| === chriadam|away is now known as chriadam | ||
| === Ursinha is now known as Ursinha-afk | ||
| === Ursinha-afk is now known as Ursinha | ||
| === chihchun_afk is now known as chihchun | ||
| === chihchun is now known as chihchun_afk | ||
| === chihchun_afk is now known as chihchun | ||
| north | Anyone who is working on ofono here ? | 06:45 |
|---|---|---|
| north | firstly I would want to know what RIL Modem is ? Is it similar to the reference-ril provided in the AOSP tree ? second how is Ubuntu's Ofono port able to talk to a QMI/MBIM enabled modems using the proprietary protocols ? | 06:52 |
| dholbach | good morning | 06:57 |
| piiramar | north: oFono in Ubuntu Touch uses a plugin called 'rilmodem', which talks to the original Android device vendor's RIL daemon. oFono does not use the QMI interface directly, but the RIL daemon does. sources at https://github.com/rilmodem/ofono | 07:01 |
| piiramar | and some info in a recent presentation by rsalveti http://events.linuxfoundation.org/sites/events/files/slides/Ubuntu%20Touch%20Internals_1.pdf | 07:02 |
| nhaines | dholbach: good morning! | 07:17 |
| JoeyChan | Hi ogra_ my screenshot script seems deprecated, pls update me a new one :) | 07:18 |
| JoeyChan | this is the current script I am using: | 07:19 |
| JoeyChan | adb shell mirscreencast -m /tmp/mir_socket -n1 | 07:19 |
| JoeyChan | adb pull /tmp/mir_screencast_768x1280.rgba /tmp/screenshot.rgba | 07:19 |
| JoeyChan | convert -size 768x1280 -alpha off -depth 8 /tmp/screenshot.rgba ./screenshot.png | 07:19 |
| nhaines | The problem with phablet-screenshot is that mir_screencast now appends "_60Hz" to the end of the filename before the extension. | 07:19 |
| nhaines | So update your adb pull line accordingly. :) | 07:19 |
| JoeyChan | adb pull /tmp/mir_screencast_768x1280_60Hz.rgba /tmp/screenshot.rgba | 07:20 |
| JoeyChan | ? | 07:20 |
| nhaines | Should work. | 07:20 |
| JoeyChan | ok trying :) | 07:21 |
| JoeyChan | lol, that works, thank you nhaines | 07:22 |
| nhaines | \o/ | 07:22 |
| dholbach | hey nhaines | 07:27 |
| === chriadam is now known as chriadam|away | ||
| north | piiramar: Also, looking at Ofono, it uses DBus, so does Ubuntu-touch uses DBus for IPC ? | 07:46 |
| north | Also is Ubuntu-touch relying on Android's kernel ? | 07:51 |
| === fmasi_afk is now known as fmasi | ||
| JamesTait | Good morning all; happy Chocolate Chip Day! :-D | 07:55 |
| bact | chocolate chip day? | 07:57 |
| bact | theres only 365 days in a year and they wanted to waste one on a ice cream flavour? | 07:57 |
| janimo | ogra_, is mountall: event failed logged by upstart something to worry about? I've seen it come up on askubuntu for desktop installs but with no clear conclusion | 09:18 |
| ogra_ | janimo, mountall is mostly a no-op for us (well, it processes fstab to crate the bind mounts on top of the already mounted fileystem) | 09:21 |
| ogra_ | our actual mounting process happens in initrd ... you should check your bind mounts though | 09:21 |
| ogra_ | (and compare it to fstab i guess) | 09:22 |
| janimo | ogra_, that error message is very non-descriptive though. | 09:24 |
| north | So the Linux box shipped along with Ubuntu-Touch provides DBus mechanism for IPC and Udev for external hardware connectivity ? | 09:28 |
| ogra_ | janimo, well, complain to mountall upstream :) | 09:32 |
| ogra_ | (i'm not sure what will happen to mountall when we switch to systemd anyway) | 09:32 |
| janimo | ogra_, or wait for systemd? Is that replacing mountall as well? | 09:32 |
| janimo | oh yes | 09:32 |
| ogra_ | pitti might know | 09:33 |
| ogra_ | he maintains most of the transition | 09:33 |
| janimo | AIUI mountall was a stopgap until upstart grew some more features | 09:33 |
| piiramar | north: yes D-Bus is heavily used, but that's not specific for Ubuntu Touch (desktop ubuntu uses it, too). | 09:58 |
| ogra_ | north, what is "the linux box" you refer to ? | 09:59 |
| === dandrader is now known as dandrader|afk | ||
| === fmasi is now known as fmasi_lunch | ||
| ogra_ | ev, having system-settings create /var/lib/apport/autoreport would not help, we need it created before "writable-paths" gets processed, else it will not be writable ... is there any reason to not discuss this on the related bug btw ... i assume the community cant read your asana board | 10:08 |
| ogra_ | ev, this is bug 1235436 btw | 10:09 |
| ubot5 | bug 1235436 in apport (Ubuntu) "/etc/init/apport-noui.conf is non-functional on the phone" [Critical,Fix released] https://launchpad.net/bugs/1235436 | 10:09 |
| ev | no reason not to discuss it there :) | 10:10 |
| ogra_ | :) | 10:11 |
| ogra_ | ev, oh, wait ... brian actually made the whole dir writable (something we usually do not do ... ) that should actually work ... but system-settings would need to grow handling of this ... i guess we need to involve Laney or seb128 into the system-settings side | 10:13 |
| seb128 | ogra_, handling of what? ev wrote that panel and it's mostly calling whoopsie interface afaik | 10:15 |
| ogra_ | seb128, for making it auto-upload /var/lib/apport/autoreport needs to exist ... | 10:16 |
| ogra_ | as i understand it is the touch equivalent of /etc/default/apport ... but in that location so it doesnt become a conffile | 10:17 |
| seb128 | shouldn't the whoopsie service create it when the settings are changed? | 10:17 |
| ogra_ | hmm, ask evan :) | 10:17 |
| Saviq | ogra_, "ass for having it merged" good one ;) | 10:17 |
| ogra_ | lol | 10:17 |
| north | orga_ I mean the Ubuntu-Touch Linux kernel. | 10:19 |
| Laney | what's wrong with SetAutomaticallyReportCrashes on WhoopsiePreferences? | 10:20 |
| Laney | the panel is supposed to set this already | 10:20 |
| north | I ask piiramar because Android kernel doesn't ship with DBus, and The Touch docs quote that Ubuntu-Kernel is based on Android Kernel, which is quiet confusing | 10:20 |
| ogra_ | north, dbus is a userspace tool | 10:21 |
| === dandrader|afk is now known as dandrader | ||
| north | But orga_ android uses Binder in place of DBus | 10:21 |
| ogra_ | north, we do not have any dbus related bits in the kernel ... in fact we use the android source (we have to, to make the binary drivers work) but with our own config and a bunch of additional patches | 10:21 |
| Laney | I think it's a non system settings issue | 10:23 |
| ogra_ | north, we have a minimal android running inside an lxc container ... which we start during the ubuntu boot process ... that container still uses binder where needed so the binary drivers can talk to their daemons ... on the ubuntu side the setup is identical to any other ubuntu install you know | 10:23 |
| Laney | we already make the right call | 10:23 |
| Laney | the state is lost after you restart the phone | 10:23 |
| ogra_ | Laney, well, then this finction needs to grow handling for the apport default setting too | 10:24 |
| Laney | what does that mean | 10:24 |
| ogra_ | *function | 10:24 |
| north | So Ubuntu-Touch essentially has Binder and DBus both running. Also, does the Ubuntu-Touch have udev support ? ogra_ | 10:24 |
