[05:56] <hikiko> hello
[08:01] <desrt> moin, desktop
[08:11] <xnox> what is gnome-software written in? and why its gapplication-service is the second largest consumer of standby RAM on my laptop in zesty?
[08:11]  * xnox is grumpy this morning
[08:12] <happyaron> morning desrt xnox seb128 pitti
[08:13] <xnox> also, i'm still questioning my intentions of updating to zesty during a sprint
[08:18] <desrt> xnox: big database of apps?
[08:18] <desrt> xnox: heavy toolkit user?
[08:19] <xnox> ¯\_(ツ)_/¯
[08:19] <xnox> i think i am just a normal laptop user, with uefi capsule support from dell
[08:19] <xnox> i only have 3.6k packages installed
[08:38] <seb128> hey desrt xnox happyaron
[08:38] <desrt> our fearless (acting) leader!
[08:38] <desrt> morning, seb :)
[08:38] <seb128> xnox, g-s is written in C
[08:38] <xnox> horum. i wonder if the memory size in system monitor is a lie
[08:39] <xnox> and it's just all the shared libraries of the world (e.g. webkit)
[08:39] <seb128> it has different columns
[08:39] <seb128> "memory" should be the correct one
[08:40] <seb128> it has different values for shared, etc
[08:40] <desrt> look for writable
[08:40] <desrt> everything else, really, is just cache....
[08:41] <desrt> (ie: non-writable segments are just bits of shared libraries from the page cache... easily evicted and read back again if needed)
[08:45] <xnox> 30M shared, 107M memory
[08:45] <xnox> so there is 100M of runtime app usage, it would seem.
[08:46] <desrt> ... look specifically for something that tells you how much "writable" memory
[08:47] <seb128> desrt, I think that's what the "memory" column is in g-s-m no?
[08:47] <desrt> in that case, don't subtract :p
[08:47] <seb128> right
[08:48] <seb128> like virtual memory for g-s is 180M here and memory is 60M
[08:48] <desrt> highest users here: chromium, telegram(!), xorg, shell...
[08:48] <desrt> as expected, more or less
[08:49] <desrt> gnome-software is only using 18MB for me...
[08:49] <seb128> xnox, but yeah it's annoying that it stays around as a service but it was needed for notifying about firmware updates and other things
[08:50]  * desrt looks slightly guilty
 run the app forever!  if it needs to, it can clean up its static state before going into the background to save memory
[08:51] <desrt> <every library ever> have some static state!  if you want to clear it, you can surely just exit(), right?
[08:51] <desrt> <c language> hey guys!  have some memory fragmentation!!
[08:52] <desrt> so ya.......
[08:52] <desrt> i underestimated how serious this problem would be
[08:52] <xnox> i will try to remove the firmware update plugin, restart gnome-software, and will check if that solves things.
[08:53] <xnox> ..
[08:53] <xnox> libgl1-mesa-glx libglapi-mesa from zesty-proposed results in ubiquity ftbfs (everything else is fine, the rest of the mesa update)
[08:54] <xnox> Segmentation fault (core dumped)
[08:54] <xnox> xauth/Xrdb output:
[08:54] <xnox> as in Xvfb fails
[08:54] <desrt> xnox: i think gnome-software also puts a listener for apt updates
[08:55] <desrt> since there is, otherwise, nothing to trigger something in the logged-in session on the basis of a system-level event
[08:55] <xnox> is this known that src:mesa in zesty-proposed regresses xvfb?
[08:56] <desrt> really, though: i think gnome-software is written poorly in terms of the never-exit thing... if it wants to sit around waiting for a couple of things, it should have a separate (tiny) process that does that, and signals the primary process when something happens
[08:56] <seb128> xnox, tjaalton might know
[08:56] <desrt> gapplication is all about activating in response to events... there really is no excuse to have this thing long-running like it is
[08:57] <desrt> of course, this is probably quite disruptive to its current plugin model :/
[08:57] <seb128> desrt, xnox, right, disabling the firmware plugin isn't going to solve anything. It has been written to start with the session and stick around as a service, it's just that we discussed changing that but the people working on g-s in Ubuntu had a consensus against because the service is sometime useful (like for firmwares) and g-s was getting in a weird state when exiting and restarting
[08:57] <seb128> It= g-s
[08:58] <desrt> imho this is something that should be solved properly
[08:58] <xnox> https://bugs.launchpad.net/ubuntu/+source/mesa/+bug/1647600
[08:59] <desrt> unfortunately, the "proper" fix probably means that every plugin, in effect, becomes two plugins
[08:59] <xnox> seb128, i'm ok with gnome-software hanging around forever
[08:59] <xnox> we already have udpate-manager hanging around forever, can that poke g-s when things change?
