[00:00] <popey> hi
=== _salem is now known as salem_
=== chihchun_afk is now known as chihchun
[03:20] <savedjuli> hi
[03:35] <X-Droid> hello ubuntu touch
[03:38] <RobbyF> hello
=== salem_ is now known as _salem
[04:00] <plars> seeing lots of =========== WLAN placed in RESET ======== in the dmesg on the latest build, and having a lot of protocol faults from adb when trying to wait for the network to come up
[04:01] <plars> http://paste.ubuntu.com/5931417/
[05:03] <murgero> hey guys!
[05:05] <murgero> Anyone online?
[05:08] <murgero> okay then...
=== tgm4883_ is now known as tgm4883
=== Namidairo`bnc is now known as Namidairo
[06:01] <AskUbuntu> Could there be a graphics card in an ubuntu touch dock | http://askubuntu.com/q/326823
[07:04] <stife> does ubuntu only work on Nexus Devices?
[07:06] <RAOF> stife: No; see https://wiki.ubuntu.com/Touch/Devices
[07:07] <stife> thanks RAOF, one last question is the OS/ROM the same or different for each device?
[07:08] <RAOF> Kinda both.
[07:08] <RAOF> You need a different base for each device, because they need different kernel drivers etc.
[07:08] <RAOF> But above that, it's the same.
[07:09] <RAOF> Although having said that, I really mean "yes, just like Android"
[07:09] <stife> Alright, thanks, i'll see what i can do on my device.
[07:11] <dholbach> good morning
[07:40] <diwic> tvoss_, good morning, do you have a moment to talk about avoiding S3 when playing back audio?
[07:40] <tvoss_> diwic, sure, shoot :)
[07:41] <diwic> tvoss_, maybe you saw my email a day ago or two, but long story short, we're trying to move over to use PulseAudio instead of Audioflinger,
[07:41] <diwic> tvoss_, and my problem is that the system goes to S3 while audio is playing back
[07:41] <diwic> tvoss_, and people said you were the right one to ask
[07:44] <diwic> tvoss_, so the question is, how is this prevented in the best way?
[07:44] <diwic> tvoss__, how much did you miss?
[07:45] <tvoss__> diwic, most, the network is kinda flaky here :)
[07:45] <diwic> <diwic> tvoss_, maybe you saw my email a day ago or two, but long story short, we're trying to move over to use PulseAudio instead of Audioflinger,
[07:45] <diwic> <diwic> tvoss_, and my problem is that the system goes to S3 while audio is playing back
[07:45] <diwic> <diwic> tvoss_, so the question is, how is this prevented in the best way?
[07:51]  * ogra_ grins about bug 1206690 and bug 1206688
[07:52] <ubot5> bug 1206690 in ubiquity-slideshow-ubuntu (Ubuntu) "lubuntu ubiquity slideshow should provide OEM slideshow to show at OEM config stage" [Undecided,New] https://launchpad.net/bugs/1206690
[07:52] <ubot5> bug 1206688 in ubiquity (Ubuntu) "Ubiquity shouldn't have a slideshow if OEM mode is selected" [Low,Confirmed] https://launchpad.net/bugs/1206688
[07:52] <ogra_> oh, xnox fixed the title, sad :)
=== rickspencer3_ is now known as rickspencer3
[07:52] <tvoss__> diwic, I would think that AudioFlinger acquires a wake-lock somewhere, we would need to look through the source code of audioflinger and see which one
[07:53] <tvoss__> diwic, I would think pulse then needs a patch to acquire the lock, too
[07:54] <diwic> tvoss__, That's my understanding too. There was some question about whether PulseAudio was the proper place to do this, which is why I was supposed to ask you, because you got all the arcitecture in your head, or something :-)
[07:55] <diwic> tvoss__, but it probably *is* the right place to do things. I mean, if people start using alternate audio stacks (JACK for pro audio etc) then they need to take the wakelock manually too
[07:55] <tvoss__> diwic, yup, that's true. I would need to look through the code and see which wakelock is acquired. In general, though: having it in a service like pulse is ok, just apps and wakelocks is something we should avoid
[07:56] <diwic> tvoss__, I can look through AudioFlinger too - but are we supposed to do the same, or talk to powerd?
[07:58] <tvoss__> diwic, powerd is the component executing policy, but the services are fine to acquire wake locks
[07:59] <diwic> tvoss__, so I should talk to the kernel's wake lock API rather than powerd?
[08:01] <Debolaz> Has there been any discussion about porting Touch to S4? *Hopefull eyes*
=== rickspencer3_ is now known as rickspencer3
[08:03] <tvoss__> diwic, I will talk with the phonedation guys, let me quickly find them :) we are on a sprint today
[08:05] <diwic> tvoss__, ok, thanks, JFTR, which people are you referring to in particular? I'm on the phonedation daily standup myself.
=== schwuk_away is now known as schwuk
[08:11] <asac> hmm
[08:11] <asac> 10:06 < asac> phablet-flash is awful
[08:11] <asac> 10:06 < asac> connected maguro and getting
[08:11] <asac> 10:06 < asac>  phablet-flash
[08:11] <asac> 10:06 < asac> Device detected as
[08:12] <asac> 10:06 < asac> Unsupported device, autodetect fails device
[08:12] <asac> 10:06 < asac> When working on flipped images, detection does not work and would require -d
[08:12] <asac> 10:06 < asac> i know that it worked onflipped images
[08:12] <asac> 10:07 < asac> a) can we remove that outdateed 'doesnt work on flipped" message
[08:12] <asac> 10:07 < asac> b) can you help me?
[08:12] <asac> 10:07 < asac> ogra_: sergiusens: ^^
[08:12] <asac> 10:09 < asac> usb info is fine:
[08:12] <asac> 10:09 < asac> [48898.588802] usb 1-1.2: New USB device found, idVendor=18d1, idProduct=d002
[08:12] <asac> 10:09 < asac> [48898.588814] usb 1-1.2: New USB device strings: Mfr=2, Product=3, SerialNumber=4
[08:12] <asac> 10:09 < asac> [48898.588821] usb 1-1.2: Product: Galaxy Nexus
[08:12] <asac> (not sure if it came through
[08:12] <asac> )
[08:13] <ogra_> asac, the udev rule broke (for whatever reason)
[08:13] <ogra_> adb kill-server && sudo adb start-server
[08:13] <ogra_> then try again it should be detected
[08:15] <asac> ogra_: same problem
[08:15]  * popey hands asac http://paste.ubuntu.com
[08:16] <asac> popey: com'on
[08:16] <asac> :)
[08:16] <asac> there was not much going on here :)
[08:17] <ogra_> asac, for the logs :P
[08:17] <ogra_> not because of the paste
[08:19] <ogra_> asac, can we have a console log ?
[08:20] <ogra_> also what state is the device in when it complains ? recovery, normal boot, bootloader screen ?
[08:20] <bfiller> ogra_: can we respin an image that has qtubuntu 0.52+13.10.20130731-0ubuntu1 in it? it's needed to make autopilot pass for webbrowser
[08:20] <ogra_> bfiller, if you can convince didrocks  to have daily release running eralier (else the normal cron build starts in 2h)
[08:21] <ogra_> *earlier
[08:21] <bfiller> ogra_: that package got released already I think, just didn't make it into image
[08:21] <ogra_> bfiller, the buildd is occupied as well, not sure how long the current build will still run
[08:21] <didrocks> yeah, that one is already in the release pocket
[08:22] <didrocks> so ogra_ can run it manually
[08:22] <ogra_> didrocks, but i'll miss the rest of daily-release
[08:22] <didrocks> (5 hours ago actually)
[08:22] <didrocks> ogra_: right, you can just run 2 of them
[08:22] <didrocks> the manual one
[08:22] <didrocks> and then the cronned one
[08:22] <ogra_> i thought i was supposed to wait to make sure all of them get in
[08:22] <didrocks> ogra_: well, it will be the first image of the day, then, you will have another one
[08:22] <ogra_> didrocks, bfiller, well, i cant ... there is an ubuntu desktop build running
[08:22] <didrocks> what is pushed to distro is coherent
[08:23] <ogra_> and livefs builds arent stoppable
[08:23] <didrocks> ogra_: so, how this is related to daily release then?
[08:23] <didrocks> as you wanted to move it back from 2 hours
[08:23] <didrocks> which was 20 minutes from now
[08:23] <ogra_> (well, they are with IS intervention)
[08:23] <bfiller> ogra_: fine with waiting till next build if it happens today, thought the build happened already
[08:23] <ogra_> didrocks, i moved it 2h forward on your request
[08:23] <didrocks> ogra_: right, but if it started 2h ago, it will be blocked right now?
[08:24] <didrocks> as we have the desktop build in progress?
[08:24] <ogra_> it would be done :)
[08:24]  * didrocks puzzled, the desktop build isn't before?
[08:24] <ogra_> there is a gap before desktop and one after desktop+server+core ... which usually run in order
[08:24] <didrocks> ok, makes more sense
[08:24] <ogra_> so i can have touch before or after this
[08:25] <didrocks> anyway, if you want to manual publishing at 3am for packaging change, I'm happy with running the iso build beforehand :)
[08:25] <ogra_> with the work of the release team last week this will all get better soon
[08:31] <asac> ogra_: which console log?
[08:31] <asac> debvice or laptop?
[08:31] <ogra_> laptop ould be a start
[08:31] <ogra_> *would
[08:31] <ogra_> just copy paste the terminal output
[08:31] <ogra_> does adb see the device ?
[08:32] <asac> ogra_: i can do adb shell
[08:32] <asac> if thats what you mean
[08:32] <ogra_> and what mode is the device in ?
[08:32] <ogra_> recovery, normal boot ?
[08:33] <asac> ogra_: not booted ... battery
[08:33] <asac> charge
[08:33] <ogra_> and you can adb shell ?
[08:33] <asac> sure i can
[08:33] <asac> i always can do it there
[08:33] <ogra_> thats weird
[08:33] <asac> once the battery icon is there i can do it
[08:33] <asac> that said, it will uncharge
[08:33] <asac> e..g maguro is never off
[08:33] <ogra_> well, thats surely not a state we support
[08:34] <ogra_> i wasnt even aware you can adb in that state since it forcefully gets stopped if it is in the charger mode
[08:34] <asac> but i cannot boot it
[08:34] <ogra_> because you drained it to 0 ?
[08:35] <asac> ogra_: sure... but its charging for 15 minutes now
[08:35] <ogra_> if there is still power in the battery, do a proper reset (remove battery and cable) and boot :) *after* you booted, re-=attach the cable
[08:36] <ogra_> thats an android problem we dont have a solution for yet ... i would actually like to rip out that shit completely
[08:36] <asac> hmm ... let me wait a few more minutes to get it charged a bi tmore
[08:37] <ogra_> the android initd has  an app called charger ... that causes the boot to lock as soon as  it detects that a cable is attached
[08:37] <ogra_> getting out of that state is horridly hard on all devices
[08:37] <asac> :(
[08:37] <asac> ok let me try in five
[08:38] <ogra_> i was planning to do a  more proper implementation in the ubuntu initrd once we have a way to use the display
[08:38] <ogra_> and rip out that stuff on the android side where we can
[08:38] <asac> +1
[08:38] <asac> can we have an ubuntu branded battery charge animation?
[08:38] <ogra_> (might be that it is hardcoded in the bootloader on some devices, there we wont be able though)
[08:39] <ogra_> if we can display something on the framebuffer from initrd (which we cant yet) we can have everything you want :)
[08:39] <ogra_> (the missing diaply support moved the issue down on the TODO)
[08:40] <asac> so maguro does this animation in the bootloader? i dont think so because i can adb shell into it
[08:40] <asac> while it shows it
[08:40] <ogra_> bug 1204630 is related btw
[08:40] <ubot5> bug 1204630 in touch-preview-images "Maguro fails to boot after command-line shutdown" [Undecided,New] https://launchpad.net/bugs/1204630
[08:40] <ogra_> asac, no, it does it in the initrd
[08:40] <ogra_> there is an app /sbin/charger
[08:41] <ogra_> you should see it running
[08:41] <ogra_> it inhibits the boot and shows the animation
[08:42] <ogra_> (btw, replacing the animation of this one should be easy ... but that wont fix the horrid user experience)
[08:43] <asac> ogra_: if i install a new kernel it was just installing the .deb?
[08:43] <ogra_> no, you need to manually run flash-touch-kernel with the path to the kernel
[08:44] <ogra_> let me fix that so it gets auto installed :)
=== dholbach_ is now known as dholbach
[08:50] <ogra_> fix uploaded :)
[08:53] <asac> ogra_: so i unpack the deb
[08:53] <asac> copy the vmlinuz over
[08:53] <asac> and use flash-touch-kernel?
[08:53] <ogra_> why so complicated
[08:53] <ogra_> dpkg -i the deb
[08:53] <asac> kk
[08:53] <ogra_> flash-touch-kernel /boot/vmlinuz-$balh
[08:53] <ogra_> *blah even
[08:54] <ogra_> (whatever the version is)
[08:54] <ogra_> my fix will call that last one automatically from next image build on
[08:54] <ogra_> so in the future dpkg -i will suffice
[08:55] <asac> ogra_: i want to kill dpkg and apt as you know :)
[08:55] <asac> guess will take 3 more weeks though :)
[08:55] <asac> j.k.
[08:55] <asac> guess not for october
[08:55] <ogra_> asac, well, but as long as we use it ... it was a cp and three chanrs to change to add support :)
[08:55] <asac> ogra_: dpkg -i doesnt work
[08:56] <ogra_> i had it on my TODO anyway ...
[08:56] <asac> so thats stupid
[08:56] <asac> i will just do the unpoacking
[08:56] <ogra_> why wouldnt dpkg -i work ?
[08:56] <asac> it wants to update initramfs etc.
[08:56] <ogra_> and ?
[08:56] <ogra_> let it
[08:56] <asac> http://paste.ubuntu.com/5932050/
[08:56] <asac> all crapware
[08:56] <asac> let me unpack and just do it without that tech
[08:57] <ogra_> asac, that looks like some more serious issue
[08:57] <ogra_> i.e. like something is wrong with your filesystem
[08:57] <asac> not sure
[08:57] <asac> i could copyu manyally to /boot
[08:57] <asac> and flash
[08:57]  * asac happy
[08:57] <asac> let me reboot
[08:57] <ogra_> well
[08:57]  * didrocks can see asac happy
[08:58] <ogra_> dpkg -i should still work
[08:58] <ogra_> is any of your disks readonly due to a fs error ?
[08:58] <asac> ogra_: dunno ... i can copy to /boot
[08:58] <asac> isnt that the main partition>
[08:58]  * asac waits for reboot to finish
[08:58] <ogra_> pastebin the output of mount
[08:59] <asac> let me waittill the system comes up
[08:59] <ogra_> that looks like your flash is borked or some such
[08:59] <asac> (if it comes up )
[08:59] <asac> http://paste.ubuntu.com/
[08:59] <asac> http://paste.ubuntu.com/5932062/
[08:59] <asac> thtas state without finishing boot
[08:59] <JamesTait> Good morning all, happy Uncommon Instrument Awareness Day! :-D
[09:00] <ogra_> hmm, looks all fine
[09:00] <asac> it doenst come up :(
[09:00] <ogra_> weird
[09:00] <asac> apw: can you gen a new kernel with the mem=! option?
