[04:45] <pitti> Good morning
[06:32] <didrocks> good morning!
[06:33]  * didrocks has been fighting for 15 minutes to get some wifi, grrr network-manager
[06:54] <larsu> good morning!
[06:55] <larsu> didrocks: hi! I'm having nm problems as well. Need to reboot constantly
[06:56] <didrocks> larsu: once I'm connected, I'm fine here, it's just to connect…
[06:56] <didrocks> hey larsu ;)
[06:58] <larsu> didrocks: ya me as well
[06:58] <larsu> but sometimes it doesn't reconnect after suspend
[06:58] <didrocks> yep, or after a new boot
[06:58] <didrocks> then, sometimes stopping the service and restarting works
[06:58] <didrocks> but some other times, you really need to reboot
[06:58]  * didrocks has this for 6 months
[06:59] <pitti> hey didrocks, hey larsu!
[07:00] <larsu> morning pitti!
[07:00] <didrocks> bonjour pitti !
[07:00] <pitti> comment allez-vous ?
[07:01] <didrocks> pitti: ça va très bien, et toi ? pas de mal de tête après ton cours de français ?
[07:02] <pitti> didrocks: en effet j'ai quelques maux de tête :-( -- alors, ce déjè commence avant la course, hier midi
[07:02] <didrocks> aïe :/
[07:02] <pitti> je ne me sense pas comme une rhume, c'est étrange
[07:03] <didrocks> pitti: peut-être de la fatigue générale ? pas assez dormi, etc.
[07:03] <pitti> didrocks: ce sont les premières nuits froide, peut-être c'est ça
[07:04] <pitti> effectivement j'ai eu beaucoup de repos
[07:16] <tjaalton> didrocks: guess I have a fix for your crasher
[07:17] <didrocks> tjaalton: \o/
[07:17] <tjaalton> will push it soon
[07:17] <tjaalton> it was caused by the update of x11proto-core
[07:17] <didrocks> tjaalton: tell me if you want to have it a try on a ppa, I still can't really trigger it reliably, but I have 2-3 crashes a day
[07:17] <tjaalton> which was meant for xserver 1.18
[07:17] <didrocks> oh
[07:17] <didrocks> interesting
[07:17] <tjaalton> I'll push it anyway, it's safe.. oneliner
[07:18] <didrocks> what can go wrong in one line, right? :p
[07:18] <didrocks> ok, keep me posted ;)
[07:18] <tjaalton> https://bugs.freedesktop.org/attachment.cgi?id=117200
[07:18]  * didrocks looks
[07:18] <didrocks> ah, all that pain for a non nullified variable
[07:18] <tjaalton> that's the patch, upstream got a different one in 1.18 but this might still land in 1.17.x
[07:19]  * didrocks likes when the rationale is 30 times longer than the fix :)
[07:22] <tjaalton> https://bugs.freedesktop.org/show_bug.cgi?id=91316 is the bug
[07:24] <didrocks> tjaalton: ah, I get now your comment here about latest xproto
[07:52] <willcooke> morning all
[07:53] <didrocks> hey willcooke
[07:54] <larsu> morning willcooke
[07:55] <tjaalton> didrocks: xserver uploaded, should arrive soon
[07:56] <didrocks> tjaalton: excellent! will keep you posted
[08:01] <Laney> hi hi
[08:03] <larsu> Laney! What up?
[08:04] <didrocks> good morning Laney
[08:05] <Laney> hey!
[08:05] <Laney> how's it going?
[08:05] <larsu> good good :)
[08:05] <didrocks> good ;)
[08:07] <Laney> GREAT!
[08:08] <pitti> hey Laney!
[08:08] <larsu> wow :)
[08:09] <Laney> it's a pitti
[08:09] <didrocks> Laney: no, "c'est un pitti" ;)
[08:09] <pitti> oui !
[08:09] <pitti> il faut parler français ici
[08:10] <didrocks> exactement !
[08:10] <Laney> kungcono pitti
[08:11] <pitti> didrocks: et tu dois défendre seb128 aussi, pendant il est en vacances !
[08:11] <pitti> Laney: Zulu?
[08:11] <Laney> or at least google's version of Zulu :P
[08:11] <didrocks> pitti: un français en vaut bien deux, non ? :)
[08:12]  * pitti donne didrocks une accolade
[08:12]  * didrocks donne une accolade en retour à pitti
[08:14]  * larsu wonders how French survived in this channel now that seb128's not here
[08:15] <didrocks> larsu: don't challenge me in summoning the french community to invade this channel :p
[08:15]  * larsu shuts up
[08:15] <Laney> larsu: not sure I dare ask... but I have a dbus question...
