| === chriadam|away is now known as chriadam | ||
| === aaron__ is now known as ahoneybun | ||
| === chriadam is now known as chriadam|away | ||
| === aaron__ is now known as ahoneybun | ||
| Mi | got any idea porting Colorfly G708 to Ubuntu-Touch guys ? it's cheapest tablet actually, but with high spec | 06:04 |
|---|---|---|
| === chihchun is now known as chihchun_afk | ||
| === chihchun_afk is now known as chihchun | ||
| === chihchun is now known as chihchun_afk | ||
| === IdleOne is now known as Guest91071 | ||
| === chihchun_afk is now known as chihchun | ||
| Mo | Hi, which window manager is used for touch? Can I find some information about the details of Ubuntu Touch before installing? | 08:28 |
| Mo | On https://developer.ubuntu.com/en/start/ubuntu-for-devices/porting-new-device/ I see, Mir is the X11 replacement. Is there some place with bundled information? | 08:31 |
| k1l | its MIR + unity8 on ubuntu touch | 08:31 |
| Mo | k1l: Can I use those with other Linux Distributions? | 08:32 |
| k1l | unity8 seems to be bundled to mir. and only ubuntu seems to be going to use mir, most other linux distros are going to use wayland after x11 | 08:34 |
| davmor2 | Mo: unity8 and mir combined provide it | 08:34 |
| k1l | i dont know how difficult it is to run mir+unity8 on other distros. maybe someone in here tried that | 08:35 |
| Mo | I'm running a All-in-one Gentoo-Desktop with Xfce4. To use the touch feature I'm looking for a better Window Manager or Display Server. | 08:36 |
| k1l | ah yeah, the regular desktops are not really usable with touch so far. | 08:37 |
| Mo | X has bad touch support, it has multi-touch afaik, but the WMs don't support it, at least Xfce doesn't. | 08:37 |
| Mo | k1l: So can't I just install ubuntu touch on that amd64? Most releases are pre-compiled/packaged for special devices like Nexus, or the inofficial ones (some day I'm going to try on my Asus Transformer Pad). but there is no universal install for the amd64 instead of Ubuntu desktop. | 08:38 |
| Mo | Or can I just just the Mir/Unity-8 touch features on Ubuntu desktop? | 08:39 |
| Mo | ..just use.. | 08:39 |
| jgdx | since there's no official way yet, there's no "just" | 08:39 |
| tvoss | Mo, we have got desktop next images, probably best to look for those | 08:39 |
| tvoss | willcooke, ^ | 08:39 |
| Mo | Moreover I would like to try as Live-Distribution first, how it feels like. | 08:39 |
| k1l | Mo: mir+unity8 is not ready for the desktop, yet. its only running on smartphones. so you can test with the "desktop next" iso. but dont expect a fully working thingy for desktops since its a development iso | 08:40 |
| Mo | desktop next? Is that some Ubuntu release? | 08:40 |
| k1l | yes | 08:41 |
| Mo | Supporting Phones would be harder than Desktops, so what works on Phones should work on Desktop as well. Then any Tablet release would fit for Desktop screen size as well, no? | 08:41 |
| k1l | Mo: you are making it to easy than it actually is. | 08:42 |
| Mo | Ok. | 08:42 |
| Mo | is there some central website for desktop next? I just find the downloads. Is desktop next a live-system already? | 08:43 |
| Mo | live meaning, an iso I can start? | 08:43 |
| k1l | yes | 08:43 |
| willcooke | What's your use case here Mo? | 08:43 |
| Mo | [10:36] <Mo> I'm running a All-in-one Gentoo-Desktop with Xfce4. To use the touch feature I'm looking for a better Window Manager or Display Server. | 08:43 |
| willcooke | Then desktop next wont help you | 08:44 |
| willcooke | If you run desktop next you also run Unity 8 and Mir | 08:44 |
| willcooke | and the relatively limited apps therein | 08:44 |
| willcooke | Unity 7 has better touch support than xfce4 | 08:45 |
| Mo | willcooke: I like to look what is possible. I like to install some Ubuntu touch if it supports amd64 desktops as well and is a full desktop featured distribution. | 08:45 |
| willcooke | oki, if you just want to give it a spin, then http://cdimage.ubuntu.com/ubuntu-desktop-next/daily-live/current/ | 08:45 |
| Mo | Why won't desktop next help me? | 08:45 |
| Mo | Ok, I'll try.. | 08:45 |
| willcooke | because its for testing and playing with Unity 8 on a desktop, not for doing actual work yet | 08:46 |
| willcooke | Like, if you're expecting a full featured desktop env. | 08:46 |
| willcooke | then you will find it missing some bits you want | 08:46 |
| Mo | Then about touch in general, thinking about replacing Android on my TF700T... This question isnt't in the FAQ: Is there some App Market? Can I use any Linux application available on Ubuntu Desktop? Can it run Android Apps? | 08:46 |
| willcooke | Yes, there is an app market | 08:47 |
| willcooke | No you cant use Android apps | 08:47 |
| willcooke | And regarding "any linux application" | 08:47 |
| willcooke | kinda | 08:47 |
| willcooke | There is a thing called Xmir (which is in a new ppa and being tested now) | 08:48 |
| Mo | willcooke: Gentoo provides Unity and Mir via the Unity-Overlay. After trying I can still use these things beside my current daily applications, that some are GTK and some Qt, some KDE. | 08:48 |
