[00:00] Ayy [01:10] Are the porting guides still recommended? I'm seeing a lot of references to 2013 and Saucy. [02:07] Also, now that clockworkmod is abandoned, should we still use that as the bootloader? [02:10] Further, am I better off grabbing the build someone else built back in '13 and try and update it, or should I just try to build for Vivid from scratch? [02:12] Is someone able to help me with a QML question? I have a skeleton of what I'm working on here: http://pastebin.ubuntu.com/12527836/ [02:13] The question is? How do I reference the Dialog element from outside the Component so that line 17 will work (or better yet so I can close the dialog from where it's created) === chriadam is now known as chriadam|away === evanwang is now known as evanwang_afk === evanwang_afk is now known as evanwang [06:13] good morning all you wonderful people! [06:14] good morning [06:14] and the same to you Kniple_! [06:14] I still have quite a lot of problems with my MX4, it charges oddly, this night, for example, it did not charge, only stayed on 40% throughout the night. :( [06:14] I remember there being something about a large fix to mx4's battery problems in OTA5, but seems there aree still problems. [06:18] Kniple_: you had it attached to a wall charger? [06:18] yes [06:18] not the original wall socket to usb though, but the cable was original. [06:19] (they sent a UK one with my order.) [06:37] hello! is there a way to get arrow keys on the virtual keyboard? === tsdgeos_ is now known as tsdgeos === chihchun_afk is now known as chihchun [08:34] DIY :P [08:50] Good morning all; happy Wednesday, and happy Restless Legs Awareness Day! ☺ [09:49] hi all! what should the telegram scope do? i get a blank page (i'm connected in the telegram app) [09:51] guest42315: have you tried pulling it down to refresh it? [09:51] davmor2, yep many times, still nothing [09:52] davmor2, i see the bussy/loading line on the buttom but when it stops.. nothing [09:52] blank page [09:53] lemme adb into the phone [09:56] davmor2, *** Error in `/usr/lib/arm-linux-gnueabihf/unity-scopes/scoperunner': corrupted double-linked list: 0xac80b680 *** [09:56] aa_getcon failed, errno = 13 === ogra2 is now known as ogra-snappy-test [09:56] is this bad? [09:59] popey, ogra: ^ is this working for you, I see stuff in my telegram scope. [10:01] it could be bug 1472755 [10:01] bug 1472755 in libqtelegram (Ubuntu) "/usr/lib/arm-linux-gnueabihf/unity-scopes/scoperunner:*** Error in `/usr/lib/arm-linux-gnueabihf/unity-scopes/scoperunner': corrupted double-linked list: ADDR ***" [High,Triaged] https://launchpad.net/bugs/1472755 [10:04] works fine here [10:06] i also get a lot of this [10:06] 2015/09/23 12:25:16.829068 ERROR session error'ed out with read tcp 91.189.88.30:443: use of closed network connection [10:06] ^^ ubuntu-push-client.log [10:06] 2015/09/23 12:29:18.111438 ERROR not connected after 1m0s; giving up [10:09] pstolowski: it seems osomon is not around, but do you think you could evaluate whether eg scopes see any problems or benefit from this recent qtbase upgrade? did you ever encounter such errors as described in https://codereview.qt-project.org/#/c/120738 's bug report like queueing multiple requests? the fix (and the previous one which it now fixes further) were urged to be included by upstream, but if t [10:09] here's no reasonable explanation for osomon's x86 unit test regression and we don't get any benefits visible, we should probably revert this fix-of-the-previous-fix to back to OTA6 level. silo 54 has the revert. [10:10] I'm thinking of running some typical https requests in a loop or such [10:12] https://bugs.launchpad.net/canonical-devices-system-image/+bug/1498841 [10:12] Ubuntu bug 1498841 in Canonical System Image "telegram scope - blank page" [Undecided,New] [10:15] davmor2: what? [10:16] popey: do you see messages in the telegrram scope [10:16] * popey enables the telegram scope to find out [10:16] no, blank [10:17] [M#k\[ 1792.236721] type=1400 audit(1443003410.450:373): apparmor="DENIED" operation="mknod" profile="com.ubuntu.telegram_sctelegram_1.3.20.114" name="/tmp/etilqs_vIXgQhqkdRkVPcE" pid=18838 comm="scoperunner" requested_mask="c" denied_mask="c" fsuid=32011 ouid=32011 [10:25] guest42315: ^ there you go [10:37] davmor2, so it's popey fault, right? [10:38] popey, broke telegram scope [10:39] guest42315: if at first you don't succeed use the fallback and blame popey ™ [10:39] :)) right [11:10] poor popey :'( [11:11] Mirv, I need to think about it a bit, will let you know. I *think* I saw 'device not open' mentioned in the bug report a couple of times in Apps/Store scope logs in the past, but it was rare and without any pattern to reproduce [11:33] pstolowski: thanks! I'm not that worried in the sense it passed QA, I've run unit tests now on many other packages, and the only known flakiness is on x86, but still the more we understand about the fix's effects the better === chihchun is now known as chihchun_afk === alan_g is now known as alan_g|lunch === pat_ is now known as Guest88770 [12:44] hi anyone around [12:45] Guest97268: best is to just ask the question [12:46] my bq aquaris e4.5 was working fine over night i woke to check time powered the button on and off to check time, this morning when i woke i pwoered the button on and screen just flashes [12:47] i thought the battery was flat at first but when i plugged into mains power it said 28% [12:48] so i tried rebooting using power button and volume buttons but nothing happens [12:50] mpt, were you working with MacSlow on the flash-the-device animation? [12:54] Guest97268: let it fully charge then try again [12:56] ok will do [12:57] Guest97268: I think there is a bug with the battery levels [12:58] only had a few updates for apps recently and it was working fine before === alan_g|lunch is now known as alan_g [12:59] Guest97268: I had something similar happen a while ago. Phone claimed 48% battery left but instant power-off at reboot due to low battery [12:59] Guest97268: phone hadn't been charged for 4--5 days so 48% battery was probably optimistic [13:00] ok will a fix for this be on the horizon or is it just a one off [13:00] Guest97268: I don't think a fix will be forthcoming unless we find a way to reproduce the bug. It hasn't happened to me since [13:01] i normally do a full charge when needed but recently i think i have only been charging to the 90%'s before having to take the phone out with me [13:02] Guest97268: I think I may have had a partial charge before my episode. I usually charge to 100% as well [13:02] its been on charge a few minutes and has not booted up [13:02] Guest97268: I had to charge mine for an hour or so [13:03] if i was out and about i'd have been real concerned my phone was brocked lol [13:03] must remember to full charge it when needed [13:03] Guest97268: Yes - hopefully your problem will be the same as mine [13:04] fingers crossed then!! absolutely love this phone!! my best one yet!! [13:10] Thanks for the guidance catch you when i'm on next === dholbach_ is now known as dholbach === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader === ogra2 is now known as ogra-snappy-test === ogra2 is now known as ogra-snappy-test === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === dandrader is now known as dandrader|lunch === alan_g is now known as alan_g|EOD [17:42] hello [17:43] i am trying to port a gaming engine to ubuntu-touch and I have an error when running an app build with the engine (cocos2d) [17:43] libust[18075/18077]: Error: Error opening shm /lttng-ust-wait-5 (in get_wait_shm() at lttng-ust-comm.c:958) [17:43] a little bit later the glCreateProgram fails and I think they are related [17:44] any hint on how to fix/work around this? [17:55] BlackJohnny, i think i had the same error with sdl/opengl [17:55] BlackJohnny, you can open a bug here https://bugs.launchpad.net/canonical-devices-system-image [17:56] ihhh [17:57] guest42315, but how come the qt apps work with opengl? [17:59] BlackJohnny, i don't know [17:59] BlackJohnny, i just used some sdl/opengl templates, just a sec [18:00] BlackJohnny, https://github.com/Sturmflut/ubuntu-touch-sdl2-gles2-template [18:00] BlackJohnny, https://github.com/Sturmflut/ubuntu-touch-sdl-template [18:01] guest42315, should that compile and run with no problems on utouch? [18:02] BlackJohnny, yes [18:02] guest42315, I will have a look, thanks! [18:02] BlackJohnny, have fun === dandrader|lunch is now known as dandrader [18:07] guest42315, it runs ok. That should help me track the issue I am currently facing [18:09] BlackJohnny, :D (the devs are usually here @ 11am - 9pm London time [18:09] that's why so quiet now in here [18:09] guest42315, i