| Laney | it is a d-bus property | 10:24 |
| ogra_ | Laney, see the bug :) ... apptort wont upload anything without that file exosting (i belive a simple touch would be enough) | 10:25 |
| Laney | if it's wrong then it is wrong in the service | 10:25 |
| Laney | no | 10:25 |
| ogra_ | *apport | 10:25 |
| Laney | clients have no business touching that file | 10:25 |
| ogra_ | north, yes, it does | 10:25 |
| Laney | and how would they touch a file in /var anyway | 10:25 |
| ogra_ | Laney, i didnt say clients :) i said the function that enables whoopsie needs to enable it too | 10:25 |
| ogra_ | else having whoopsie running is pointless | 10:26 |
| north | Then, in that case, we could use Modem Manager which can be optimized for different devices. (I am assuming Ubuntu supports Glib too, correct me if am wrong) ogra_ | 10:26 |
| ogra_ | i think we use modemmanager already | 10:26 |
| ogra_ | ah, no, we dont | 10:27 |
| north | no mm is not used | 10:27 |
| north | Network manager is used | 10:27 |
| ogra_ | but ofono seems to use libmm-glib0 | 10:27 |
| north | Modem Manager speaks all the protocols. The current design could cut the RIL dependency inside the LXC container | 10:28 |
| ogra_ | but we need ril | 10:28 |
| ogra_ | you wont manage all features of the modems without it | 10:28 |
| north | Modem Manager is the RIL ogra_ | 10:28 |
| ogra_ | and how does that handle the proprietary settings only the binary rild from the vendor provides ? | 10:29 |
| north | What RIL is doing on android based phones/devices is done using modem manager on the desktop | 10:30 |
| ogra_ | <- not a telephony guy ... you better wait for awe to wake up in a few hours to discuss such stuff ... | 10:30 |
| north | mm doesn't depend on the rild | 10:30 |
| north | it itself provides the functionality of rild | 10:30 |
| ogra_ | north, well, the rild's we use comes from qualcomm ... and only this binary daemon is able to use all HW features the modem has | 10:30 |
| north | I agree. I got no doubt in that, libqmi can do whatever the qualcomm ril can do. | 10:31 |
| ogra_ | the HW manufacturer uses proprietary protocols internally ... modemmmanger would not know about these features | 10:31 |
| ogra_ | and we dont have the manpower to reverse engineer each and every binary rild out there | 10:32 |
| north | yes yes, exactly, I always feel Android also can cut the usage of rild and start using mm somehow (if it can support udev) | 10:32 |
| north | You don't have to | 10:32 |
| north | mm already supports so many modems | 10:32 |
| north | out there | 10:32 |
| ogra_ | sure, but only the open bits | 10:33 |
| north | No. | 10:33 |
| Laney | I think that it's just a bug in whoopsie-preferences | 10:33 |
| ogra_ | Laney, fine with me :) | 10:33 |
| north | it follows the same mechanism, like rild but can talk the gpled libqmi and libmbim | 10:33 |
| ogra_ | Laney, i dont really care *where* it gets fixed ... just saying something needs to touch that file ... ev suggested that could be system-settings | 10:34 |
| north | sorry can the talk the proprietary protocols using the gpled libqmi and libmbim | 10:34 |
| north | also, does ubuntu-touch support glib ? | 10:34 |
| Laney | it is already supposed to | 10:34 |
| Laney | whoopsie-preferences | 10:34 |
| Laney | but really it could never be anything other than the service itself which owns this | 10:35 |
| ogra_ | north, i think awe tried modemmanager and it didnt work for the two nexus phones that we supported back then because their rild implementation differen so massively already | 10:35 |
| ogra_ | *differed | 10:35 |
| ogra_ | north, some bits use glib on the low leve, yes | 10:35 |
| ogra_ | *level | 10:36 |
| Stskeeps | not to mention the whole mess of audio routing and how that ties into the stack.. | 10:36 |
| north | so glib is not fully functional ? | 10:36 |
| ogra_ | oh yeah | 10:36 |
| Stskeeps | s/audio/call audio/g | 10:36 |
| north | O_o | 10:36 |
| north | where can I get the ubuntu-touch kernel ? | 10:37 |
| ogra_ | glib is fully functional but we only use it underneath ... we have no glib support at the UI level everything goes through Qt | 10:37 |
| ogra_ | from kernel.ubuntu.com or from the archive via apt-get | 10:37 |
| ogra_ | ask in #ubuntu-kernel for more details :) | 10:38 |
| north | sure ogra_ | 10:38 |
| ogra_ | i only know that mm was ruled out by the guys that know more about telephony than me ... you need to talk to them ... i remember it wouldnt have been as easily portable to all the other android devices and i remember there were missing features you would have had to do a lot of reverse engineering for | 10:40 |
| ogra_ | which made us pick the binary rild's instead | 10:40 |
| north | Yes, cheap and dirty workaround and completely not the open source way except for ofono | 10:41 |
| north | :p | 10:41 |
| north | and neat too | 10:41 |
| ogra_ | well, surely not cheap and dirty | 10:41 |
| Stskeeps | you can choose to ship a modern smartphone product that sells and gets more money for you to be sustainable, or you can choose to do the right thing and not be sustainable.. | 10:41 |
| ogra_ | (that was enough hard work to not call it cheap and dirty ... and actually i think thats one of the most collaborative pieces where jolla, firefoxOS and ubuntu work very closely together) | 10:43 |
| Stskeeps | well, not firefox os but yes, it makes life a lot easier that we can rely on a factually api-known ril implementation | 10:43 |
| Stskeeps | which is massively less investment than implementing a new modem stack in practice | 10:43 |
| ogra_ | oh ? i thought FFOS uses it the same way | 10:44 |
| ogra_ | just natively (since they are mostly android under the hood) | 10:44 |
| Stskeeps | yeah, true | 10:44 |
| north | oh, FFOS doesn't use ofono | 10:44 |
| ogra_ | it surely does | 10:45 |
| Stskeeps | it doesn't, they speak ril directly | 10:45 |
| ogra_ | we just merged some code from them afaik | 10:45 |
| Stskeeps | hmm | 10:45 |
| Laney | lolz | 10:45 |
| Laney | ok I found the error | 10:45 |
| Stskeeps | news to me, i'll go check my sources ;) | 10:45 |
| ogra_ | or are reviewing something to merge ... i think i heard tony tallk abot some MMS bits he reviews | 10:45 |
| ogra_ | probably he just seeks inspiration though and didnt men to merge actually :) | 10:46 |
| ogra_ | *mean | 10:46 |
| north | I guess FFOS talks without ofono | 10:47 |
| north | I am pretty sure about it actually | 10:47 |
| north | Ubuntu-touch is the only one AFAIK uses ofono port | 10:47 |
| ogra_ | i highly doubt that | 10:48 |
| Stskeeps | (sailfishos, openwebos uses ofono rilmodem, at least..) | 10:48 |
| Stskeeps | maybe more out in field we don't know about | 10:48 |
| ogra_ | right | 10:48 |
| ogra_ | and i'm pretty sure FFOS too | 10:48 |
| ogra_ | but afaik in their completely own fork | 10:48 |
| ogra_ | (though as i said, FFOS is mostly android under the hood anyway ... ) | 10:49 |
