[01:01] <twoj__> Anybody have a good idea how to reset a 'frozen' session?  I'm running TeamViewer to one account/desktop and that session is frozen, where I get X for cursor
[01:02] <twoj__> However in second session I can log in via RSH and get console
[01:02] <twoj__> No process is really pegging the CPU or memory?
[01:06] <twoj__> Hello....   anybody free?
[05:48] <oSoMoN> good morning desktoppers!
[05:52] <duflu> Hello oSoMoN
[05:52] <duflu> Hmm, curious how many Gnome Tweaks don't actually work...
[05:53] <oSoMoN> hey duflu, good afternoon!
[06:34] <didrocks> good morning
[07:01] <duflu> Morning didrocks
[07:01] <didrocks> hey duflu
[07:02] <seb128> hey duflu didrocks desktopers
[07:03] <didrocks> re seb128
[07:03] <duflu> lo seb128
[07:03] <duflu> hi seb128
[07:04] <seb128> duflu, how are you? having a good friday? ready for the w.e?
[07:06] <duflu> seb128: More VAAPI debugging. Not fun. And I'm not sure about being ready for the weekend :) How are you?
[07:07] <seb128> I had a good 7 hours sleep, feeling good
[07:07] <seb128> let's see what friday brings now
[07:26] <oSoMoN> salut didrocks, seb128
[07:27] <didrocks> joyeux vendredi oSoMoN
[07:27] <oSoMoN> :)
[07:32] <oSoMoN> seb128, I’m reconsidering the move to a classic snap for LO, thinking it’s not such a good idea after all. Built a strict snap with the desktop-gtk3 launcher, and that fixes most of the known issues \o/
[07:32] <oSoMoN> only bug #1665106 requires a patch, but it’s trivial, and I’ll see if I can upstream it, given how similar it is to https://github.com/LibreOffice/core/commit/28a03248b1d1649e157b788e43dfe8326f165379
[07:32] <ubot5> bug 1665106 in libreoffice (Ubuntu) "[Snap] Ctrl+click on a link doesn't open the link" [Medium,In progress] https://launchpad.net/bugs/1665106
[07:33] <seb128> lut didrocks, bon vendredi !
[07:33] <seb128> oSoMoN, oh, nice
[07:34] <seb128> oSoMoN, oh, right, we also have a fake xdg-open calling a dbus method, only it's under /usr/local/bin/xdg-open for us
[07:35] <didrocks> oSoMoN: I'm opening a donation page for the launcher ;)
[07:35] <seb128> lol
[07:36] <oSoMoN> seb128, yep, so the patch looks like this: http://paste.ubuntu.com/24987313/
[07:36] <seb128> oSoMoN, looks good, the other alternative would be to tell them to call "xdg-open" rather than "/usr/bin/xdg-open"
[07:36] <oSoMoN> didrocks, will the donations pay for all those extra hours you put into it? ;)
[07:37] <seb128> so local versions take over
[07:37] <seb128> including the snap one
[07:37] <didrocks> oSoMoN: hem, "funds repartitions need to be decided and voted on" :p
[07:38] <oSoMoN> seb128, yes that’s another option, I’ll see what upstream prefers
[07:38] <didrocks> hum
[07:38] <didrocks> /usr/local/bin was in $PATH
[07:38] <didrocks> then removed by upstream
[07:38] <didrocks> but ogra told he would reinclude it IIRC
[07:38] <didrocks> (some months ago)
[07:39] <seb128> that's a snap issue to be resolved yeah
[07:39] <seb128> I was talking about the libreoffice side
[07:39] <didrocks> that was in the snap-confine related change IIRC (the move to snap run)
[07:40] <andyrock> good morning
[07:41] <oSoMoN> buon giorno andyrock
[07:41] <seb128> hey andyrock, happy friday!
[07:42] <didrocks> hey andyrock
[07:43] <oSoMoN> tjaalton, when do you think llvm-toolchain-4.0 will migrate to xenial-updates? I’ve tested a build of chromium-browser (dev branch) with clang-4.0 from -proposed, and that works well
[07:44] <tjaalton> oSoMoN: along with the rest of the 16.04.3 hwe stack
[07:44] <tjaalton> next month
[07:44] <tjaalton> best guess
[07:44] <oSoMoN> good, looking forward to it
[07:46] <ogra_> didrocks, it should be in the launchers afaik ...
[07:46] <ogra_> also ... we moved xdg-open inside core to a sane path
[07:46] <didrocks> ogra_: changing PATH to a system one?
