should ensure that a new image gest produced right after [17:48] the only change is promoting an image ... and thats usually only handled by me anyway ... rarely by rsalveti [17:48] ah [17:48] ogra_: so we pipeclean the infra and confirm that stuff continues to beat [17:48] ogra_: dont wait for cron [17:48] just kick it so we can react before stgraber etc. go to bed :) [17:49] but tell landing folks [17:49] well, there is a new image pending if unity8 has built [17:49] ogra_: your call, but dont wait too long as you know that we dotn want stgraber and folks to do a night shift [17:49] rather early NO or GO :) [17:49] if stgraber does his change before we can both tested in one go [17:49] right [17:49] so hurry up [17:49] you mean he is more likely to do a nigghtshift than me ? [17:49] :P [17:50] (the unity8 build is planned to happen before cron ... no worries) [17:51] ogra_: I'm doing the change now, waiting for a full backup first of system-image (just in case), should be all done in ~10min [17:51] great, thanks [17:51] asac: i'll do a double check on our end [17:53] stgraber: ogra_: so keep doanac` in the pipecleaning loop :) [17:53] thanks [17:53] * ogra_ hands doanac` a spare cloth [17:53] whee ! [17:53] symmetry http://ci.ubuntu.com/smokeng/trusty/touch/mako/240:20140317.1:20140304/7193/ [17:54] only the bottom ones fail the scheme [17:54] (and UITK) [17:54] ogra_: I downloaded and extracted trusty-preinstalled-recovery-armel+manta.tar.gz (http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/20140317/). Then I used unmkbootimg on boot.img and extracted initramfs and again I see a shell script. Am I missing something? [17:55] hedz09, ignore the device specific files, these are the ones your android build spit out [17:57] take a look at rootstock-ng ... you want to put recovery.img and boot.img from your CM based build in place, then use http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/20140317/trusty-preinstalled-touch-armhf.tar.gz together with the CM system.img to create a proper rootfs [17:59] ogra_: server updated now, test run in progress [17:59] great === alan_g is now known as alan_g|EOD [18:03] barry: thanks next time I hit it I'll have a look :) [18:04] ogra_: simple update worked fine [18:04] davmor2: ack. there's a new s-i-dbus silo on its way, though it should have no effect on auto downloads [18:04] re-activing the import-images cron [18:05] ha ! [18:05] * ogra_ has his first fallout on the image watcher bot :P [18:05] and fixed [18:05] ogra_: ah, was it using hardcoded paths? :) [18:05] ogra_: so you say ;) [18:05] stgraber, heh, yeah === jhodapp|doc is now known as jhodapp [18:08] mhall119, no, still working on it [18:08] ogra_: do you know when the next image should be publishing? I'll want to make sure that the new version tarball is correct at that point (it should be and it was on the staging server, but I prefer to be 100% sure :)) [18:08] bfiller_afk, renato__: hi! About this branch: https://code.launchpad.net/~mardy/account-plugins/lp1252751/+merge/198689 [18:09] bfiller_afk, renato__: me and dbarth think it's better if you take care of landing that, since you are the ones who need it [18:09] bfiller_afk, renato__: you can land it together with any other stuff you have [18:09] stgraber, a build is planned for later today (once unity8 is in the archive, not sure when exactly, cyphermox has that in his hands) ... promotion will not happen until we have a green one though [18:09] which can still take a few days [18:10] ogra_: yeah, I'm not worried about promotion, I just want to triple-check the image once it hits trusty-proposed [18:10] stgraber, there is an imagebot in #ubuntu-ci-eng now that will announce start and finish of a touch build [18:20] ogra_: ahh, okay. So rootstock-ng is not device specific? [18:20] no, its a script for people developing their own images [18:20] its not really designed for soemthing like a beagle (since it uses adb and such) ... but you should get an impression what you need to do by looking at the code [18:33] Hi! [18:33] I am trying to get batman-adv running on Ubuntu Touch, but I am running into issues that I am so far unable to resolv. The module