| Laney | ev: would you consider moving lp:whoopsie-preferences to a team branch? :) | 10:49 |
| north | awe is a person ? I thought awe is an expression :p silly me | 10:49 |
| ogra_ | north, awe and abeato maintain our ofono stack | 10:50 |
| ogra_ | mainly at least | 10:50 |
| Laney | shock and awe | 10:50 |
| ogra_ | haha, yeah | 10:50 |
| === MacSlow is now known as MacSlow|lunch | ||
| ev | Laney: I would :-P | 10:54 |
| * ogra_ waits for the price negotiation to start | 10:54 | |
| ogra_ | :) | 10:54 |
| ev | Laney: done: https://code.launchpad.net/~daisy-pluckers/whoopsie-preferences/trunk | 10:54 |
| Laney | one patch complete with upload and that's my final offer! | 10:54 |
| ev | ha! damn | 10:54 |
| ev | missed opportunity | 10:54 |
| Laney | :P | 10:55 |
| ogra_ | heh | 10:55 |
| Laney | maybe you'll take it to add me to that team? | 10:55 |
| north | So in a sense Ubuntu is nice because it supports glib, has udev and DBus (probably can switch to kdbus soon) when compared to android | 10:55 |
| north | :p | 10:55 |
| Laney | ev: ^ team joining request | 10:56 |
| ogra_ | unlikely that you will see us going to kdbus | 10:56 |
| ogra_ | unless android does | 10:56 |
| piiramar | I was under the impression that these two components https://mxr.mozilla.org/mozilla-central/source/dom/system/gonk/ril_worker.js , https://github.com/mozilla-b2g/rilproxy sortof are Firefox's equivalents to oFono and the rilmodem driver, respectively | 10:56 |
| piiramar | stskeeps: ^^ | 10:56 |
| ev | Laney: done :) | 10:56 |
| Laney | merci | 10:56 |
| ogra_ | north, we are ... and will have to go on ... using the android HAL layer to get full access to the hardware ... so we are kind of bound to use the android kernel source too and not modify it in a way that would break the HAL | 10:57 |
| ogra_ | which binds us to certain kernel versions after all | 10:58 |
| north | Yea, Ubuntu can be more of a developer friendly OS, I mean low level. | 10:59 |
| ogra_ | well, ubuntu is ubuntu ... as i said ... the android bits all live inside an lxc container ... outside of that container there is a "normal" ubuntu | 10:59 |
| ogra_ | but to run that container we are bound to use the richt kernel source ... kdbus would be a pretty intrusive thing to backport ... | 11:00 |
| ogra_ | s</richt/right/ | 11:00 |
| ogra_ | i'm not even sure kdbus will be used on the desktop/server ... even though we will switch to systemd i think our security team has objections to the current kdbus design | 11:02 |
| ogra_ | so at least you wont see it used unmodified (unless the design changes before being implemented) | 11:02 |
| ogra_ | i also wouldnt expect kdbus to be in a usable state for being a default thing within the next two years ... | 11:03 |
| north | I see | 11:03 |
| north | and is systemd an option for future Ubuntu-Touch ? | 11:04 |
| ogra_ | no | 11:04 |
| ogra_ | its a requirement ... not optional | 11:04 |
| ogra_ | touch will follow suit whatever desktop does | 11:05 |
| north | But it will break the current android reliance of Ubuntu | 11:05 |
| ogra_ | and desktop already works on the switch ... at least for init | 11:05 |
| north | because android still uses the oooooooooold init design | 11:05 |
| ogra_ | yeah, no issue with that | 11:05 |
| north | if you know what I mean | 11:06 |
| ogra_ | well, not the old one ... it uses some completely self designed | 11:06 |
| ogra_ | but since android runs inside a container that doesnt matter at all for us | 11:06 |
| ogra_ | it can use what it likes ... as long as that can run under lxc | 11:06 |
| north | haha in a way, ubuntu depends and isolates android :'D | 11:07 |
| ogra_ | right | 11:07 |
| ogra_ | well, ubuntu depends on libhybris with many HW related bits ... hybris talks to the container | 11:07 |
| ogra_ | if some vendor would release a phone with system "foople" we could just port hybris to tallk to that and use "foople" inside the container ... | 11:08 |
| ogra_ | its like a black box | 11:08 |
| ogra_ | you can replace the content as long as your communication layer keeps working | 11:08 |
| ogra_ | (theoretically you could use the windows HAL and run on win8 phones ... if you had tghe source and could make hybris talk to it) | 11:10 |
| ogra_ | (or an iphone) ... | 11:10 |
| north | I see some light... looking at your comments | 11:11 |
| ogra_ | indeed thats all hypothetical ... nobody will ever give us the win8 or IOS source :) | 11:13 |
| === fmasi_lunch is now known as fmasi | ||
| ogra_ | just trying to line out our design here ... the container is a black box that can contain anything you like as long as hybris can talk to it | 11:14 |
| ogra_ | which gives us the opptotunity to keep a totally normal ubuntu outside of the container | 11:14 |
| north | and btw, is ubuntu-touch compiled against glibc ? | 11:15 |
| ogra_ | sure | 11:15 |
| north | which means it is pure linux | 11:16 |
| ogra_ | it uses all the normal armhf packages from the ubuntu archive ... the same ones you find on armhf server installs | 11:16 |
| north | just just some workarounds for a mobile platform | 11:16 |
| ogra_ | (for example) | 11:16 |
| ogra_ | right | 11:16 |
| north | wow | 11:17 |
| ogra_ | well, ubuntu-touch has a specific design ... like the system-image install ... like app convergence (which puts way stricter and more secure contraints on apps than IOS or android have) etc | 11:17 |
| ogra_ | sigh | 11:17 |
| ogra_ | not convergence | 11:17 |
| ogra_ | confinement | 11:17 |
| ogra_ | too many similar terms nowadays :P | 11:17 |
| north | hehe | 11:17 |
| ogra_ | there we differ a lot from i.e. a normal desktop install ... we dont use debs except for building the readonly system-image for example | 11:18 |
| wellsb | How can I sign into Ubuntu One using account-console from the terminal? I keep getting network error when I try from Online Accounts in System Settings, so I can't perform updates or see apps in click store | 11:40 |
| === alan_g is now known as alan_g|lunch | ||
| === dednick is now known as dednick|lunch | ||
| ogra_ | wellsb, probably dobey knows ... i think he is involved with that | 11:42 |
| ogra_ | (not sure that is possible at all) | 11:42 |
| dobey | it's not. account-console doesn't let you sign into anything. it's a debugging tool | 11:43 |
| dobey | (and i'm not really here at the moment) | 11:44 |
| ogra_ | wellsb, what image number are you on ? we had a few broken ones recently ... you probably need to use ubuntu-device-flash to get to one without network issues | 11:45 |
| sergiusens | ogra_: did you do the seed MP I gave you? | 11:52 |
| ogra_ | sergiusens, yep, in image 29 | 11:56 |
| === MacSlow|lunch is now known as MacSlow | ||
| === psivaa is now known as psivaa-afk | ||
| pepoo | hi ? | 12:06 |
| wellsb | ogra_, I'm on r27, but I've been having this issue in various versions for about the last month | 12:16 |
| ogra_ | sounds weird, what device is that ? | 12:17 |
| wellsb | n4 | 12:17 |