| willcooke | Eventually Gtk and Qt/Kde apps will run natively on Mir | 08:49 |
| Mo | willcooke: So applications must support Mir to work on Mir (without the new Xmir)? So Qt applications for X won't work in general on Mir? | 08:49 |
| willcooke | for everything else there's Xmir | 08:49 |
| Mo | ok | 08:49 |
| willcooke | We have Mir backends for Qt and Gtk so they'll work once well | 08:50 |
| k1l | willcooke: btw: do you already know when snapp packages will be enabled on the desktop next image? | 08:50 |
| willcooke | s/once/<null> | 08:50 |
| willcooke | k1l, So the current desktop next image will stop being updated to be replaced with a Snap version Real Soon Now (probably a few weeks to a month) | 08:50 |
| willcooke | and then Snaps will work | 08:51 |
| k1l | willcooke: ok, thanks. i am curious to test the snappy setup :) | 08:51 |
| willcooke | k1l, in the early days it will be one big image, where apps etc are in the image, but then as we mature the platform then those will turn in to snaps | 08:52 |
| JamesTait | Good morning all; happy No Homework Day! 😃 | 09:01 |
| mr2515_ | JamesTait: Morning! | 09:16 |
| === mr2515_ is now known as sturmflut-work | ||
| leonmortiba | good morning | 09:17 |
| sturmflut-work | Comments welcome: https://sturmflut.github.io/ubuntu/touch/2015/05/06/hacking-ubuntu-touch-part-2-devices-and-images/ | 09:18 |
| JamesTait | Morning, sturmflut-work! | 09:18 |
| mcphail | Is there an easy way to contribute a ringtone to UT? | 09:59 |
| jgdx | mcphail, propose against lp:ubuntu/vivid/ubuntu-touch-sounds | 10:06 |
| ogra_ | mcphail, file a wishlist bug against the ubuntu-touch-sounds package | 10:06 |
| jgdx | that's the more diplomatic route :p | 10:06 |
| mcphail | jgdx: ogra_: Thanks :) | 10:06 |
| ogra_ | yeah, both will work :) | 10:06 |
| jgdx | mcphail, here's some info on the current tones: bug 1239612 | 10:07 |
| ubot5 | bug 1239612 in ubuntu-touch-sounds (Ubuntu) "New ringtone and notification sounds for 13.10 mobile" [Low,Fix released] https://launchpad.net/bugs/1239612 | 10:07 |
| jgdx | pete-woods, sometimes my sim is not unlocked, and I can't unlock it manually (from indicator or settings). | 10:15 |
| jgdx | it's like the private connectivity service silently fails | 10:16 |
| pete-woods | jgdx: the SIM unlock codepaths are pretty scary, so sadly this doesn't surprise me :( | 10:16 |
| === greyback__ is now known as greyback | ||
| * jgdx puts fingers in ears and hums | 10:16 | |
| pete-woods | it's like we talk to an inobvious dbus api (ofono), pump out a message to the notification service saying, "here's a gmenu export and some actions", then unity8 connects to the gmenu endpoints, shoves some stuff in, hopefully calls all the right actions, then we ship this data back off to ofono | 10:17 |
| jgdx | pete-woods, same as the indicator pre your refactor? Not using qtdbus? | 10:17 |
| pete-woods | jgdx: the problem really lies in the complexity of the architecture | 10:18 |
| pete-woods | and because it's gmenu, we have a mix of qdbus and gdbus | 10:18 |
| pete-woods | it makes it so hard to know you're doing the right thing | 10:18 |
| pete-woods | like, I'd really like to say "give me a SIM unlock dialogue" so some pre-canned nice bit of Qt / QML API | 10:19 |
| pete-woods | but the reality is it spans across like 5 classes | 10:19 |
| pete-woods | there's probably a bug in the code, sure | 10:19 |
| pete-woods | and it's much simpler post-qdbus | 10:19 |
| pete-woods | but still there's a lot of room for screwups | 10:19 |
| pete-woods | writing the tests for SIM unlock took literally over a week | 10:20 |
| pete-woods | just to figure out how it all hangs together | 10:20 |
| pete-woods | and we still can't work out how to test the cancel button | 10:20 |
| jgdx | does it help that the dbusmock ofono template is really spartan? :p | 10:20 |
| pete-woods | well we had to fill in a bunch of missing stuff there | 10:21 |
| pete-woods | which is fine | 10:21 |
| jgdx | who'd you ask for the unlock dialogue anyway? | 10:21 |
| pete-woods | it pre-dates me working on the project | 10:21 |
| pete-woods | I've only been on it a month | 10:22 |
| jgdx | ideally, where would that lie? | 10:22 |
| pete-woods | the SDK maybe? | 10:22 |
| pete-woods | I just don't like having to construct it via "convention" using magic gmenu incantations | 10:22 |
| pete-woods | which the shell interprets as SIM unlock please | 10:22 |
| pete-woods | I want a nice async qt API for it | 10:23 |
| pete-woods | like you'd expect for opening a Qt dialogue box | 10:23 |
| jgdx | this probablt falls outside the scope of the sdk | 10:24 |
| pete-woods | yeah, that's fair enough | 10:24 |
| pete-woods | we've basically created the API I want to talk to, inside the indicator itself | 10:24 |
| pete-woods | realistically my views about notifications / dialogues are not going to be addresses, anyway | 10:25 |
| pete-woods | *addressed | 10:25 |