[08:15] <Laney> http://irclogs.ubuntu.com/2015/09/07/%23ubuntu-desktop.html#t16:45 please to look if you have time
[08:24] <larsu> Laney: the dbus test runner failure?
[08:25] <Laney> ya
[08:25] <pitti> I retried http://autopkgtest.ubuntu.com/packages/d/dbus-test-runner/ an hour or so ago, if you mean that
[08:25] <pitti> it succeeded again
[08:25] <pitti> yay flaky tests
[08:25] <Laney> it is a real failure though
[08:25] <pitti> http://autopkgtest.ubuntu.com/packages/d/dbus-test-runner/wily/ppc64el/ you mean?
[08:25] <Laney> yes
[08:26] <Laney> a real race condition
[08:27] <Laney> although I didn't get it on my desktop with 99999 or so iterations
[08:27] <Laney> the ppc64el porter machine gets it in < 50
[08:34] <larsu> ugh, this looks nasty
[09:01] <Sweet5hark> this reminds me of some project, cant point my finger on which: http://devopsreactions.tumblr.com/post/128322627867/its-a-slow-oversized-and-ugly-framework-but-we
[09:01] <Sweet5hark> it is friday yet?
[09:02] <larsu> KDE?
[09:04] <Sweet5hark> larsu: close. ;)
[09:04]  * Sweet5hark hands larsu a cookie.
[09:05]  * larsu chews
[09:11] <Laney> larsu: is it a race in g_bus_watch_name?
[09:12] <Laney> that's what dbus-test-runner uses
[09:13] <larsu> doesn't look like a race to me
[09:13] <Laney> then what?
[09:13] <larsu> well, it's complicated
[09:13] <larsu> I'm trying to reproduce to no avail
[09:13] <Laney> go on the porter machine
[09:13] <larsu> looking at the logs, it seems like the service appears and disappears faster than the method returns
[09:13] <larsu> which is a legit thing to do for dbus-1
[09:13] <pitti> Laney: perhaps it helps to run it in a container which you just give 5% CPU or so?
[09:14] <larsu> hm actually you're right - it's not ok for watch_name() to miss this
[09:14] <larsu> pitti: I've long wanted a dbus proxy for debugging that inserts random timeouts and sometimes throws an error and such
[09:15] <larsu> Laney: my first hunch was that it shouldn't have to look at the signal as well, because who cares about the name if it only exists for such a short time
[09:16] <larsu> but then, bus_watch_name() guarantees to fire when the name appears...
[09:16]  * larsu checks how it is implemented
[09:19] <larsu> Laney: I don't have access to that
[09:23] <Laney> ah, this can be fixed!
[09:26] <larsu> ooh thanks Laney
[09:34] <larsu> I don't understand why dbus-test-runner is using g_bus_watch_name() for this
[09:34] <larsu> it knows for a fact that the name doesn't exist yet (doesn't it??)
[09:34] <didrocks> tjaalton: ok, installed, waiting for next crash to get the server restarted
[09:34] <larsu> and simply waits for it to appear
[09:35] <larsu> btw, gdbus ignores the nameownerchanged signal until getnameowner returns
[09:35] <tjaalton> didrocks: ok
[10:16] <desrt> Laney: you must always watch the signal first, for this reason
[10:16] <desrt> Laney: but gdbus provides for you here -- g_bus_watch_name()
[10:22] <didrocks> and hopefully, that was my last Xorg crash :p
[10:22] <Laney> desrt: it's dbus-test-runner here, which is indeed using that
[10:22] <Laney> but larsu is looking into this now so let's not keep too many brains on it :P
[10:22] <larsu> desrt: morning. You catch me at a bad timing - about to meet dholbach at sahara
[10:23] <larsu> (I'm sure you understand)
[10:23] <desrt> ya.  of course :)
[10:23] <desrt> freaking sahara
[10:23] <Laney> bit gritty for me
[10:23] <larsu> the problem is that g_ubs_watch_name() ignores the signal until GetNameOwner returns
[10:23] <Laney> (HAHAHAHAHHAHAHAOHAOHFOASHFASH)
[10:23] <larsu> Laney: I have half a patch that *might* work
[10:24] <larsu> using the signal directly
[10:24] <larsu> I still think gdbus is wrong here
[10:24] <larsu> desrt: ^^ but lets discuss after lunch
[10:25] <Laney> someone has "taken" the ticket to give you access so hopefully that is done when you get back
[10:45] <qengho> Anyone know of a good visual testing tool? Like, fire up arbitrary GUI program, test that somewhere on the screen looks like _this_?
