=== mibofra is now known as Guest94531 === mibofra is now known as Guest62711 [03:56] Hm. Has anyone else noticed the arale backlight not being turned off? [04:11] RAOF: Historically when that happens on any device most people don't notice :) [04:12] But it's not good [04:18] Oh, well. LP# 1460898 filed. === chihchun_afk is now known as chihchun [07:29] Oh, hey. [07:29] Our GMainLoop alarm implementation blatantly ignores the specified behaviour of various calls. [07:32] I'm pretty sure I relied on that behaviour in the FrameDroppingPolicy... [07:35] hm? [07:35] RAOF: I think something like SceneObserver could be replaced in a nicer way with signal/slot [07:36] anpok: This is indeed what I was thinking, yes. [07:36] anpok: Oh, AlarmImpl unconditionally returns true on cancel() (even if it didn't actually cancel), and likewise on reschedule. [07:37] While the documentation says that they return true iff they cancelled something, rescheduled a pending alarm respectively. [07:37] in a previous system signal(event) / slot registrations with functor objects were the #1 memory consumers [07:37] but there we had hundreds of them [07:37] and most of them did not fit into the functor payload.. [07:38] oh === mibofra is now known as Guest41543 === mibofra is now known as Guest55129 [10:34] alf_: happy now? https://code.launchpad.net/~alan-griffiths/mir/helptext-mentions-config-file/+merge/260439 === mzanetti is now known as mzanetti|run [11:19] greyback: tvoss do we want to reschedule the WM discussion? When? [11:20] alan_g: today is bad for me, am traveling. Wed & Thurs my timetable is uncertain, lots of meetings. Friday traveling again! [11:21] greyback: OK. np. [11:22] best I can do is next week, Wed the earliest [11:26] Fine by me. Today I'm more interested in how lp:~alan-griffiths/mir/only-visible-surfaces-change-scene/ might affect Qt clients and qtmir. [11:28] alan_g: let me have a look and I'll get back to you [11:30] alan_g, we definitely should, next Wednesday is fine by me, too === tvoss is now known as tvoss|lunch === chihchun is now known as chihchun_afk === mzanetti|run is now known as mzanetti === dandrader_ is now known as dandrader|afk === dandrader|afk is now known as dandrader [14:08] alan_g, fixed the test name on: https://code.launchpad.net/~kdub/mir/multistream-protobuf-additions/+merge/259970 [14:11] kdub: TA'd [14:12] thanks alan_g === mzanetti is now known as mzanetti|food [15:09] kdub: I've also run in to some problems with the event RPC lately... === mzanetti|food is now known as mzanetti [15:35] racarr, which sort of problems? [15:37] kdub: oh I guess just that the memcpy to a flat buffer doesn't work for [15:37] some of the event cleanup I am doing as well [15:37] whic hsort of relates to your can't pass FD's over events perhaps [15:38] racarr, right, so if i'm scrubbing may as well scrub in a way thats helpful to that concern too === dandrader is now known as dandrader|lunch === alan_g is now known as alan_g|EOD [17:06] vogons any running mir on nouveau lately ? [17:07] intel for me [17:07] kgunn, yes I am on nouveau [17:08] popey: ^ [17:08] camako: i think popey is trying to run unity8-desktop-session-mir on nouveau [17:08] gah, just logging into unity7 with nouveau gives me 1280x1024 not 1920x1080 :( [17:08] i was telling him, mir + nouvea might be ok [17:09] camako: so i figure your probably running mir+nouvea frequently [17:09] kgunn, I'm not running U8 dektop session recently [17:09] right...so problem might lie in the unity8 part [17:09] kgunn, just mir [17:09] and examples [17:09] sure [17:10] so popey, that answers that ^ [17:10] hmm, sudo mir_demo_server fails [17:10] failed to load libraries from path /usr/lib/x86_64-linux-gnu/mir/server-platform [17:11] following http://unity.ubuntu.com/mir/using_mir_on_pc.html Running Mir natively [17:12] kgunn: so if I want to actually test Unity8 (specifically apps) what am I to do? Buy an intel laptop / tablet? [17:13] popey: that might be easier :-P but seriously, let's see if we can get you some help....what's your gpu ? [17:13] camako: ^ can i leave popey in your capable hands... [17:13] nVidia GeFprce GTX 460 [17:13] *GeForce [17:15] kgunn, don't really work on U8 desktop, so not sure how helpful I can be [17:15] camako: he's failing on mir-demo-server [17:15] please read [17:16] kgunn, ah ok [17:16] sorry [17:16] popey, are you running in a vt? [17:16] http://paste.ubuntu.com/11524763 [17:16] yes [17:17] popey, let's step back, did you install the packages yourself? Or running from binaries you built from the tree? [17:17] popey: is this mir 0.13? [17:17] i have a clientplatform in /usr/lib/x86_64-linux-gnu/mir but not a server-platform [17:17] this is on wily, from repo [17:18] popey: apt-get install mir-platform-graphics-mesa2 [17:18] maybe I'm missing a package? [17:18] ah yeah..isn't there a dumb issue about having to install graphics platform [17:18] ok, it fails differently now :) [17:19] http://paste.ubuntu.com/11524864/ [17:21] popey: i had to do this the other day, i did apt-get install mir-graphics-drivers-desktop [17:21] ok [17:21] same error after doing that [17:22] popey do you have drm devices in your system? [17:22] ls -l /dev/dri <<--- [17:23] ls: cannot access /dev/dri: No such file or directory [17:23] popey that's the problem [17:23] camako@camako-MacBookPro:~$ ls /dev/dri/ [17:23] card0 card1 controlD64 controlD65 renderD128 renderD129 [17:23] popey ^ ... mir needs card* devices to run [17:24] popey, kernel version? [17:24] Linux wopr.popey.com 3.19.0-20-generic #20-Ubuntu SMP Fri May 29 10:10:47 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux [17:29] popey at the top of the "Running Mir natively" page there are checks for hw [17:29] what do they return? [17:30] oh, i skipped that bit because kgunn told me to do the Running Mir natively bit, sorry. [17:30] http://paste.ubuntu.com/11524937/ [17:33] popey, so I guess you don't really have a nouveau driver [17:33] camako@camako-MacBookPro:~$ lsmod | grep drm [17:33] drm_kms_helper 122880 2 i915,nouveau [17:33] drm 344064 9 ttm,i915,drm_kms_helper,nouveau [17:33] camako@camako-MacBookPro:~$ sudo pmap `pidof X` | grep dri.so [17:33] [sudo] password for camako: [17:33] 00007fb9b76e5000 7620K r-x-- nouveau_dri.so [17:33] 00007fb9b7e56000 2048K ----- nouveau_dri.so [17:33] 00007fb9b8056000 348K r---- nouveau_dri.so [17:33] 00007fb9b80ad000 48K rw--- nouveau_dri.so [17:34] :( [17:36] popey, I'm surprised your distro doesn't have it.... I guess it should be as simple as installing xserver-xorg-video-nouveau package... But proceed cautiously as you're meesing with the video driver [17:36] ii xserver-xorg-video-nouveau 1:1.0.11-1ubuntu2b amd64 X.Org X server -- Nouveau display driver [17:36] it _is_ installed [17:37] I'm on Ubuntu 15.10 (wily) [17:38] aha, it's not being used. It's using FBDEV! [17:38] popey, I'm still on vivid... perhaps this is a new (wily) nouveau issue [17:38] that's why the resolution is wonky I guess. [17:38] hmmm [17:38] popey, some help here : http://nouveau.freedesktop.org/wiki/UbuntuPackages/ [17:38] aha https://wiki.ubuntu.com/X/Troubleshooting/VideoDriverDetection#Problem:__Need_to_fully_remove_-nvidia_and_reinstall_-nouveau_from_scratch [17:39] need to _re_ install nouveau [17:40] there you go... [17:40] thanks for the help! [17:40] yw [17:40] silly software :) [17:56] has mir_socket moved? I don't see it in /tmp [17:57] aha! /run/user/1000/mir_socket [17:58] Ok, so I have managed to get mir running, and ran a mir demo (mir_demo_client_eglplasma) [17:59] so we have proved the video driver works now. [18:00] is there a way to poke for the ABI version of mir? [18:00] with the new mir lib, i need to check what version of code I compile depending on the ABI version [18:02] rebooting to lightdm, and choosing unity8 lxc session fails though. [18:02] bunch of apparmor failures... http://paste.ubuntu.com/11525449/ [18:04] balloons: you tried this? [18:05] bschaefer, mir version is printed to stdout (by default) [18:06] camako, ill need it through configure.ac though, is there a function that prints out the version? [18:07] there seems to be: mir_toolkit/version.h:65: MIR_VERSION_NUMBER(MIR_CLIENT_MAJOR_VERSION, \ [18:07] bschaefer, there is a cmake variable MIR_VERSION [18:07] camako, o cool, thanks! [18:08] yw [18:08] that works for cmake :), ill have to hack something for autogen === dandrader|lunch is now known as dandrader [18:49] camako: alan_g|EOD: so the qtubuntu updates to match mir 0.13 made it to wily right? [18:49] shouldn't the branch be merged then? [18:49] AlbertA, kgunn said that LT was investigating why it didn't get merged [18:50] yes it should be merged [18:59] after the buffer semantics changes, we dont really have a stream anymore... but will cross that bridge later [19:00] camako: so should we merge it manually? or do we need to ask somebody? [19:01] AlbertA, I am assuming LT will do it once they are through with their investigation. kgunn [19:01] kgunn ^^ [19:02] camako: ok...who is LT though? [19:02] camako: AlbertA so we added that MP to the rotation silo4, mzanetti is working on landing it....which should merge it [19:02] AlbertA, landing team [19:02] is there another issue currently ? [19:02] camako: kgunn: ok... [19:02] kgunn: no just want to rebase my branch === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader [22:41] @vogons: since technically we broke client API in mir 0.13.1 (like mir_surface_set_event_handler), the client/mir_toolkit/version.h should have bumped it's major number right? [22:41] Yes. I thought we did the backwards-compatibility symver trick for that, though? [22:42] RAOF: yes we did [22:44] so I'll be attempting a mir 0.13.2 for entirely unrelated reasons (have to get the vegetahd quirk in)... [22:44] was wondering if I should fix it then. [22:44] Oh, but we don't expose both in the API, right. [22:44] Yeah, we should have bumped version. [22:44] Otherwise there's no good way for client code to know what to build against. [22:44] RAOF: right which is what bschaefer wanted [22:45] and robert_ancell! [22:45] Ding! [22:45] yep [22:46] vogons: ok so mir 0.13.2 = client 2.0.0, and lp:mir: client 3.0.0 since we broke for good there ? [22:48] thanks! [23:31] @vogons: just FYI I've requested a silo for mir 0.13.2 - with the intention of landing to vivid+overlay and wily [23:32] AlbertA, sounds good. The freeze should be over soon. [23:32] AlbertA, is only Mir in the silo? [23:32] camako: yes. with only the vegetahd quirk and the client API version fix [23:33] great [23:33] camako: https://code.launchpad.net/~mir-team/mir/0.13/+merge/260902 [23:35] ok [23:48] strange crash... http://paste.ubuntu.com/11531043/ only seems to fail on SDL2 tests (if i run my own SDL2 app it works fine) [23:51] something in dynamic loader must be doing something strange...hmm