[08:59] <xnox> i'm wondering if the firmware plugin is the memory hog.
[08:59] <seb128> xnox, bug #1646064 might be the same?
[08:59] <desrt> xnox: the memory hog is gtk
[08:59] <desrt> (and webkit, and ....)
[08:59] <xnox> if gnome-software would use less memory i would be fine with it.
[08:59] <seb128> right, what desrt is saying
[08:59] <xnox> seb128, no, because infinity fixed glibc yesterday and glibc is irrelevant.
[09:00] <desrt> the problem is, after the first time you show the UI, there is an awful lot of memory that gets used and never freed
[09:00] <seb128> we just keep a service which is the full application including all the UI librarires
[09:00] <desrt> all of our UI stack is made with the assumption that apps exit when they exit
[09:00] <desrt> so basically, the process that links webkit, gtk, etc. needs to call exit()
[09:00] <flexiondotorg> Good morning seb128 desrt happyaron hikiko :-)
[09:00] <seb128> hey flexiondotorg, how are you?
[09:00] <desrt> flexiondotorg: morning :D
[09:00] <hikiko> hi all :p
[09:01] <seb128> hey hikiko
[09:01] <desrt> hikiko: hey :)
[09:01] <flexiondotorg> seb128, Very well thanks.
[09:01] <desrt> i think i'll take a note to have a chat with hughsie later today
[09:01] <seb128> xnox, see the comment from dobey on the glibc bug, I meant that one looks like your issue
[09:01] <flexiondotorg> Despite staying up too late to tinker with Pi projects :-)
[09:01] <desrt> this has been something that has been bothering me for a while...
[09:01] <seb128> xnox, which might be different from whatever glibc issue there was
[09:01] <seb128> desrt, thanks, let me know how the discussion goes
[09:02] <seb128> flexiondotorg, late night hacking is fun, next mornings are less right?
[09:02] <seb128> :-)
[09:02] <desrt> would be interesting to have a simple activation framework for monitoring system events in a generic way
[09:02] <desrt> like if you could ask systemd to start your process if /var/apt/whatever got an inotify event, or in response to a particular signal sent to the system dbus
[09:02] <flexiondotorg> seb128, Yeah, I didn't make the school run this morning.
[09:03] <seb128> desrt, can't you already do that sort of things with systemd jobs?
[09:03] <flexiondotorg> And hacking ended up being playing classic arcade games from the 80s :-D
[09:03] <xnox> desrt, there is .path units in systemd, and thus in user-session systemd can totally spawn / poke things.
[09:03] <seb128> lol
[09:03] <Laney> morning
[09:03] <seb128> hey Laney
[09:03] <desrt> Laney: hey :)
[09:03] <xnox> note it may not know which session one is in =) or which display to use for popups
[09:03] <desrt> ya... that's one problem
[09:03] <Laney> just went to pick up my laptop but was holding a tea and poured it out
[09:03] <Laney> /o\
[09:04] <desrt> this user vs. session thing is a drag
[09:04] <desrt> Laney: floor or laptop?
[09:04] <flexiondotorg> Morning Laney
[09:04] <seb128> Laney, you didn't poor it on the laptop at least?
[09:04] <seb128> pour
[09:04] <flexiondotorg> Poor laptop?
[09:04] <desrt> xnox: so like, i bet we could use this to replace whatever monitoring the apt backend is doing
[09:04] <seb128> hehe
[09:05] <Laney> no
[09:05] <Laney> other hand
[09:06] <desrt> .... better than laptop. at least?
[09:06] <Laney> sofa / floor / me
[09:06] <desrt> (unless you fancied your hand, but wanted a new laptop...)
[09:06] <Laney> could be worse
[09:06] <Laney> gnome-software calls a 'refresh' function on each plugin
[09:06] <Laney> at an interval
[09:07] <desrt> ...
[09:07] <Laney> you should go have this chat in #gnome-software rather than #ubuntu-desktop
[09:07] <desrt> (╯°□°）╯︵ ┻━┻
[09:22] <davmor2> Morning all
[09:23] <seb128> hey davmor2
[09:52] <flexiondotorg> davmor2, Morning
[10:24] <flexiondotorg> seb128, Laney Have you got a few minutes to discuss a couple of HiDPI issues raised by System 76?
[10:24] <seb128> flexiondotorg, sure
[10:25] <flexiondotorg> I'll get to Unity scale-factor in bit. I'll start with console fonts.