[07:46] <didrocks> I don't understand why the PATH was removed without warning btw
[07:46] <ogra_> yeah
[07:46] <didrocks> it's a behavior change
[07:46] <didrocks> and core 16 was already released
[07:47] <didrocks> (the fact that it was independent of snapd was requested by the snap team btw when designed)
[07:47] <oSoMoN> ogra_, what do you mean by a sane path? it's currently /usr/local/bin/xdg-open, right? is this going to change?
[07:47] <ogra_> https://github.com/snapcore/core/pull/35/files
[07:48] <ogra_> nothing changed PATH though ... the in-core xdg-open was just moved around so the PATH hack wouldnt be needed anymore
[07:48] <oSoMoN> oh! that means I can drop my LO snap patch then, that’s good \o/
[07:48] <ogra_> yeah :)
[07:49] <oSoMoN> ogra_, when is that landing in a snap store near me?
[07:49] <ogra_> now ... thats a merge to master though
[07:49] <ogra_> not sure if thats already in a relese
[07:49] <ogra_> i would guess with the next stable core snap ... mvo should know when that happens
[07:49] <oSoMoN> good, I’ll ask mvo
[07:50] <oSoMoN> hrm, he doesn't appear to be around
[07:51] <ogra_> he sits next to me ... one sec
[07:51] <oSoMoN> heh, even better :)
[07:51] <ogra_> he says next week
[07:51] <oSoMoN> excellent, say hi for me btw
[08:02] <Laney> hi, happy friday!
[08:02] <pitti> happy Friday to you too!
[08:03] <didrocks> hey Laney! morning pitti
[08:03] <Laney> ah, pitti, a nice Friday treat!
[08:03] <Laney> how's it going?
[08:04] <Laney> hey didrocks, how are you?
[08:04] <didrocks> Laney: good, thanks! Yourself?
[08:07] <Laney> didrocks: good!
[08:08] <Laney> off on holiday next week ;-)
[08:08] <didrocks> nice! :)
[08:10] <seb128> hey Laney, how are you?
[08:10] <seb128> oh, off next week, any nice plan ?
[08:11] <seb128> going to surf? ;-)
[08:16] <Laney> seb128: I'm grrrrrrrrrrrrreat, what about you?
[08:16] <Laney> maybe, it is near a surfing place
[08:41] <seb128> Laney, I'm great as well, just got a nice cappuccino :-)
[08:41] <seb128> there is a new coffee place that opened just next street
[08:42] <seb128> easy walk to get some coffee to go, it's nice
[08:48] <willcooke> morning all
[08:48] <willcooke> broken down train today \o/
[08:49]  * willcooke also has an ethernet cable today 
[08:49] <flexiondotorg> tintou Hi
[08:49] <flexiondotorg> didrocks Morning
[08:50] <flexiondotorg> didrocks Have you seen any reports about desktop-gtk3 snaps not using localisation?
[08:50] <didrocks> good morning flexiondotorg
[08:50] <didrocks> flexiondotorg: I didn't, but if it's related to gettext, we know the issue and nothing that desktop-gtk3 can fix
[08:51] <flexiondotorg> OK. related to gettext how? Bug?
[08:51] <flexiondotorg> Anything an application author can work around?
[08:51] <didrocks> gettext doesn't let you using relocatable path
[08:51] <didrocks> it's based on --prefix
[08:51] <flexiondotorg> Oooh
[08:51] <didrocks> so, you can --prefix=/snap/<snap_name>/current
[08:51] <didrocks> we raised that multiple times on the ML
[08:52] <flexiondotorg> Right, I remember now.
[08:52] <didrocks> apparently, this path is supposed to be stay stable ^
[08:52] <flexiondotorg> I've used that myself.
[08:52] <didrocks> stay*
[08:52] <flexiondotorg> I'm sat with the elementary guys.
[08:52] <flexiondotorg> They've bumped into this.
[08:52] <didrocks> I hope it's cross-distro, we were told that shouldn't change
[08:52] <didrocks> and don't have any other way to make it work
[08:56] <flexiondotorg> didrocks Well, seem to be consistent on GNOME, Unity, MATE and Pantheon so far. I'll check with the KDE guys.
[09:21] <robert_ancell> Laney, great blog post!
[09:21] <robert_ancell> Sorry for ignoring everyone in here this week, been crazy busy here
[09:23] <seb128> robert_ancell, hey
[09:24] <seb128> good decisions?
[09:24] <seb128> what are the news on the desktop side
[09:30] <robert_ancell> seb128, we have a title field from snapd, so snaps should start looking nicer
[09:30] <robert_ancell> We have a plan for the license field to pass through, though we need to wait for support from the store
[09:31] <robert_ancell> We've fixed a bunch of small issues that block us from having Snap support in Fedora. These are mostly solved so just tidying up loose ends.