| wellsb | dobey, can you tell me what fields should be populated (and their values) when doing account-console show <id> | 12:17 |
| ogra_ | 26 and 27 had the flight mode issues | 12:17 |
| ogra_ | 28 has them fixed | 12:17 |
| ogra_ | so make sure to run at least 28 ... | 12:18 |
| ogra_ | 29 and 30 had other issues ... dont bother with them | 12:18 |
| ogra_ | (or simply just run the promoted ones instead of the untested ones ...) | 12:18 |
| wellsb | Thanks ogra_ . I'll try another upgrade. I hesitate because it's so painful to have to download 300+mb every time using ubuntu-device-flash | 12:19 |
| ogra_ | we usually dont promote the broken ones ... and images from -proposed can break at any time | 12:19 |
| ogra_ | so use the ubuntu-touch/devel channel this time | 12:19 |
| ogra_ | that will make sure you get fully tested images | 12:19 |
| sergiusens | ogra_: heh, polling you was easier than polling the system :-P | 12:20 |
| ogra_ | heh | 12:20 |
| sergiusens | mandel: are you really on holidays? | 12:24 |
| sergiusens | :-) | 12:24 |
| === jhodapp|afk is now known as jhodapp | ||
| === dandrader is now known as dandrader|lunch | ||
| === tedg is now known as ted | ||
| === alan_g|lunch is now known as alan_g | ||
| === ted is now known as tedg | ||
| ev | Laney: thanks for the fix! | 13:02 |
| Laney | ev: yw | 13:03 |
| === dandrader|lunch is now known as dandrader | ||
| rsalveti | ogra_: firefox os uses ril directly, over socket | 13:32 |
| rsalveti | but not ofono :-) | 13:32 |
| ogra_ | ah, thanks | 13:32 |
| rsalveti | they implemented a modem manager on java script or similar | 13:33 |
| ogra_ | finally someone who understands the matter :) | 13:33 |
| rsalveti | crazy dudes | 13:33 |
| ogra_ | yeah | 13:33 |
| === psivaa-afk is now known as psivaa | ||
| rsalveti | north: see http://events.linuxfoundation.org/sites/events/files/slides/Ubuntu%20Touch%20Internals_1.pdf | 13:33 |
| ogra_ | https://mxr.mozilla.org/mozilla-central/source/dom/system/gonk/ril_worker.js then | 13:33 |
| rsalveti | right | 13:35 |
| ogra_ | that means they still use the binary rild from andrpid though | 13:35 |
| rsalveti | yeah, the use android for everything | 13:36 |
| ogra_ | just with a different mechanism on top | 13:36 |
| ogra_ | right | 13:36 |
| rsalveti | they basically replace java with javascript | 13:36 |
| rsalveti | :-) | 13:36 |
| ogra_ | haha, yeah | 13:36 |
| === gnu is now known as Guest28314 | ||
| rickspencer3 | tasty | 13:49 |
| rickspencer3 | I think the phone app crashed during a call | 13:49 |
| rickspencer3 | but the call kept going, which was good | 13:50 |
| === pete-woods is now known as pete-woods-lunch | ||
| dobey | bzoltan: hi. who's the best person to ping about things that should be in the sdk ppa? | 14:00 |
| dobey | dpm: hi, is there a way to request an immediate sync of translations back to the project in launchpad? | 14:03 |
| dpm | dobey, unfortunately not. IIRC it happens only once a day very early in the European mornings and it cannot be manually requested | 14:04 |
| dpm | dobey, for the automatic exports (commits) | 14:04 |
| dpm | if you are talking about the other way round (.po file contents to be imported to LP) that happens all the time if you enable it, or you can request a manual sync in that direction, yes. But that's generally not used, as most Ubuntu translators, if not all, translate directly from the web UI | 14:05 |
| === boiko_ is now known as boiko | ||
| dobey | no, i mean the commits | 14:06 |
| dpm | ok | 14:06 |
| dpm | dobey, the alternative is to request a download of all the .pos in Launchpad and then do a manual commit of them | 14:07 |
| dobey | right | 14:07 |
| dpm | that'd effectively be an immediate, totally manual, sync | 14:07 |
| dpm | but that's probably not what you're looking for | 14:07 |
| dobey | dpm: right, i'd prefer to have a button on launchpad to do an automatic export back to the branch | 14:08 |
| dpm | yeah :) | 14:08 |
| bzoltan | dobey: me, I guess :) shoot what do you need. | 14:18 |
| === mhr3_ is now known as mhr3 | ||
| dobey | bzoltan: it seems like unity-scopes-api, unity-api, and ubuntu-download-manager should be in it, no? | 14:19 |
| bzoltan | dobey: sounds logical, yes | 14:20 |
| dobey | great | 14:20 |
| wellsb | ogra_, it was my ISP after all. Surprise surprise. I guess I need to begin working on a vpn client for this platform | 14:23 |
| bzoltan | dobey: would you mind to file a bug against the ubuntu-sdk package? | 14:24 |
| dobey | bzoltan: sure | 14:24 |
| ogra_ | wellsb, oh my ... | 14:24 |
| ogra_ | great that you found the issue though :) | 14:25 |
| dobey | wellsb: just install openvpn (though there's no UI for it yet) | 14:27 |
| dobey | you should be able to manually configure it | 14:27 |
| dobey | wellsb: how is your isp the issue though? they block login.ubuntu.com? | 14:27 |
| === alan_g is now known as alan_g|tea | ||
| dobey | bzoltan: against ubuntu-touch-meta you mean? | 14:33 |
| bzoltan | dobey: yes | 14:34 |
| dobey | bzoltan: https://bugs.launchpad.net/ubuntu/+source/ubuntu-touch-meta/+bug/1319867 | 14:37 |
| ubot5 | Ubuntu bug 1319867 in ubuntu-touch-meta (Ubuntu) "Missing packages in ubuntu-sdk-team ppa" [Undecided,New] | 14:37 |
| === alan_g|tea is now known as alan_g | ||
| bzoltan | ogra_: could you please take that request -> https://bugs.launchpad.net/ubuntu/+source/ubuntu-touch-meta/+bug/1319867 ? | 14:43 |
| ubot5 | Ubuntu bug 1319867 in ubuntu-touch-meta (Ubuntu) "Missing packages in ubuntu-sdk-team ppa" [Undecided,Confirmed] | 14:43 |
| awe_ | davmor2, did you file a new bug for 3g/wifi being disabled by default? | 14:43 |
| davmor2 | awe_: I did let me grab it | 14:48 |
| davmor2 | bug #1319812 | 14:49 |
| ubot5 | bug 1319812 in indicator-network (Ubuntu) "Indicator-network is doing odd things" [Critical,New] https://launchpad.net/bugs/1319812 | 14:49 |
| davmor2 | awe_: ^ | 14:49 |
| ogra_ | bzoltan, any particular reason that you filed that against the meta package ? or was that just a placeholder for something ? | 14:51 |
| ogra_ | davmor2, no, the other one ... i dont think you know about it yet | 14:51 |
| ogra_ | davmor2, since yesterday mako comes up with wifi off by default | 14:51 |
| bzoltan | ogra_: is not that package should list those new stuff? | 14:51 |
| awe_ | davmor2, ogra_ mentioned another bug, but if it hasn't been filed yet, he said he would do so... | 14:52 |
| ogra_ | bzoltan, well, i'm not sure what you want me to do there ... | 14:52 |
| ogra_ | bzoltan, do you want these packages somewhere in the seeds so that ubuntu-touch directly depends on them ? | 14:52 |
| bzoltan | ogra_: I think we need couple of extra packages in the sdk meta package ... but educate me please if you think it should be handled differently | 14:53 |
| ogra_ | plars, do you know if we had a bug open for the mantas coming up offline by default ? | 14:53 |
| bzoltan | dobey: ^^ | 14:53 |
| davmor2 | ogra_: awe_ I added it in that bug | 14:54 |
| davmor2 | item 2 | 14:54 |