[10:26] <flexiondotorg> System 76 have tweaked the console fonts in 16.04 to used the Terminus font at FONTSIZE="16x32"
[10:26] <flexiondotorg> Rather than the Fixed font at FONTSIZE="8x16"
[10:27] <flexiondotorg> Would it be possible to make Termius at the higher size the default?
[10:28] <flexiondotorg> Is it possible to retrospectively adjust this in 16.04?
[10:28] <flexiondotorg> My concern is, this will also affect server.
[10:28] <flexiondotorg> And may not be desirable.
[10:29] <seb128> you/they should open a bug about that with the suggested changes and the rational
[10:29] <seb128> not sure who would be right to comment on the issue but I don't know enough about that to reply without reading/investigating
[10:29] <flexiondotorg> OK.
[10:29] <seb128> Gunnar might have an opinion
[10:30] <flexiondotorg> I'll ask them to file a bug with rationale.
[10:31] <seb128> thanks
[10:31] <flexiondotorg> seb128, We chatted earlier about the MR for Unity7 that will set the scale-factor on HiDPI systems.
[10:31] <flexiondotorg> https://code.launchpad.net/~kaihengfeng/unity/set-hidpi-scale-factor/+merge/299380
[10:32] <flexiondotorg> That has landed in 17.04.
[10:32] <seb128> right
[10:32] <flexiondotorg> I'#m going to test it this morning.
[10:32] <seb128> great
[10:32] <flexiondotorg> seb128, What concerns do you have about back porting this to 16.04?
[10:33] <seb128> not much, I was just wondering if unity was the right place to do it where unity-settings-daemon had already code for that which got disabled
[10:33] <seb128> but the unity scaling is per screen
[10:34] <seb128> though at the end we only have one value for gtk
[10:34] <seb128> but if andyrock & Trevinho judged it's right to do it in unity and it works no objection from me
[10:34] <seb128> so yeah, please test, if that works well let's land that rather than do more work
[10:34] <flexiondotorg> So I've asked System 76 to test current 17.04 images.
[10:35] <seb128> great
[10:35] <Trevinho> seb128, flexiondotorg: hey!
[10:35] <flexiondotorg> I'm going to do that same on my XPS 15.
[10:35] <seb128> hey Trevinho :-)
[10:35] <flexiondotorg> Trevinho, o/
[10:35] <Trevinho> seb128: so, I would have preferred to keep some things inside ups but...
[10:35] <seb128> usd you mean?
[10:35] <Trevinho> yeah, sorrry
[10:35] <Trevinho> but... That's a global option as said
[10:36] <flexiondotorg> I can see the benefit of having the logic in usd.
[10:36] <flexiondotorg> For example, OEM first boot doesn't currently auto scale on HiDPI systems.
[10:36] <Trevinho> and at the same time we probably should have moved all this kind of stuff to usd in the first place... But at this point we accpeted it to be inside unity
[10:36] <seb128> right need to look at that
[10:36] <flexiondotorg> But usd is laucnhed via ubiquity-dm
[10:36] <seb128> usd has code for it and we use it in the greeter
[10:36] <Trevinho> flexiondotorg: that's a different thing I think
[10:36] <seb128> we should perhaps reverse the logic to be !unity
[10:36] <andyrock> morning
[10:36] <Trevinho> flexiondotorg: as OEM install isn't using unity, isn't it?
[10:37] <seb128> hey andyrock
[10:37] <flexiondotorg> Trevinho, correct.
[10:37] <Laney> only-ubiquity mode does scale though
[10:37] <flexiondotorg> Ubiquity has it's own display manager and executes usd.
[10:37] <andyrock> does anyone know how to block spam on freenode?
[10:37] <andyrock> :D
[10:38] <seb128> sorry but need to step out for a bit
[10:38] <seb128> going to read the backlog and comment once I'm back
[10:38] <flexiondotorg> andyrock, o/
[10:38] <seb128> (if there anything else needed from me in the discussion)
[10:38] <flexiondotorg> seb128, OK
[10:38] <Laney> If usd doesn't set up scaling, how does that work?
[10:39] <flexiondotorg> Laney, I'm saying OEM first boot doesn't scale.
[10:39] <flexiondotorg> But if the scale-factor logic were in usd, then OEM first boot would scale.
[10:40] <Laney> I know
[10:40] <Laney> I'm asking why it does work for the only-ubiquity mode on live CDs
[10:41] <flexiondotorg> Yes ^
[10:41] <flexiondotorg> What I'm going to test.