| pete-woods | so I just need to shut up and fix the bug in the indicator, wherever it lies | 10:25 |
| jgdx | so is the private connectivity api put in the indicator just because it was easy? It's already running and is privileged etc | 10:30 |
| jgdx | anyway, if I can help, let me know pete-woods. I see it from time to time, and I am happy to run experimental/debugging indicator-network code | 10:32 |
| === Guest91071 is now known as IdleOne | ||
| pete-woods | jgdx: thanks. if there's anything you can find that increases the chances of it happening, that'd be great to know | 10:34 |
| === dandrader is now known as dandrader|afk | ||
| === JMulholland_ is now known as JMulholland | ||
| === MacSlow is now known as MacSlow|lunch | ||
| === dandrader|afk is now known as dandrader | ||
| === chihchun is now known as chihchun_afk | ||
| === _salem is now known as salem_ | ||
| === MacSlow|lunch is now known as MacSlow | ||
| jgdx | pete-woods, i've seen it a couple of times when ofono needs to reprovision apns. Not sure if that pertinent though. | 12:31 |
| pete-woods | jgdx: could well be. for all I know the bug is in qofono. but without being able to reproduce it, I've had a very hard time figuring out what's wrong | 12:36 |
| sturmflut-work | barry: Ping | 12:56 |
| sturmflut-work | Can anybody explain to me how http://paste.ubuntu.com/10997080/ is to be interpreted? It's the flash partition layout of the bq E4.5, why do eleven partitions point to the block device itself (mmcblk0)? Are those flash regions at the beginning of the block device, before the "actual" partitions? Or are these fake entries for something special? | 13:02 |
| barry | sturmflut-work: pong | 13:14 |
| lotuspsychje | http://news.softpedia.com/news/Ubuntu-Developers-Working-to-Use-Systemd-on-Ubuntu-Touch-480326.shtml | 13:19 |
| peat-psuwit | Why NetworkManager says that wlan interface is 'managed' if it appear after NetworkManager is started? | 13:20 |
| brunch875 | Hello! | 13:23 |
| === dholbach_ is now known as dholbach | ||
| sturmflut-work | ogra_: Ping | 13:35 |
| ogra_ | whats up ? | 13:35 |
| sturmflut-work | ogra_: I want to look at the initrd that is part of the krillin recovery.img. So I went an used "abootimg -x recovery.img", and the initrd is extracted to initrd.img, but it doesn't have any known file format. It's suposed to be gzip, but it's not. | 13:37 |
| sturmflut-work | ogra_: Any idea? | 13:37 |
| ogra_ | sturmflut-work, not really, ask janimo or ondra, i'm not sure how these files are post processed ... the recovery initrd is an android one though | 13:37 |
| sturmflut-work | ondra, janimo: Ping | 13:38 |
| sturmflut-work | ogra_: Thanks | 13:38 |
| ondra | sturmflut-work use android tools | 13:38 |
| ondra | sturmflut-work recovery image is packed as android, so you need to split it from kernel and the gzip + cpio to get recovery ramdisk out | 13:39 |
| janimo | sturmflut-work, dd if=initrd.img of=newinitrd.img bs=512 skip=1 | 13:40 |
| ogra_ | ah, a header :) | 13:40 |
| janimo | ogra_, right, as some uboot images used to append on dev arm boards supported by ubuntu | 13:40 |
| janimo | prepend actually | 13:41 |
| ogra_ | well, these are usually recognized by file | 13:41 |
| sturmflut-work | janimo: Argh, thanks. One would expect that "abootimg-unpack-initrd" knows about this. | 13:41 |
| ogra_ | sturmflut-work, no, it expects plain ubuntu initrds | 13:41 |
| ogra_ | what you have there is actually a uInitrd file for u-boot | 13:42 |
| janimo | the android build tools in this specific tree add some extra headers, like ROOTFS/RECOVERY/BOOTIMG, not sure what else is there though | 13:42 |
| === dandrader is now known as dandrader|afk | ||
| brunch875 | Does the browser keep session when downloading file? Downloads fail for a website which requires login. | 14:05 |
| brunch875 | I think it's a browser issue since it doesn't work on the desktop version of the browser either. | 14:07 |
| === dandrader|afk is now known as dandrader | ||
| === brunch876 is now known as brunch875 | ||
| === sturmflut-work is now known as sturmflut2 | ||
| soothran | hey people! | 15:53 |
| soothran | i am thinking about starting a porting project just for fun.. | 15:53 |
| soothran | or is a port available fro moto e? | 15:54 |
| soothran | *for | 15:54 |
| soothran | how does ubuntu phone booting work? | 15:56 |
| soothran | the different steps, i mean | 15:56 |
| tvoss | ogra_, do you happen to have the porting guide link handy for soothran? | 15:58 |
| ogra_ | no, but the channel topic does ;) | 15:58 |
| * genii considers https://developer.ubuntu.com/en/start/ubuntu-for-devices/porting-new-device/ | 15:59 | |
| soothran | thanks | 16:01 |
| soothran | but i was more interested in my second questions - which is details about booting process in ubuntu phone | 16:02 |
| ogra_ | what do you want to know ? | 16:03 |
| kissiel | popey, hello, still online? | 16:05 |
| popey | kissiel: yeah, I'm in a hangout session right now | 16:06 |
| soothran | the whole process of booting if possible | 16:07 |