| ogra_ | bzoltan, the sdk-libs meta should just contain packages that define the framework ... i.e. for ubuntu-download-manager we have an entry for qtdeclarative5-ubuntu-download-manager0.1 in sdk-libs ... | 14:54 |
| dobey | bzoltan: i don't know that the ubuntu-sdk package needs to list them (maybe it does), but we need the newer versions in the PPA, so that people on trusty can build things | 14:54 |
| ogra_ | bzoltan, for developers we have all the headers in sdk-libs-dev ... | 14:55 |
| davmor2 | ogra_, awe_ : I bundled all the issues together in one | 14:55 |
| ogra_ | bzoltan, and for image building we have ubuntu-touch which depends on actual binaries shipped in the image ... | 14:55 |
| ogra_ | davmor2, well, the manta issue happened before | 14:55 |
| ogra_ | davmor2, which might point to the possibility that there are two bugs | 14:55 |
| davmor2 | ogra_: yeah so that one you are on your own with :) | 14:56 |
| awe_ | we'll see... | 14:56 |
| ogra_ | but it looks like nobody filed yesterdays bug | 14:56 |
| davmor2 | ogra_: I wasn't here was I ;) | 14:56 |
| ogra_ | thats what we get if you are not around for a day | 14:56 |
| dobey | bzoltan: so it seems like the meta package has the deps already | 14:56 |
| ogra_ | nobod files bugs anymore | 14:56 |
| dobey | bzoltan: and ubuntu-touch-meta doesn't seem to be in the PPA anyway | 14:57 |
| ogra_ | dobey, bzoltan, we definitely dont have either of these packages seeded directly ... for scopes there is libunity-scopes-dev in sdk-libs-dev ... for download manager we have the above in sdk-libs ... and system-image seeded in ubuntu-touch depending on download manager in ubuntu-touch ... | 14:58 |
| ogra_ | there is nothing for unity-api | 14:58 |
| ogra_ | and i dont think we want that seeded in either ubuntu-touch or sdk-libs ... since these define what goes into the image | 14:58 |
| dobey | ogra_: sure. i'm not picky about whether they are necessarily seeded or not. i just think they should be in the ubuntu-sdk-team ppa | 14:59 |
| ogra_ | right, is the meta package in that PPA ? | 14:59 |
| dobey | no | 14:59 |
| dobey | at least, not for trusty | 14:59 |
| dobey | but launchpad doesn't have bugs against archives, so no way to file a bug against the ppa itself. and bzoltan asked me to file the bug there, so i did | 15:00 |
| ogra_ | the two -api packages smell like they might be good to be shipped with the headers, not sure if everyone developing afainst sdk-libs wants or should get that installed thogh | 15:00 |
| ogra_ | *against | 15:00 |
| dobey | ogra_: well, anyone developing a scope will need it, i guess | 15:00 |
| ogra_ | ok, so we should seed these two in sdk-libs-dev | 15:01 |
| ogra_ | which is what you should have installed when buiolding against the framework | 15:01 |
| dobey | or well, at lest, unity-scopes-api requires unity-api | 15:01 |
| ogra_ | but that wont fix the PPA at all indeed | 15:01 |
| ogra_ | dobey, is that properly reflected in package deps ? | 15:01 |
| dobey | right, i mostly care about the ppa for this | 15:01 |
| dobey | ogra_: afaik, yes | 15:01 |
| ogra_ | then we would only need to seed unity-scopes-api and get the other one for free | 15:02 |
| ogra_ | but again, that wont fix the PPA | 15:02 |
| bzoltan | ogra_: I see | 15:03 |
| dobey | right. i'm asking for them to be in the PPA | 15:03 |
| * bzoltan just passed the daily minimal self humiliation | 15:03 | |
| dobey | i haven't looked at any possible issues with their relation to the meta packages yet | 15:03 |
| bzoltan | dobey: Okey .. now I get it. So you ask a backport of the meta package to the PPA | 15:04 |
| dobey | bzoltan: well the meta package could go in the PPA too i guess. what i'm aksing for is a backport of unity-api, unity-scopes-api, and ubuntu-download-manager, in the PPA | 15:05 |
| bzoltan | dobey: ohh... that is a different exercise. And a different problem. | 15:10 |
| bzoltan | dobey: according to our new policy the APIs should come from the framework in the chroots, so the click chroot must have these APIs installed. But the click chroots use the archive not the PPA. Hmmm... | 15:11 |
| bzoltan | dobey: because we stopped backporting runtime stuff to dekstop | 15:11 |
| dobey | hmmm | 15:12 |
| dobey | :-/ | 15:12 |
| bzoltan | dobey: are you sure you do not want to SRU them to Trusty? | 15:13 |
| === gatox is now known as gatox_lunch | ||
| dobey | bzoltan1: well, that's not up to me. if it's feasible to do so, i would love for that to happen too | 15:25 |
| dobey | bzoltan1: but i think such packages do need to be in a reasonably available manner in a ppa until then at least (and probably after then too) | 15:26 |
| === vying is now known as Guest30643 | ||
| bzoltan1 | dobey: these pacages are needed to build the scopes, right? | 15:27 |
| dobey | bzoltan1: yes | 15:29 |
| bzoltan1 | dobey: but not for running them | 15:35 |
| dobey | bzoltan1: well the new versions of the libs are needed for anything that's developed against the newer versions of the libs | 15:36 |
| davmor2 | ogra_, awe_: issue one sorted icon is the wrong name, onto issue two, why the wifi is disabled by default | 15:36 |
| ogra_ | davmor2, can you file a new bug for the disabled wifi | 15:37 |
| dobey | ah well, need to get lunch | 15:42 |
| === pete-woods-lunch is now known as pete-woods | ||
| === jhodapp is now known as jhodapp|lunch | ||
| north | rsalveti: Then why did Ubuntu-telephony team had to go with another abstraction layer again ? | 15:53 |
| rsalveti | north: to use ofono | 15:54 |
| north | Well while that is a simple answer, why use Ofono, when you have RIL already ? | 15:54 |
| rsalveti | north: because we don't want to depend on ril-only | 15:54 |
| rsalveti | with ofono you could later on use whatever phone that ofono supports | 15:54 |
| rsalveti | like on the desktop, with at modems and so on | 15:55 |
| rsalveti | remember we also need to have a compatible solution for the desktop | 15:55 |
| north | Yea kinda single RIL multiple modem support | 15:55 |
| davmor2 | awe_, cyphermox, ogra_: https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/1319906 | 15:55 |
| north | Have you tried porting Modem Manager to Ubuntu-Touch ? | 15:55 |
| ubot5 | Ubuntu bug 1319906 in network-manager (Ubuntu) "indicator-network show wifi disabled on a bootstrap install or ota update on 31" [Undecided,New] | 15:55 |
| rsalveti | right, but ofono also supports a bunch of additional modems | 15:55 |
| ogra_ | davmor2, thanks ... | 15:55 |
| ogra_ | awe_, ^^^^^ | 15:56 |
| rsalveti | north: we discussed that already, before we started the ofono rilmodem project, and ofono was way more complete | 15:56 |
| rsalveti | voice calls, sim, etc etc | 15:56 |
| north | Yes, but look at this perspective, while ofono supports legacy modems, mm supports newer modems which speak newer protocols | 15:56 |
| rsalveti | modemmanager could only manage data at that point | 15:56 |
| north | ok | 15:56 |
| rsalveti | well, it's a price you pay, the features we needed were already in ofono | 15:56 |