[10:41] <Laney> Maybe metacity sets this up itself
[10:41] <flexiondotorg> I'd like to understand the difference and try to find a solution.
[10:43]  * desrt gets upset with ld -pie
[10:47] <flexiondotorg> Laney, so metacity is in ubuntu-live
[10:47] <Laney> That's right
[10:47] <flexiondotorg> And in ubiquity-dm metacity will be executed in preference to compiz.
[10:48] <Laney> Correct
[10:48] <flexiondotorg> Do you know the history on that?
[10:48] <Laney> Not really, just that compiz didn't work properly for some reason and metacity was ok
[10:48] <flexiondotorg> OK, so something to test is if Compiz does work now.
[10:48] <Laney> you will notice that I did some hidpi fixes in metacity a little while ago
[10:49] <flexiondotorg> Yep.
[10:50] <Laney> Not sure what's to be gained by trying to change that
[10:51] <Laney> but if you want to own any resulting problems...
[10:51] <flexiondotorg> Unity 7 (in Zesty) can set scale-factor when HiDPI is detected.
[10:52] <Laney> It works in metacity already.
[10:52] <flexiondotorg> Except is OEM first-boot.
[10:52] <Laney> Yes, but it works in only-ubiquity
[10:52] <Laney> so there is no point in ripping it out of there
[10:52] <Laney> find out *why* it does and then replicate that
[10:52] <flexiondotorg> Agreed.
[10:53] <flexiondotorg> But I'm curious to see if executing compiz from ubiquity-dm "does the right thing".
[10:53] <flexiondotorg> In OEM first-boot with the scale patch applied.
[10:56] <Laney> If you really want to, then that's up to you
[11:01]  * Laney needz beanz
[12:31] <Laney> flexiondotorg: u-s-d does the autocalculation on non-Unity
[12:32] <flexiondotorg> Laney, OK.
[12:33] <flexiondotorg> Will test after lunch.
[15:09] <dobey> seb128, xnox: any luck on the mesa issue?
[15:09] <xnox> dobey, i know it breaks things
[15:09] <xnox> dobey, seb128: i want src:mesa to be removed from proposed
[15:10] <xnox> it is blocking a lot of things
[15:10] <dobey> yeah, it breaks anything that uses gtk+ at least; i'm sure not everything it breaks has an autopkgtest that hits that issue
[15:11] <seb128> tjaalton, hey, is ^ a known issue/being worked on?
[15:12] <dobey> it's blocking us landing a fix for (at least) indicator-keyboard for the u8 snap too
[15:12] <seb128> well, proper way is to fix the bug
[15:13] <seb128> not to delete an update that is going to need to be redone/debugged at some point anyway
[15:13] <dobey> of course
[15:13] <dobey> i got a stack trace pointing to where it's breaking, thought someone who deals with mesa would be fixing it
[15:25] <seb128> dobey, that someone is tjaalton and he doesn't seem to be around, doesn't mean others can't step up and try to fix it though
[15:30] <desrt> hihi
[15:30] <qengho> hi!
[15:30] <qengho> Will won't.
[15:30] <seb128> #startmeeting
[15:30] <meetingology> Meeting started Tue Dec  6 15:30:53 2016 UTC.  The chair is seb128. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
[15:30] <meetingology> Available commands: action commands idea info link nick
[15:30] <desrt> it's on seb128 to save us!
[15:32] <seb128> hey everyone
[15:32] <desrt> seb128: rollcall helps
[15:32] <seb128> desrt, I know, was checking email to know who is out ;-)
[15:32] <seb128> Roll call: andyrock, attente, desrt,  dgadomski, fjkong (out), flexiondotorg, happyaron, hikiko, laney, qengho, seb128, sweet5hark, themuso (out), tkamppeter, trevinho, robert_ancell (out)
[15:32] <Sweet5hark> .
[15:32] <flexiondotorg> 'elo
[15:32] <andyrock> hey
[15:33]  * FJKong wave hello
[15:33] <seb128> indeed, rollcall works!
[15:33] <Laney> hi
[15:33] <seb128> FJKong, hey, didn't you say you were out on tuesday?