| soothran | like what happens in each step.. | 16:07 |
| soothran | etc | 16:07 |
| kissiel | popey, I got another checkbox-touch I want to push to the store :) https://launchpad.net/checkbox-touch/trunk/1.1.2/+download/com.ubuntu.checkbox_1.1.2_multi.click relnotes: http://paste.ubuntu.com/10997977/ | 16:08 |
| popey | ok | 16:09 |
| kissiel | popey, thx, no rush | 16:10 |
| popey | kissiel: it fails | 16:10 |
| kissiel | popey, fails what? | 16:10 |
| popey | http://paste.ubuntu.com/10997988/ | 16:10 |
| kissiel | popey, click-review? | 16:10 |
| popey | click-reviewer-tools | 16:10 |
| ogra_ | soothran, its not different from an ubuntu PC ... bootloader loads kernel, kernel loads initrd ... the slight differnce to a PC is the mounting of the rootfs that happens from the initrd to gain a few writable files on top of the readonly rootfs | 16:11 |
| popey | kissiel: https://bugs.launchpad.net/click-reviewers-tools/+bug/1395204 | 16:11 |
| ubot5 | Ubuntu bug 1395204 in Canonical Click Reviewers tools "Click review rejects multi-arch clicks" [High,Fix released] | 16:11 |
| kissiel | interesting | 16:11 |
| popey | dholbach: ^ | 16:12 |
| ogra_ | soothran, oh, and i forgot ... preparing the mounts for the android container that gets started later in the boot | 16:13 |
| dholbach | popey, what am I looking at? a multi click failing review with the most current review tools? | 16:14 |
| popey | yes | 16:14 |
| popey | the click is linked above | 16:14 |
| dholbach | popey, I don't think you're using the latest | 16:15 |
| popey | Now on revision 457. | 16:16 |
| popey | i am | 16:16 |
| popey | my script pulls latest from bzr every time I run it | 16:16 |
| * popey updates the chroot in which he does this | 16:17 | |
| popey | no updates | 16:17 |
| soothran | @ogra thanks! is there a documentation or something on it? | 16:17 |
| dholbach | popey, here's what I see: http://pastebin.ubuntu.com/10998030/ | 16:17 |
| ogra_ | soothran, linked from the porting guide i think | 16:18 |
| kissiel | popey, I've run click-review from r457 and it doesn't complain about that problem... | 16:19 |
| popey | hmmm | 16:19 |
| * popey re-runs | 16:19 | |
| popey | still fails | 16:19 |
| kissiel | um, weird :D | 16:20 |
| popey | on utopic | 16:20 |
| dholbach | i can create a utopic chroot and test it there if you want.... | 16:20 |
| === dandrader is now known as dandrader|lunch | ||
| popey | or i can upgrade mine to vivid | 16:20 |
| soothran | @orga how about the Ubuntu phone? guess it is same like in Ubuntu PC? | 16:20 |
| kissiel | popey, I ran the r457 on utopic as well | 16:21 |
| * popey upgrades | 16:21 | |
| popey | oh | 16:21 |
| popey | wtf | 16:21 |
| kissiel | popey, tho I might have a mess with imports and it pulls some stuff from the installed version | 16:21 |
| kissiel | gimme a sec | 16:22 |
| popey | ok, running it on my vivid laptop not in a chroot works | 16:22 |
| popey | ignore me | 16:22 |
| dholbach | ok, cool | 16:23 |
| dholbach | glad it works | 16:23 |
| popey | thanks guys | 16:24 |
| kissiel | popey, so we're good? | 16:25 |
| popey | kissiel: published | 16:25 |
| kissiel | popey, yuppiedo | 16:26 |
| popey | :) | 16:26 |
| popey | how did you make that multi-package kissiel ? | 16:26 |
| popey | manually or automagically? | 16:26 |
| kissiel | popey, $ click build :) | 16:27 |
| popey | you use qmake? | 16:27 |
| kissiel | popey, nope, just click build $path | 16:27 |
| kissiel | popey, we just have qml + python | 16:28 |
| popey | uh | 16:28 |
| ogra_ | soothran, hmm, i thought https://wiki.ubuntu.com/Touch/ContainerArchitecture was actually linked from the portin guide ... seems it is not | 16:28 |
| popey | do you ship with python binaries? | 16:28 |
| popey | hang on, easier if I just rip the click apart and look at it :) | 16:28 |
| kissiel | yep ... | 16:28 |
| soothran | so does ubuntu phone use Android LXC container? | 16:33 |
| soothran | @ ogra Android LXC container | 16:33 |
| dobey | yes, there's a container which contains some android components which must be run | 16:34 |
| ogra_ | soothran, yes ? | 16:34 |
| ogra_ | well, it uses lxc to run a container that contains we few bits of the android HAL ... the bits that are needed to make use of the binary drivers | 16:35 |
| soothran | @ogra so does ubuntu phone use Android LXC container? | 16:35 |
| ogra_ | it doesnt use android-lxc or some such | 16:35 |
| ogra_ | it is a normal ubuntu system that fires up a container to use the sensors, gps, camera and graphics drivers ... you could run the phone witrhout it but then you couldnt use any of these components ... (i.e. as a headless server or whatever) | 16:37 |
| dobey | well i guess wifi might not work without, so might be difficult to have a server with no network :) | 16:38 |
| dobey | depending on hardware of course | 16:38 |
| ogra_ | wifi works without the container up on all our devices currently | 16:38 |
| popey | usb to go ethernet adapter :) | 16:38 |
| awe | ogra_, not true... hybris is required for power control on MTK devices | 16:39 |