[09:31] <robert_ancell> KDE Discover is now using snapd-qt, and the MATE Software Botique is using snapd-glib GIR bindings. These shook out some bugs in snapd-glib.
[09:32] <seb128> nice
[09:32] <robert_ancell> We are all in agreement that snapd / store should return translated fields, though no ETA for it
[09:32] <robert_ancell> We have a plan to get basic AppStream information through the store and snapd so G-S etc can show things better.
[09:33] <robert_ancell> We'll make a snapcraft plugin to pull out that data and populate the Snap YAML - so it should be easier to snap existing projects
[09:33] <Laney> hey robert_ancell
[09:33] <Laney> thanks!
[09:34] <Laney> a summary mail next week would be nice ;-)
[09:34] <robert_ancell> We have a plan to make an snap interfaces tool that will be shipped as a Snap (i.e. available for all systems) and will be linked to from G-S and/or G-C-C. It's a highly questionable UI so there will be ongoing discussions I'm sure...
[09:34] <robert_ancell> Laney, yep, will do
[09:35] <robert_ancell> We have a plan to install/remove snaps on a GUI without a store login, which is awaiting confirmation by the Snappy designer
[09:40] <jamesh> willcooke: do you know if xdg-desktop-portal issues have been discussed at the sprint? (allowing per-user bind mounts)
[09:41] <willcooke> jamesh, not yet I'm afraid.
[09:41] <willcooke> I'll try and grab m_vo later on and see if he's got any thoughts
[09:42] <jamesh> willcooke: the main things I'd want are an agreement that snapd should switch to per-user per-snap mount namespaces, and a basic sketch of how they should be managed.
[09:43] <jamesh> willcooke: with that, it we might be able to help with the implementation
[09:43] <willcooke> jamesh, ack, thanls
[10:52] <jbicha> robert_ancell: gnome-software/zesty was just pushed to updates and got stuck in phased-updates again: https://errors.ubuntu.com/problem/f4d12cec512c23a979e3ccd1f0dc94fd8b406007
[10:53] <robert_ancell> jbicha, yeah, I saw :(
[10:53] <jbicha> anyway, for next week :|
[10:54] <robert_ancell> oh weird. An icon issue.
[10:54] <robert_ancell> I wonder if that's always existed
[10:55] <robert_ancell> Perhaps there's an out of memory issue and I'm not checking for the buffer being NULL
[10:56] <jbicha> I had someone from Azerbaijan tell me the polkit message was not Azerbaijani, it was Turkish
[10:58] <Laney> I think the problem we saw before is that it was picking the last message from the file
[10:59] <Laney> that was in Fedora - in Ubuntu we had some patch somewhere (awesome level of recall) that made the translations work
[11:10] <robert_ancell> Laney, I've been looking through the Polkit docs and existing programs and still scratching my head how the locale is supposed to work there
[11:10] <robert_ancell> It either "just works" for other prompts or they aren't currently translated
[11:19] <Laney> arhghghaoih
[11:19] <Laney> rob taB TAB TAB T
[12:29] <ricotz> oSoMoN, hi, fun fact about https://bugs.launchpad.net/bugs/1700692 -- it built on trusty ;)
[12:29] <ubot5> Ubuntu bug 1700692 in libreoffice (Ubuntu) "FTBFS on i386: dbaccess_RowSetClones unit test segfaults" [Critical,New]
[13:09] <oSoMoN> ricotz, interesting, do you know what kernel was used for the build?
[13:09] <oSoMoN> jbicha, btw did you see my comments on bug #1700692 ?
[13:09] <ubot5> bug 1700692 in libreoffice (Ubuntu) "FTBFS on i386: dbaccess_RowSetClones unit test segfaults" [Critical,New] https://launchpad.net/bugs/1700692
[13:15] <Laney> bah
[13:15] <Laney> this gst <-> packagekit stuff is annoying
[13:15] <Laney> gstreamer1.0-packagekit feeds different fields into the backend compared to sessioninstaller
[13:16]  * Laney cries at ximion 
[13:16] <ximion> Laney: is that a problem? as long as packagekitd does the right thing, it shouldn't be an issue...
[13:17] <Laney> it ends up in that regex of doom
[13:17] <ximion> or is the gstreamer plugin itself also broken?
[13:17] <ximion> yeah, urgh...
