=== chihchun_afk is now known as chihchun === anpok_ is now known as anpok [08:40] tjaalton: RAOF has put the Mir SRU into https://launchpad.net/ubuntu/xenial/+queue?queue_state=1, but someone else needs to approve. I see you named on https://launchpad.net/~ubuntu-sru/+members#active - does that mean you can help? [08:40] alan_g: sure [08:40] while it's not my sru day today but can have a look [08:40] because I need this too :) [08:41] That was exactly my thought. 8^) [10:14] anpok: I know you talked to bschaefer about the libinput patch. That's all that's happened right? [10:22] I hope those get either dropped or upstreamed [10:26] I know, it is one of the embarrassments I need to get cleaned up. [10:26] i'm updating it to 1.7.2 for debian experimental [10:26] and then merge for artful [10:33] alan_g: I see the mir backport adds new packages, does it need a transition? [10:35] tjaalton: I don't think so. It installs alongside just fine [10:37] one thing I noticed is that all the bugs that this SRU references have the upstream tasks as "fix committed" and not fix released [10:38] tjaalton: yeah... [10:38] We were in the middle of a 0.27 release when things happened [10:40] That will get picked up again, but I've fewer hands than before [10:40] okay [10:41] They are all low risk changes, and we're the only real consumers of that stuff [10:42] alright [10:55] alan_g, mind taking a look at https://forum.snapcraft.io/t/bug-in-core-running-on-rpi3/657 ? (was mir-kiosk/mir-libs recently updated ?) [10:57] ogra_: sure. I'm not aware of any recent updates. (Near future however...) [10:58] we didnt really change anything in core either, the drivers should still be the saem as always ... [10:58] snaps are built and linked against the debs though ... [10:59] so if the archive has something that doesnt work with mir-libs or mir-kiosk, that might have some impact [13:15] could the libinput patch to keep compatibility with old mir servers be dropped from artful? [13:23] tjaalton: I know that would be nice, but I think that would likely break the mir source currently in archive. [13:30] ogra_: I got the same symptoms on dragonboard after a "snap refresh". Looked like the interfaces were screwed up (they'd been frigged manually in the past). Uninstalling and reinstalling libs, kiosk and apps seemed to resolve whatever went wrong. Hopefully, will work for rIl3 too. [13:31] rPi3 [13:38] alan_g, ah, cool, thanks for researching that! [13:39] It is my post if I can help with debugging it? [14:24] AlbertA: you've a rPi3 haven't you? Could you have a look into https://forum.snapcraft.io/t/bug-in-core-running-on-rpi3/657/11 [14:33] alan_g: I do [14:35] umm "Loading module: 'libubuntu_application_api_test.so.3.0.0'" seems wrong [14:43] ogra@pi3:~$ find /snap/mir-kiosk-apps/16/ -name '*libubuntu_application_api_test*' [14:43] oops [14:43] /snap/mir-kiosk-apps/16/usr/lib/arm-linux-gnueabihf/libubuntu_application_api_test.so.3 [14:43] /snap/mir-kiosk-apps/16/usr/lib/arm-linux-gnueabihf/libubuntu_application_api_test.so.3.0.0 [14:43] definitely existing though [14:46] ogra_: right but it's just a stub used for testing [14:46] ah [14:46] ogra_: interesting part, I booted my rpi3 with some old core version I suppose... [14:47] mir-kiosk-apps started fine.... [14:47] then auto-updated... [14:47] weird [14:47] and I see the issue [14:47] well, mir-kiosk itself starts for me just fine [14:47] just not the apps [14:47] ogra_: yeah [14:47] same here [14:47] so i wonder how a system update could relate to the breakage here [14:48] was this an old core with no mir snaps installed ? [14:48] (initially) [14:48] ogra_: no the snaps were installed already [14:48] ah [15:02] ogra_: ummm I'm not getting anything in syslog... is there a command to enable output now? or did logging change? [15:02] AlbertA, sudo journalctl --no-pager [15:02] we dropped rsyslog [15:03] ogra_: ahh thanks! [15:03] (kills SD cards ... journalctl uses a ringbuffer) [15:03] s/journalctl/journald/ [15:14] ogra_: so yeah seeing the same thing "Unable to load selected module, using dummy." a bit useless... [15:17] well [15:18] looks like it falls over in the inotifywait call of /snap/mir-kiosk-apps/16/bin/mir-kiosk-app-daemon [15:18] but it all originates from papi not being able to load the module [15:18] which ... if i read that right ... is supposed to wait til a config file has been created [15:21] ogra_: yeah it [15:21] it's used to wait for changes to the confif file [15:22] but the main issue is that papi fails to load the module (so dlopen failed on it) for some reason [15:23] but since nothing changed about the snaps... only core (at least here) probably a new security related feature? [15:24] or a snapd change ... [15:26] ogra_: right... [15:29] ogra_: yeah just a dlopen: http://bazaar.launchpad.net/~phablet-team/platform-api/trunk/view/head:/src/ubuntu/application/base_module.h#L113 [15:29] mind noting that in the thread ? [15:40] ogra_: done [15:45] AlbertA, thanks! [15:46] ogra_: happens on VM with an amd64 core image so should make it easier to debug [15:47] yeah === chihchun is now known as chihchun_afk [17:03] ogra_: so turns out newer core snap revisions do not have libjson-c.so.2 anymore, which libubuntu_application_api_desktop_mirclient.so.3.0.0 depended on. [17:04] IIRC snapcraft has some smarts to not include "known" system libraries or some such... perhaps that's why it was not included in the mir-kiosk-apps snap originally... [17:04] AlbertA: thanks for tracking that down [17:05] alan_g: no prob [17:30] hey guys [17:30] allah is doing [17:31] sun is not doing allah is doing [17:31] to accept Islam say that i bear witness that there is no deity worthy of worship except Allah and Muhammad peace be upon him is his slave and messenger === JanC_ is now known as JanC