| ogra_ | but yeah, really depends if your WIFI firmware perhaps lives inside the container for your port | 16:39 |
| ogra_ | awe, i can ssh into my phone if the container isnt up | 16:39 |
| dobey | over usb, or over wifi? | 16:39 |
| ogra_ | well, i could once ... havent tried that in a long time | 16:39 |
| awe | sure, if you mean "works good enough for development" | 16:39 |
| ogra_ | dobey, wifi | 16:39 |
| awe | but if you want full power control ( ie. for airplane mode, battery savings, hybris may be required on some devices ) | 16:40 |
| ogra_ | iirc there was someone who ran a tomcat server on a headless N4 :) | 16:40 |
| ogra_ | with the whole container removed ... | 16:40 |
| awe | that sounds wrong | 16:40 |
| awe | ;D | 16:40 |
| ogra_ | haha, yeah | 16:40 |
| ogra_ | people are crazy like that | 16:40 |
| dobey | well, i guess one could maybe run on a freerunner or something with latest kernel and no container as full real ubuntu arm server maybe | 16:41 |
| ogra_ | yeah | 16:42 |
| soothran | hmm | 16:42 |
| dobey | android phones are a pain though, because it depends on manufacturers releasing the driver kit for the phone | 16:43 |
| ogra_ | soothran, so if you do a port you effectively need to port the driver framework of your phone that we then put into a container so ubuntu can make use of the drivers (once the container is started in the boot process) | 16:50 |
| pundir | cyphermox, hi, how tightly bound this mtp implementation https://launchpad.net/ubuntu/+source/mtp is to the Android USB gadget driver? | 16:54 |
| pundir | can it work with a generic MTP configfs driver as such? or is it too much to ask for :)? | 16:54 |
| Se7 | hellou all | 16:55 |
| soothran | @ogra thanks! | 16:57 |
| === alan_g is now known as alan_g|EOD | ||
| cyphermox | pundir: I think it should work as long as you have /dev/mtp pointing to some usb device that can do mtp (or if you edit the code to use another path), but best is to test it | 17:00 |
| pundir | cyphermox, so there is no hard dependency of ubuntu's mtp-server on Android's sys class, /sys/class/android_usb/android0, entries? | 17:03 |
| pundir | and what about Android usb uevents? does mtp-server use those uevents in any way? | 17:03 |
| ogra_ | there definitely is such a dependency in the upstart jobs | 17:04 |
| ogra_ | you would have to adjust them | 17:04 |
| pundir | and finally what about sys.config.usb property dependencies? | 17:04 |
| ogra_ | same thing | 17:04 |
| pundir | ogra_, ok | 17:04 |
| cyphermox | pundir: not that I remember, except for the upstart job, as ogra mentioned | 17:04 |
| cyphermox | pundir: that said, it expects some property stuff to start up, yes | 17:04 |
| ogra_ | not in the daemon itself though | 17:05 |
| cyphermox | I think it will possibly silently ignore things | 17:05 |
| ogra_ | you should be able to use the binary in non android context ... but will have to use your own upstart/init jobs | 17:05 |
| cyphermox | right | 17:05 |
| cyphermox | so, as long as property_get doesn't outright crash as it may have done in the past, you should be fine | 17:06 |
| ogra_ | oh, and i think the binary is tied into the greeter, no ? | 17:06 |
| cyphermox | I mean, as long as it doesn't crash on non-hybris, x86 or whatever and instead respects that there is a default value set | 17:06 |
| cyphermox | ogra_: yes | 17:06 |
| ogra_ | doing a dbus call | 17:06 |
| ogra_ | so you are kind of bound to unity8 | 17:07 |
| cyphermox | well, a lot of it can be easly ripped out | 17:07 |
| ogra_ | sure | 17:07 |
| cyphermox | (or better yet, made into optional features with a runtime switch) | 17:07 |
| cyphermox | pundir: all of it (the path to the device, dbus calls, property_get) are in http://bazaar.launchpad.net/~phablet-team/mtp/trunk/view/head:/server/server.cpp | 17:09 |
| cyphermox | so, yay, only one place to change things. :) | 17:09 |
| pundir | Great!! so i'll take it as something which may not be difficult to pull off then. | 17:10 |
| pundir | thanks for the link. | 17:10 |
| pundir | cyphermox, ogra_ so the background is that Android USB composite gadget is on its way out from 3.14 kernel onwards | 17:10 |
| pundir | and aosp already have mtp/ptp/audio_accessory.. drivers implemented as configfs gadget. | 17:10 |
| pundir | it works on android with some userspace hacking | 17:10 |
| cyphermox | pundir: since it's all in one file, you should probably be able to add some argument parsing in main() to toggle things, and pass what you want to d->run() | 17:10 |
| ogra_ | pundir, well, once we switch to a newer android base we will likely have to adapt then | 17:11 |
| pundir | I just want to make sure that it is generic enough to be able to go to upstream kernel | 17:11 |
| pundir | and can work on generic linux like ubuntu-touch (or mer I believe?) | 17:12 |
| ogra_ | upstream as in linus ... or as in android ? | 17:12 |
| pundir | ogra_, linus. it is already in aosp/android-3.14 and aosp/android-3.18 trees | 17:12 |