[09:00] <asac> i think the one you uploaded is not compatible anymore
[09:00] <ogra_> asac, where does it hang ?
[09:00] <asac> [   94.383422] PVR_K:(Error): BridgedDispatchKM: Initialisation failed.  Driver unusable. [4783, /home/apw/build/ubuntu-saucy/ubuntu-saucy/drivers/gpu/pvr/bridged_pvr_bridge.c]
[09:00] <asac> ogra_: ^^
[09:00] <ogra_> do yoou have adb to debug ?
[09:00] <asac> Google on screen
[09:00] <asac> yes i am in
[09:00] <asac> see above
[09:01] <asac> i use http://people.canonical.com/~apw/maguro-saucy/
[09:01] <asac> but didnt change boot args
[09:01] <ogra_> well, i'm not sure thats not just a symptom
[09:01] <ogra_> check with /system/bin/logcat -d ...
[09:01] <asac> ogra_: should i try to reflash?
[09:01] <ogra_> do you see a lot of sensorservice messages ?
[09:02] <asac> ogra_: where would i see those?
[09:02] <ogra_> in the logcat
[09:02] <asac> ogra_: note: i am still on 26th image
[09:02] <asac> or wait
[09:02] <asac> le tme really reflash fresh
[09:02] <asac> who knows what ChickenCutlass did
[09:02] <ogra_> cat /var/log/installer/media-info
[09:02] <ogra_> :)
[09:02] <asac> (he installed the image i have on it now)
[09:03] <asac> cat /var/log/installer/media-info
[09:03] <asac> Ubuntu Saucy Salamander (development branch) - armhf (20130726.1)
[09:03] <ogra_> yeah
[09:03] <ogra_> so you are on 26th
[09:03] <popey> stgraber: the image based update is missing /var/log/installer/media-info - is that intentional? want a bug filed?
[09:03] <ChickenCutlass> asac, ogra_  what did I do?
[09:03] <asac> ogra_: logcat is not avail on ubuntu root
[09:03] <ogra_> popey, i think davmor2 filed one yesterday
[09:03] <asac> ChickenCutlass: ignore :)
[09:03] <ogra_> asac, huh ?
[09:03] <asac> ChickenCutlass: you instlaled my current install... and that is not busted
[09:03] <ogra_> asac, copy paste the command i gave you
[09:04] <ogra_> it definitely is there
[09:04] <ogra_> else something else is seriously broken
[09:04] <asac> ogra_: which command?
[09:04] <asac> oh nevermind
[09:04] <ogra_> <ogra_> well, i'm not sure thats not just a symptom
[09:04] <ogra_> <ogra_> check with /system/bin/logcat -d ...
[09:04] <ogra_> it isnt in $PATH, but it is there
[09:04] <asac> ogra_: http://paste.ubuntu.com/5932080/
[09:05] <asac> i realy think the dpkg -i just trashed my initrd
[09:05] <asac> or something
[09:05] <asac> let me do a fresh flash
[09:05] <asac> who kinows what this failing dpkg -i did to my thing
[09:05] <ogra_> it shouldnt do any harm
[09:05] <asac> wlel. i tworked
[09:05] <asac> i flashed the kernel from apw (which workd 1-2 weeks ago)
[09:05] <asac> now i have what you see in paste
[09:06] <ogra_> asac, yeah, thats the kernel
[09:06] <ogra_> definitely
[09:06] <ogra_> if the initrd would be broken you wouldnt be in a full filesystem with adb
=== Namidairo is now known as Namidairo`bnc
[09:06] <asac> ogra_: did we change the kernel for maguro at all?
[09:06] <ogra_> it would stop on boot and give you a busybox shell (like on x86)
[09:06] <asac> in last 2-3 weeks?
[09:06] <ogra_> i dont think we did, let me check -changes
[09:06] <asac> ogra_: it is at googlke prompt and i can  adbh shell
[09:07] <asac> apw: help :)
[09:07] <asac> wait ... let me really reflash
[09:07] <ogra_> 22h was the last upload
[09:07] <asac> i believe its all busted because of drainage to zero or something
[09:07] <ogra_> right, the google prompt just shows that surfaceflinger cant start
[09:07] <asac> ogra_: of maguro? but we dont have ne pvr stuff?
[09:07] <ogra_> and the PVR error above agrees that this is a kernel issue
[09:08] <ogra_> maguro is OMAP4, we sure do
[09:08] <asac> ogra_: did we get a new pvr binary or something?
[09:08] <apw> asac, i will make you a more current kernel
=== Namidairo`bnc is now known as Namidairo
[09:08] <asac> apw: did something change on maguro that might cause incompatibilty with pvr?
[09:08] <ogra_> asac, i think rsalveti updated the android bits to 4.2.2 recently
[09:08] <asac> apw: thanks a bunch ... wnat to demo the mem=!300m thing to jane this afternoon
[09:08] <apw> nothing changed that we know of
[09:08] <ogra_> but i also think that was before the 22nd ... not sure
[09:09] <asac> right
[09:09] <asac> let me first do a fresh reflash
[09:09] <asac> i believe ogra is to blame in general in any case :)
[09:09] <asac> j.k. :)
[09:09] <ogra_> yeah, i'm easy to blame ... just blame me :)
[09:09] <asac> ogra_ regressed to cowboy mode :)
[09:10] <asac> oki am reflashing /home/asac/Downloads/phablet-flash/ubuntu-touch/20130726.1/saucy-preinstalled-touch-armel+maguro.zip
[09:10] <ogra_>   * rebase to cm-10.1.2
[09:10] <ogra_>     Variable GPU OC: sysfs interface to cycle through different top gpu speeds
[09:10] <ogra_> ....
[09:10] <ogra_>     gpu: pvr: Update to DDK 1.8@2198402
[09:10] <ogra_> ....
[09:10] <ogra_> from the last upload
[09:10] <asac> ogra_: did that land before 26.1 ?
[09:10] <ogra_> (linux-maguro)
[09:10] <ogra_> that landed on the 22nd
[09:10] <asac> cool
[09:10] <asac> apw: so yeah i guess i need a new one
[09:11] <stgraber> popey: hmm, I'm not removing that file from the standard rootfs, so I'm not sure what creates it on standard flipped
[09:11] <asac> apw: so i think i want to file the bug properly now... where do i file it again? just linux?
[09:11] <ogra_> stgraber, live-build
[09:11] <ogra_> stgraber, it should be in the tarball as well as in the zip
[09:11] <apw> asac, linux-maguro
[09:13] <stgraber> ogra: yeah, so it should be in the read-only images... weird... I'll have to look at that when I'm back on Friday
[09:13] <ogra_> stgraber, you dont repack the tarball beyond making it xz, right ?
[09:13] <ogra_> hmm, or do you handle /var/log in a special way ?
[09:14] <didrocks> stgraber: leave, run away before it's too late! :)
[09:14] <ogra_> (during boot)
[09:14] <stgraber> ah, yeah, that's it... /var/log is a bind-mount from writable storage
[09:14] <ogra_> ah
=== vrruiz_ is now known as rvr
[09:14] <ogra_> so you might want to copy that file over or some such
[09:15] <asac> apw: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1206835
[09:15] <stgraber> I'll have to tweak that so we can still access it
[09:15] <ubot5> Launchpad bug 1206835 in linux (Ubuntu) "please add support for forceful mem=!XXXm for phablet kernels" [Undecided,New]
[09:15] <popey> stgraber: also, no swap on the new image based updates, but I believe you know about that?
=== freeflyi1g is now known as freeflying
[09:16] <stgraber> popey: yeah, no swap is known
[09:16] <popey> k
[09:16] <ogra_> stgraber, i could put the file elsewhere ... but i konw ubuntu-bug and whoopsie use it ... so a link would be required at least
[09:17]  * popey gets bitten by bug 1190792
[09:17] <ubot5> bug 1190792 in touch-preview-images "ueventd in a busy loop on container-flipped image" [High,In progress] https://launchpad.net/bugs/1190792
[09:17] <stgraber> right, moving it would be a problem
[09:17] <ogra_> popey, try adding a "sleep 3" to /etc/init/lxc-android-config.conf ... right above the lxc-wait
[09:18] <asac> apw: will be back in 20-30min
[09:18] <asac> etc.
[09:18] <asac> thanks for your help
[09:18] <ogra_> i'll upload that today wit a bunch of changes needed for galxy S2 support
[09:19] <popey> ogra_: this is on a device i flash every day so that won't help really
[09:19] <ogra_> stgraber, yeah, but a link seems easier than having to copy it around , i can ship it as /etc/media-info for example
[09:19] <ogra_> popey, well, tomorrows image will have the sleep
[09:19] <popey> great!
[09:19] <ogra_> just to prevent you from running into it today :)
=== schwuk is now known as schwuk_away
[09:21] <apw> asac, kernel update in http://people.canonical.com/~apw/maguro-saucy/
[09:21] <ogra_> apw, i need some kernel cmdline advice ... the SGS2 (galaxy S2)  i'm just trying to port to has hardcoded teh cmdline in ATAGs .in the bootloader ... it uses console=ram there which makes upstart fall over, the kernel options i see only seem to allow to prepend to the hardcoded cmdline ....
[09:22] <ogra_> do you know a way how to append to it ?
[09:22] <ogra_> i can boot fine if i force it to ignore the ATAG line but then i need to hardcode it in the kernel binary (there is a hw seriall that i need from the atags though)
[09:23] <ogra_> which in turn breaks recovery mode ... which gets another option prepended when active
[09:25] <apw> ogra_, kernel command line processing is completely boot loader specific really ... right?
=== schwuk_away is now known as schwuk
[09:25] <apw> ogra_, and you are saying it is appending something bad and that is out of your control ?
[09:25] <ogra_> well, kernel cmdline is hardcoded in the unchangeable part of the bootloader
[09:26] <ogra_> and the normal samsung devices dont use an actual boot.img like we know it
[09:26] <apw> you said you could prepend to it?
[09:26] <ogra_> so i cant change it there
[09:26] <ogra_> if i set the kernel option to prepend to it, the hardcoded one from the kernel gets prepended, yes
[09:26] <ogra_> but that still leaves me with console=ram as the last console= arg
[09:27] <ogra_> i need a way to keep the ATAG cmdline and to append a new console=tty1 or so
[09:27] <apw> you can add what you want on the front, but end up with fix stuff shoved on the end, and you need some of it
[09:27] <apw> but not all of it ?
[09:27] <ogra_> i just need a console= at the end
[09:28] <apw> right which is impossible with your bootloader as i understand it
[09:28] <ogra_> the rest should come from the ATAGs (which dynamically changes if oyu press the key combo for recovery for examlple)
[09:28] <apw> can you send me a full command line with something prepended
[09:28] <ogra_> right
[09:28] <apw> and mark in it with | where the split is
[09:28] <apw> and i'll see what we can do
[09:29] <apw> i may be able to offer you a like console=NOMORECONSOLES which stops later ones working or something
[09:29] <ogra_> that will take a while, since i cant really boot (need to create a kernel with the right hardcoded cmdline runes to pick it out of the initrd shell)
[09:29] <apw> ogra_, well type a sample you can make it up ... i want to see what want the command line to look like
[09:30] <ogra_> note that this is a community port btw ... we dont have the kernel in kernel.u.c
[09:30] <apw> ogra_, and a simulation of how it looks as close as you would get it with the bootloader adding extra junk
=== alan_g is now known as alan_g|tea
[09:30] <apw> ogra_, then i would need to know where the kernel is as well
[09:33] <ogra_> apw, like that http://paste.ubuntu.com/5932151/ ?
[09:35] <apw> ogra_, so can we add multiple things at the front, say console=tty1 console=STOP
[09:36] <ogra_> the kernel tree is on github afaik https://wiki.ubuntu.com/Touch/Devices/i9100
[09:36] <ogra_> apw, that should be possible
[09:36] <ogra_>  <project name="Ubuntu-i9100/android_kernel_samsung_smdk4412" path="kernel/samsung/smdk4412" remote="github" />
[09:36] <ogra_> (from the xml entry)
=== alan_g|tea is now known as alan_g
[09:37] <apw> ogra_, so then what i would so is mod the console handler to lock the console list so the next one cannot add
=== b0bben_ is now known as b0bben
[09:37] <ogra_> yeah, that sounds like it could work
[09:38] <asac> apw: rock on ... let me try
[09:39] <ogra_> apw, ah, found it https://github.com/Ubuntu-i9100/android_kernel_samsung_smdk4412
[09:42] <asac> ogra_: abootimg -u $(find /dev -name boot|head -1) -c 'mem=!386m'
[09:42] <asac> gives me bad config entry
[09:42] <ogra_> apw, what i dont get is why the CONFIG_CMDLINE_EXTEND can only prepend ...
[09:42] <ogra_> is there any technical reason for that ?
[09:42] <asac> abootimg -u $(find /dev -name boot|head -1) -c 'mem=!386m'
[09:42] <asac> ogra_: wasnt that the way to set it?>
[09:42] <ogra_> asac, close :)
[09:42] <ogra_> abootimg -u $(find /dev -name boot|head -1) -c 'cmdline=mem=!386m'
[09:42] <asac> aye :)
[09:43] <asac> me goes for 300m
[09:43] <ogra_> brave
[09:44] <asac> guess 250m would be even cooler for demo
[09:44] <asac> 128m :)
[09:44] <ogra_> well, luckily we wont go for such low specs
[09:44] <apw> ogra_, what is this tag command line, where is it getting it from i wonder
[09:44] <asac> when will we kill swap?
[09:44] <asac> ogra_: can you just do it?
[09:44] <ogra_> apw, bootloader in ROM
[09:44] <asac> or when will we do it?
[09:45] <ogra_> asac, with the readonly images i belive
[09:45]  * apw reminds people that turning off swap has negative ramification when we are close to running out of memory
[09:45] <apw> you need to be very very careful around the no-memory boundary if you have no swap
[09:46] <apw> swap is more than a buffer space it also allows the kernel to pemptivly make more memory throwawayable
[09:46] <ogra_> apw, http://paste.ubuntu.com/5932193/
[09:46] <apw> which make it more likely to survive when there is none
[09:46] <ogra_> thats s/proc/last_kmsg 9from several boots, it doesnt seem to flush it)
[09:47] <ogra_> (scroll to the bottom)
[09:47] <apw> ogra_, so there is no technical reason that this is append only, that is just what the code does
[09:47] <ogra_> samsung uses a weird three stage bootloader
[09:47] <apw> though doing anything other than appending is much much harder
[09:48] <ogra_> "Welcome to Samsung Primitive Bootloader."