[15:33] <desrt> 'sup
[15:33] <seb128> k, let's get started
[15:33] <seb128> #topic andyrock
[15:33] <seb128> andyrock, hey
[15:33] <andyrock> hey
[15:34] <andyrock> I'm still working on adding the windows list to the unity8 quicklist
[15:34] <andyrock> the code is almost there but I'm having issues testing it
[15:34] <andyrock> coz I can no longer install webbrowser-app
[15:35] <andyrock> did some reviews also
[15:35] <andyrock> eow
[15:35] <seb128> thanks andyrock
[15:36] <seb128> andyrock, is that using the overlay ppa? I think they had an issue with oxide and a security update overriding it but a silo fixing those
[15:36] <seb128> #topic attente
[15:36] <seb128> attente, hey
[15:36] <attente> hi seb128
[15:36] <attente> bug fixes to the mir backend upstream for gtk-4
[15:37] <attente> also tried debugging a libtool/rpath problem but not much luck there
[15:37] <attente> filed an RT to fix an issue with the integration tests in the snapcraft jhbuild plugin
[15:37] <attente> title bars on separate child windows discussion with upstream
[15:37] <attente> content-hub-glib branch landed in z, so could use someone to review wip/attente/mir-content-hub on git.gnome.org. considering just merging it though...
[15:37] <andyrock> seb128: do you know which silo?
[15:38] <seb128> andyrock, no, but maybe it landed since?
[15:38] <seb128> attente, did you try desrt / robert_ancell / Laney for reviewing the gtk-mir change?
[15:39] <seb128> one of them might be able to find some review cycles for you?
[15:39] <attente> i haven't asked anyone yet :)
[15:39] <seb128> k, well I sort of just did for you :p
[15:39] <seb128> thanks attente!
[15:39] <attente> desrt: do you want to try? ^
[15:39] <seb128> #topic desrt
[15:39] <seb128> desrt, hey
[15:39] <desrt> hey.  not a lot to say.  work continues on the proxy.  got some cleanups done vs. last week
[15:39] <desrt> that's all
[15:40] <andyrock> https://www.irccloud.com/pastebin/irRFZIiQ/
[15:40] <andyrock> seb128: ^^^
[15:40] <seb128> andyrock, let's move that to #ubuntu-unity
[15:40] <seb128> thanks desrt
[15:40] <seb128> #topic dgadomski
[15:40] <dgadomski> `hey
[15:40] <seb128> dgadomski, hey
[15:40] <dgadomski> * focusing on #1550983. Looks like the problem is in gtk itself or it's packaging. I would appreciate an opinion on this (last comment in the bug).
[15:40] <dgadomski> bug #155098 that is :)
[15:41] <dgadomski> sorry, bug #1550983
[15:41] <dgadomski> that's the one
[15:41] <dgadomski> eof
[15:41] <seb128> dgadomski, let's have a look after the meeting
[15:41] <seb128> dgadomski, thanks
[15:41] <dgadomski> thanks
[15:41] <seb128> #topic FJKong
[15:41] <FJKong> hi
[15:41] <FJKong> bug fixed:
[15:41] <FJKong> #1575350 #847087 Gnome Software catalog entry missing for GNU Radio
[15:41] <FJKong> #1634226 Gnome Software catalog entry missing for D-rats
[15:42] <FJKong> still in progress
[15:42] <FJKong> #1498972 Change icon to 256x256 version to support HiDPI
[15:42] <FJKong> #1639863 Firefox and Thunderbird don't appear in the (new) appstream metadata
[15:42] <FJKong> eof
[15:42] <seb128> thanks FJKong
[15:42] <seb128> #topic flexiondotorg
[15:42] <seb128> flexiondotorg, hey
[15:42] <flexiondotorg> Time appropriate greetings!
[15:42] <flexiondotorg> · Researched this weeks candidates for the Snap Upstream Blitz and update Trello
[15:42] <flexiondotorg> · Completed new joiner induction at BlueFin on Thursday.
[15:43] <flexiondotorg> · Worked on fixes for `source-type: deb` in Snapcraft. Discovered while working with an ISV.
[15:43] <flexiondotorg> · Video conference with System 76 to understand their issues with HiDPI on 16.04. They have started tagging bugs here https://bugs.launchpad.net/ubuntu/+bugs?field.tag=hidpi
[15:43] <flexiondotorg> · Working on some follow up actions as a result of the call with System 76, including how the can complete early testing and QA for the 16.04.2 HWE kernel on their range of computers.
[15:43] <flexiondotorg> · Thanks to andyrock and trevinho for the Unity 7 scale factor merge in Zesty. System 76 are testing.
[15:43] <flexiondotorg> · Snap advocacy video conference with Plone and out reach to many other ISVs and upstreams. Daily standup video conferences now.
[15:43] <flexiondotorg> 🎬
[15:44] <seb128> flexiondotorg, thanks!