| ogra_ | well, we are currently bound to kitkat with all phones ... | 17:13 |
| ogra_ | once we switch to lollipop we will have to adjust ... but we dont have any device using a mainline linus kernel currently | 17:14 |
| pundir | cyphermox, ogra_ sorry for this noob question but which kernel is currently supported on ubuntu-touch platform? | 17:18 |
| pundir | And how much effort is it to get ubuntu-touch working on a newer device or dev board lets say? | 17:18 |
| ogra_ | pundir, whatever the android tree of a device provides | 17:19 |
| pundir | for this mtp-server experiment lets say | 17:19 |
| cyphermox | I think it currently depends on which device? | 17:19 |
| ogra_ | pundir, we are bound to the binary drivers from android ... some hardcode oaths or kernel interfaces yoou only find in that particular kernel code | 17:19 |
| pundir | cyphermox, ogra_ is it possible to boot ubuntu-touch with a minimal shell only profile? | 17:20 |
| ogra_ | so we usually use the AOSP vendor tree, rip out 95% (everythin but the HAL), add libhybris and are done ... | 17:20 |
| pundir | and still be able to do mtp stuff? | 17:20 |
| ogra_ | pundir, sure | 17:20 |
| pundir | great | 17:20 |
| pundir | thats what i'm looking for then. | 17:20 |
| ogra_ | as i raved about ~1h ago :) you can run ubuntu phones completely without the android container | 17:21 |
| ogra_ | as a headless system | 17:21 |
| pundir | yay!! | 17:21 |
| pundir | and if I'm running it without android container then it pretty much guarantees to run on any recent kernel version as well, right? | 17:23 |
| pundir | I dont need a stable system to do any UI or fancy stuff.. i'll just boot to shell and fire up mtp-daemon/service | 17:23 |
| ogra_ | sure ... even with the container ... | 17:23 |
| ogra_ | the container would likely just fail to start gracefully | 17:24 |
| ogra_ | lxc-init can do such stuff ;) | 17:24 |
| pundir | Great!! so how do I get started then? | 17:24 |
| ogra_ | grab a nexus4 ... i think thats the easiest device to tinker with ... build a kernel you get to boot ... then ... well ... do what you want :) | 17:25 |
| pundir | is it possible to run it on a development board just like a regular linux distro? | 17:29 |
| pundir | i dont have a Nexus4.. | 17:29 |
| pundir | I have a Galaxy Nexus, Nexus9/7 if it helps? | 17:29 |
| ogra_ | the nexus7 has an image (2013 version only) that you could use to start from | 17:31 |
| Z3 | Hi, is there a release date for the "docking mode" on Ubuntu Touch? | 17:31 |
| ogra_ | Z3, end of the year there should be a first device with the first iteration of that feature | 17:32 |
| Z3 | ogra_ and for Nexus 4? | 17:33 |
| ogra_ | (as announced on monday by marks opening talk of the ubuntu online summit that is currently going on) | 17:33 |
| ogra_ | not sure anyone will build such an image for the N4, but once the phone is released you can surely hack up the image to work on the N$ (and send patches etc) | 17:34 |
| ogra_ | i doubt anyone from canonical will actually work on this on the N4 but community patches will indeed be accepted | 17:34 |
| pundir | ogra_, thanks i'l try it on Nexus 7 | 17:35 |
| Z3 | thank you very much. I have another question: I don't like Unity, so in Ubuntu desktop I can install the ubuntu classic desktop and use and "old" desktop. Will something like this be possible on Android? A "classic" desktop like Android | 17:35 |
| Z3 | sorry, I mean: | 17:36 |
| Z3 | will something like this be possible on Ubuntu Touch? | 17:36 |
| Z3 | modify desktops and install whatever you like, just like you do in regular pcs | 17:36 |
| ogra_ | the phone uses Mir ... | 17:37 |
| ogra_ | so you would need a desktop env that can make use of it | 17:37 |
| Z3 | I understand, it would be nice a "classic" alternative for people that don't like Unity | 17:38 |
| ogra_ | i could imagine once the lxde Qt implementation is done you could perhaps run it instead of unity8 | 17:39 |
| ogra_ | notr sure though | 17:39 |
| dobey | the converged phone won't run a different environment on the external display. it will be the same interface. i doubt you'd want to use a "classic" alternative as the main interface when trying to use your phone as a phone | 17:39 |
| === dandrader|lunch is now known as dandrader | ||
| Z3 | dobey I think in a "classic" desktop like Android. Just a desktop and icons. I just want that | 17:40 |
| dobey | if lxde qt had a decent interface in phone mode and worked with the system to be a client for scopes and such, it could maybe be useful | 17:40 |
| ogra_ | dobey, oh, i didnt notice anything about "useful" in the question :) | 17:41 |
| dobey | i guess you could write a shell that was like android and build your own images if you wanted | 17:41 |
| dobey | ogra_: well, i presume if one wants a phone to be usable as a pc as well, they'd want it to be somewhat useful in both modes :) | 17:41 |
| dobey | otherwise you can just install an armhf build of the ubuntu ISO today if you really wanted to :) | 17:41 |