do this after work :) [18:09] you can also try on #unity-mir [18:09] thanks for the hint [18:09] oh not that [18:09] #ubuntu-mir [18:10] :D [18:10] every piece of hardware i work on is linux :) [18:11] and ... "landroid" [18:11] :)) [18:21] BlackJohnny, if you're having trouble porting a toolkit to Mir, try poking bschaefer he did the SDL port and may know a few things [18:22] bregma, thanks for the info [18:23] can anyone tell me where I need to look on my phone to see logs written during the previous failed boot? [18:23] I can connect using "adb sehll" while in recovery mode, regular boot hangs and I want to find out why [18:23] I tried mounting system.img and ubuntu.img but they don't seem to harbour anything that looks likely [18:24] * bschaefer reads backlog [18:25] I know ogra_ in particular could tell me but he's forgiven if he's not around right now [18:25] i sadly am around and super busy [18:25] bregma: i can tell you where I found mine [18:25] popey, that would please me [18:26] /proc/last_kmsg [18:26] thats the dmesg output from exactly the last boot ... [18:26] mm, nothing helpful in there, it looks normal to me [18:26] bregma: where did you find the system.img? [18:26] any other logs, like upstart or anything get preserved? [18:26] BlackJohnny, im not sure what cocos2d is off the top of my head :), does it use SDL2? (sounds like something mac related) [18:27] popey, I found them in /data [18:27] bschaefer, cocos2d is a game engine initially created for ios but now is for may platfoms including linux ... but not for touch [18:28] hm, well this is odd, they're missing here [18:28] BlackJohnny, now SDL2 has a cocoa backend [18:28] but ... im not sure if thats the same *cocoa* [18:28] popey, do a 'mount -a' first to get /data mounted? [18:28] and if it was, it wouldnt work on mir [18:28] * bschaefer reads the cocoa backend [18:28] oh, duh [18:29] bschaefer, I don't think cocoa and cocs2d are the same [18:29] bschaefer, the engine does not seem to support sdl [18:29] BlackJohnny, if it doesnt support SDL2 then you'll have to write your own native window for the opengl context [18:29] and create your own Mir native window for it to work [18:29] bregma: so /data/system-data/var/log/syslog ? [18:29] * bschaefer isnt sure how open the cocoa2d system is [18:29] bregma, yeah i figured but wanted to double check [18:30] BlackJohnny, https://hg.libsdl.org/SDL/file/63f03a567fe3/src/video/mir [18:30] the backend for SDL2 for mir [18:31] popey, that looks hopeful [18:31] Sep 22 16:22:00 ubuntu-phablet kernel: [ 36.015321] init: lxc-android-config post-start process (704) terminated with status 1 [18:31] which sets up creation of an opengl context through EGL, or uses software renderering, as well as handles input events and feeds it into the SDL2 backend [18:31] bschaefer, I was thinking that I will have to somehow replace the X11 code but had the feeling that glCreateProgram is called before having anyting to do with that [18:32] BlackJohnny, yeah, not 100% sure how cocos2d handles the video driver backend situation [18:32] but yes the x11 code will have to be replaced with mir code [18:32] to talk to the mir api bits, that allows for rendering on the mir display server [18:32] a lot of it is simply plumbing [18:33] bregma: yeah, that's what I had [18:33] depends on how much work cocos would need to fit to how mir works ie. EGL2 support (is the bigger depends) [18:33] which if it works on the phone, should already be expected in the backends [18:34] it uses egl [18:34] but if its only for X11 then im not sure about the EGL2 support [18:34] thats good [18:34] popey, I don't have the mesa links in /etc/alternatives you had, as far as I can tell [18:34] that makes things easier to do [18:34] as far as I've understood it only uses xopendisplay and close from x11 [18:35] BlackJohnny, these are the three examples of ported libraries to MIR: SDL1.2, SDL2, GLFW3 [18:35] https://github.com/glfw/glfw/tree/master/src [18:35] https://code.launchpad.net/~brandontschaefer/+junk/sdl1.2-mir [18:35] I am new to GL programming so pls dont judge me :) [18:35] https://hg.libsdl.org/SDL/file/63f03a567fe3/src/video/mir [18:35] BlackJohnny, :) no worries [18:35] BlackJohnny, i've not seen cocos source code [18:35] ok thanks [18:35] but ... i would assume it has some plumbing of the X11 code to the