[15:44] <seb128> #topic happyaron
[15:44] <seb128> happyaron, hey
[15:45] <seb128> oh, he sent me notes
[15:45] <seb128> 1. ibus-table/1.9.1-4
[15:45] <seb128> 2. libpinyin/1.3.0-3
[15:45] <seb128> 3. iptraf-ng RC bug fix
[15:45] <seb128> 4. opencc SRU to yakkety
[15:45] <seb128> 5. landed n-m in zesty
[15:45] <seb128> 6. prepared n-m SRU for xenial
[15:45] <seb128> thanks happyaron ;-)
[15:45] <seb128> #topic hikiko
[15:45] <seb128> hikiko, hey!
[15:45] <hikiko> hi
[15:45] <hikiko> worked on chromium + lowgfx, but mostly on chromium:
[15:45] <hikiko> * learned how to use ninja, gn and other google stuff,
[15:45] <hikiko> * started the code in the mir backend,
[15:45] <hikiko> * I am still studying ozone, the mir client API and the wayland backend: my first "goal" is to be able to draw on a mir surface.
[15:46] <qengho> So sorry.
[15:46] <hikiko> oh also spent a several amount of time trying to accelerate the build
[15:46] <qengho> Also sorry.
[15:46] <hikiko> lol qengho
[15:46] <hikiko> eof :)
[15:46] <seb128> thanks hikiko
[15:46] <seb128> #topic Laney
[15:46] <seb128> Laney, hey
[15:46] <Laney> yo
[15:46] <Laney> • lame week
[15:46] <Laney> • mostly worked on a big transition which keeps getting set back, now I'm casting around in #ubuntu-release for help, hard to tell if it's close or not
[15:46] <Laney> • some random updates: gstreamer 1.10.2, appstream-glib
[15:47] <Laney> • fixed a g-s 3.22 bug breaking installation of packages
[15:47] <Laney> 🙇
[15:47] <tjaalton> seb128: i'm out until thursday
[15:48] <seb128> tjaalton, k, enjoy the time off!
[15:48] <seb128> Laney, sorry I was in middle of other things yesterday and didn't really understand your request and then forgot to get back to you and earlier infinity was active so I though he would do it, remind me what you need after the meeting and I can do it
[15:48] <seb128> Laney, thanks
[15:49] <seb128> #topic qengho
[15:49] <seb128> qengho, hey
[15:49] <qengho> Hi!
[15:49] <qengho> * Building and testing new upstream Cr release. Cr 55.
[15:49] <qengho> * printf-debugging a crasher.
[15:49] <qengho> * Refreshed tor snap. Upstream release. Snaps are so nice.
[15:49] <qengho> EOF
[15:49] <Laney> seb128: no worry, I listed the request earlier in #ubuntu-release
[15:49] <seb128> thanks qengho
[15:49] <Laney> doesn't have to be you all the time
[15:50] <seb128> #topic seb128
[15:50] <Laney> thanks for your previous help ;-)
[15:50] <seb128> yw!
[15:50] <seb128> • looked at some network-manager issues, including zesty and xenial regressions, talked to people who know about those changes to try to get things moving
[15:50] <seb128> • triaged a bit nautilus bugs and talked to upstream about some of the issues
[15:50] <seb128> • reviewed recent launchpad bugs & e.u.c reports
[15:50] <seb128> • some archive admin work
[15:50] <seb128> • starting a sponsoring shift
[15:50] <seb128> • snappy work
[15:50] <seb128> ∘ tried to build yakkety snaps on launchpad to backport on xenial but hit build issues, wasted some time trying to debug/workaround those, turned out to be a known launchpad bug and not easy to workaround
[15:50] <seb128> ∘ discussed several issues & howtodothings with Didier
[15:50] <seb128> ∘ looked at bit more at what kde did with their platform sharing

[15:50] <seb128> #topic Sweet5hark
[15:50] <seb128> Sweet5hark, hey
[15:50] <Sweet5hark> - 5.3 beta1 snap build starts ...
[15:50] <Sweet5hark> - ... but fails on weird linking issue: missing some harfbuzz/icu symbols, need to investigate
[15:50] <Sweet5hark> - writer refactoring and cleanup (got 7 patches in, 3 more pending)
[15:50] <Sweet5hark> - reviewed and merged some patches by others
[15:50] <Sweet5hark> - TDF: 2017 budgeting
[15:50] <Sweet5hark> - some team management (travel, events)
[15:50] <Sweet5hark> - some FOSDEM prep
[15:50] <Sweet5hark> EOF
[15:51] <seb128> thanks Sweet5hark
[15:51] <seb128> #topic TheMuso
[15:51] <seb128> * Had someone comment on bug 1574324 about the latest puse SRU fix for that bug breaking bluetooth for them... *sigh* there had to be some hardware that would be worse with that change wouldn't there. :S Need to look at the log they provided, and go from there. Not going to mark as regression yet, because this fixes things more than it does break things, based on user feedback.