| north | Yes, I would like to port mm to ubuntu | 15:57 |
| north | and test with it | 15:57 |
| ogra_ | we do have mm in the archive | 15:57 |
| rsalveti | mm already works on ubuntu | 15:57 |
| rsalveti | desktop uses mm with nm | 15:57 |
| ogra_ | we just dont use it on the phones | 15:57 |
| ogra_ | on desktop we use it for 3G dongles etc | 15:57 |
| awe_ | north, MM doesn't support voice | 15:57 |
| * rsalveti lunch | 15:58 | |
| north | But using mm we can remove the dependency on libhybris | 15:58 |
| ogra_ | no | 15:58 |
| ogra_ | you still need to talk to the container | 15:58 |
| awe_ | north, we don't use libhybris for telephony | 15:58 |
| awe_ | north, we talk to rild thru a socket | 15:58 |
| ogra_ | (for audio etc) | 15:58 |
| north | oh yea. sorry awe_ | 15:58 |
| awe_ | np | 15:58 |
| ogra_ | so while you have no hybris in the actual telephony stack there are enough bits that still requite hybris | 15:59 |
| awe_ | well sure...most of the image requires hybris | 16:00 |
| awe_ | ;D | 16:00 |
| awe_ | but again the main reason no to use MM was complete last of voice related functionality | 16:00 |
| north | Hmmm, | 16:00 |
| awe_ | damn, my typing sucks this morning | 16:00 |
| north | telephony is a quite complex area to work on, where noOEM goes the open source way :p | 16:01 |
| north | and that OEM is Qualcomm most of the times :'D | 16:01 |
| awe_ | and by voice, I mean basic call support, call-waiting, conf calling, call forwarding, ... | 16:01 |
| north | Yes, I understand | 16:02 |
| * awe_ has to take-off; bbl | 16:02 | |
| SolarAquarion | I'm getting 404's when i update via apt-get | 16:02 |
| === chihchun is now known as chihchun_afk | ||
| AlbertA | Saviq: ping | 16:13 |
| === gatox_lunch is now known as gatox | ||
| Saviq | AlbertA, pong | 16:13 |
| AlbertA | Saviq: hey so I just responded to the e-mail thread about powerd/usc rework | 16:14 |
| AlbertA | Saviq: so one questions is do we expect any other non-mir sources | 16:14 |
| AlbertA | Saviq: for brightness/inactivity timer values? | 16:14 |
| Saviq | AlbertA, I don't think so | 16:15 |
| AlbertA | Saviq: otherwise I'll just get started on putting the protobuf messages and reusing the mir socket | 16:15 |
| Saviq | AlbertA, the only sources would be u-s-c sessions | 16:15 |
| Saviq | AlbertA, so they'd all have a Mir session | 16:15 |
| Saviq | AlbertA, yeah, that's fine by me | 16:15 |
| AlbertA | Saviq: ok and the other question is the proximity stuff I wasn't clear | 16:16 |
| AlbertA | Saviq: my main concern is breaking things as they are right now | 16:17 |
| AlbertA | Saviq: proximity is not handled by USC | 16:17 |
| Saviq | AlbertA, it's handled in powerd right now, are you moving it out of there? | 16:18 |
| AlbertA | Saviq: the handling was entrenched with display state changes which are out of powerd | 16:18 |
| Saviq | AlbertA, right, so we need to find a place for it between greeter, usc and unity8 | 16:18 |
| AlbertA | Saviq: we could modify powerd slightly | 16:19 |
| AlbertA | Saviq: because it still has the ofono stuff in there | 16:19 |
| AlbertA | Saviq: and it could just emit a proximity event in that case, at least until the policy rearchitecture comes | 16:19 |
| Saviq | AlbertA, makes sense | 16:20 |
| Saviq | AlbertA, I'll need a clear head to think where I see it in the end, and that (clear head) is not gonna happen today | 16:21 |
| AlbertA | Saviq: ok, just send me your thoughts on the e-mail thread | 16:21 |
| Saviq | AlbertA, will do, but for now, since we want to keep the ofono things still in powerd, the above sounds good | 16:22 |
| AlbertA | Saviq: ok | 16:23 |
| north | also rsalveti, why did ubuntu-touch have to go with Cyanogenmod ? and Not AOSP ? | 16:24 |
| ogra_ | north, because at that time CM supported more devices so people could port more easily | 16:25 |
| ogra_ | north, we switched to AOSP kitkat a while ago for the supported devices | 16:25 |
| north | Which means this statement should go into the wiki pages | 16:25 |
| === chihchun_afk is now known as chihchun | ||
| north | also, I would suggest the maintainers to update the wiki with more illustrations :) Just a suggestion :) | 16:26 |
| ogra_ | if we would find the time to update the wiki we would :P | 16:27 |
| north | Sure ogra_ :) | 16:27 |
| ogra_ | to do that properly for all the new stuff doing a port yourself is kind of required to add/remove all the new steps | 16:28 |
| ogra_ | and that takes time | 16:28 |
| ogra_ | rsalveti, ricmm_, what is libconnectivity-cpp doing ? we are just trying to nail down the network breakage and that seems to be pulled in a new package | 16:28 |
| ogra_ | cyphermox, ^^^ do you knwo ? | 16:29 |
| cyphermox | Wellark knows exactly | 16:30 |
| cyphermox | it would allow things like indicator-network to do things with network devices, in theory | 16:30 |
| ogra_ | oh, i thought that was in the gcc4.8 changes | 16:30 |
| ogra_ | i see | 16:31 |
| ogra_ | so it is a new thing that came in with indicator-network | 16:31 |
| Wellark | ogra_: yes. it just wraps NM, ofono and urfkill api's | 16:32 |
| Wellark | only used by indicator-network atm | 16:32 |
| ogra_ | Wellark, does it alread yhook in to anything ? | 16:33 |
| Wellark | the version in the repo only hooks up with NM | 16:33 |
| ogra_ | and could it interfere with anything we had before ? | 16:33 |
| Wellark | nope. | 16:33 |
| Wellark | does not change anything or keep any states of it's own | 16:34 |
| Wellark | on system level | 16:34 |
| ogra_ | we can clearly nail donw the network breakage to a few certain images and your landing is the most suspicious in there | 16:34 |
| ogra_ | well, actually the only suspicious one | 16:34 |
| Wellark | ogra_: what are the symptoms? | 16:34 |
| Wellark | did you have a bug filed? | 16:34 |
| ogra_ | http://people.canonical.com/~ogra/touch-image-stats/29.changes http://people.canonical.com/~ogra/touch-image-stats/30.changes http://people.canonical.com/~ogra/touch-image-stats/31.changes are the related changelogs | 16:35 |
| ogra_ | Wellark, no network on boot | 16:35 |
| ogra_ | we discussed that before :) | 16:35 |
| Wellark | ogra_: will visit grocery store quickly | 16:35 |
| Wellark | I will double check after taht | 16:35 |
| Wellark | *that | 16:35 |
| ogra_ | and on flo and manta actually not even the wifi options are shown in the system-settings app | 16:35 |
| ogra_ | Wellark, davmor2 is just doing some tests and will update the bug | 16:36 |
| davmor2 | I might | 16:36 |
| Wellark | ogra_: which bug? | 16:36 |
| ogra_ | Wellark, bug 1319906 | 16:36 |
| ubot5 | bug 1319906 in network-manager (Ubuntu) "indicator-network show wifi disabled on a bootstrap install or ota update on 31" [Critical,Confirmed] https://launchpad.net/bugs/1319906 | 16:36 |
| === alan_g is now known as alan_g|EOD | ||
| Wellark | ogra_: what does "nmcli nm" show ? | 16:37 |
| ogra_ | (sorry, i already assigned to cyphermox ... might actually be libconnectivity-cpp0 related) | 16:37 |