[09:48] <apw> as you have nowhere to handle the bits
[09:48] <ogra_> :)
[09:48] <ogra_> well appending instead of prepending would be totally enough
[09:49] <ogra_> the ATAGs can dynamically change based on what buttons on the phone you press
[09:49] <ogra_> currently i can only get it to boot by copying the full atag cmdline into the kernel ... but that breaks i.e. recovery mode
[09:50] <apw> yep
[09:50] <apw> ogra_, so what kernel base is this heap
[09:50] <ogra_> because the dynamic bit is missing
[09:50] <apw> oh god 3.0
[09:50] <apw> this isn't a kernel
[09:50] <ogra_> root@ubuntu-phablet:/# uname -a
[09:50] <ogra_> Linux ubuntu-phablet 3.0.64-CM-g27bbb67-dirty #10 SMP PREEMPT Tue Jul 30 12:45:16 CEST 2013 armv7l armv7l armv7l GNU/Linux
[09:50] <ogra_> .64
[09:50] <ogra_> :)
[09:53] <kalikiana> bah, I'm making an alias "killall adb; adb" now
[09:53] <kalikiana> it's so annoying to keep running into random errors
[09:54] <ogra_> kalikiana, make sure to file bugs :)
[09:55] <kalikiana> ogra_: I did. aside from the fact it's so glaringly obvious
[09:57] <asac> wow our 26 image is really not working well :)
[09:57] <ogra_> really ?
[09:57] <asac> yeah taking photos dont save the pics on the disk
[09:57]  * ogra_ guesses most people use --pendind anyway 
[09:58] <asac> ogra_: btw, we have to kill that option or randomly rename it
[09:58] <ogra_> *pending
[09:58] <asac> because we want the majority of folks to not use --pending
[09:58] <asac> but not now :)_
[09:58] <ogra_> we surely want the option to verify th epending image
[09:58] <asac> the idea is that not having something in our published image creates back pressure
[09:58] <asac> ogra_: we want an option
[09:58] <ogra_> its not our fault if people use it
[09:58] <asac> but since everybody is using that weshould raname it
[09:58] <asac> regularly :)
[09:59] <ogra_> how would that change anything
[09:59] <ogra_> people will just use it with the new name
[09:59] <asac> ogra_: can we at least not have --pending documented?
[09:59] <asac> on the wiki?
[09:59] <asac> hide it from --help?
[09:59] <ogra_> convince them to not always need the latest and greatest :)
[09:59] <asac> or make it an env? GANGNAM_STYLE=true phablet-flash :)
[10:00] <ogra_> see the mailing list, people complain and are confused about not getting the last from cdimage by default
[10:00] <popey> change --pending to --broken-crack
[10:00] <asac> i think thats fine. just explain to them that it doesnt pass our rising, but still pretty low quality bar
[10:00] <ogra_> heh
[10:00] <asac> i really want GANGNAME_STYLE :)
[10:01] <asac> or --cowboys-for-the-world
[10:01] <asac> OGRA_STYLE
[10:01] <popey> hah
[10:01] <ogra_> --pending-and-i-know-this-has-not-been-tested-yet-i-will-olnly-file-bugs-that-are-valid
[10:01] <popey> </bikeshed>
[10:01] <asac> aye
[10:01]  * ogra_ votes for that ^^
[10:01] <popey> me too
[10:01] <asac> ack
[10:01] <asac> do it
[10:01] <asac> with typos plz :)
[10:01] <popey> although I'd spell olnly as "only" ☻
[10:01] <ogra_> and no tab complettion allowed :)
[10:01] <popey> hah
[10:02] <asac> --penidng-...
[10:02] <cjwatson> I would just hide it from --help if I were you
[10:02] <asac> and from wiki for sure
[10:02] <ogra_> after all that will just raisee the noise though
[10:02] <cjwatson> There isn't much point embarking on an arms race
[10:02] <ogra_> more people will ask on the ML then
[10:02] <popey> change the parameter each day
[10:02] <ogra_> i would just leave it as is
[10:03] <asac> sure. just remove doc
[10:03] <asac> and dont hand it out to everyone
[10:03] <ogra_> no
[10:03] <cjwatson> Compare the "hidden" URLs that we use during pre-publication of images just before release
[10:03] <ogra_> just leave it as is :)
[10:03] <cjwatson> We know a bunch of people know where they are - but honestly, it doesn't matter that much
[10:03] <popey> yeah, nature finds a way
[10:03] <ogra_> else you will have a thread about "how do i get the latest image" once a week on the ML
[10:03] <cjwatson> So we just learned to be relaxed about it
[10:03] <asac> "latest image is what you get with phablet-flash" is the answer
[10:03] <ogra_> right
[10:04] <ogra_> asac, no, it isnt ... "i want to test feature XY that got in yesterday, but phablet-flash only gets me something from last week, should i file a bug ?"
=== gusch_ is now known as gusch|lunch
[10:04] <asac> ogra_: "feature XY didnt make it throug our gate, sorry"
[10:04] <ogra_> thats what we have on the ML a lot
[10:05] <ogra_> dropping the option from help will just increase that noise
[10:05] <cjwatson> ogra_: And the solution to that is not to let current get so far behind pending
[10:05] <cjwatson> It's just straight technical debt
[10:05] <asac> exactly
[10:05] <ogra_> cjwatson, exactly
[10:05] <ogra_> so lets fix the cause, not the symptom :)
[10:06] <ogra_> btw, seems slangasek respun the images and didnt notify QA ... the tests still need a lot of handholding
[10:06] <ogra_> so nothing passed since nobody resterted the broken ones
[10:06] <cjwatson> Surely QA are autonotified
[10:06] <cjwatson> At least the automatic tests
[10:06] <ogra_> i'm not sure
[10:06] <ogra_> yeah
[10:06] <cjwatson> If automatic QA is relying on images appearing at a particular time then that's a problem
[10:06] <ogra_> but they need someone watching them through the whole process
[10:06] <cjwatson> Not so automatic then :)
[10:07] <ogra_> since they fequently just hang or dont even start
[10:07] <ogra_> *frequently
[10:07] <asac> ogra_: you are supposed to poke jcollado to rekick failed jobs
[10:07] <asac> (like everyone)
[10:07] <ogra_> well, semi automatic, but racy and crashy
[10:07] <ogra_> asac, sorr, not at 4am
[10:07] <ogra_> i have to sleep sometimes :P
[10:08] <cjwatson> Spot the signs of a project relying on heroism ...
[10:08] <ogra_> asac, if someone does a re-spin at night he needs to notify QA to watch the test run :)
[10:08] <ogra_> i guess i should write a mail :)
[10:08] <cjwatson> QA might not be around either depending on the time.  I expect they need to sleep too
[10:09] <ogra_> they get up at 4pm every day :P
[10:09] <cjwatson> Perhaps it'd be better to say that if we have multiple image builds per day then maybe only one of them gets fully tested
[10:09] <ogra_> if someone sleeps that long they surely can work later ... (... americans....)
[10:09] <asac> we need to scale up and fix fix fix stuff
[10:09] <ogra_> :)
[10:09] <ogra_> asac, yeah, known to everyone ...
[10:10] <ogra_> cjwatson, yeah, the regular one usually gets tested
[10:10] <asac> ogra_: do you have maguro?
[10:10] <ogra_> and the broken respin clearly shows we still have a long way to go
[10:10] <asac> can you check if taking a photo saves that in /home/phablet/Pictures for you?
[10:10] <asac> i tried 26.1 and --pending
[10:10] <asac> both seem to have that issue
[10:11] <ogra_> asac, works, let me check which image i have here
[10:12] <popey> image20130731_0001.jpg
[10:12] <popey> works here
[10:12] <popey> Ubuntu Saucy Salamander (development branch) - armhf (20130731)
[10:12] <ogra_> hmm
[10:12] <ogra_> 20130725.1
[10:13] <ogra_> asac, is your disk full ?
[10:14] <ogra_> (that would also explain the initramfs issues you saw)
[10:16] <stgraber> ogra_: ah yeah, that'd be easier
[10:16] <ogra_> stgraber, great, i'll ship it there then and put the linnk in place during build, you can do the same from initrd then
[10:17] <asac> ogra_: dont think so
[10:17] <asac> /dev/mmcblk0p12  13975888 1499432  12476456  11% /
[10:17] <asac> /dev/mmcblk0p12  13975888 1499432  12476456  11% /data
[10:18] <ogra_> yeah. looks fine
[10:18] <asac> ogra_: ^^
[10:18] <asac> it doesnt save images
[10:18] <asac> even without hacking mem=...
[10:18] <asac> what i observe on maguro:
[10:18] <asac> 1. start camera app
[10:18] <ogra_> can you touch a file in /home/phablet/Pictures ?
[10:18] <asac> 2. see what the camera sees live (good)
[10:18] <asac> 3. take photo
[10:18] <asac> 4. photo slides down and in background i see the moving picture
[10:19] <asac> 5., aftrer slide is done it shows the pic taken
[10:19] <asac> and not the live picture (e.g. video style)
[10:19] <asac>  /home/phablet/pictures -> empy
[10:19] <ogra_> well, the camera stores the raw image in ram afaik
[10:19] <ogra_> its a capital P
[10:19] <asac> i know
[10:19] <ogra_> k
[10:20] <asac> phablet@ubuntu-phablet:~$ touch Pictures/asda
[10:20] <asac> phablet@ubuntu-phablet:~$
[10:20] <ogra_> hmm, seems fine
[10:21]  * ogra_ syncs the last images to test here 
[10:21] <asac> as i said... doesnt matter whether i poke mem or not
[10:21] <asac> doesnt work
[10:21] <asac> ogra_: had the same issue with 26.1 and --pending from 5 minutes ago
[10:21] <ogra_> it definitely works on my installed image
[10:21] <ogra_> but thats 25.1
[10:21] <asac> yeah
[10:21] <popey> i tested fine on mako
[10:21] <asac> ogra_: is that still somewhere so i can use it for demo>
[10:21] <ogra_> gimme 30min torsync
[10:21] <asac> popey: mako is different yes
[10:21] <ogra_> *to rsync
[10:21] <popey> kk
[10:21] <asac> ogra_: 30 minutes
[10:21] <asac> ?
[10:22] <ogra_> 2Mbit :)
[10:22] <asac> it takes like 3 minutes to download this stuff on decent bandwidth
[10:22] <asac> omg
[10:22] <ogra_> thats the poit, decent :)
[10:22] <asac> kill that contract, get 30 or 100
[10:22] <asac> with 6m up
[10:22] <ogra_> not contract realted
[10:22] <asac> MOVE to an area with proper infrastructuure
[10:22] <ogra_> the wiring doesnt allow more
[10:23] <ogra_> i wish i could move :P
[10:23] <asac> file a complain with the human rights court in kassel :)
[10:23] <ogra_> haha
[10:23] <asac> its cruel
[10:23] <ogra_> its the wiring on my property
[10:23] <asac> just put a cable overland
[10:23] <asac> overground
[10:23] <ogra_> replacing the wiring from the street to the house is needed ... the wires are from 1960something
[10:23] <asac> :)
[10:23] <asac> right. you just take a cable, go to the street and tell them to connect that
[10:23] <ogra_> i'm not american, thanks :P
[10:24] <asac> maybe you can rent a shared office?
[10:24] <asac> and not work from home (might have other positive effects on your lifestyle and health :))
[10:24] <ogra_> if you raise my salary to pay the difference :P
[10:25]  * ogra_ is happy how he works ... and i'll get a new line at some point ... just not right now 
[10:25] <murgero> Morning from chicago!
[10:25] <asac> ogra_: you can offer ubuntu support for the office :)
[10:25] <asac> then you get if for free
[10:25] <ogra_> great, and become the support bitch
[10:25] <ogra_> awesome idea
[10:25] <asac> hehe
[10:25] <ogra_> :)
[10:25] <murgero> Woah guys lets calm down a bit lol
[10:26] <murgero> how is everyone?
[10:26] <ogra_> murgero, great, no worries
[10:26] <ogra_> we're just joking
[10:26] <murgero> I figured lmao
[10:26] <asac> ogra_: where can i get 25.1 ?
[10:26] <davmor2> ogra_: you can't be anybody else support bitch you are already mine, now how are the images today support ;)   /me runs and hides :)
[10:27] <asac> man we have really sucky retention
=== alan_g is now known as alan_g|afk
[10:27] <ogra_> asac, i think popey keeps a mirror
[10:27] <asac> yoiu cannot really work that way
[10:27] <asac> popey: :)
[10:27] <asac> gimme secret url
[10:27] <murgero> Wow
[10:27] <asac> ogra_: ho do i install a special image
[10:27] <asac> ?
[10:27] <asac> popey: /msg
[10:27] <ogra_> i would do it manually :)
[10:27] <AskUbuntu> Ubuntu Touch Apps don't recognize Ubuntu MultiTouch, utouch | http://askubuntu.com/q/326911
[10:27] <davmor2> ogra_: of course he has a mirror how do you expect him to shave without one ;)
[10:27] <ogra_> but i always install manually anyway
[10:28] <asac> ogra_: we still dont support phablet-flash -i URL ?
[10:28] <murgero> I shave without a mirror all the time.
[10:28] <ogra_> davmor2, he does shave ? i thought he has little critters in a box that chew off the hair
[10:28] <popey> asac: http://popey.mooo.com/mirror/cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/
[10:28] <murgero> Manual install is usually safer isnt it?
[10:28] <asac> ogra_: anyway. i think i can handle puttings tuff to autodeploy
[10:28] <asac> ogra_: so i just take the zip and push it to /sdcard/autodeploy.zip and reboot?
[10:29] <ogra_> murgero, not really, if you need additional setup or so the tools usually do better
[10:29] <murgero> Ah gotcha
[10:29] <ogra_> asac, right, first the armel+subarch.zip, then the armhf.zip
[10:29] <murgero> asac: You also need to flash it lmao
[10:29] <murgero> asac: but yes
[10:29] <asac> ogra_: two reboots?
[10:29] <ogra_> asac, and reboot in the middle
[10:29] <ogra_> yeah
[10:30] <asac> ogra_: is there a "reboot command" that auto goes into recovery?
[10:30] <ogra_> adb reboot recovery
[10:30] <asac> ok so i do that
[10:30] <ogra_> but the zip will try to boot normally
[10:30] <murgero> ttyl guys!
[10:30] <asac> after pushing to autodeploy
[10:30] <ogra_> so youo need to pull the battery before armhf
[10:30] <asac> hell
[10:30] <ogra_> or use sideload or something
[10:31] <ogra_> to load then in succession in the same session
[10:31] <ogra_> i always found pulling the battery easiest
[10:31] <asac> i have no idea what i shall do
[10:31] <asac> its not possible for me
[10:31] <asac> can we add phablet-flash support?
[10:31] <asac> ok me tries ... starts pushing latest maguro.zip as autodeploy.zip
[10:31] <ogra_> i guess there is phablet-flash support as long as you have the imgs on cdimage
[10:31] <asac> and reboot into recovery
[10:32] <ogra_> cjwatson, whould elmo cry if we kept touch images for a week ?
[10:32] <asac> i guess i will just go to chicken
[10:32] <w-flo> isn't there some "choose zip to flash" feature in the Ubuntu recovery?
[10:32] <ogra_> we should probably just change it there
[10:32] <ogra_> asac, well, your start was fine :)
[10:32] <asac> i dont know
[10:32] <asac> its far too complicated
[10:32] <asac> i will have chicken do it
[10:32] <ogra_> its two commands
[10:33]  * ogra_ thinks mgmt doesnt do any good to asac's coding skills :P
[10:33] <ogra_> MaxWallstedt, !