[15:51] <seb128> * Discussions about input events and accessibility begun, but decided to talk to upstrea accessibility devs first, and based on that conversation, it is something that needs to be written up in a spec, and will likely require accessibility dev involvement in upstrea mprojects to implement it. I've got some questions to run by them, but other than that I will have to look at writing the spec myself, since we are needing it sooner rather than l
[15:51] <seb128> ater.
[15:51] <seb128> * Migration to the use of espeak-ng in main continues, most of the work being done in Debian, at which point we can sync/merge when everything is ready, and it will be a matter of filing a MIR once everything is built and available in universe... Part of the process has been to separate out as much as possible into separate packages, which brings multiple benefits anyway.
[15:51] <seb128> * Further testing for bug 1641954, and can reproduce, so will proceed with the SRU in the coming week. THis is only a problem in Xenial, since the fix is in newer versions of pulseaudio.
[15:51] <seb128> #topic tkamppeter
[15:52] <tkamppeter> - cups-filters: Investigation of crashes of cups-browsed, probably will need locks due to parallel tasks/threads.
[15:52] <tkamppeter> - CUPS: Investigated a problem of shutdown with systemd.
[15:52] <tkamppeter> - OpenPrinting: Investigated a bug in adding new PPDs and printer entries.
[15:52] <tkamppeter> - Bugs.
[15:52] <seb128> tkamppeter, hey, you sent me a summary by email but seems you are online? should I copy it or...
[15:52] <seb128> seems you are around
[15:52] <seb128> tkamppeter, thanks
[15:52] <tkamppeter> No, it is the same. Thanks.
[15:52] <seb128> yw!
[15:52] <seb128> #topic Trevinho
[15:52] <seb128> Trevinho, hello
[15:52] <Trevinho> · More work in unity8 indicators pointer/touch interactions
[15:52] <Trevinho> · Fixed a failure in ubuntu settings components tests due to JS date issues
[15:52] <Trevinho> · Fixed unity7 to compile with gtest-1.8
[15:52] <Trevinho> · Landed new unity7 with various fixes
[15:52] <Trevinho> · Fixed libappindicator to use proper realpath for icons and themes in $SNAP
[15:52] <Trevinho> · Various tunings, new tests and cleanups to my snapcraft PRs (now all merged)
[15:52] <Trevinho> · Few tunings to the Remmina snap
[15:53] <Trevinho> · Written a snapcraft ML post to explain how to use app-indicators from $SNAP
[15:53] <Trevinho> EOF
[15:53] <seb128> Trevinho, thanks!
[15:53] <seb128> #topic robert_ancell
[15:53] <seb128> - snapd-glib 1.4 released and SRU'd
[15:53] <seb128> - Relocation improvements in LightDM
[15:53] <seb128> - (Away two days)
[15:53] <seb128> #topic aob
[15:53] <seb128> other topics anyone?
[15:53] <seb128> did I forget anybody?
[15:55] <seb128> k, seems everybody is quiet ;-)
[15:55] <seb128> should do for this week then
[15:55] <seb128> let's call it a wrap, thanks everyone
[15:55] <attente> thanks seb128
[15:55] <Laney> great meeting
[15:55] <attente> 5/5
[15:55]  * Laney gets a tattoo of 2016-12-06
[15:56] <seb128> #endmeeting
[15:56] <meetingology> Meeting ended Tue Dec  6 15:56:01 2016 UTC.
[15:56] <meetingology> Minutes:        http://ubottu.com/meetingology/logs/ubuntu-desktop/2016/ubuntu-desktop.2016-12-06-15.30.moin.txt
[15:56] <seb128> Laney, k, let me have a look to your earlier requests
[15:56] <seb128> dgadomski, then I look at your bug
[15:56] <dgadomski> seb128: great, thank you!
[16:05] <seb128> Laney, doing the requests in order, gyoto deleted but I let you deal with doing a copy or upload of the old version
[16:07] <Laney> seb128: ok will do, if you're doing it in order then qgis was first ;-)
[16:07] <seb128> Laney, right, did that but I had no comment about that one ;-)
[16:07] <Laney> k, just wanted to make sure it wasn't missed
[16:08] <seb128> right
[16:16] <seb128> Laney, k, I think I got the things you asked for now, let's see at the next round
[16:17] <Laney> seb128: thanks!