| Wellark | run inside adb-shell | 16:37 |
| ogra_ | Wellark, i dont have an broken install here ... davmor2 ^^^^ | 16:37 |
| Wellark | indicator-network shows wifi as disabled if NetworkManager tells it that WirelessEnabled property is false | 16:37 |
| davmor2 | ogra_: I don't have a broken install yet 28 worked remember | 16:37 |
| cyphermox | all the information is already on the bug | 16:38 |
| davmor2 | Wellark: I filed it against network-manager as it seemed to be specific issue with wifi rather than 3g or the indicator | 16:39 |
| cyphermox | it's the indicator | 16:39 |
| cyphermox | in that bug, we see the output of nmcli con, nmcli dev, nmcli nm. the wireless device is not disabled, but it's disconnected | 16:40 |
| cyphermox | Wellark: is anything you know setting an initial state in indicator-network as disconnected ? | 16:40 |
| davmor2 | cyphermox: I wasn't sure I assumed they were all related initially hence I added them all to one bug. but there seems to be a new issue in that the settings app is blank on manta and flo on image 31 | 16:41 |
| mterry | Wellark, I have a branch attached to bug 1319580 btw | 16:41 |
| ubot5 | bug 1319580 in indicator-network (Ubuntu) "Secret agent should watch NetworkManager lifecycle" [Undecided,New] https://launchpad.net/bugs/1319580 | 16:41 |
| mterry | Wellark, I was testing it fine yesterday but I went to test a merge from trunk I did on it, and now I'm hitting the "no network indicator" bug | 16:42 |
| mterry | Wellark, I thought I just had to update unity8 to fix that? But didn't seem to work for me | 16:42 |
| cyphermox | *sigh* I'm going to half-trash my bluetooth work and reflash my device and try to reproduce the issue | 16:44 |
| cyphermox | davmor2: please, never assume two things are the same bug. I much rather we have hundreds of bug we can close off than one that is open for 3 different issues, that will never be resolved | 16:45 |
| mterry | doanac, can you fill out the unity8 checklist (https://wiki.ubuntu.com/Process/Merges/Checklists/Unity8) for your https://code.launchpad.net/~doanac/unity8/unlock-device-emulator/+merge/219573 branch? | 17:10 |
| Wellark | cyphermox: _nothing_ in indicator-network should be setting any initial state | 17:11 |
| Wellark | otherwise I would get disconnects each time I restart indicator-network | 17:12 |
| Wellark | does anyone have such a system running where this problem is showing? | 17:12 |
| Wellark | + could someone please start looking into making the test scripts to enable networking if it's disabled | 17:13 |
| Wellark | that we need anyway | 17:14 |
| polyl | is there a terminal on the ubuntu phone ? | 17:14 |
| polyl | to connect to a server ? | 17:14 |
| doanac | mterry: looks like you beat me to it? | 17:15 |
| Wellark | mterry: is there a critical bug depending on that secret-agent bug? | 17:15 |
| Wellark | I would like to take a look at it tomorrow if at all possible | 17:15 |
| mterry | doanac, I did for the review side, but the proposer has their own checklist | 17:15 |
| ogra_ | Wellark, we're fiddling to much atm ... i think davmor2 will do a fresh install later and can give you info | 17:15 |
| Wellark | it's already 8pm here | 17:15 |
| ogra_ | heh, you are ahead 1h :) | 17:15 |
| mterry | Wellark, not five alarm critical, just a blocker for split greeter landing | 17:16 |
| mterry | Wellark, i.e. take your time | 17:16 |
| Wellark | mterry: how close is it to land? | 17:16 |
| mterry | Wellark, or rather, tomorrow would be nice :) | 17:16 |
| Wellark | mterry: please send me an email on your EOD | 17:16 |
| mterry | Wellark, close, but I've been saying that for weeks | 17:16 |
| Wellark | so I don't forget | 17:16 |
| Wellark | cyphermox: where does NM store the WirelessEnabled ? | 17:17 |
| cyphermox | /var/lib/NetworkManager/NetworkManager.state. you shouldn't have to worry about that | 17:17 |
| ogra_ | Wellark, see #ubuntu-ci-eng ... | 17:18 |
| Wellark | cyphermox: yep, I should not. just wanted to check | 17:18 |
| === beidl_ is now known as beidl | ||
| === jhodapp|lunch is now known as jhodapp | ||
| === tedg is now known as ted | ||
| doanac | rsalveti: you have a link/list of what apps are available for x86 emulator. for example seems like gallery_app and notes_app are missing? | 18:52 |
| dobey | are we not building fat packages yet? | 18:56 |
| rsalveti | doanac: I believe the native apps that are click are not yet available, we're waiting cjwatson to create the chroot for us to be able to build them for x86, then sergiusens will take care of creating and uploading them next week | 18:59 |
| rsalveti | atm I think just qml and webbrowser apps are available, as they are arch independent | 18:59 |
| sergiusens | doanac: native clicks are pending, and I think I didn;t loop you in, but we are going to enable them next week with fginther | 19:00 |
| sergiusens | rsalveti: webbrowser is a deb | 19:00 |
| rsalveti | sergiusens: sure, I said web apps :-) | 19:00 |
| sergiusens | rsalveti: lol, yeah, it's hard to read into that :-P | 19:01 |
| sergiusens | core community apps would be there as well except for filemanager, music and terminal... as soon as music rids itself from grilo, it should also land there | 19:01 |
| SolarAquarion | I'm getting 404's by the ubuntu repo's | 19:07 |
| dobey | SolarAquarion: what repo? the main archive for deb packages? | 19:09 |
| SolarAquarion | dobey: the main ones | 19:09 |
| dobey | SolarAquarion: #ubuntu is the help channel for ubuntu, this channel is for discussion about ubuntu on phones | 19:10 |
| SolarAquarion | dobey: i'm discussing ubuntu touch | 19:10 |
| SolarAquarion | for armhf | 19:10 |
| SolarAquarion | or something | 19:10 |
| dobey | can you pastebin a specific error then? | 19:12 |
| SolarAquarion | dobey ports.ubuntu.com 404 not found | 19:15 |
| SolarAquarion | for binaries | 19:15 |
| dobey | SolarAquarion: i think you need to ask for more help in #ubuntu then. | 19:15 |
| SolarAquarion | dobey: this is for discussing bugs in the software? | 19:16 |
| dobey | SolarAquarion: this channel is for discussion of development of ubuntu touch | 19:16 |
| SolarAquarion | ok | 19:17 |
| slvn_ | hello ! just wondering if the branch of ted is going to be merged? https://code.launchpad.net/~ted/upstart-app-launch/ld-library-path/+merge/217832 | 19:18 |
| cjwatson | rsalveti: I have a patch for that from Michael in my queue to review, I think | 20:02 |
| notilew | hi, someone porting ubuntu touch for Motox devices? | 20:03 |
| BigWhale | Greetings... | 20:21 |
| popey | hello BigWhale | 20:30 |
| BigWhale | popey, hey! :) | 20:32 |
| popey | hows you? | 20:32 |
| BigWhale | popey, Overworked. But ok. :) | 20:32 |
| BigWhale | but I just installed ubuntu on my nexus | 20:33 |
| BigWhale | :D | 20:33 |
| popey | \o/ | 20:33 |
| kenvandine | hey BigWhale! | 20:33 |
| BigWhale | hey ken! | 20:34 |
| BigWhale | Ok, KeePass needs porting to Ubuntu :) | 20:38 |
| BigWhale | Phone | 20:38 |