compiles fine, but when I try to load it, I get this error message: [18:33] $modprobe batman-adv [18:33] modprobe: ERROR: could not insert 'batman_adv': Exec format error [18:33] $modprobe batman-adv -f [18:33] modprobe: ERROR: could not insert 'batman_adv': Unknown symbol in module, or unknown parameter (see dmesg) [18:33] dmesg gives me this: [18:33] [ 7203.530456] batman_adv: no symbol version for module_layout [18:34] [ 7209.188629] batman_adv: Unknown symbol try_module_get (err 0) [18:34] [ 7209.188659] batman_adv: Unknown symbol sysfs_create_file_ns (err 0) [18:34] [ 7209.188751] batman_adv: Unknown symbol prepare_to_wait_event (err 0) [18:34] [ 7209.188842] batman_adv: Unknown symbol sysfs_remove_file_ns (err 0) [18:34] [ 7209.188903] batman_adv: Unknown symbol prandom_u32 (err 0) [18:34] [ 7209.188934] batman_adv: Unknown symbol netdev_upper_dev_unlink (err 0) [18:34] [ 7209.188964] batman_adv: Unknown symbol __pskb_copy (err 0) [18:34] [ 7209.188995] batman_adv: Unknown symbol netdev_master_upper_dev_link (err 0) [18:34] [ 7209.189086] batman_adv: Unknown symbol netdev_master_upper_dev_get (err 0) [18:34] [ 7209.189117] batman_adv: Unknown symbol delayed_work_timer_fn (err 0) [18:34] [ 7209.189147] batman_adv: Unknown symbol kfree_call_rcu (err 0) [18:34] Where do I go from here? The module works fine on all other hardware I have run it on... [18:36] groggster, pretty please use paste.ubuntu.com for such huge pastes [18:36] bfiller_afk: renato__, mardy: +1 [18:37] ogra_, sorry, I will be sure to do that the next time. [18:37] thanks :) [18:40] sergiusens: ping -- was wondering why on goget-ubuntu-touch, the resulting binaries end up in debian/tmp/usr/bin instead of debian/goget-ubuntu-touch/ [18:43] ERROR:phablet-flash:https://system-image.ubuntu.com/trusty-proposed/grouper/index.json cannot be retrieved [18:43] ^^ is 2012 n7 now totally unusable with touch? [18:45] dobey, it hasn't been on proposed for a while I believe; let me check (you should use ubuntu-device-flash btw) [18:45] sergiusens: it was there late last week [18:45] cwayne, is it because I have multiple packages instead of just one? [18:46] ah, i think if igured it out -- it's cus i had the source in src/ so it was getting compiled as src [18:46] dobey, I think it's just stgraber migrating now (should be temporary); try --channel ubuntu-touch/trusty-proposed [18:47] well, ubuntu-device-flash seems to be downloading an image [18:47] yeah, stop using these old dos like scripts :) [18:48] when you migrate the scripts, add deprecation warnings that specify what the new thing to use is [18:48] rather than just having it fail [18:48] sergiusens: hmm, didn't you say phablet-flash would follow the redirect fine? [18:48] stgraber, that was ubuntu-device-flash [18:49] ok, is phablet-flash still around? if so, it may be worth burning it with fire or something [18:49] the other tool is deprecated; and would land as soon as I get an FFe [18:49] because now that the change is live, /trusty-proposed doesn't exist anymore [18:49] dobey, nah, we want to keep in touch with our users ... that way they show up on IRC once in a while :P [18:49] stgraber, are the links coming though? [18:49] sergiusens: the "links" are there, as in, entries in channels.json [18:49] ogra_, there is a deprecation warning; can't land it though [18:49] but if you ignore channels.json, you'll fail [18:50] the redirects are channel entries in channels.json pointing to the renamed index files (as per the spec, you're never supposed to assume the location of the indexes...) [18:52] stgraber, I don't see any redirect entry here: https://wiki.ubuntu.com/ImageBasedUpgrades/Server [18:53] sergiusens: it's because it's a standard channel.json entry, it just happens to point to the exact same files as the new one [18:53] so nothing special to support on the client side, as long as you already support parsing channels.json [18:54] stgraber, I do parse it; but I never saw a redirect key which I'm seeing now; do I need to do anything with it? [18:54] sergiusens: no, the redirect key is only there as a flag to be used by the server (to prevent publishing new images to a redirect channel) [18:54] sergiusens: on the client side it can be treated as any regular channel [18:54] stgraber, then we are fine [18:55] we're fine with ubuntu-device-flash, but it may be worth reminding everyone that phablet-flash is deprecated [18:55] although I need to ping doanac` [18:55] stgraber: alias in "ubuntu-touch/devel" is just "trusty" and not "ubuntu-touch/trusty", is that intentional? [18:56] ogra_: ok, this is making more sense to me. So sbin/init is the init file that the kernel should look for? And is var/lib/lxc/android/pre-start.sh the install script you were referring to (to find out how the android system.img gets mounted?) [18:56] hedz09, no, i was referring to the rootstock-ng installer script [18:56] Tassadar: hmm, not intentional, I didn't think of those when I wrote rename_channel apparently (quite a corner case...). I'll update them by hand now. [18:57] Tassadar: thanks for noticing. They technically should work, but we don't want them to break when we deprecated the redirects :) [18:57] hedz09, rootstock-touch-install in the bzr branch i gave you above [18:57] no hurry, just noticed my app wasn't correctly handling the aliases because it is doing strcmp(alias, channel_name), it isn't breaking anything or something like that [18:57] thanks [18:58] stgraber, sent an email [18:58] for deprecation [18:59] plars, fyi I can't reach doanac` ^^ [19:00] hmm, ubuntu-device-flash couldn't push the tar.xz to the device apparently :( [19:01] Tassadar: fixed [19:01] thanks [19:02] sergiusens: this is about the download location change right? [19:02] dobey, clean your /cache [19:02] sergiusens: we have a bug open for it to change things on our end, but as I understand it, we have a bit of time thanks to the redirects right? [19:02] plars, yeah, I never fixed phablet-flash; so there are two options [19:03] plars, not with phablet-flash [19:03] plars, where are the scripts; I can propose an MR for you [19:03] sergiusens: lp:ubuntu-test-cases/touch === dandrader is now known as dandrader|afk [19:05] sergiusens: ok, we are looking at updating that system to saucy on Friday. I was planning to wait until doing that to make the changes to use ubuntu-device-flash, but we can try to do something sooner if needs be. Will just need to download the package and install it [19:05] plars, I have a quick fix for phablet-flash, don't worry [19:06] sergiusens: ok, so we could apply that locally on our host system for this week? [19:06] plars, more like right now [19:07] sergiusens: shoot me the patch then and we'll take a look [19:09] plars, yeah, already wrote it; just testing it now :-) [19:17] plars, here's the quick fix https://code.launchpad.net/~sergiusens/ubuntu-test-cases/pf_channel/+merge/211394 [19:18] sergiusens: ok, and the changes are already in place for that to work ok? [19:19] plars, let me put it this way; without that change it won't work [19:19] plars, the phablet-flash dep was an oversight; I think this is the quickest way given that raring is unuploadable [19:19] sergiusens: understand, just confirming that this isn't something that's about to happen, but unsure [19:19] yep [19:20] plars, not sure if the parameter is overrided on jenkins (given it's an $IMAGE_OPT with default params) === dandrader|afk is now known as dandrader [19:21] sergiusens: it is for the custom image - does the change affect those also? [19:21] jdstrand, ping [19:21] plars, yes; where is that triggered from? [19:22] sergiusens: under jenkins dir, there are some more changes needed probably [19:22] ogra_: I entered "bzr branch lp:rootstock-ng" and it said "bzr: ERROR: Not a branch: "bzr+ssh://bazaar.launchpad.net/+branch/rootstock-ng/"." Any thoughts? [19:22] plars, ah, let me fix [19:24] ogra_: actually, using "bzr branch lp:project-rootstock-ng" worked [19:24] heroops, sorry,, yeah, projects starting with "root" in the name are not allowed on launchpad ... i walways forget the silly prefix [19:24] kenvandine: hey [19:25] jdstrand, hey! [19:25] jdstrand, i have a branch of content-hub that attempts to use a hardlink instead of copy for exchanges that aren't set to persist (in the HubIncoming dir) [19:25] but i don't want the destination app to be able to edit the