[13:17] <Laney> which isn't that tolerant
[13:17] <Laney> although in this case it's actually nicer
[13:17] <Laney> because gst-packagekit gives you "64bit" so we are able to use that to filter out some results
[13:20] <ximion> I wonder whether using libappstream + the new components in there would make some sense
[13:20] <ximion> (bypassing PKs own detection)
[13:20] <ximion> alternatively, making the regex more tolerant or using string operations instead might make sense
[13:22] <Laney> we need to specify AS so that type=codec is good enough to do this
[13:23] <Laney> have a look at apt show gstreamer1.0-plugins-bad - those caps are the things it searches on
[13:24] <jbicha> oSoMoN: could you ask someone that can update the LP builders about that then?
[13:25] <jbicha> oh, that's the stack clash regression
[13:25] <jbicha> maybe next week then
[13:30] <ricotz> oSoMoN, of course 4.4.0-81 like all builders ;)
[13:30] <ricotz> oSoMoN, maybe there is even another factor like glibc, but I guess the kernel is still the culprit after all
[13:31] <jbicha> ricotz: yes, there's a known regression affecting java
[13:32] <ricotz> jbicha, I know, surprisingly a build on trusty/i386 succeeded
[15:55] <flexiondotorg> seb128 didrocks Can you point me at the yaml for the current GNOME platform snap?
[15:55] <flexiondotorg> ANd an example application snapcraft.yaml that uses it please?
[15:56] <seb128> flexiondotorg, ken has been working on that but
[15:56] <seb128> http://bazaar.launchpad.net/~ubuntu-desktop/+junk/snap-gnome-3-24/view/head:/snapcraft.yaml
[15:57] <seb128> and http://bazaar.launchpad.net/~ubuntu-desktop/+junk/gedit/view/head:/snapcraft.yaml
[15:57] <didrocks> and using it is like any other traditional app using the launcher: http://bazaar.launchpad.net/~ubuntu-desktop/+junk/ghex-snap-gnome-3-24/view/head:/snapcraft.yaml
[15:59] <didrocks> hum, ken is adding another path for gsettings: http://bazaar.launchpad.net/~ubuntu-desktop/+junk/gedit/view/head:/snapcraft.yaml#L44
[15:59] <oSoMoN> chrisccoulson, hey, have you had a chance to test chromium-browser 59.0.3071.109 in https://launchpad.net/~canonical-chromium-builds/+archive/ubuntu/stage/+packages ?
[15:59]  * didrocks looks at the launcher, we should probably factor that in
[15:59] <didrocks> GS_SCHEMA_DIR=$XDG_DATA_HOME/glib-2.0/schemas
[16:00] <didrocks> I bet the additional export doesn't work ^
[16:00] <didrocks> kenvandine[m][m]: FYI ^
[16:00] <flexiondotorg> seb128 didrocks Thanks! Sat here with elementary.
[16:00] <seb128> are they snapping things?
[16:09] <flexiondotorg> seb128 Yes.
[16:09] <flexiondotorg> ANd we are now looking at using a content snap for their SDK.
[16:33] <kenvandine> seb128, problems with the gedit snap?
[16:39]  * oSoMoN ⒺⓄⓌ
[16:39] <oSoMoN> have a great week-end everyone!
[17:10] <andyrock> have a nice we :D
[17:16] <seb128> have a nice w.e everyone
[17:16] <seb128> & good holidays Laney
[17:16] <Laney> laters seb128!
[17:16] <Laney> happy week to you
[17:16] <seb128> kenvandine, not that I know, flexiondotorg was asking for an example for snap using the platform
[17:16] <kenvandine> seb128, ok
[17:17] <seb128> Didier was pointing something, maybe ask him on monday if you don't understand what he meant
[17:17] <seb128> I'm not sure what was the issue
[17:17] <kenvandine> it was working fine
[17:17] <kenvandine> yeah, saw that but didn't understand
[17:17] <seb128> I think he just pointed out that your env hack was unrequired/probably not doing anythin
[17:17] <kenvandine> it was needed :)
[17:17] <seb128> like the desktop helper set that
[17:17] <seb128> or should set
[17:17] <seb128> or override your def
[17:17] <seb128> well then the launcher should be fixe
[17:17] <seb128> d
[17:17] <kenvandine> desktop helper probably would be better
[17:17] <seb128> yes
[17:18] <seb128> that's probably what he meant
[17:18] <seb128> that's supposed to work without the hack
[17:18] <seb128> so if that's needed there is a bug in the helper we should look at
[17:23] <Laney> night! see you in a week!
[17:23] <Laney> break all the things
=== Cimi_ is now known as cimi
[20:04] <kenvandine> sarnold, any updates on the gdm3 security review?
[22:38] <twoj__> Hello,   I have a unity desktop question...  any takers?
=== JanC_ is now known as JanC
[23:46] <sarnold> hey kenvandine, still in progress. I gotta say I wish we were sticking with lightdm.