[10:49] <willcooke> qengho, I've seen Jenkins produce stuff like that
[10:49] <willcooke> Trevinho, do you know how we do that for U7>? ^^
[10:50] <qengho> About once every season Sikuli breaks, and chromium-browser tests use it. Last week, some dependency of Sikuli updated, probably openjdk-8-jdk and now Jython library doesn't parse python programs that don't have final indentation lines at zerotOMGWTFthis has nothing to do with chromium
[10:50] <Trevinho> qengho: we use autopilot for that
[10:50] <Trevinho> ah... well no
[10:50] <Trevinho> you want to check pixel-perfection of things, right?
[10:52] <didrocks> jenkins isn't targeting that use case
[10:52] <didrocks> if you want to do visual comparison, you can use visualdiff
[10:53] <didrocks> (and introduce some "noise" to accept slight changes)
[10:54] <Trevinho> Mh, ma ha levato il post?
[10:54] <Trevinho> err wrong chat :D
[10:55] <qengho> Trevinho: I don't require pixel comparison, but I don't have access to turn things back into utf8 for comparison.
[11:00] <qengho> didrocks: I don't know of "visualdiff". Neither does "apt-cache search".
[11:06] <didrocks> qengho: not sure if it's packaged: https://www.mediawiki.org/wiki/Visual_Diff
[12:10] <larsu> desrt: any ideas?
[12:11] <Laney> larsu: good falafel?
[12:12] <larsu> Laney: yes! Also some boring errands. Had to buy a dust mop, too
[12:14] <Laney> yin and yang
[12:14] <ogra_> planning to do some gtk-1.2 work ?
[12:15]  * larsu runs
[12:40] <Trevinho> Laney: for landing https://code.launchpad.net/~3v1n0/ubuntu-themes/osd-scrollbars-improvements/+merge/269245 can i join your silo 031 or is it too late?
[12:41] <willcooke> Sweet5hark, tkamppeter - just been to the post office.  You should have presents in the next week
[12:41] <Sweet5hark> willcooke: \o/ awesome.
[12:42] <davmor2> Sweet5hark: By presents he means some empty raspberry pi boxes don't think he sent on the boards ;)
[12:43] <Laney> Trevinho: late, just put it in your unity one or something
[12:43] <willcooke> sush, davmor2 :)
[12:43] <Trevinho> Laney: that's what I was going to do... Should I use some other option in order to make this depend on that or... the train is smart enough to handle it?
[12:43] <Sweet5hark> davmor2: you mean the board will have had "accidents" on the way?
[12:44] <Laney> Trevinho: it's already in
[12:44] <davmor2> Sweet5hark: don't they always ?
[12:44] <Laney> ag, boottest
[12:44]  * Laney stabs that
[12:44] <Laney> Trevinho: I'll just make it merge that one to trunk
[12:44] <Laney> then you should be ok
[12:45] <Trevinho> Laney: ah, ok... I thought it was still moving to trunk...
[12:45] <Laney> yeah but you can override that
[12:57] <ricotz> willcooke, oh, Pi2 ;P
[13:15] <tedg> Laney: Did you get anywhere with the dbustestrunner PPC64 failure?
[13:15] <Laney> larsu is on the case
[13:16] <tedg> Woot! Go larsu!
[13:23] <Sweet5hark> cyphermox: ahhh, you are back? LibreOffice 5.0 FFE is greenlighted from ubuntu-release, do you think you might review/sponsor the package for wily?
[13:31]  * larsu goes and makes some tea
[13:46] <cyphermox> Sweet5hark: where is it?
[13:48] <Sweet5hark> cyphermox: http://people.canonical.com/~bjoern/wily/5.0.1/libreoffice_5.0.1-0ubuntu1_source.changes and http://people.canonical.com/~bjoern/wily/5.0.1/libreoffice-l10n_5.0.1-0ubuntu1_source.changes -- ./debian changes/DCVS at http://anonscm.debian.org/cgit/pkg-openoffice/libreoffice.git/log/?h=ubuntu-wily-5.0
[13:49] <Sweet5hark> cyphermox: nlpsolver and writer2latex should be updated along with that: http://people.canonical.com/~bjoern/wily/nlpsolver_0.9~beta1-10ubuntu2_source.changes and http://people.canonical.com/~bjoern/wily/writer2latex_1.0.2-10ubuntu2_source.changes
[13:50] <Sweet5hark> cyphermox: https://bugs.launchpad.net/ubuntu/+source/libreoffice/+bug/1491964 has more links, including buildlog etc.