[10:33] <ogra_> good to see you here
[10:33] <asac> it must be super easy to go back and forward
[10:33] <asac> for images
[10:33] <asac> locally
[10:33] <asac> cdimage
[10:33]  * ogra_ is just working on making flipped ikmmages work on the SGS2
[10:33] <asac> popey url
[10:33] <MaxWallstedt> Thanks, thought I'd check it out
[10:34] <asac> from everywhere
[10:34] <asac> thats our whole strategy of bisecting
[10:34] <asac> to figure where we regressed etc.,
[10:34] <asac> anyway :)
[10:34]  * asac goes back to normal mode
[10:34] <ogra_> asac, well, it will become easier once we have system images
[10:34] <popey> asac: http://popey.mooo.com/mirror/cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/
[10:34] <ogra_> i wouldnt invest to much time into the current setup, since the world will completely change again soon
[10:34] <asac> ogra_: i just dont get why folks didnt add it from day one :)
[10:34] <asac> it feels so obvious that folks want to install stuff etc.
[10:35] <ogra_> asac, heh, because they needed to make the HW work first
[10:35] <asac> popey: yeah ... was saying that phablet-flash should support installing easily from everywhere
[10:35] <ogra_> then we needed to flip the container model
[10:35] <asac> your URL, cdimage, local stuff
[10:35] <popey> oh
[10:35] <ogra_> now we need to wait for all bits going through logind
[10:35] <ogra_> and *then* we can switch
[10:35] <asac> ogra_: well, that doesnt explain why folks didnt just do it... feels like zero work: yuou start with installing from specific locations
[10:35] <ogra_> it simply needs some preparation
[10:35] <asac> then you add --current and --pending tracking
[10:36] <sil2100> bfiller: hi! I just re-ran the apps stack, so it should be published soon
[10:36] <cjwatson> ogra_: We should have adequate space; I'm happy to up the limit
[10:36] <ogra_> oh
[10:36] <sil2100> bfiller: sorry for the lag
[10:36] <ogra_> asac, i thought you meant system images from the start
[10:36] <bfiller> sil2100: thanks
[10:36] <ogra_> cjwatson, please do it then :) (i would have to dig where it is set nowadays)
[10:36] <asac> nevermind :)
[10:37] <asac> ChickenCutlass: i need your help
[10:37] <asac> :
[10:37] <asac> i will come over
[10:37] <ogra_> asac, the issue with pulling arbitrary versions simply revolves around using jenkins in the beginning
[10:37] <ogra_> asac, https://wiki.ubuntu.com/Touch/Install#Manual_Installation
[10:38] <ogra_> seriously not that hard :)
[10:38] <slangasek> ogra_: well, yes, I didn't notify QA, didn't know that was expected
[10:38] <ogra_> slangasek, right, i'm writing a mail
[10:38] <ogra_> though there is hope the automated tests will simply run automated (and restart themselves on hang) at some point
[10:39] <ogra_> (or even stop hanging)
[10:39] <cjwatson> ogra_,asac: Done.  https://bazaar.launchpad.net/+branch/ubuntu-cdimage/revision/1325
=== alan_g|afk is now known as alan_g
[10:39] <ogra_> cjwatson, aweesome, thanks !
[10:40]  * ogra_ cant belive there is still an ubuntu-mid entry ... funny
=== chriadam is now known as chriadam|away
[10:41] <cjwatson> *shrug* I tend not to delete stuff; it's not a major cognitive load, and it's surprising how much I need to refer to how things used to be
[10:41] <ogra_> heh, yeah
[10:41] <ogra_> i doubt you could build them though
[10:41] <ogra_> they were closely tied to lpia iirc
[10:42] <ogra_> MaxWallstedt, do you have any deeper knowledge of the SGS2 bootloader ? fro flipped container images i need to get rid of the console=ram on the cmdline ... this seem to come from ATAGs in the bootloader
[10:42] <davmor2> ogra_: to save me the annoyance of digging through all the back logs on the server, what bug did I file that you were telling popey about?
[10:43] <ogra_> (someone from our kernel team is just looking into a patch to do it from the kernel ... but if you know a way to modify the ATAGs that wouldnt be needed)
[10:43] <ogra_> davmor2, dunno, let me dig through the backlogs  on the server :P
[10:43] <ogra_> davmor2, the media-info one :)
[10:44] <MaxWallstedt> ogra_: I'm afraid I'm of no help regarding that
[10:44] <ogra_> MaxWallstedt, thanks, i doubt anyone except samsung is ... but i thought i should ask :)
[10:46] <asac> cjwatson: can we get a month?
[10:46] <asac> who do i need to talk to to get us more disk space (if thtas the blocker)
[10:46] <asac> cjwatson: thx for the quick patch btw
[10:47] <cjwatson> asac: That would probably screw over the mirrors of cdimage.ubuntu.com
[10:47] <cjwatson> Not all of which we operate
[10:47] <asac> cjwatson: hmm. maybe system-images.ubuntu.com is where we can do that?
[10:48] <asac> or whatever that URL i was shown today is
[10:48] <cjwatson> Our own mirrors have well over 400G free so that probably isn't a problem, but inflating cdimage by 30G+ would require discussion
[10:48] <cjwatson> asac: Sure, that probably doesn't have the same mirroring constraints
[10:48] <asac> cjwatson: yeah. but can we just move all our phablet images to system-images? i guess we can avoid that then
[10:48] <asac> for a while
[10:48] <cjwatson> Eventually, I suppose
[10:48] <asac> kk
[10:48] <asac> lets keep it on the radar
[10:48] <asac> maybe after the system update stuff has settled
[10:49] <cjwatson> I question whether a month would actually be useful though
[10:49] <ogra_> we cant just move all of it to system images
[10:49] <cjwatson> If you have to refer that far back then normally you're doing something wrong
[10:49] <ogra_> the ports need the interim products
[10:49] <cjwatson> We keep all the logs so you can see what versions different images had; that's enough for archaeology
[10:49] <asac> cjwatson: well, our gates are not perfect yet. issues can go unseen for a long time
[10:49] <asac> but then having a way to narrow down the regresison window is key
[10:49] <cjwatson> asac: Yeah, but a month is a long time!
[10:49] <ogra_> (unless we move these too)
[10:49] <asac> i think its not that long :)
[10:50] <asac> its just better be safe
[10:50] <cjwatson> Think what state our images were in a month ago
[10:50] <asac> and disk is checap
[10:50] <cjwatson> Do you really care about investigating anything that old? :)
[10:50] <asac> while developers are precious
[10:50] <asac> cjwatson: if i see a bug which someone says we havent had before
[10:50] <ogra_> asac, the datacenter doesnt use a ton of cheapo USB didks on a big HUB :P
[10:50] <asac> i want to be able to find when it regressed before having developers deep dvice/debug etc.
[10:50] <ogra_> disk isnt as cheap as you might think
[10:50] <asac> ogra_: its still cheap
[10:51] <cjwatson> asac: That's actually kind of my point.  I think having developers investigate month-old issues in something moving as fast as touch is would be a waste of time.  At that point my experience is that it's generally easier to investigate directly rather than trying to bisect.
[10:51] <cjwatson> Bisection is one tool but it can be overused
[10:51] <ogra_> ++
[10:51] <asac> cjwatson: it might or it might not. i feel safer to have both options
[10:51]  * ogra_ thinks a week is really enouogh 
[10:51] <asac> ogra_: 25.1 is already close to that barrier
[10:51] <ogra_> i'd rather have more buolds per day than wasting space for old cruft
[10:51] <asac> two days ahead i would have 25.1 anymore
[10:52] <asac> wouldnt
[10:52] <cjwatson> asac: Except that the image that current points to is always kept
[10:52] <ogra_> righ
[10:52] <asac> i know
[10:52] <ogra_> t
[10:52] <asac> but the current image has a bug
[10:52] <asac> that ogra doesnt see on 25.1
[10:53] <asac> current is 26.1
[10:53] <asac> so i download from popey now
[10:53] <cjwatson> Ah right
[10:53] <asac> so i have a problem that camera doesnt work for me
[10:53] <asac> i would prefer to be able to find what changed
[10:53] <cjwatson> Still, I worry that this is instituting policies to cope with growing technical debt, which IMO is the wrong direction
[10:53] <asac> rather than asking one of our overloaded engineers to figure this
[10:55] <rickspencer3> cjwatson, are you saying that we should rather be paying down the technical debt?
[10:55] <asac> cjwatson: i dont think saying we just hide the technical debt would help :)
[10:55] <cjwatson> I know touch is different, but I've been investigating issues of one kind or another in Ubuntu images for nine years, and I can count the times that it would have been useful to have >week-old images on the fingers of one hand
[10:55] <asac> having longer history of images would help me and rick to narrow down the regression window... which might gain us 1h engineering time or 10h
[10:55] <asac> we dont know
[10:55] <asac> every minute is precious on our current schedule :)
[10:55] <rickspencer3> asac, I'm not so sure
[10:56] <rickspencer3> I think serious things usually had changed in the last day or so
[10:56] <rickspencer3> or are pretty obvious that it's from "the last xorg update" for exmaple
[10:56] <asac> rickspencer3: pictures not getting saved?
[10:56] <asac> when did that start?
[10:56] <asac> on maguro?
[10:56] <cjwatson> rickspencer3: I'm concerned that having an enormous safety net of lots of old images will encourage people to use debugging techniques that I think are likely to be very inefficient over such time windows, but that look superficially attractive
[10:56] <asac> do you know?
[10:56] <rickspencer3> cjwatson, is that the kind of thing you are referring to?
[10:56] <rickspencer3> asac, I don' t know or care, tbh
[10:56] <cjwatson> Usually over that kind of window you're better off just attacking directly, *maybe* up/downgrading single packages
[10:57] <rickspencer3> the *developer* might care so that they can figure out what might have gone wrong
[10:57] <rickspencer3> but I am not typically going to tell them how to debug, they probably know better than me
[10:57] <rickspencer3> that said, I would like to be able to quickly attribute bugs to changes so that we can revert those changes
[10:57] <rickspencer3> anyway, I guess if you phablet-flash daily, you have every image locally anyway ;)
[10:58] <cjwatson> But if the change is over a week old, the chances that simply reverting it would be an improvement are already getting pretty slim
[10:58] <rickspencer3> cjwatson, right, I think it's more a matter of the last day
[10:58] <asac> you dont know if the change was done in a component touched since then
[10:58] <cjwatson> Since people will have built things on top
[10:58] <asac> in some cases it is not appropriate, in others it might be
[10:58] <rickspencer3> I do think it's good to have a "last known good" for demo and other purposes
[10:58] <rickspencer3> sort of what we used to need alphas and such for
[10:59] <davmor2> rickspencer3: no I have to knock them off from time to time E:TOOMUCHHDDSPACEUSED
[10:59] <popey> +1000
[10:59] <cjwatson> rickspencer3: Right, the last-known-good "current" image thing is definitely an improvement and I have no issues with that
[10:59] <popey> i was demonstrating touch at OSCON last week, and frequently had the phone wedge on me
[10:59] <popey> glad I had 3 devices
[10:59] <rickspencer3> honestly, in my experience, if issues are seriousness for me to want a revert ...
[10:59] <popey> and on more than one occasion i had to use all three devices to demo touch to someone!
[11:00] <rickspencer3> the issue was introduced in the last 24 hours
[11:00] <davmor2> popey: when you say wedge is it that the apps menu froze or disappeared?
[11:00] <davmor2> popey: For that one I have written a bug
[11:00] <popey> any one of a number of issues I am trying to reproduce and file bugs on
[11:01] <popey> had the shell lock up, restart, apps lock up
[11:01] <popey> luckily people were forgiving
[11:01] <rickspencer3> popey, well, we are still going fast
[11:01] <popey> but it's hard to say "hey, give us $775" in one breath and "look at this... *crash*" in the other
[11:01] <rickspencer3> once we pull out a real "release" you'll be in good shape ;)
[11:01] <popey> sure, I understand, completely, but I do like the idea of a "known good" now and then
[11:02] <popey> cue rickspencer3 saying every day should be "good" ☻
[11:02] <rickspencer3> lol
[11:02] <rickspencer3> exactly
[11:02] <rickspencer3> we are getting there!
[11:02] <popey> indeed
[11:02] <rickspencer3> http://reports.qa.ubuntu.com/smokeng/saucy/image/3270/
[11:03]  * popey wonders where the welcome screen gets the data for "1 photo taken today"
[11:03] <rickspencer3> popey, there is a new API for htat
[11:03] <rickspencer3> apps need to start feeding the API
[11:03] <asac> i just dont get why you would explicitely prevent the ability to quickly narrow down the time window of when a bug was introduced ...serious or UI glitch
[11:04] <davmor2> popey: https://bugs.launchpad.net/unity8/+bug/1202794, https://bugs.launchpad.net/touch-preview-images/+bug/1204942 were my too latest
[11:04] <asac> if it only helps us once, it was worth it
[11:04] <popey> i shoved a month full of photos into ~/Pictures hoping it would find them, and it doesnt
[11:04] <ubot5> Launchpad bug 1202794 in touch-preview-images "Maguro: Apps lens sometimes keeps scrolling" [Undecided,Confirmed]
[11:04] <ubot5> Launchpad bug 1204942 in touch-preview-images "tilt sends qmlscene rocketing to 100% cpu" [Undecided,New]
[11:04] <popey> davmor2: will take a look, thanks
[11:04] <asac> but well, we can talk later
=== gusch|lunch is now known as gusch
[11:04] <rickspencer3> asac, I think what cjwatson was saying is that capturing and storing the images is not free, or maybe not particular cheap *and*
[11:05] <rickspencer3> bisecting through previous versions is not usually a very effective debugging method, so it's probably not worth the cost
[11:05] <rickspencer3> so, let's rather invest that cost in things that will have more benefit
[11:05]  * rickspencer3 tries not to put words in cjwatson's mouth ;)
[11:06] <ogra_> as long as it are the right words ...
[11:06] <ogra_> :)
[11:06] <ogra_> asac, i'm nearly done flashing
[11:06] <davmor2> rickspencer3: I wouldn't try that cjwatson is likely to bite, honest ;)
[11:06]  * ogra_ waits for the final reboot
[11:07]  * rickspencer3 steps away
[11:07] <ogra_> and i really cant bear that robot with rotating guts anymore ... we need our own animation one day
[11:07] <rickspencer3> ogra_, a teddy bear with a chainsaw?
[11:07] <ogra_> ++
[11:07] <ogra_> child friendly FTW !
[11:07] <davmor2> ogra_: a rotating ubuntu logo
[11:07]  * rickspencer3 really steps away
[11:08]  * davmor2 now doesn't believe that rickspencer3 ever really steps away
[11:08] <sergiusens> ogra_: we have the anim, we just need to put it there, the idea was cut a while back until image based upgrades was complete, luckily it's not a rewrite
[11:09] <ogra_> great !