content, since it is really a link to the source apps content. [19:25] how would you feel about adding a rule like this: [19:25] deny @{HOME}/.cache/@{APP_PKGNAME}/HubIncoming/** w, [19:26] sergiusens: I think the only update needed there would be in that template - production.py [19:26] plars, I updated custom-demo.py as well just in case [19:26] plars, just pushed [19:27] kenvandine: so let me understand, app foo asked the content-hub for some non-persistent file. the content-hub coordinates with a provider and eventually puts it in @{HOME}/.cache/foo/HubIncoming/... [19:27] kenvandine: now the app has read access to it, cause that is in one of its directories [19:28] yup [19:28] it has rw now [19:28] if i hardlink it, the app would be able to write to the file [19:28] kenvandine: cause it is a hard link, writes to the file cause the 'source' to be modified [19:28] yeah [19:28] which is actually owned by the source [19:28] yep [19:28] got it [19:28] so this prevents that [19:29] kenvandine: did you confirm that it prevents that? [19:29] yes [19:29] nice [19:29] tested it [19:29] i'm much happier with the link than copy [19:29] I thought it would, but I wanted to double check (ie, it is a bug if it didn't) [19:29] :) [19:29] copy is fairly icky [19:29] i also talked to some of your team last week while you were out [19:29] kenvandine, ugh, careful with hardlinks on touch [19:30] ogra_, not across filesystems [19:30] we span across multiple loop devices [19:30] it'll be in /home [19:30] link has its own issues. eventually, we will have some nice fd delegation that we can use to help here [19:30] ah, k [19:30] but that is down the line [19:30] yeah... [19:31] kenvandine: I'm guessing by the name that HubIncoming will never have anything else in it, correct? [19:31] just stuff that the content-hub puts there [19:31] right [19:31] and the hub purges it [19:32] what happens if the app puts something there just before the content-hub does? will content-hub behave? [19:32] well, it doesn't matter if we have that rule [19:32] right :) [19:33] right now there are numbered subdirs [19:33] based on the transfer.id [19:33] so HubIncoming/1 [19:33] we purge 1 when it's finalized or if the hub exits [19:33] kenvandine: what will content do if the app removes HubIncoming or chmod 0's it? [19:34] s/content/content-hub/ [19:34] sergiusens: pushed, thanks [19:34] just fail to remove it [19:34] but it won't have write access, so yay! [19:34] kenvandine: and before content-hub puts it there? [19:35] it makes the directory if it doesn't exist [19:35] but what if it is chmod 0? [19:35] transfer will abort [19:35] ok [19:36] kenvandine: so, these were just things I thought of otoh with an app trying to subvert the content hub in some way. want to make sure the hub is robust in the face of these sorts of things [19:36] kenvandine: I think the rule you have is good. we'll want to revisit with fd delegation comes about, but that's for another day [19:36] indeed [19:36] kenvandine: can you file a bug report against apparmor-easyprof-ubuntu? [19:37] this is a nice interim improvement [19:37] sure [19:37] thanks! [19:37] kenvandine: is this what you talked about with mdes laur, et al last week when I was away? [19:37] yes [19:38] ok, I had a note to follow up with you, but didn't last week cause of uds [19:38] so, nice, get to check that off my list :) [19:38] same here :) [19:38] kenvandine: are you blocked on this rule being added? (ie, when do you need me to upload it?) [19:39] it'll block me merging this branch [19:40] but we have a pile of branches to merge, this isn't urgent [19:40] jdstrand, ^^ [19:41] kenvandine: ok, I'll queue it up for apparmor-easyprof-ubuntu 1.1.9. if for some reason that isn't in the archive by the time you are landing, holler [19:41] cool [19:41] jdstrand, thx! [19:41] np [19:43] kenvandine: one quick question. can you think of any reason why I wouldn't want to put this in the content_exchange policy group? (context: http://paste.ubuntu.com/7110042/) [19:44] (ie, I want to put it with the other content hub rules) [19:46] nope [19:46] in fact i'd rather it not be there [19:46] oh [19:46] sorry... inverse :) [19:46] ok [19:46] I was confused for a minute [19:46] you *do* want it with the others :) [19:46] if you put it in content_exchange, then an app could do the chmod 0 [19:47] but... in theory those apps wouldn't be requesting content [19:47] those apps can't request content [19:47] the dbus api is unavailable to them [19:47] i guess the only reason is paranoia then :) [19:48] deny rules are always evaluated last [19:48] putting it there is fine, up to you :) [19:48] so it it is in the policy at all, the app won't be able to write there [19:48] ok, good enough for me [19:49] kenvandine: ok, I'll work on it now actually-- I see I had another fix queued. I will add it to the 1.1 version of the policy, so make sure when testing, you use some ubuntu-sdk-14.04* framework in your click manifest [19:49] ok, will do [19:49] still need a bug report? [19:49] kenvandine: yes please, I'd like to reference it in the policy [19:52] jdstrand, bug 1293771 [19:52] bug 1293771 in apparmor-easyprof-ubuntu (Ubuntu) "Add deny rule in content_exchange" [Undecided,New] https://launchpad.net/bugs/1293771 [19:52] jdstrand, thanks! === bfiller_afk is now known as bfiller === dandrader is now known as dandrader|afk [20:39] hmm.. phablet-tools.settings has system_image_uri = 'https://system-image.ubuntu.com'. Should that be updated to https://system-image.ubuntu.com/ubuntu-touch? [20:50] stgraber: ack; thanks === chihchun is now known as chihchun_afk [21:15] ogra_, I'd like to remove a divert of NetworkManager's polkit policy file in lxc-android-config. I can't remember what the process is for dropping a divert. Add a if-upgrade-from-old-version dpkg-divert --remove call? [21:15] (to preinst?) [21:16] mterry, just drop it from the package [21:16] we dont support apt-get upgrade ... [21:16] and lxc-android-config is only installed on touch [21:16] ogra_, fair... especially not for that package [21:16] lxc-android-config fails to update 100% of the time anyway [21:16] that too [21:16] mterry, but please test that [21:17] ogra_, after testing, any theoretical objections to me dropping the divert? (it's actually a pointless divert these days) [21:17] iirc there is another mangling of the polkit setup in lvecd-rootfs at build time [21:17] (for NM) [21:17] ogra_, curious, will look [21:18] if you tested and dont see any issues with it (even with AP tests, run at least one) then yeah, just go ahead [21:18] (something like webbrowser-app ... that sets up a loopback webserver iirc) [21:18] ogra_, ah yeah, that livecd thing works around the fact that the lxc divert didn't do what it was supposed to [21:19] * ogra_ is happy to get rid of both hacks if we know dropping them doesnt do harm [21:19] (in fact i'd love to get rid of 90% of the hooks stuff in livecd-rootfs ... its butt ugly) [21:20] ogra_, well we may still want the livecd one... Or maybe move the hack into session-manager-touch in a different way [21:20] k [21:21] ogra_, the livecd hack is just to avoid having to challenge the user if they are admin [21:22] ah, right [21:22] (and now lightdm user will want that same privilege) [21:22] for network stuff anyway [21:23] do we use the lightdm user for anything on the touch images ? [21:23] ogra_, not today [21:23] (apart from firing up usc) [21:23] ogra_, not even firing up usc [21:23] ogra_, lightdm user is only for greeters [21:23] ok [21:30] mterry: surely the old package's prerm should've removed the diversion anyway [21:31] or postrm [21:31] cjwatson, on upgrades? I don't think so. Well, at any rate, it didn't [21:31] ah, yeah, only on remove|purge [21:31] so yeah, something would have to be in the new preinst if you wanted to handle upgrades [21:32] cjwatson, makes sense. Was hoping that would be fully automatic these days, but still pretty close to automatic :) === dandrader|afk is now known as dandrader === chihchun_afk is now known as chihchun === Ursinha is now known as Ursinha-afk [23:44] ogra_: I've been reading through the installer script. I'm confused by this line "do_shell "cd /cache/system/lib && ln -s /system/lib/modules modules" ". How does /system/lib/modules exist if boot.img and recovery,img are the only partitions flashed as per the README?