[13:53] <cyphermox> ok
[13:54] <cyphermox> Sweet5hark: looks rather big so I'll first finish other things and then review it this PM
[13:56] <Sweet5hark> cyphermox: great, having feedback/uploading it on wednesday would be awesome: as you say its rather large and has many deps, so having it finished building before the weekend would be great.
[13:57] <cyphermox> ok
[14:00] <Sweet5hark> cyphermox: thanks a lot!
[14:40] <larsu> desrt: hi. did you see Laney's paste?
[14:40] <desrt> yes.  let me see it again.
[14:40] <Laney> it is a good paste
[14:40] <larsu> I managed to work around this issue by connecting to NameOwnerChanged() directly
[14:41] <larsu> http://paste.ubuntu.com/12308415/
[14:41] <larsu> desrt: the problem is that g_bus_watch_name() ignores that signal until GetNameOwner() returns
[14:41] <larsu> (see the initialized flag in the struct)
[14:41] <desrt> that's correct.
[14:41] <larsu> why?
[14:41] <desrt> i think i advocated that behaviour, in fact
[14:42] <larsu> it's wrong in this case, though
[14:42] <desrt> because it will always give the initial callback based on teh result of GetNameOwner call
[14:42] <desrt> and the GetNameOwner call and reply are handled atomically with respect to bus ordering (since it's the daemon processing it)
[14:42] <larsu> right, that is totally fine
[14:42] <desrt> so the only possibility for something to intercede here is if a name owner change happens before i requested who has the current name
[14:42] <larsu> the problem is that we're missing the name appearing and disappearing
[14:43] <desrt> in which case it's totally uninteresting, since i will get the "correct" answer later anyway
[14:43] <desrt> but before the name owner is requested, you have no right to know
[14:43] <larsu> this change happens *after* though
[14:43] <larsu> and before the call returns
[14:43] <larsu> it's already gone
[14:43] <desrt> not possible
[14:43] <desrt> the request isn't made until dbus sees it
[14:44] <larsu> which request?
[14:44] <desrt> you can't do a watch and then (out of band) fire off some other thing and expect to be guaranteed to see it
[14:44] <desrt> GetNameOwner
[14:44] <desrt> but for that matter, let's talk about the AddMatch
[14:44] <desrt> there is no guarantee that even the AddMatch would make it to the bus before your process owns/disowns the name
[14:44] <desrt> so if you wanted to fix this, you'd made it slightly more reliable, but still not really reliable
[14:45] <larsu> hm, I don't think we're talking about the same thing
[14:45] <desrt> please explain to me what is happening
[14:45] <desrt> because my best guess is that you're trying to use name ownership to figure out if some process has started or not
[14:45] <larsu> yes
[14:45] <larsu> indeed
[14:46] <desrt> and you spawn this process out-of-band of dbus
[14:46] <desrt> like via fork/exec or some other method
[14:46] <larsu> in the worst way possible even: using gdbus-tool to call RequestName
[14:46] <desrt> that's... disturbing
[14:46] <desrt> okay
[14:46] <desrt> here's the real problem
[14:46] <larsu> which is why the name disappears so quickly
[14:46] <desrt> dbus worker thread is what it is, right?
[14:46] <desrt> it can process things after an arbitrary delay
[14:47] <desrt> so if i queue up some AddMatch and GetNameOwner... there is no guarantee that those will be delivered "soon"
[14:47] <desrt> so it could be that gdbus-tool runs before even AddMatch gets seen by the bus
[14:47] <desrt> in which case i will be completely and totally in the dark
[14:47] <larsu> ya
[14:47] <desrt> i won't even have a chance to ignore the signal -- i won't ever even see it
[14:47] <larsu> but this is not what is happening here
[14:47] <desrt> so you can't fix what you're trying to fix here
[14:47] <larsu> we *do* get both signals
[14:47] <desrt> i know
[14:47] <desrt> but you might not
[14:48] <desrt> you only happen to get it because AddMatch made it in time, but not the reply for GetNameOwner
[14:48] <larsu> right
[14:48] <desrt> i'm saying that it could be just as easily the case that AddMatch itself failed to make it
[14:48] <larsu> indeed
[14:48] <desrt> so, in general, you can't do what you're doing any expect it to work, regardless of gdbus
[14:48] <larsu> so we don't fix that other bug becuase this might happen anyway?