[16:19] <seb128> yw!
[16:19] <seb128> andyrock, got things to work without proposed?
[16:25] <seb128> dgadomski, can you reproduce that libgl issue?
[16:28] <dgadomski> seb128: yes, easily, my only concern is whether gtk should have an explicit runtime dependency on libGL.so or should it check in runtime before making the mentioned call
[16:28] <seb128> dgadomski, weird, is that amd64 specific?
[16:29] <seb128> I'm on i386 but with a pbuilder I don't have libGL.so.1 installed and gtk binaries start without issue
[16:30] <dgadomski> seb128: the way I was able to reproduce it was to run it remotely (ssh -X) on xenial from a Windows host (cygwin with X)
[16:30] <dgadomski> from a linux client it worked flawlessly - is it possible that it was due to my local libGL.so?
[16:31] <seb128> well, started from ssh it has access to the same fs no?
[16:32] <dgadomski> tbh I'm not sure what happens under the good of gtk with ssh -X
[16:35] <dgadomski> but installing libgl1-mesa-glx on the remote host fixes it for non-Linux hosts (I was told that it also happens on OS X)
[16:36] <seb128> k, could be a gtk upstream bug
[16:45] <seb128> dgadomski, but yeah, that symbol seems to be from libGL so unsure why there is no depends generated
[16:47] <dgadomski> seb128: there is a decent number of libGL calls inside gdk/x11/gdkglcontext-x11.c so it's really weird that nothing is linked against it
[16:48] <seb128> dgadomski, libGL is diverted by e.g the nvidia drivers so I don't know if that has to do with it...
[16:50] <dgadomski> I'll file an upstream gtk bug and see what's their opinion about this
[16:51] <seb128> it might not be an upstream issue if it's a missing depends
[16:51] <seb128> you can maybe check if Debian has the same issue and report it there
[16:52] <dgadomski> seb128: ok, I'll try to reproduce it with Debian, thanks for taking a look
[17:04] <seb128> dgadomski, yw
[17:39] <seb128> cyphermox, hey, just you case you didn't know you got a follow up on bug #1629611 including debug log
[17:56] <seb128> Laney, update_excuses refreshed but unsure what bits you are watching? gdal is still blocked due to hdf being not considered on failing tests
[17:56] <Laney> not looking at tests yet
[17:57] <Laney> the queue is mad
[17:57] <seb128> k
[17:57] <Laney> been working from http://people.canonical.com/~ubuntu-archive/proposed-migration/zesty/update_output_notest.txt
[17:57] <seb128> k
[17:57] <Laney> although the generated hint misses some packages
[17:57] <Laney> so you have to start the hint tester and add those
[17:57] <seb128> the situation is too complex for me to understand
[17:57] <Laney> glewmx and r-base
[17:58] <seb128> but let me know if you need more pocking at things
[17:58] <Laney> yeah it's great!
[17:58] <Laney> will do, thanks
[17:58] <seb128> yw
[17:58] <Laney> gyoto didn't come up yet
[17:59] <Laney> see how it is tomorrow
[18:02]  * Laney wants to move autopkgtest to lxd where possible (not isolation-machine)
[18:02] <Laney> check armhf http://autopkgtest.ubuntu.com/running
[18:02] <Laney> way faster
[18:05] <Laney> goodnight!
[18:35] <seb128> have a nice evening desktopers see you tomorrow
[23:18] <attente> robert_ancell: hi! not sure if you saw the meeting notes, but i'm looking for a reviewer for https://bugzilla.gnome.org/show_bug.cgi?id=775732. do you think you'd be able to do review it? if not i might ask Lan3y or Tr3vinho
[23:19] <robert_ancell> attente, sure, I can review
[23:19] <attente> robert_ancell: great, thanks!
[23:21] <robert_ancell> attente, you can call g_hash_table_new_full with an empty hash/equal func?
[23:23] <attente> robert_ancell: i think so
[23:23]  * attente checks
[23:23] <robert_ancell> attente, we should update the glib docs
[23:23] <robert_ancell> I see other GTK+ code does it
[23:25] <attente> yeah, no hash function means g_direct_hash, and no equal func means literal ==
[23:27] <attente> robert_ancell: i have those patches on the branch wip/attente/mir-content-hub, which is pre-gtk-4
[23:27] <attente> so you can test it against gedit, or gtk3-demo
[23:27] <attente> you just need libcontent-hub-glib-dev
[23:52] <robert_ancell> attente, reviewed
[23:57] <attente> robert_ancell: thanks