[11:09] <cjwatson> rickspencer3: While I do have some concerns about the storage costs, I'm actually more worried about inadvertently socially-engineering developers into debugging approaches that will probably waste their time, IYSWIM
[11:10] <cjwatson> It's not worth it if it helps us once but in ten other cases it causes developers to waste a couple of hours bisecting through old images when they would have been faster debugging it directly (harder and less tempting, but quicker)
[11:11] <ogra_> asac, no issues with the camera app here on latest pending
[11:11] <cjwatson> To clarify further: I think even fairly long-range bisection on individual components is an excellent tool, just not on full images
[11:11] <ogra_> the image is in the Pictures folder as well as in the gallery app
[11:12] <asac> cjwatson: developers dont need to bisect... but i can bisect and help giving developers a very good start
[11:12] <popey> davmor2: confirmed 1204942
[11:12] <popey> can't confirm the other
[11:12] <ogra_> davmor2, are you on the latest pending image with your maguro and could you test asac's issue ?
[11:13] <asac> i nour current schedule, every engineer day we can save is far more valuable than what the engineer costs
[11:13] <davmor2> ogra_: I'm just installing it current what is asacs issue
[11:13] <ogra_> asac, in our current schedule we block far more engineers than we unblock with the tests
[11:14] <cjwatson> asac: And I'm saying that my experience is that over long enough time periods, image bisection simply generates false alleyways that engineers have to follow up
[11:14] <ogra_> davmor2, asac doesnt see photos show up in /home/phablet/Pictures after taking them
[11:14] <davmor2> ogra_: give me about 2 minutes robot cogs whirling currently
[11:15] <ogra_> asac, which is the reason why everyone uses --pending
[11:15] <cjwatson> I've been here before and largely found it to be a waste of time even if somebody else was doing the bisection for me
[11:15] <cjwatson> It's a very tempting waste of time, I understand, but ...
[11:15] <ogra_> davmor2, yeah, no hurry ... it works fine for me with the latest image
[11:16] <ogra_> asac, are you using the modified kernel in both test cases ? might be related
[11:16] <apw> asac, which of the phablet systems are we going to apply this testing to ?
[11:17] <davmor2> asac: I try to remember to do to things when I write a bug add the image that broke and the phone it broke on.  I tend to play on my phone fairly heavily on an evening once I knock off work, but give it a quick blast through post flash
[11:17] <davmor2> s/to things/ two things
[11:17] <davmor2> asac: and I flash most days
[11:18] <asac> ogra_: no... tried fresh install and had the same, can try again
[11:18] <ogra_> well, i cant reproduce it
[11:18] <davmor2> asac, ogra_: are you both doing --wipe or not?
[11:18] <asac> apw: dunno. think might change over time
[11:18] <ogra_> davmor2, i dont
[11:18] <asac> based on availability of devices in lab etc.
[11:19] <ogra_> well, we wont change devices for 13.10
[11:19] <ogra_> so you can safely say we'll test on the currently supported devices
[11:20] <davmor2> asac, ogra_: I've done a --pending --flash so I'm on 20130731 I have image20130731_001.jpg in .home/phablet/Pictures
[11:21] <ogra_> davmor2, yeah, same here
[11:21] <ogra_> i now even have 001.jpg :)
[11:21] <ogra_> err
[11:21] <ogra_> 002
[11:21] <ogra_> asac, so it must be soemthing with your device or with the old phablet homedir
[11:22] <ogra_> try a flash with --wipe
[11:22] <ogra_> and see if that solves anything
[11:22] <davmor2> ogra_, asac: and that is showing up in file manager, terminal and gallery
[11:23] <apw> asac, thats not in the plan document for this work ?
[11:23] <sergiusens> jdstrand: regarding your email about stagefright, you might want to talk to jhodapp, in a nutshell it will be used over gstreamer (e.g.; recording in the camera-app, which today is done using qtubuntu-media iirc)
[11:27] <asac> apw: i dont know. you feel resistant to add it to mako and maguro and want me to pick one?
[11:28] <asac> if so i can have serious talk where we want it
[11:28] <ogra_> argh !
[11:28] <ogra_> seems i broke the builds with the last change i did
[11:28]  * ogra_ reverts and will re-trigger a build
[11:29] <slangasek> ogra_: well, don't do that !
[11:29] <davmor2> ogra_: see that's not the news I need to hear after being asked to flash my device :P
[11:29] <ogra_> slangasek, breaking or re-triggering ?
[11:29] <asac> ogra regressed to cowboy mode :)
[11:29] <ogra_> davmor2, i broke the build process
[11:30] <ogra_> not the images :)
[11:30] <ogra_> no worries
[11:30] <davmor2> ogra_: Phew
[11:31] <ogra_> i was relying on the fact that we set FLASH_KERNEL_SKIP in all our builds ... but seems where the new code kicks in this isnt set anymore
[11:31] <slangasek> ogra_: breaking :)
[11:31] <slangasek> ogra_: sorry for the ambiguity
[11:31] <ogra_> heh
[11:31] <popey> davmor2: if you get a moment can you try and reproduce https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/1206870
[11:31] <ubot5> Launchpad bug 1206870 in Ubuntu UI Toolkit "OSK doesn't automatically hide when fields no longer on screen" [Undecided,New]
[11:32] <ogra_> slangasek, well, already unbreaking .... and it showed a flaw in the build process , so it wasnt completely useless
=== schwuk is now known as schwuk_away
=== schwuk_away is now known as schwuk
[11:35] <ogra_> cjwatson, http://paste.ubuntu.com/5932466/ does that look ok to zou (i'm not so sure about the chroot, will the env end up in there ?)
[11:35] <davmor2> meh that's not good
=== frals_ is now known as frals
[11:37] <sil2100> bfiller: apps published o/
[11:40] <davmor2> ogra_, popey: https://bugs.launchpad.net/touch-preview-images/+bug/1206888
[11:40] <ubot5> Launchpad bug 1206888 in touch-preview-images "Maguro: 20130731 image no gprs" [Undecided,New]
[11:41] <ogra_> davmor2, oh, i thought awe worked on that yesterday
[11:41] <davmor2> ogra_: that's why I did a wipe today to check
[11:42]  * ogra_ pushes the livecd-rootfs change ... i guess we'll see if it works
=== ashams_ is now known as ashams
[11:43] <davmor2> ogra_: the weird thing is list-modems say attached and powered
[11:44] <ogra_> davmor2, yeah, i think thats the one awe was working on
[11:44] <ogra_> wait for boston to get up :)
[11:44] <ogra_> i think it was even merged .... the 31 image was simply built to early to pick it up
[11:45] <davmor2> ogra_: that's why I included the dpkg -l info for version numbers :)
[11:45] <ogra_> :)
[11:50] <sergiusens> davmor2: try to see if you can get the new ofono from apt-get
[11:50] <ogra_> yeah
=== alan_g is now known as alan_g|lunch
[11:59] <davmor2> sergiusens, ogra_: I don't see ofono in the list of updates
=== hikiko is now known as hikiko|lunch
[12:37] <asac> nice
[12:37] <asac> all works with apw kernel and 386m
[12:37] <asac> camera/gallery
[12:37] <asac> etc.
[12:38] <ogra_> and your photose end up where they shuld ?
[12:38] <ogra_> *should
[12:38] <asac> yeah
[12:38] <asac> it didnt work a few times
[12:38] <asac> rebooting a few times solved the problem
[12:38] <ogra_> any idea what it was ?
[12:38] <asac> now it works even if i reboot
[12:38] <asac> nope
[12:38] <asac> just flakiness i guess
[12:39] <asac> and my incompetence
[12:39] <asac> howeve,r i am now on --pending
[12:39] <asac> not current
[12:39] <asac> but since that didnt work before either, I believe it was really something wird
[12:39] <ogra_> as most of us
[12:39] <asac> like my filesystem needed multiple overrides
[12:39] <asac> etc.
[12:39] <ogra_> well, the initramfs issue was weird as well
[12:39] <ogra_> probably your flash is about to give up
[12:40] <asac> i am sure we have memory leaks
[12:40] <asac> everytime i open and close apps
[12:40] <asac> i loose memory
[12:40] <asac> kind of 5m per start
[12:40] <asac> goes away forever
[12:41] <seb128> asac, how do you measure? top?
[12:41] <asac> seb128: free and meminfo
=== chihchun is now known as chihchun_afk
[12:41] <asac> but ... dont take it as the truth
[12:41] <seb128> asac, is that accurate? or is it just the kernel just giving it back since you don't need it?
[12:41] <asac> i might just be confused:)
[12:41] <asac> seb128: well, you see how much buffer mem you have free
[12:41] <seb128> *not* giving it back
[12:42] <asac> so if you sum up the free
[12:42] <asac> plus the free buffer
[12:42] <asac> you get the total free
[12:42] <asac> ...
[12:42] <seb128> asac, I though the kernel wouldn't give them back until they were needed (or at least not automatically)
[12:42] <asac> talking to didrocks i believe it might be just qml garbage collection
[12:42] <asac> seb128: yeah but you see which buffers it could give back
[12:42] <asac> but again, i might be wrong :)
[12:42] <asac> will check with tvoss
[12:42] <seb128> asac, I'm not sure those tools provide reliable informations
[12:43] <asac> seb128: free and meminfo dumbs numbers from the kernel
[12:43] <asac> but if qml is like java
[12:43] <asac> it might claim more and release stuff
[12:44] <seb128> asac, try with a pure qt app (if we have some, the phone-app is at least part cpp)?
[12:44] <asac> seb128: http://paste.ubuntu.com/5932649/
[12:44] <asac> thats what the kernel konws
[12:44] <asac> the kernel knows nothing else
[12:45] <asac> well. it works :)
[12:45] <asac> i will leave the rest to the knowing
[12:46] <psypher246> hello all, I am trying to install ubuntu-touch on my nexus 4 and getting very stuck. the phablet-flash fails due to connections being refused to cdimage.ubuntu.com so I am following the manual instructions and they do not match up to what I am seeing at all, quite a few inconsistencies
[12:47] <popey> psypher246: can you pastebin the output of phablet-flash ?
[12:47] <popey> psypher246: also, what version of phablet flash are you using? find out with "apt-cache policy phablet-tools"
[12:47] <popey> 0.15+13.10.20130730-0ubuntu1 is what I'm on
[12:48] <psypher246> popey i think it's coporate policies blocking rsync from dl the image so I am stuck with the manul process whihc isn;t well documented
[12:49] <psypher246> popey, i get stuck with uploading autodeploy.zip
[12:49] <ogra_> well, the manual process kind of assumes that you have a properly working recovery
[12:49] <ogra_> beyond that it never failed for me following the instructions step by step
[12:49] <psypher246> ah thats not clear at all
[12:50] <psypher246> adb reboot recovery puts me in recovery,can'ytrun adb poush then
[12:50] <ogra_> oh ?
[12:50] <ogra_> whats the error you get ?
[12:50] <psypher246> error device not found
[12:50] <ogra_> ah, thats a problem on the PC side
[12:51] <ogra_> adb kill-server; sudo adb start-server
[12:51] <ogra_> try that
[12:51] <psypher246> but device is found when I am not in recovery mode
=== hikiko|lunch is now known as hikiko
[12:51] <psypher246> and can copy the autodeploy then
[12:52] <psypher246> but then when I go back into recovery mode I am not sure how to run that file
[12:52] <popey> you don't run it, recovery finds it and uses it
[12:53] <popey> psypher246: have you tried as ogra_ suggested? adb kill-server; sudo adb start-server  ?
[12:53] <psypher246> yes
=== alan_g|lunch is now known as alan_g
[12:53] <ogra_> and you still can not use adb in recovery ?
[12:53] <psypher246> restgart adb, adb devices shows device, go into recovery and adb commands says device not found
[12:54] <ogra_> well, i meant to do the restart when the device is in recovery
[12:55] <ogra_> not before you reboot itno recovery :)
[12:55] <psypher246> ok trying that. btw to be in recovery I first get android with red triangle, I am then supposed to hit power and up to properly be in recovery correct>
[12:56] <ogra_> right
[12:56] <ogra_> and in recovery you should see a menu on screen
[12:57] <ogra_> err, wait
[12:57] <ogra_> red triangle sounds not right
[12:57] <psypher246> pls add that part about power and volume up to the instructions popey
[12:57] <psypher246> i only see the menu after hitting power and volume up
[12:57] <ogra_> you seem to be in bootloader mode somewhere, not in recovery
[12:58] <psypher246> says android system recovery
[12:58] <ogra_> ok
[12:58] <psypher246> ok adb restarted
[12:58] <psypher246> no run adb push?
[12:58] <ogra_> can you now get in with adb shell ?
[12:58] <psypher246> soz what does that mean?
[12:58] <ogra_> if that works, log out of adb again and do the push
[12:58] <ogra_> adb shell
[12:58] <ogra_> execute that
[12:59] <psypher246> device not found
[12:59] <ogra_> do you end up in an android shell
[12:59] <ogra_> hmm
[12:59] <ogra_> that smells a bit like you didnt properly unlock the device before starting
[12:59] <diwic> awe, how do you unlock your SIM? I wanted to try a phone call.
[13:00] <diwic> awe, or at least, I think that's why it isn't working, but I really don't know. :-)
[13:00] <psypher246> ogra_, think I might still need to do a facvtory reset, thought the rest was ddone since after unlocking the device looked flashed
[13:00] <ogra_> do you see an open lock under the google logo when booting ?
[13:00] <ogra_> thats a clear indicator
[13:00] <ogra_> if this isnt there you are not unlocked
=== jhodapp is now known as jhodapp|afk
[13:02] <awe> diwic, hold on a sec... I'll fwd you the email
[13:02] <diwic> thanks
[13:02] <awe> you just need to use a script that's part of ofono-scripts after you boot the device
[13:03] <awe> diwic, just sent you the email...
=== _salem is now known as salem_
[13:03] <awe> diwic, are you sure the SIM is locked?
[13:04] <diwic> awe, no. I'm not sure of anything, really. :-)
[13:04] <awe> is it a new SIM?
[13:04] <ogra_> if it is a new european SIm it is definitely locked
[13:04] <diwic> awe, just that nothing happens when I press the dial button in the phone app
[13:04]  * awe hopes diwc has the PIN # then
[13:04] <psypher246> ogra_, did a wipe, the lock is unlocked, but was unlocked anyway
[13:05] <ogra_> psypher246, so lets speed this up a bit :)
[13:05] <ogra_> psypher246, http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/pending/saucy-preinstalled-recovery-armel+mako.img download that file
[13:05] <awe> diwc, grep'ing syslog for 'ofonod' should be helpful is anything is wrong
[13:05] <psypher246> adb shell still not working
[13:05]  * w-flo recently bought a sim card from congstar and now his android phone no longer wants to know the PIN on boot
[13:06] <ogra_> psypher246, then boot to the bootloader mode (teh red triangle part) and stay there
[13:06] <awe> diwic, also... the script 'list-modems' in /usr/share/ofono/scripts is very helpful too...