| ogra_ | right | 17:42 |
| Z3 | I can't do that alone ... I was thinking in the equivalent of desktop world, where you have gnome, kde, lxde and others ... and you can choose whatever you like more | 17:43 |
| Z3 | I don't see we can't have that on mobiles too | 17:43 |
| Z3 | thank you all for the information | 17:45 |
| Z3 | :) | 17:45 |
| === aaron__ is now known as ahoneybun | ||
| === ljp is now known as lpotter | ||
| === om26er_ is now known as om26er | ||
| morphis | rsalveti: what is the best way to rebuild the kernel for the n4 with ubuntu touch? is this done in the aosp tree or is the kernel part of the archive? | 18:23 |
| ogra_ | morphis, in the archive | 18:23 |
| ogra_ | search for linux mako | 18:23 |
| ogra_ | just apt-get source it (there is a git tree at kernel.ubuntu.com too somewhere) | 18:24 |
| morphis | ogra_: and then just cross build the package? | 18:24 |
| rsalveti | yeah, easier is getting the sources and cross building it | 18:24 |
| rsalveti | the android package pulls the kernel binaries (from the package), and then makes that available as part of the image | 18:25 |
| morphis | rsalveti: ah | 18:25 |
| morphis | rsalveti, ogra_: so https://code.launchpad.net/~ubuntu-branches/ubuntu/vivid/linux-mako/vivid should be the latest, right? | 18:26 |
| rsalveti | yeah, that's the auto-import branch | 18:26 |
| rsalveti | the real source code for it is in a git repo | 18:27 |
| rsalveti | https://launchpad.net/ubuntu/+source/linux-mako | 18:27 |
| rsalveti | the src package | 18:27 |
| rsalveti | hm, where is the gitweb for it | 18:27 |
| rsalveti | http://kernel.ubuntu.com/git/ | 18:27 |
| rsalveti | changed but not showing the vivid kernel | 18:27 |
| rsalveti | actually, it is | 18:28 |
| rsalveti | http://kernel.ubuntu.com/git/ubuntu/ubuntu-vivid.git/log/?h=mako | 18:28 |
| rsalveti | there you go | 18:28 |
| morphis | ah good | 18:28 |
| morphis | and that will generate me a fastboot'able image? | 18:32 |
| ogra_ | no, that will generate a zImage | 18:34 |
| ogra_ | (or vmlinuz or how you want to call it) | 18:35 |
| ogra_ | you can grab the boot.img from the device and use abootimg from the archive to replace the kernel in there | 18:35 |
| morphis | ogra_: ok, that was my alternate | 18:38 |
| morphis | ogra_: but how is the fastboot image then generated for the deployment? | 18:38 |
| ogra_ | to actually prodcue the boot.img you would have to build the android tree | 18:38 |
| morphis | is that a special job somewhere? | 18:38 |
| morphis | ah right | 18:39 |
| ogra_ | apt-get source android | 18:39 |
| ogra_ | ;) | 18:39 |
| morphis | so you still take the the android initrd and put that into the kernel? | 18:39 |
| ogra_ | no, we use an ubuntu initrd for booting | 18:39 |
| morphis | but bundling that in the android-build ... I see | 18:39 |
| ogra_ | there is an android initrd that we unpack on the fs to use for the lxc container | 18:39 |
| morphis | yeah I know | 18:39 |
| ogra_ | (since in android the initrd is actually the rootfs) | 18:40 |
| morphis | the one which has the init.rc files | 18:40 |
| ogra_ | right, it is shipped in the device tarball and unpacked to /var/lib/lxc/android/rootfs/ | 18:40 |
| morphis | ok | 18:41 |
| dobey | anyone else seen issues with SMS where messages would be delivered way late, and out of order? | 18:45 |
| ChloeWolfieGirl | Question, I heard click packages where moving to snappy, what does this mean and why are they changing? | 18:46 |
| ogra_ | ChloeWolfieGirl, thats still far out and it will hapen transparently in the store i was told | 18:48 |
| dobey | snappy is the new format that will replace the click format. so at some point they will change. but should be relatively transparent to users/developers when it happens | 18:48 |
| ChloeWolfieGirl | Whats different between click and snappy? | 18:48 |
| dobey | ogra_: well, part of it already happened, with the change in namespaces which happened for clicks too | 18:48 |
| ogra_ | right | 18:48 |
| ogra_ | ChloeWolfieGirl, they are integrated with snappy :) | 18:49 |
| ogra_ | click packages are not | 18:49 |
| ogra_ | ChloeWolfieGirl, and since all of ubuntu will eventually move to snappy that means you can use snappy server or desktop packages on your phone | 18:49 |
| ogra_ | (without making it writable or having to use debs) | 18:50 |
| ChloeWolfieGirl | What does that mean :P I feel like I've missed out because I thought snappy was just for like servers and for small smart devices not including desktops and phones xD | 18:50 |
| ogra_ | snappy is the new ubuntu framework, replacing the old ways ... | 18:51 |
| dobey | the problem there is that "snappy" is being used to refer to multiple things | 18:51 |
| dobey | sort of like "chips" in australia. you have to use context to know which thing you're talking about :) | 18:51 |
| ogra_ | snap packages are kind of an evolution of the click packages ... with more abilities, more fine grained control etc | 18:51 |