cocos API [18:35] such that the users are not aware of X11 existing [18:35] indeed [18:35] when using the toolkit [18:36] soo all you have to do in a sense is replug those lower level API calls with calls that match up to mir [18:36] such that mir will know how to render all the pixels :) [18:36] so for mir, XDisplayOpen is equal to mir_connect_sync [18:36] bschaefer, I need to consult the links you gave me. Many thanks! [18:37] BlackJohnny, just poke me with any questions! (Im on PST time) [18:37] bschaefer, ok, thanks [18:38] np! [18:57] kenvandine, mind looking at the two mps in silo 22? [18:58] jgdx, sure [18:58] jgdx, green in jenkins... i love it! [18:58] kenvandine, thanks. Two passes in a row among them [18:59] am I dreaming [18:59] :p [18:59] jgdx, approved [19:01] jgdx: yes you're dreaming [19:01] jgdx: now wake up and do some work ;) [19:01] davmor2, okay, i'll guess I just have to enjoy it while it lasts then [19:01] :) [19:01] :D [19:03] Wellark, hey... i just found https://bugreports.qt.io/browse/QTBUG-33123 [19:04] Wellark, you have patches for that, any chance of getting someone to merge those? [19:14] Wellark, although your fix hard codes the fallback to /usr/share/pixmaps [19:15] Wellark, i'm trying to get icons from a libertine container, so i want it to find icons in $appdir/pixmaps, for example to find the firefox icon :) [19:16] Wellark, where $appdir is the path to /usr/share inside the container [19:24] Hey all, question: is anyone working on mesh networking for Ubuntu Phone? [19:24] being able to transfer files or send text-based messages in a p2p way would be awesome, without needing centralized infrastructure [19:31] BlackJohnny, looking at the code, it looks like it actually uses GLFW [19:31] though it might have a hard coded depends on X11 for Linux [19:31] (for getting the DPI) [19:32] for desktop yes but what I have done is to mirror the android implementation for ubuntu-touch [19:32] BlackJohnny, and it looks like it *can* use GLFW3 soo all you need to do is grab this source code: https://github.com/glfw/glfw [19:32] o i see [19:32] glfw3 *should* work on the phone as well [19:32] since it uses egl [19:32] I can go on this path though [19:33] IIRC [19:34] i think you might just have to enable DESKTOP (for ubuntu touch + glesv2 as well as disable the X11 bits) [19:34] BlackJohnny, but idk with https://github.com/glfw/glfw you need -DGLFW_USE_MIR=ON [19:34] so having the desktop implementation and use mir instead of x11 [19:34] to enable the mir parts of GLFW (since it doesnt do runtime video driver switches) [19:35] BlackJohnny, right, you just need to recompile GLFW with MIR enabled [19:35] and not X11 enabled [19:35] BlackJohnny, there is some hard coded X11 in the linux part of the platform [19:35] that will have to commented out/disabled [19:35] that would be awesome [19:35] as that wont work :) [19:35] BlackJohnny, it should*just* work with a bit of hacking [19:36] i was studying the sdl code [19:37] BlackJohnny, well it turns out you wont need to implement that since i've done that backend work in GLFW already :) [19:37] kenvandine: I don [19:38] 't have the git branches anymore [19:38] cloning glwf [19:38] Wellark, i see part of it was merged [19:38] but not the base dirs part [19:38] which is what i think i need [19:38] but they should be available from qt repositories [19:38] BlackJohnny, yup, just make sure you have -DGLFW_USE_MIR=ON :) [19:38] Wellark, they are :) [19:38] kenvandine: you can cook up a distro patch [19:38] (you might need some libraries not sure) [19:39] kenvandine: and have Mirv to include them in overlay ppa [19:39] Wellark, i'll look at it, after thinking about it i need a slight variation of your patch [19:39] Wellark, thanks for the info [19:40] bschaefer: I may need to poke you for some advice about getting the Godot engine running on the phone. Have been spending a bit of time playing with Mir code. Think I can crack it, but hope you don't mind if I call on your expertise at some point? [19:41] mcphail, anytime! [19:41] bschaefer: cheers! [19:41] you can always send me an email if off time zones :) [19:42] kenvandine: see https://codereview.qt-project.org/#/c/63987/3/src/platformsupport/themes/genericunix/qgenericunixthemes.cpp [19:42] if the code has not changed in later Qt Releases [19:42] you can see the line