[13:06] <psypher246> ogra_, Iok busy
[13:06] <psypher246> ogra_, done
[13:06] <ogra_> psypher246, then do: sudo fastboot flash recovery saucy-preinstalled-recovery-armel+mako.img
[13:07] <awe> diwic, I'm around for another ~20m or so, then will be off for awhile.  ogra_ or sergiusens should be able to assist if you still need help
[13:07] <ogra_> now you should have a proper recovery mode
[13:07] <psypher246> ogra_, what mode am I in on the phone when running that?
[13:07] <ogra_> bootloader mode
[13:07] <ogra_> the thing with the red triangle on screen
[13:08] <ogra_> you can also do: adb reboot-bootloader  from the running android
[13:08] <ogra_> should get you to the same screen
[13:08] <diwic> awe, nothing at all happens in syslog when trying to dial
[13:09] <diwic> awe, "enter-pin sim xxxx" fails with a python stacktrace ("Argument format is not recognized")
[13:09] <diwic> awe, "enter-sim pin xxxx" - there is no "enter-sim" executable
[13:10] <diwic> awe, ah, it's "enter-pin pin xxxx"
=== Caine^^ is now known as C^^
[13:10] <diwic> awe, ok, now it's unlocked at least, according to list-modems
[13:10] <ogra_> diwic, /usr/share/ofono/scripts/enter-pin
[13:10] <ogra_> its not in path
[13:11] <diwic> \o/
[13:11] <ogra_> :)
[13:11] <diwic> dialing now works. I rejected the call, want to fixup the audio first :-)
[13:12] <ogra_> yeah, it also isnt so exciting to talk to yoourself all day
[13:13] <awe> diwic, ;#
[13:13]  * awe is amused by his typo emoticon
[13:13] <ogra_> beardy
[13:14] <psypher246> ogra_, ok got an ubuntu recovery screen now, do I continue with the adb push commands?
[13:14] <ogra_> yeah, that should all work now
[13:15] <plars> ogra_: not sure if there's anything helpful in it, but http://paste.ubuntu.com/5931417/ has a pastebin from a run that was happening last night, it was failing when trying to adb to the device to run a script that just does a small ping loop to detect when the network is up, after reboot and adb wait-for-device returned ok
[13:15] <plars> ogra_: this was another one of those "adb protocol fault" errors that keep killing us
[13:16] <ogra_> plars, hmm, i guess a syslog would have been more informative
[13:17] <plars> ogra_: I'll grab that too when I see it again and the system is idle right after. This was one where I had an easy window to get on the device right after witnessing the failure
[13:18] <plars> ogra_: I hadn't remembered seeing some of those wifi errors previously though
[13:18] <ogra_> plars, i see adb coming up just fine between line 460 and 470
[13:18] <ogra_> and it doesnt seem to be torn down again or restart or anything
[13:18] <ogra_> at least nothing in that dmesg reflects it dying
[13:19] <plars> ok
[13:19] <psypher246> ogra_, ok deployed the suacy-preinstalled-touch-armel+maku.zip and it ran, now I the black screen with Google and the unlocked padlock. do i run recvery mode again and now push saucy-preinstalled-touch-armhf.zip?
[13:19] <ogra_> yes
[13:20] <ogra_> note that thiss is 350M big, it will take longer than the first zip ... try to be patient :)
[13:20] <psypher246> ok busy d/l that file. thanks a lot for the help so far.
[13:20] <psypher246> yeah 30 mins
[13:20] <ogra_> np
[13:21] <psypher246> so was something wrong in my steps or is the docs missing the fact you need to load the ubuntu recovery first?
[13:21] <ogra_> well, the manual instructions kind of assume you have the recovery mode already ...
[13:21] <ogra_> that should probably be added as info
[13:30]  * ogra_ curses ... i had i never uploaded that fix for asac's kernel install ... grmbl
[13:30] <ogra_> s/i/oh/
[13:32] <robert_ancell> mterry, hey, got some info from Saviq - trying greeter now. Will tell you after restaret
[13:34] <robert_ancell> mterry, no luck :(
[13:34] <mterry> robert_ancell, you're talking about the pure-client branch?
[13:34] <robert_ancell> mterry, ok, point 1 - unity8-greeter.desktop has /usr/bin/ubuntu-touch-session which doesn't seem to exist - should it?
[13:34] <robert_ancell> mterry, split.client
[13:35] <ogra_> robert_ancell, it exists in ubuntu-touch installs
[13:35] <mterry> robert_ancell, ah...
[13:35] <robert_ancell> mterry, so what do I need for that..
[13:35] <robert_ancell> what package
[13:35] <mterry> robert_ancell, yeah, but will need my ubuntu-touch-session branch for that...
[13:35] <ogra_> robert_ancell, oh, i lied, it doesnt anymore
[13:35] <ogra_> robert_ancell, that was moved to plain upstart user session jobs
[13:35] <mterry> robert_ancell, ogra: oh, it got lost since I updated those branches
[13:36] <ogra_> well, we dropped the script a while ago
[13:36] <mterry> robert_ancell, it's goal was to set up a bunch of environment variables that ubuntu-touch requires for its shells
[13:36] <ogra_> right, thats now done from the respective upstart jobs
[13:36] <ogra_> ubuntu-touch-session-setup.conf abd ubuntu-touch-session.conf
[13:37] <ogra_> *and
[13:37] <robert_ancell> mterry, ok, so I disabled it and I get "QXcbConnection: Could not connect to display" - It seems Qt is trying to connect to a non-existent X server
[13:37] <ogra_> you will likely need the env vars
[13:37] <robert_ancell> Saviq suggested I set QT_QPA_PLUGIN=mirclient before. I did that and it still tries to connect to X then segfaults now
[13:38] <mterry> robert_ancell, I think my ubuntu-touch-session branch splits the upstart job into a script again
[13:38] <mterry> robert_ancell, let me grab you the url
[13:38] <robert_ancell> mterry, ok
[13:38]  * didrocks waits in the queue once robert_ancell will release mterry's lock :p
[13:39] <robert_ancell> mterry didn't want to do anything for himself today, right?
[13:39] <didrocks> he better not!
[13:39] <mterry> robert_ancell, it's here: ~mterry/session-manager-touch/lightdm-switch but I'm merging from trunk right now
[13:39] <robert_ancell> k
[13:39] <mterry> robert_ancell, there are two other branches you might want....  will get to those next
[13:40] <robert_ancell> mterry, any nice PPAs with this stuff?
[13:43] <mterry> robert_ancell, no, I hadn't expected use yet  :)
[13:44] <mterry> though I did have plenty of warning you wanted it
[13:44] <robert_ancell> mterry, heh, I have all the support in LightDM in theory, want to see a real greeter working :)
[13:44] <ogra_> pfft, who cares about greeters ... we dont need them on phones
[13:45] <ogra_> (and who cares about tablet people anyway)
[13:45] <davmor2> awe: https://bugs.launchpad.net/touch-preview-images/+bug/1206888 :(
[13:45] <ubot5> Launchpad bug 1206888 in touch-preview-images "Maguro: 20130731 image no gprs" [Undecided,New]
[13:45] <seb128> ogra_, you probably want a lock screen on your phone...
[13:45] <ogra_> seb128, nah, the shlle does that for me :P
[13:45] <ogra_> *shell
[13:45] <didrocks> seb128: you're so old-fashioned, you hate freedom! lockscreen? :)
[13:45] <ogra_> ++
[13:45] <mterry> robert_ancell, ok, that branch updated.  Let me get you others
[13:46] <seb128> didrocks, because having your leg dialing when the phone is in your pocket is more fun? :p
[13:46] <didrocks> seb128: especially if it's a foreign country :)
[13:46] <seb128> ;-)
[13:46] <didrocks> ogra_: just FYI, in fact, stuff are being separated from the shell to the greeter actually
[13:46] <mterry> robert_ancell, others are just config changes actually.  Let me tell you them and you can just make them directly.  One is a seed change for lightdm and the greeter, which you probably already installed
[13:46] <ogra_> insane !
[13:46] <ogra_> :)
[13:47] <lool> barry: you saw stéphane's reply?
[13:47] <lool> barry: what's really weird is that slangasek tried upgrading another device to the same target image and it worked!
[13:47] <barry> lool: i did
[13:48] <lool> sorry, it failed, but differently
[13:48] <mterry> robert_ancell, the other is to change UID_MIN in /etc/login.defs to 1001 (if you're on the phone)
[13:48] <lool> (later)
[13:48] <barry> lool: i did a cli update this morning of my nexus 7 and it worked ;)
[13:48] <mterry> Because it has a 'system' user as 1000 that we don't want to show up
[13:48] <robert_ancell> mterry, no, this is desktoop
[13:48] <barry> lool: what device/channel are you on?
[13:48] <mterry> robert_ancell, ok, then you're fine.  Just know that if you have more than one user, you will get tablet view
[13:48] <lool> barry: nexus 7
[13:48] <lool> barry: default channel
[13:49] <robert_ancell> that's fine
[13:49] <lool> barry: are there more channels?
[13:49] <barry> lool: there are, but daily is the default channel, and it's what my nexus7 is on
[13:49] <barry> lool: how very odd
[13:50] <lool> barry: did you manage to figure something from the traceback?
=== jhodapp|afk is now known as jhodapp
=== annerajb is now known as annerajb_brb
[13:51] <barry> lool: not yet.  i'm working on a different bug atm.  soon as i fix that one, i'm going to dive deeper into yours
[13:51] <slangasek> ogra_: so it looks like your rebuild attempt also failed?  Why are linux image packages being pulled into the livefs build?
[13:51] <lool> barry: k, thanks
[13:52] <ogra_> slangasek, to create the boot.img
[13:52] <robert_ancell> mterry, debuild on lp:~mterry/session-manager-touch/lightdm-switch doesn't work
[13:52] <slangasek> ogra_: mmm?  I thought the boot.img was coming from jenkins
[13:52] <ogra_> slangasek, i have my livecd-rootfs branch that pulls in xnox' packages nearly ready
[13:52] <robert_ancell> "cp: cannot stat ‘debian/tmp/lxc-overrides/init.rc’: No such file or directory"
[13:52] <ogra_> then we wont need that step anymore
[13:52] <slangasek> we don't need which step?
[13:52] <ogra_> slangasek, the android boot.img comes from jenkins, not the ubuntu one
[13:53] <mterry> robert_ancell, ok, my merge from trunk must have borked.  Give me a few minutes
[13:53] <slangasek> that sounds like confusing terminology
[13:53] <slangasek> why would there be both android and ubuntu boot.img?
[13:53] <slangasek> we are only booting once :P
[13:53] <ogra_> slangasek, jenkins as well as the android binary package now have ubuntu boot.img's  livecd-rootfs needs updating that i didnt get to yet
[13:54] <ogra_> slangasek, wrong, we are booting two systems during one boot ;)
[13:54] <barry> lool: one thing that would be interesting is if you can also do a cli update.  e.g. system-image-cli -v
[13:54] <ogra_> slangasek, the initrd of the android boot.img is the android rootfs :)
[13:54] <ogra_> (to add some more confusion)
[13:55] <mterry> robert_ancell, try now?   hopefully it works
[13:55] <ogra_> slangasek, so we need both ramdisks for a successfull boot ... ubuntus and androids
[13:56] <slangasek> ogra_: so the livefs builder does what? pulls the kernels in one by one, uses them for generating Ubuntu (-Touch) initramfs, pushes those into a boot.img, and uninstalls the kernels again?
[13:56] <ogra_> the livefs builder code predates all of this bootimg stuff and rolls a boot.img on the fly during build atm
[13:57] <xnox> ogra_: \o/
[13:57] <robert_ancell> mterry, worked. gtg, I'll be back later
[13:57] <slangasek> ogra_: it's clearly not just "on the fly", I'm looking at a livefs build log where a linux image package is failing to configure
[13:57] <ogra_> slangasek, now that we have the android packaging ,that code snippet will be replaced by pulling in the android binary package and copying it's bits in place
[13:57] <mterry> robert_ancell, well, built anyway.  Working is another thing  :)
[13:57] <robert_ancell> heh :)
[13:58] <xnox> ogra_: slangasek: in my package I generate both android & ubuntu bootimages if that helps.....
[13:58] <ogra_> slangasek, yes, asac complained that the packages wouldnt flash automatically this morning, so i added flashing support ... kind of assuming that FLASH_KERNEL_SKIP would be set during builds
[13:58] <ogra_> slangasek, which is apparently not the case ... so it tries to flash even when being in a chroot
[13:59] <xnox> ogra_: ....unless they don't work (worked for me on grouper...)
[13:59] <slangasek> ogra_: "the packageS"?
[13:59] <ogra_> slangasek, i just backed out the whole thing again
[13:59] <ogra_> slangasek, kernel packages
[13:59] <slangasek> ogra_: well, but the second build failure shows the exact same error.  Timing problem?
[13:59] <slangasek> or did the failure mail send the wrong log?
[14:00] <ogra_> no, i exported FLASH_KERNEL_SKIP in a livecd-rootfs upload but the activity happens inside a chroot, so that export was moot
[14:00] <ogra_> which resulted in the same build failure
[14:00] <slangasek> ok, so you've *now* reverted it in the kernel packages
[14:00] <ogra_> no
[14:01] <ogra_> i didnt touch any kernel packages ... but yeah, i reverted the breaking change completely
[14:01] <ogra_> (it was in initramfs-tools-ubuntu-touch)
[14:02] <ogra_> it was a pointless change anyway, since we usually dont use kernel packages ... was just to make asac happy
[14:02] <ogra_> since he didnt like to have to call a separate script for flashing
[14:03] <ogra_> (i thought i'd do him a favour ... but that kind of ended in disaster ... )
[14:08] <slangasek> ogra_: heh ;)
[14:08] <ogra_> next build will be fine again
[14:08] <ogra_> and from tomorrow on i hope to use the android packages , so all that kernel package crap will be gone from the build
[14:12] <slangasek> ogra_: yeah... you made me reflash my phone again though, I was counting on a new update being available for another system-image demo this afternoon ;)
[14:13] <ogra_> slangasek, well, waiting for -proposed migration of initramfs-tools-ubuntu-touch before i can start a new build, sorry
[14:16] <ogra_> now for publisher ...
[14:17] <ogra_> slangasek, new build running (wow, that package went through fast) ... 50min
[14:18] <ogra_> go mir !