| ogra_ | snappy images are kind of an evolution of the phone OS installation ... readonly rootfs with some writable bits ... buut also rollback functionality etc | 18:52 |
| ChloeWolfieGirl | It just confused me cos I thought clicks where new and the future xD | 18:52 |
| ogra_ | so the phone simply will move to the next point in evolution | 18:52 |
| ogra_ | they were ... at some point :) | 18:52 |
| ChloeWolfieGirl | Will snappy images affect devices ubuntu's ported to? | 18:53 |
| ChloeWolfieGirl | Like will there be port issues? | 18:53 |
| ogra_ | porting will become easier | 18:53 |
| ogra_ | the device bits of a snappy install live as snap packages in the store today | 18:54 |
| ChloeWolfieGirl | Thats awesome :D But would it negativly affect previously ported devices? | 18:54 |
| ogra_ | so making a port (for snappy currently, phones might be a bit different) means you roll a kernel and device bits into a snap package | 18:55 |
| ogra_ | and just upload it to the store | 18:55 |
| rsalveti | morphis: you could as well fetch the repos from https://code-review.phablet.ubuntu.com/#/q/status:open,n,z | 18:56 |
| rsalveti | which is basically how we create the android src package | 18:56 |
| rsalveti | as part of the android build we produce a boot.img | 18:56 |
| ChloeWolfieGirl | ogra_, that sounds pretty awesom :P | 18:56 |
| rsalveti | that cdimage then makes it public at http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/pending/ | 18:57 |
| ogra_ | ChloeWolfieGirl, so it is ... but thats still a bit away | 18:57 |
| ogra_ | and it is the reason the phone wont switch to wily | 18:57 |
| ogra_ | wily will become the playground to create a snappy based phone | 18:57 |
| ogra_ | while all UI development and bug fixes for the actual phone will happen on vivid | 18:58 |
| ChloeWolfieGirl | ogra_, Aww, ok but making porting easier and better, I'm excited :P | 18:59 |
| ogra_ | well, it is also a requirement for the desktop mode and convergence | 19:00 |
| ogra_ | not only porting | 19:00 |
| dobey | i'm both tempted, and afraid, to get snappy running on my wifi routers | 19:00 |
| ogra_ | it will be awesome | 19:00 |
| ogra_ | you can just install my fhem snap and make it control your heating alongside ;) | 19:01 |
| ogra_ | (or once we have it ... someones "nest" snap) | 19:01 |
| ogra_ | (or the VOIP server snap to turn it into a phone controller) | 19:02 |
| lotuspsychje | https://erlerobotics.com/blog/erle-robotics-launches-ubuntu-snappy-core-drone/ | 19:04 |
| dobey | i think i might have a vonage voip box laying around somewhere | 19:04 |
| dobey | unless i dumped it off to recycling | 19:04 |
| ChloeWolfieGirl | This is gonna be interesting | 19:12 |
| SturmFlut | Why not install the VOIP server snap on Erle drones and build our own mobile network | 19:14 |
| Tassadar | because you probably want your network to last more than 15 minutes | 19:16 |
| dobey | Tassadar: depends on what you're using it for ;) | 19:24 |
| Tassadar | if you land it on a building, that could be okay) | 19:25 |
| dobey | building will probably interfere with radio though | 19:28 |
| dobey | but if you want an encrypted mesh wireless comms network that's only up when you need it, then drones that pop up for 10 minutes would probably be a reasonable thing | 19:29 |
| Tassadar | I admit I have little experience with building imaginary drone-based networks) | 19:29 |
| === boiko_ is now known as boiko | ||
| dobey | mandel: wtf do you have 3 0-line "rebuild" branches for the click scope? :P | 19:50 |
| ogra_ | he is on vacation this week ... | 19:50 |
| dobey | ah | 20:01 |
| taiebot | Thank you for the web-browser app not turning off the screen on vivid devel proposed \o/ when watching videos. Currently watching uos videos without having to tap every 30sec on the screen | 20:01 |
| === salem_ is now known as _salem | ||
| mariogrip | ChloeWolfieGirl: Hi! | 21:59 |
| ChloeWolfieGirl | mariogrip, Heya | 22:03 |
| mariogrip | ChloeWolfieGirl: are you still running ubuntu on you opo? | 22:03 |
| ChloeWolfieGirl | mariogrip, yeah haha xD | 22:05 |
| mariogrip | ChloeWolfieGirl: awesome! how is battery life in ub vs android? | 22:06 |
| mariogrip | ChloeWolfieGirl: hehe, i haven't run ubuntu for more then a hour without rebooting and re flashing :P | 22:07 |
| mariogrip | that's why im asking | 22:07 |
| ChloeWolfieGirl | mariogrip, Not to well tbh >.< It also sometimes gets quite warm, and if you turn it off by the pop-up it reboots after a few minuets and if you don't realise your battery has drained quite a bit | 22:08 |
| mariogrip | ChloeWolfieGirl: okay, Thanks! :D well i have to take a look at that. i did't think it was that bad... :P | 22:11 |
| ChloeWolfieGirl | Well I haven't used it a great amount outside of reddit since you cant install apps from the store yet, but its at 89% right now, so I'll record this time and I'll record how long it takes to go down by 10% screen off (only on for checking) so you have a better idea | 22:13 |
| Perracomax | 7001Hi | 23:25 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!