| BigWhale | all my passports are 30+ characters long and entering them with on screen keyboard is a pain :D | 20:39 |
| kenvandine | BigWhale, i feel your pain | 20:40 |
| BigWhale | :)) | 20:40 |
| veladem | Hello :D | 21:33 |
| beidl_ | is someone here who could answer a tiny technical question regarding ubuntu touch? | 21:44 |
| === beidl_ is now known as beidl | ||
| beidl | I'm planning to port one of my qt apps to ubuntu touch. however, it requires some way to get access to usb devices. | 21:46 |
| beidl | it's a music learning program. main feature is the ability to connect MIDI devices to the device... and play some music. | 21:47 |
| beidl | currently the application works on windows, linux (rtmidi) and android (libusb) | 21:47 |
| beidl | are there any plans to give applications a way to access usb devices? | 21:48 |
| t1mp | beidl: I think if you can access the usb devices via cpp you can expose those to qml and write an ubuntu touch app | 21:49 |
| t1mp | maybe there is already a qt lib for that? | 21:50 |
| beidl | t1mp: I'm more concerned about how opening the usb devices file descriptor from within the app works in the sandbox world | 21:50 |
| t1mp | beidl: seems like libusb is already included in the imag libusb-1.0-0: Installed: 2:1.0.17-1ubuntu2 | 21:52 |
| t1mp | beidl: as far as I know you need to configure your click package to request the access that you need, and then the user needs to allow it when installing the package | 21:53 |
| beidl | t1mp: interesting. I'd have to test it, but I'd be quite surprised if access to usb devices is not restricted on ubuntu touch | 21:53 |
| t1mp | everything is restricted, so probably you need to study how the click packages work and request access to hardware | 21:54 |
| t1mp | beidl: see http://bazaar.launchpad.net/~phablet-team/camera-app/trunk/view/head:/click/camera-apparmor.json for an example of the camera-app | 21:54 |
| t1mp | http://bazaar.launchpad.net/~phablet-team/camera-app/trunk/files/head:/click/ | 21:55 |
| t1mp | beidl: I don't know much about it, but until you catch someone here who does, maybe http://developer.ubuntu.com/publish/apps/security-policy-for-click-packages/ helps | 21:56 |
| * t1mp off now | 21:56 | |
| beidl | t1mp: not sure if there is an apparmor policy that would allow access to usb devices. I'm aware of the fact that it's a quite exotic use case ^^ | 21:56 |
| beidl | t1mp: anyway, thanks a lot! | 21:57 |
| beidl | if there was a policy for my use case, the application would probably run without changes to the source. it's a miracle that it works on windows, desktop linux and android using the same code base. :) | 21:58 |
| beidl | it was pretty hard though... for example, on windows a QMutex sometimes fails to lock, causing serious issues related to concurrency | 21:59 |
| Saviq | jdstrand, hey, should pkcon install-local generate a profile for a .click? | 22:03 |
| beidl | the thing is, on android, as soon as you connect a USB device to the tablet/phone using a USB to go adapter, the android framework asks the user if he'd like to give the application access to the USB device | 22:03 |
| jdstrand | Saviq: yes. sudo -H -u phablet pkcon install-local /path/to/click | 22:03 |
| Saviq | jdstrand, I'm ssh'd in, good enough? | 22:04 |
| beidl | in an ubuntu touch world, I could see hooking this functionality up with udev. but I assume there are no design plans for something like this | 22:04 |
| beidl | for the UI | 22:04 |
| belkinsa | Who is the Evernote developer sandbox for? | 22:04 |
| jdstrand | Saviq: the security manifest will be placed in /var/lib/apparmor/clicks, then aa-clickhook will take that manifest and drop the appparmor policy in to /var/lib/apparmor/profiles, then the profile is loaded in to the kernel and the apparmor policy cache file is put in /var/cache/apparmor | 22:05 |
| jdstrand | Saviq: should be. you just want to be phablet and not root | 22:05 |
| Saviq | jdstrand, yeah, I'm not seeing it in /var/lib/apparmor/profiles | 22:05 |
| Saviq | jdstrand, what do I do to debug? | 22:05 |
| Saviq | jdstrand, want me to send you the .click? | 22:05 |
| belkinsa | And is the way to close/kill an app? | 22:05 |
| jdstrand | Saviq: what does 'sudo aa-clickhook' have to say? | 22:05 |
| jdstrand | Saviq: you can send me the click | 22:06 |
| Saviq | jdstrand, ah, invalid policy version, probably didn't update the policy for new framework | 22:06 |
| jdstrand | Saviq: fyi, you can install click-reviewers-tools and run 'click-run-checks /path/to/click' | 22:07 |
| jdstrand | that should catch that sort of thing | 22:07 |
| Saviq | jdstrand, thanks, will dig more | 22:08 |
| jdstrand | (though it is limited in that it won't know what click frameworks are on the target device) | 22:08 |
| === Ursinha is now known as Ursinha-afk | ||
| Saviq | jdstrand, I added "location" to my webapp, when trying to access location, getting http://pastebin.ubuntu.com/7470164/ - is that expected? | 22:15 |
| nethical | there are any news about ubuntu-touch for Galaxy s3? | 22:19 |
| === Ursinha-afk is now known as Ursinha | ||
| jdstrand | Saviq: we don't let apps talk to nm at this time, so if it is trying to figure out something from there, then yes it is expected | 22:36 |
| Saviq | jdstrand, well, it's just a webapp asking the browser for location info | 22:37 |
| jdstrand | Saviq: if there is an updated api that is safe to use, please file a bug against apparmor-easyprof-ubuntu with the output of 'grep DEN /var/log/syslog' | 22:37 |
| Saviq | jdstrand, I'll have a chat with tvoss tomorrow on the location service then | 22:37 |
| jdstrand | what is the output of that command? ^ | 22:37 |
| jdstrand | (I know people have been working on a connection api, but I haven't seen it yet) | 22:38 |
| Saviq | jdstrand, there's actually no DEN for the location API: http://pastebin.ubuntu.com/7470252/ | 22:42 |
| === jhodapp is now known as jhodapp|afk | ||
| jdstrand | looks like you need the audio policy group, btw | 22:43 |
| Saviq | jdstrand, but std{out,err} has: http://pastebin.ubuntu.com/7470260/ | 22:43 |
| Saviq | jdstrand, well, not really, there's no sound coming from that webapp - and if there is, I don't want it ;) | 22:43 |
| Saviq | because it'd be ads | 22:43 |
| jdstrand | Saviq: ok, then you are hitting the explicit denials in /usr/share/apparmor/easyprof/policygroups/ubuntu/1.1/networking, so yes, that is expected | 22:45 |
| Saviq | jdstrand, ok, will talk to tvoss tomorrow - this seems to be trying to figure out location from surrounding wifi networks | 22:45 |
| jdstrand | there is nothing in syslog because we silenced them using 'deny' | 22:45 |
| Saviq | yup | 22:45 |
| jdstrand | it very well might be. the nm api however reveals way too much | 22:46 |
| jdstrand | (which is why we deny it) | 22:46 |
| jdstrand | he probably knows how this will all relate to the new connection api | 22:46 |
| jdstrand | ok, gotta run! | 22:47 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!