[14:21] <mkozjak> hi
=== ejat- is now known as ejat
[14:25] <slangasek> ogra_: well, the new image is no longer a blocker for me, so don't worry about it on my account
[14:25] <slangasek> but, good to know the fixed image will be available soon
[14:25] <ogra_> i worry about it in general :)
=== chihchun_afk is now known as chihchun
[14:27] <nexwave-mat> Hey guys, trying to port a device with latest tools and repos but I keep getting this error: target thumb C++: gralloc.msm8660 <= hardware/qcom/display-caf/libgralloc/framebuffer.cpp
[14:27] <nexwave-mat> hardware/qcom/display-caf/libgralloc/framebuffer.cpp: In function 'int fb_post(framebuffer_device_t*, buffer_handle_t)':
[14:27] <nexwave-mat> hardware/qcom/display-caf/libgralloc/framebuffer.cpp:93:25: error: 'MDP_DISPLAY_COMMIT_OVERLAY' was not declared in this scope
[14:27] <nexwave-mat> make: *** [/home/mdaniel/ubuntutouch/out/target/product/p930/obj/SHARED_LIBRARIES/gralloc.msm8660_intermediates/framebuffer.o] Error 1
=== alan_g is now known as alan_g|tea
=== Namidairo is now known as Namidairo`bnc
=== alan_g|tea is now known as alan_g
=== Namidairo`bnc is now known as Namidairo
=== oreneeshy_ is now known as oreneeshy
[14:59] <ogra_> nexwave-mat, looks like you are missing headers or binariews for the graphics stack
[15:05] <mfisch> ogra_: do you know where session error messages end up on the phone? stuff that would be in .xsession-errors
[15:06] <seb128> mfisch, .cache/upstart/<service>.log
[15:06] <ogra_> mfisch,  /home/phablet/.cache/upstart/
[15:06] <seb128> mfisch, same on the desktop btw, since we use upstart user sessions
[15:06] <ogra_> ah, seb128 beats me
[15:07] <mfisch> hmmm, usually I'd expect to see some complaints from dconf about the changes I made
[15:16] <tsdgeos> renato_: you there?
[15:17] <diwic> rsalveti, remove dependency on qtubuntu-media: https://code.launchpad.net/~diwic/mediaplayer-app/remove-qtubuntu-media/+merge/177719 and https://code.launchpad.net/~diwic/gallery-app/remove-qtubuntu-media/+merge/177718
[15:17] <diwic> rsalveti, notice that there is no merge proposal to add it to the relevant seeds because I don't know what those seeds are
[15:23] <ogra_> yay
[15:23]  * ogra_ sees a 20130731.3
[15:23] <ogra_> ophew
[15:24] <popey> ooh
[15:26] <nexwave-mat> make -C kernel/lge/iproj O=/home/mdaniel/ubuntutouch/out/target/product/p930/obj/KERNEL_OBJ ARCH=arm CROSS_COMPILE=" /home/mdaniel/ubuntutouch/prebuilt/linux-x86/toolchain/arm-eabi-4.4.3/bin/arm-eabi-" zImage
[15:26] <nexwave-mat> make[1]: Entering directory `/home/mdaniel/ubuntutouch/kernel/lge/iproj'
[15:26] <nexwave-mat> make[2]: /home/mdaniel/ubuntutouch/prebuilt/linux-x86/toolchain/arm-eabi-4.4.3/bin/arm-eabi-gcc: Command not found
[15:27] <nexwave-mat> I think I fixed my previous issue, now I get that.
[15:29] <plars> https://bugs.launchpad.net/touch-preview-images/+bug/1206969
[15:29] <ubot5> Launchpad bug 1206969 in touch-preview-images "selecting videos does not bring up any data or thumbnails" [Undecided,New]
[15:33] <asac> ogra_: there?
[15:33] <ogra_> asac, half (standup meeting)
[15:33] <asac> ogra_: more important here :)
[15:33] <asac> ogra_: just quick: did you kickoff new image for bill?
[15:33] <asac> where is that right now?
[15:33] <ogra_> http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/20130731.3/
[15:34] <asac> we either want to publish the /current
[15:34] <asac> err the last
[15:34] <asac> but would like to see if this one is better
[15:34] <asac> and then push that...
[15:34] <asac> then we add all the other tests that we wanted to add
[15:35] <ogra_> asac, wont be better
[15:35] <ogra_> video lens is completely empty for example
[15:36] <asac> ogra_: thtas fine
[15:36] <asac> ogra_: we want webbrowser green
[15:36] <asac> that one is fixed... so wanted to give it a chance
[15:36] <asac> ...
[15:36] <asac> lets wait for results
[15:36] <asac> otherwise lets go for 31
[15:39] <asac> doanac: where is the .3 image in test?
[15:39] <asac> can you check?
[15:39] <asac> plars: ?
[15:39] <asac> (read above)
[15:39] <ogra_> asac, its only a few minutes old
[15:39] <asac> yeah
[15:39] <ogra_> didnt reach testing
[15:39] <asac> 30+ minutes
[15:39] <asac> should be going off soon
[15:40] <plars> asac: looks like the jobs are running right now
[15:40] <asac> nice!!
[15:40] <doanac> asac: they are running
[15:40] <asac> doanac: gema: so we want to rename the Smoke tab to "Image Testing"
[15:41] <asac> plars: ^^
[15:41] <rsalveti> diwic: https://code.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/ubuntu-touch.saucy for the seeds
[15:41] <asac> doanac: gema: plars: not sure if we can just piggyback that with next dashboard role out
[15:41] <asac> we hoped so :)
[15:41] <doanac> asac: should be easy i think
[15:41] <diwic> rsalveti, will this then spread to other experimental images (e g the mir image) ?
[15:42] <asac> doanac: ok... can you note that somewhere so it just happens at some point?
[15:43] <asac> thx
[15:44] <rsalveti> diwic: it should, yeah
[15:44] <diwic> rsalveti, ok, but still not my pulseaudio image?
[15:45] <doanac> asac: cjohnston is making the change now and we'll roll it out in a bit
[15:46] <cjohnston> asac: are you wanting 'Image testing' or just 'Image'? due to spacing, 'Image' would work better for us
[15:46] <diwic> rsalveti, I mean, if there's some auto-spreading from the main seed to the mir seed, is there also one from the main seed to the pulseaudio seed?
[15:46] <rsalveti> diwic: yeah, because there's a custom seeds for that one specifically
[15:46] <rsalveti> diwic: right, I don't think the mir one is using a custom seeds at this point, like we do for pulse
[15:46] <rsalveti> so for pulse you'd need to change the specific seeds for the pulseimage as well
[15:47] <rsalveti> and upload a new meta
[15:47] <diwic> rsalveti, okay. I'll just trust you on that and submit a MR for the branch you suggested, a bit later tonight
[15:47] <rsalveti> diwic: sure, just give me the mr link and I'll take care of it
[15:47] <diwic> rsalveti, thanks
[15:48] <cjwatson> ogra_: I doubt that FLASH_KERNEL_SKIP=1 change you pasted earlier will do anything useful.  You need to do   Chroot chroot "env FLASH_KERNEL_SKIP=1 ..."   for the relevant comments, in much the same way as it's done in the ac100/nexus7 case.
[15:49] <ogra_> cjwatson, yeah, i learned that the hard way
[15:49] <ogra_> i just backed out the change again ... after all it was just for asac's conveninece that i added it initially
[15:49] <cjwatson> Sorry I wasn't around earlier; it's my wife's birthday tomorrow so I had an entirely predictable OMG-need-to-buy-presents lunchtime
[15:49] <ogra_> hah
[15:49] <ogra_> yeah, no worries
[15:50] <ogra_> i just backed out all the stuff again, its not important
[15:50] <asac> ogra_: what did you back out?
[15:50] <ogra_> asac, autiomatic flashing of the kernel if you install a deb
[15:50] <asac> for me? :)
[15:50] <ogra_> i added that for you since i had the code around already
[15:50] <asac> ogra_: i dont need it... i will not install debs :)
[15:51] <ogra_> but didnt think about the fact that we dont set FLASH_KERNEL_SKIP anymore during image builds
[15:51] <ogra_> so it yried to flash the maguro kernel on the panda buildd ... didnt work out :)
[15:52] <doanac> cjohnston, asac: how about "Images" instead of "Image Testing"
[15:52] <asac> doanac: hold off until tomorrow :)
[15:52] <asac> cjohnston: :)
[15:52] <asac> we have to agree on what we want to do
[15:53] <asac> guess we redo the whole dashboard as it sounds :)
[15:53] <asac> (j.k.)
[15:53] <cjohnston> ok
=== schwuk is now known as schwuk_away
[15:59] <awe> davmor2, sigh...
=== schwuk_away is now known as schwuk
[16:01] <davmor2> awe: I thought that would make your day right :)
[16:02] <awe> davmor2, I have a meeting now, I'll get to debug when I'm back
[16:03] <awe> can you attach the output of 'list-contexts' and 'nmcli dev'?  Also, grabbing the syslog and filtering out all the NM messages would be super-useful ( eg. 'grep "NetworkManager" /var/log/syslog' )
[16:03] <PaoloRotolo> Hi all!
[16:05] <davmor2> awe: yeap no worries
[16:05] <awe> davmor2, pretty sure this is NM not doing the right thing
[16:06] <davmor2> awe: from list-modems it is attached and powered so it just fried my head completely when it didn't connect  :(
=== Namidairo is now known as Namidairo`bnc
[16:07] <awe> davmor2, note due to the way NM works, there may be a delay of ~5m or so before the connection is activated
[16:08] <davmor2> awe: this has been since this morning I still see no rmnet only wlan0 and local
[16:08] <awe> k
[16:08] <awe> then the NM messages from syslog should tell us what's going on
=== Ursinha is now known as Ursinha-afk
[16:11] <Joe_B> \join #1310wallpaper
[16:14] <wilee-nilee> Joe_B, all by ur self
=== Ursinha-afk is now known as Ursinha
[16:29] <davmor2> awe: NetworkManager syslog in and the other two you asked for are now added to the bug
[16:32] <awe> davmor2, thanks!
[16:41] <mfisch> seb128: is the system-level upstart gone? I don't have a log folder for it (or maybe nothing's been logged)
[16:41] <seb128> no it's not
[16:41] <seb128> what folder?
[16:42] <seb128>  /var/log/upstart ?
[16:42] <popey> ogra_: did you expect the ueventd spinning to be fixed in the latest pending image?
[16:43] <mfisch> seb128: yeah
[16:43] <ogra_> popey, no, that will only be fixed by jodh's magic hands
[16:43] <seb128> mfisch, works here
[16:43] <mfisch> I'm trying to figure out why I can
[16:43] <ogra_> popey, i expect it to happen less often though
[16:43] <mfisch> I'm trying to figure out why I can't get past the google logo :(
[16:43] <awe> mfisch, what phone/image?
[16:44] <mfisch> awe: yesterday's image, but I'm doing some settings stuff. I locked the wallpaper which is probably the cause, but I'd like to know why
[16:44] <mfisch> I'm going to revert that change real quick
[16:44] <awe> which phone?
=== dandrader_ is now known as dandrader|lunch
[16:46] <mfisch> awe: nexus4
=== annerajb_brb is now known as annerajb
[16:49] <jdstrand> sergiusens: thanks
[16:53] <awe> mfisch, OK... I ran into a can't boot problem last night on nexus4
[16:53] <awe> wifi drivers failed to load
[16:59] <mfisch> Laney: you around? I had some settings questions
[16:59] <Laney> mfisch: I'm going in a minute
[16:59] <Laney> ask though - one of the others should be able to help
[16:59] <mfisch> Laney: is there an architecture specified for the system settings?
[16:59] <Laney> in what sense?
[17:00] <mfisch> I'm curious about the layout of /usr/share/settings
[17:01] <ogra_> mfisch, if you cant get past the google logo it means your container didnt start
[17:01] <ogra_> (for whatever reason)
[17:02] <Laney> the pcfile has variables for locations you should use
[17:02] <mfisch> ogra_: would the lxc logs be the best place to debug that?
[17:02] <ogra_> lxc logs, /system/bin/logcat as well
[17:02] <ogra_> and the upstart logs
[17:02] <mfisch> I always forget about logcat
[17:04] <mfisch> Laney: I'll catch up with you more tomorrow, thanks
[17:04] <Laney> mfisch: kenvandine will be around for a bit longer :-)
[17:04] <mfisch> ogra_: I do not have a /var/log/upstart, even on a functional build
[17:04] <mfisch> kenvandine: morning
[17:05] <ogra_> mfisch, hmm, thats is definitely shipped inside the tarball
[17:05] <mfisch> ogra_: I'm using the system image update images, perhaps thats why
[17:05] <ogra_> so either your flashing went badly or the /data partition is full or some such
[17:05] <ogra_> oh
[17:06] <ogra_> yeah, that might be
[17:06] <mfisch> /var/log itself is mounted rw even on that image
[17:06] <ogra_> yes, but it doesnt copy the existing readonly stuff in
[17:06] <ogra_> before bind mounting the rw space
[17:07] <mfisch> and upstart won't make the folder?
[17:07] <ogra_> which is slightly broken i think, since some apps might expect their logdirs in place
=== LarrySteeze is now known as LarrySteeze|Away
[17:07] <ogra_> dunno, i hoestly never tried deleting it :)
[17:08] <ogra_> and i'm still using normal flipped everywhere since i try to get ports working properly
[17:08] <mfisch> ogra_: yep, upstart won't make it, but if I do, I get logs
[17:08] <mfisch> thats a bug
[17:08] <ogra_> mfisch, well, there is a bug in the design here
[17:08] <mfisch> these logs will make my life a lot easier
[17:09] <ogra_> stephane is out until next week though
[17:09] <mfisch> ogra_: when we did RO rootfs in UfA for awhile we were sure to cp everythnig out of the rootfs first
[17:09] <ogra_> yeah
[17:09] <ogra_> i think thats whay should happen here too
[17:09] <mfisch> and my unpacked tarball has the folder, odd
[17:09] <ogra_> we used that setup in LTSP years ago
[17:10] <mfisch> i'm also missing a folder called fsck and one called apt
[17:10] <ogra_> i kind of expected stgraber to copy that functions
[17:10] <mfisch> ogra_: I'll email him, thanks for the pointer
[17:10] <ogra_> the prob here is that it will make the first boot really slw ... unless we pre-populate rw spaces
[17:10] <ogra_> i.e. during build time
=== salem_ is now known as _salem
[17:11] <ogra_> *slow
[17:13] <mfisch> ogra_: cp -a /var/log foo is going to be really slow?
[17:13] <mfisch> hopefully there's not much there on first boot
[17:13] <ogra_> mfisch, doing that for all dirs and files in all rw spaces in use
[17:14] <ogra_> i wouldnt just do it for logs, i'm sure there are other spaces where it makes sense
=== dandrader|lunch is now known as dandrader
=== _salem is now known as salem_
[17:30] <mfisch> kenvandine: do the settings allow for a background image that lives outside of /usr/share/settings/system/qml-plugins/background?
=== schwuk is now known as schwuk_away
[17:46] <seb128> mfisch, yes, the settings app write to gsettings, like on the desktop, the path can be anything ... why?
[17:46] <seb128> mfisch, https://code.launchpad.net/~unity-team/unity8/unity8.background_gsettings/+merge/174958 didn't get merged yet, so the key is not used on the touch image
[18:00] <stgraber> ogra_: not quite, I'm actually working on Friday this week and the bind-mount logic rework is planned for that day
[18:01] <stgraber> ogra_: (then I'm off again on Monday, Tuesday and Wednesday)
[18:03] <ogra_> stgraber, heh, ok
[18:05] <ogra_> stgraber, what i always loved about mdz's old bind mount code was how he used tar to actually speed up cp :)
[18:06] <stgraber> ogra_: :) yeah, I think my code is currently just a good old cp -a which tends to be easier for people to understand (and as it's a one-time thing on very small dirs, shouldn't be a big deal anyway)
[18:07] <ogra_> yeah, we dont run it on every boot
[18:22] <diwic> rsalveti, qtubuntu-media: https://code.launchpad.net/~diwic/mediaplayer-app/remove-qtubuntu-media/+merge/177719 and https://code.launchpad.net/~diwic/gallery-app/remove-qtubuntu-media/+merge/177718 and https://code.launchpad.net/~diwic/ubuntu-seeds/add-qtubuntu-media/+merge/177917
[18:22] <rsalveti> diwic: great, thanks
=== iveand is now known as rik-shaw
[18:36] <rik-shaw> is there a better irc channel for ubuntu edge?