QString xdgDirString = QFile::decodeName(qgetenv("XDG_DATA_DIRS")); [19:43] so, you might not need a patch, just a clever append to XDG_DATA_DIRS [19:43] Wellark, yeah [19:43] bschaefer: I may do that. Thanks! If you're interested, I'll be working from https://github.com/mcphail/godot/tree/try_mir . Will be digging in again in a week or so when I finish my travels [19:44] mcphail, cool, i check it out [19:46] kenvandine: oh, also. firefox ships a .xpm so you need this too: https://codereview.qt-project.org/63988 [19:46] mcphail, nice, looks like a good skeleton setup, a lot of those functions you can check SDL2 out for [19:47] Wellark, actually my container has a png for firefox [19:47] um actually base it off this SDL2 branch (since new mir API/ABI changes) [19:47] Wellark, but... support for xpm should be there :) [19:47] https://code.launchpad.net/~brandontschaefer/+junk/SDL2-new-mir-ABI [19:47] * bschaefer needs to make sure thats up-to-date with mir 0.16 [19:47] bschaefer: I'm working off the older API just now as I want it to work with the phone :) [19:48] kenvandine: ok. good. firefox used to ship only xpm [19:48] mcphail, o right [19:48] mcphail, one thing i dont see is surface/opengl/events [19:48] yeah... finally getting more modern, but still ships the icon in the pixmaps dir [19:48] kenvandine: feel free to pick up the upstream work and get those pull requests updated and merged to qt proper :) [19:48] is that to come in with the skeleton? [19:48] bschaefer: yes, that's what I'm experimenting with just now [19:48] Wellark, i'll take a look [19:48] kenvandine: thanks! [19:48] mcphail, nice! Yeah just looked at x11 [19:49] Wellark, the only one of the apps on the immediate radar that's causing a problem is firefox [19:49] bschaefer: yes - that's what I'm copying [19:49] mcphail, the first thing i start with is getting the main mir connect, then opengl, then render SOMETHING [19:49] then i worry about events [19:49] nice to see that pixel flip :) [19:49] bschaefer: ha! [19:50] * bschaefer needs to go back and do SFLW ... it was lacking EGL/GLESv2 support IIRC [19:50] bschaefer: this has been a fun learning process [19:50] mcphail, it is very fun! [19:50] also nice to see how much smaller the x11 bits are to the mir bits [19:50] err [19:50] other way around [19:51] how much smaller the mir code ends up being vs the x11 code [19:51] bschaefer: I hope so, anyway [19:51] :) [19:52] bregma, do you remember that list of toolkits we had for things that needed porting to mir? [19:52] * bschaefer cant find that old oc [19:52] doc* [19:52] kenvandine: good luck! :) [19:55] nm found it === dandrader is now known as dandrader|afk [20:14] elimisteve: aquarius made soonsnap which kinda does that, but does use a central server, but kinda transfers between machines === dandrader|afk is now known as dandrader [21:13] bschaefer, hi again [21:13] BlackJohnny, hello! [21:14] bschaefer, before i go to sleep :) I cant compile because some MirBufferStream type missing [21:14] is it some work in progress ? [21:14] glfw [21:14] BlackJohnny, hmm what version of mir do you have? [21:14] i have installed the sdk with it installed by default [21:14] IIRC Installed: 0.15.1+15.10.20150903-0ubuntu1 [21:15] didn checked [21:15] you'll need 0.15 or more [21:15] BlackJohnny, new API/ABI [21:15] BlackJohnny, apt-cache policy libmirclient-dev [21:16] 0.12.1+15.04.20150324-0ubuntu1 0 [21:16] ok [21:16] BlackJohnny, you'll need to https://github.com/glfw/glfw/commit/9527c93a1c9a43ab3b2a2cfb56389cd3619b90f3 [21:16] remove that commit [21:16] from GLFW3 [21:16] and it should fine :) [21:16] ok, thanks [21:16] np! [21:16] * bschaefer has glfw3 upstream aimed at trunk mir :) [21:18] bschaefer: what is the Mir versioning convention when they break the interface? It would be nice if they'd bump the major version [21:18] mcphail, 0.14/0.15 is the major API/ABI breakage [21:19] from there its pretty clean, just new API functions (and no ABI breakage) [21:19] bschaefer: surely the project is worthy of a 1.x release with major version bump on each breakage? [21:19] mcphail, the two big changes are mir events 2.0, and buffer stream vs surface [21:19] mcphail, i've no clue :), i think thats a goal [21:19] Cheers :) [21:20] * mcphail thinks OS projects are too recticent about achieving a 1.x release [21:22] :) [21:35] bye bye