[18:39] <nexwave-mat> I managed to make a build but the screen stayed black. I casted int to make it build but that broke. I was wondering if someone could set me on the path for the correct fix. http://pastebin.com/tbbjR1sC
=== Ursinha is now known as Ursinha-afk
=== Ursinha-afk is now known as Ursinha
[19:02] <mfisch> rik-shaw: I think this is the place
[19:03] <mfisch> seb128: interesting that the setting is not used. When I set it in a secondary dconf db, I see the setting pointing to /usr/share/settings/...
[19:03] <mfisch> seb128: and when I lock that setting with a dconf lock, I don't get a UI
[19:10] <mfisch> Ubuntu for Android AMA on reddit: http://www.reddit.com/r/Ubuntu/comments/1jfqpf/we_are_the_engineering_team_behind_ubuntu_for/
[19:22] <HumanUserX64> Hi, could someone design a flyer, which can be spread in towns and cities?
[19:22] <HumanUserX64> ^^ for Ubuntu Edge
[19:23] <wilee-nilee> HumanUserX64, It's not even been released, you looking for a bad rap, say if it does not get there.
[19:23] <doanac> sergiusens: any updates on the lxc-android-config stuff needed for rndis?
[19:24] <wilee-nilee> has to be funded to be released
[19:24] <troyready> mfisch, what happened to the AMA?
[19:24] <mfisch> sorry about the AMA, it was rescheduled for Monday
[19:24] <mfisch> troyready: ^^^
[19:24] <mfisch> so it could be announced better
[19:24] <seb128> mfisch, it's used by the settings app, but not by the shell
[19:25] <troyready> Makes sense
[19:27] <rik-shaw> regarding ubuntu edge, wondering if anyone knows a bit more about possibility of using w/ Verizon.  I know that Verizon will move to LTE, but I just read that "VoLTE" (voice) will not come to Verizon until "sometime 2014".
[19:28] <sergiusens> doanac: I think I need consensus with ogra_ and stgraber regarding the _where_
[19:28] <rik-shaw> Meaning, quite possible late 2014, which means in April 2014 Verizon users receive a shiny new Ubuntu Edge that won't be able to handle voice calls for several months.  Is the possible true?
[19:28]  * ogra_ looks up
[19:28] <rik-shaw> So, if a Verizon only customer, then don't get too excited about getting an Edge just yet.
[19:29] <ogra_> sergiusens, doanac, was the the sed in sed thing ?
[19:30] <sergiusens> ogra_: yes, it's wrt that
[19:30] <doanac> ogra_: yeah, basically you can't update this file in the new image style because its read-only
[19:30] <doanac> so its 2 related issues, i'd say
[19:30] <HumanUserX64> wilee-nilee: The people are advertising it anyway via the social networks. So why don't make a flyer for the post-boxes
[19:30] <ogra_> well, lets add a pre-start.d dir and source it ... i can add that with the next upload... you need stgraber for having it writable in the system images
[19:31] <sergiusens> ogra_: exactly, that is the thing, do we want it writable on all images?
[19:31] <ogra_> no, in dev mode only
[19:31] <sergiusens> ogra_: but if you are in dev mode, it's already writable
[19:32] <ogra_> if there is something we want to live constantly in it this should come from a package at build time
[19:32] <ogra_> oh, indeed
[19:32] <ogra_> so we dont need stgraber :)
[19:32] <ogra_> i assume the tests will always (or at least for a long time) have to use dev mode
[19:33] <sergiusens> ogra_: the target is to not ;-)
[19:33] <ogra_> heh
[19:33] <HumanUserX64> wilee-nilee: something like "Support us to establish the next generation of personal computing"
[19:34] <ogra_> ok, if thats actually a short term goal lets make it always writable
[19:34] <ogra_> hmm
[19:34] <ogra_> or how about we make it writable on demand ... from the host side ?
[19:35] <ogra_> (create a tmpfs bindmount for the dir before we copy something in)
[19:35] <ogra_> do we need it persistent across reboots ?
[19:36] <sergiusens> ogra_: yeah, needs to reboot (from doanac's script at least)
[19:36] <doanac> we reboot the device between tests in the lab, so that would probably be nice
[19:36] <ogra_> i assume the tests will run as root, or is that also planned to go away ?
[19:36] <ogra_> or at least the instance that fires off the tests will
[19:36] <doanac> we need to be root at least now, to install packages
[19:37] <sergiusens> doanac: not in that image
[19:37] <sergiusens> can't install unless it's click
[19:37] <ogra_> well, that wil go away with rw capability
[19:37] <sergiusens> and click can be installed by the user
[19:37] <doanac> true
[19:37] <ogra_> right, so we cant do that ...
[19:38] <ogra_> ell, then its an stgraber thing ... and i'd say lets always have the dir rw then
[19:38] <sergiusens> ogra_: doanac we can also setup the container to do whatever the sed is trying to do...
[19:39] <wilee-nilee> HumanUserX64, Well as a owner operator of a carpet cleaning business, I would get only a 3% return on mail outs to peoples home offering a service for cost, not even a support, so my advice is to not waste your time. It was profitable for me, but in your scenario, I doubt you get much response.
=== gustavold1 is now known as gustavold
[19:39] <doanac> sergiusens: that's fine with me, unless that breaks some usage you guys need
[19:39] <sergiusens> doanac: you were disabling adb in the container, right?
[19:39] <ogra_> hmm, looking closer the whole function does a lot of stuff that needs root
[19:40] <sergiusens> ogra_: you are still root on the images, it's just ro
[19:41] <ogra_> yes, but if that goes away too most of these bits wont work
[19:41]  * sergiusens needs to pull the MR again
[19:41] <ogra_> neither the sed, nor the chmod or even writing to the /data/property dir
[19:42] <ogra_> "sleep 4  # give time to reboot" ... now thats  naive ... add a 4 in front
[19:43] <ogra_> reboots are definitely more than 40sec
[19:43] <sergiusens> ogra_: yup, but adb wait-for-device works there
[19:43] <ogra_> ok
[19:43] <ogra_> well, then you dont need a sleep ;)
[19:43] <sergiusens> ogra_: it hangs when going from recovery to system or viceversa
[19:43] <sergiusens> ogra_: wait-for-device is not perfect ;-)
[19:44] <ogra_> kk
[19:45] <ogra_> well, so since we are root and clearly need to be in this function, doing a bind mount ourselves can work as well
[19:45] <doanac> sergiusens, ogra_: i came up with a new hack yesterday for dealing with a "wait for reboot" type operation. basically  you run "reboot; stop adbd" (adb dies) then run "adb wait-for-device"
[19:45] <doanac> its a bit more deterministic that way
[19:45] <doanac> might be a decent utility to add the phablet-tools
[19:45] <sergiusens> doanac: does that really work?
[19:46] <sergiusens> doanac: well I'm going to add it to the stuff I'm doing now then
[19:46] <doanac> sergiusens: its working in a utah branch where I needed to do the same thing
[19:48] <ogra_> ok, so i'll do a rework of pre-start and will actually put all the snippets from the existing pre-start.sh into that dir ... which means we will have something like /var/lib/lxc/android/pre-start.d/adbd ... that you can just rm
[19:48] <ogra_> how does that sound ?
[19:48] <ogra_> (should save you from push/pull and sed
[19:49] <ogra_> )
[19:50] <sergiusens> ogra_: that sounds good, but would need rw for the rm ...
[19:50] <ogra_> yeah, we need it rw in either case
[20:42] <nexwave-mat> I managed to boot unflipped image, what needs to be done to boot flipped image?
[20:52] <w-flo> nexwave-mat, update the android_build repo and make sure you're using a recent ubuntu rootfs zip
[20:53] <w-flo> Depending on your device you might have to do more than that.. and to enable sensors/GPU drivers/sound/.. you should create a udev rules file, see https://wiki.ubuntu.com/Touch/PortingFlippedInProgress
[20:57] <w-flo> oh, and you should probably make sure you're using the phablet-saucy branch in the repo manifest, nexwave-mat
[20:58] <fartface> I've just installed the latest daily preinstalled on my Galaxy Nexus--but it says there are no data sources available.  How do I set up my SIM to work?
[20:58] <w-flo> fartface, you can ignore that message. It means there's nothing important to tell you about (like "20 new tweets" or "missed 10 calls")
[20:59] <fartface> But none of my data works either?
[20:59] <fartface> Oh hold on, maybe I'm wrong.
[21:00] <fartface> Nope.  No data.
[21:01] <fartface> Is there a way to set APNs yet?
[21:01] <fartface> It seems to recognize my SIM and my carrier, but I'm not able to get a data connection
[21:01] <dixeflatline> @fartface.  last i was using ubuntu touch i had to create an APN connect, touch the file, then nmcli con up id "file connection"
[21:01]  * fartface blinks
[21:02] <fartface> So what you're saying is "flash Android back, it's not set up"
[21:02] <w-flo> fartface, is your SIM locked / does it need a PIN?
[21:02] <fartface> No, no pin necessary
[21:02] <dixeflatline> no create one from a linux box
[21:02] <fartface> And it's unlocked, I'm able to use it anywhere
[21:02] <dixeflatline> then adb push it to your phone
[21:03] <fartface> Don't have a linux box at work unfortunately.
[21:03] <dixeflatline> make a virtualbox.
[21:03] <dixeflatline> who is your carrier?
[21:03] <fartface> Fido
[21:04] <dixeflatline> hmmm... i dont know Canada's bands
[21:05] <dixeflatline> GSM right?
[21:05] <fartface> Yes, they're the same as the states
[21:05] <dixeflatline> you have a linux box at home
[21:06] <fartface> Yeah, but this is nowhere near daily usage at this point.
[21:06] <fartface> I'm flashing android back.
[21:08] <dixeflatline> i think its a pretty awesome OS
[21:09] <dixeflatline> it can be a day to day but you just need to be ready to use that terminal like a boss
[21:10] <dixeflatline> the N4 has battery drain issues so i am waiting for beta
[21:10] <w-flo> he's long gone :(
[21:10] <dixeflatline> lol
[21:11] <dixeflatline> what device/s are you using w-flo
[21:12] <w-flo> Desire Z dixeflatline
[21:12] <w-flo> my daily use phone is a nexus 4, but that's currently android only
[21:12] <dixeflatline> yeah.  I have the N4 daily (android)
[21:13] <dixeflatline> Nexus S Ubuntu Touch and N7 dual boot Ubuntu / Android
[21:14] <w-flo> I won't switch before there's whatsapp for ubuntu. and my bank adds a authentification method other than mTAN or "QR TAN" which relies on an android app..
[21:14] <dixeflatline> hahaha i no what you mean.
[21:15] <dixeflatline> i thought that was a whatsapp port
[21:15] <KHendrik> dixeflatline: how did you manage to setup dual boot?
[21:15] <w-flo> It seems like it was not very active lately
[21:16] <dixeflatline> it was whosthere in place of whatsapp
[21:16] <Tassadar_> KHendrik: http://forum.xda-developers.com/showthread.php?t=2011403
[21:16] <KHendrik> I hope once tox is released i can get my friends to use it
[21:16] <KHendrik> that would be a great whatsapp replacement
[21:17] <dixeflatline> thanks <Tassadar_>
[21:17] <w-flo> KHendrik, the problem is getting everyone to use it. there's lots of messenger I personally prefer over whatsapp, but as long as people don't use them, it's pretty useless :D
[21:18] <Tassadar_> Am I allowed to promote <_< ?
[21:18] <Tassadar_> would be great to get it to some other devices than N7
[21:18] <dixeflatline> like the N10
[21:18] <dixeflatline> lol
[21:18] <dixeflatline> and the N4
[21:18] <Tassadar_> http://igg.me/at/multirom-n7 thinking of the new N7 first)
[21:18] <dixeflatline> your right :)
[21:18] <w-flo> yeah, the N4 :D
[21:19] <KHendrik> Tassadar_: thanks looks promising will give it aspin this weekend
[21:20] <dixeflatline> brb
[21:20] <KHendrik> w-flo: you're absolutely right hope prism got them a little more aware of the need for secure messaging but i doubt it
[21:20] <dixeflatline> VPN
[21:20] <dixeflatline> TOR
[21:21] <dixeflatline> that would be a good port for ubuntu touch... TOR-touch
[21:23] <Fishscene> I like combining words... Ubuntu TORch
[21:23] <w-flo> whatsapp got pretty much every aspect of security wrong (and then later fixed some of it).. and they might be the most expensive (?) IM app available for android.. but still they're the most successful, simply because it is very easy to use.
[21:24] <KHendrik> w-flo: yeah ease of use is always the most important things for a programm to become widespread
[21:25] <KHendrik> tox seams to be easy but also secure though my cryptography knowledge is not very good so i could be wrong
[21:25] <dixeflatline> im off to grab lunch.  ill come bug you all in a bit
[21:26] <KHendrik> and I'll go to sleep its already late and i have to be up early
[21:27] <KHendrik> cya
=== roasted_ is now known as roasted
=== kentb is now known as kentb-out
[23:33] <cpatrick08> I was wondering if the saucy-preinstalled-touch-armhf.tar.gz was a tar.gz file of thr saucy-preinstalled-touch-armhf.zip file
[23:34] <Tassadar_> no, saucy-preinstalled-touch-armhf.tar.gz is inside the zip file)
[23:34] <Tassadar_> zip file installs contents the the .tar.gz archive
[23:35] <cpatrick08> ok thanks I was going to multirom it on my nexus 7 and liked the .zsync option of that file. multirom requires the ,zip file, could you make a .zsync file of the .zip file please
[23:36] <Tassadar_> hmm, you can probably just download the zip file once and keep tar.gz zpdated
[23:36] <Tassadar_> *updated
[23:36] <Tassadar_> it belongs to the root of the zip file
[23:37] <cpatrick08> ok thanks
[23:38] <cpatrick08> here is the multirom instructions if you are interested, it is #2 under adding roms http://forum.xda-developers.com/showthread.php?t=2011403
[23:38] <Tassadar_> yeah, I know - I made it)
[23:39] <Tassadar_> and, if _you_ are interested, I'm trying to get MultiROM to more devices, the first beign 2nd gen Nexus7 - http://igg.me/at/multirom-n7
[23:40] <cpatrick08> ok so would i just unzip the file .zsync the tar.gz file and add the new file to a zip file with saucy-preinstalled-touch-armhf.zi p as the file name?
[23:40] <cpatrick08> I have the 1 gen nexus 7
[23:41] <Tassadar_> cpatrick08: yes. you should see the .tar.gz inside the ZIP file
[23:41] <cpatrick08> ok i will do that, thanks for your help