[14:48] <desrt> it's not a bug
[14:48] <larsu> desrt: s/what you're doing/what dbus-test-runner is doing/
[14:49] <desrt> so dbus-test-runner has the bug.  fix it there.
[14:49] <larsu> I fixed it by listening to the signal
[14:49] <desrt> not a good enough fix, unless your signal match is synchronous
[14:49] <larsu> haven't tested how well that works, but by your logic the same bug applies
[14:49] <larsu> but then g_bus_watch_name() has really weird semantics....
[14:50] <larsu> "I'll start watching that name some point in the future"
[14:50] <desrt> you have to understand that all of these things are made for an async event-driven world
[14:50] <desrt> where the first answer need not be correct as long as the eventual result is
[14:50] <desrt> it's like how GMenuModel will lie to you at first
[14:51] <desrt> and in this case, even, the first answer _is_ correct
[14:51] <desrt> ...at the time it is given
[14:51] <larsu> there's no answer at all...
[14:51] <larsu> true
[14:51] <desrt> well, it will tell you either "vanished" or "appeared" one way or the other
[14:51] <larsu> no
[14:51] <larsu> it never gives you vanished
[14:51] <larsu> if it's never been there
[14:51] <desrt> it ought to
[14:51] <larsu> ?
[14:52] <larsu> I wouldn't expect it to
[14:52] <desrt> i would...
[14:52] <desrt> You are guaranteed that one of the handlers will be invoked after calling this function. When you are done watching the name, just call g_bus_unwatch_name() with the watcher id this function returns.
[14:52] <larsu> oh indeed it does according to docs
[14:52] <larsu> weird
[14:52] <larsu> in any case using name watching for this is wrong
[14:52] <desrt> no kidding
[14:53] <larsu> as is using gdbus-tool
[14:53] <larsu> what do you suggest? Listen to the signal and write a thing that owns the name for a bit longer?
[14:53] <desrt> using gdbus-tool to grab a name as a way to alert [something] of [something] is a very ... creative idea
[14:53] <desrt> but not a good one
[14:53] <desrt> why not just listen to a normal signal and have gdbus tool send that signal?
[14:53] <desrt> or register an object and have it call a method on that object
[14:54] <desrt> that's even better
[14:54] <larsu> because this is a test for the --wait-for parameter
[14:54] <desrt> ahh.  i see.
[14:54] <larsu> which waits for a name before doing something
[14:54] <desrt> so ping the bus
[14:54] <desrt> actually, even better, just wait for the vanished call
[14:54] <larsu> that might happen immediately
[14:54] <desrt> which is an indication that watching has been successfully established, and the thing doesn't exist
[14:55] <larsu> ah
[14:55] <larsu> indeed
[14:55] <desrt> and then it's safe to launch the thing in a way that you will be guaranteed to see the result
[14:55] <larsu> we might still miss it, no?
[14:55] <desrt> pretty sure it's safe by this point
[14:55] <larsu> ah that's what you mean...
[14:55] <desrt> you don't get vanished until GetNameOwner returns
[14:55] <desrt> and this is after the watch
[14:55] <larsu> indeed
[14:55] <desrt> so if this doesn't work, it would almost have to be a bug...
[14:56] <larsu> we'd still miss this specific thing
[14:56] <larsu> oh wait, no
[14:56] <larsu> sorry
[14:56] <desrt> :)
[14:56] <larsu> desrt: I'll try that instead of listening to NameOwnerChanged. Thanks for your help
[15:05] <desrt> willcooke: might be hard for me to make it to the sprint on time :(
[15:06] <willcooke> desrt, no worries
[15:07] <larsu> desrt: red-eye flight!
[15:07] <desrt> larsu: i was thinking the opposite, in fact
[15:07] <larsu> on Monday?
[15:07] <desrt> willcooke: btw: would not mind london again, if this turns out to be the easiest option
[15:07] <desrt> the earliest i could leave it early monday... and there is a non-redeye flight to london on this day, i think
[15:08] <Trevinho> larsu: so... I'm testing the silo with new scrollbars, but it seems that there's something wrong with terminal. Also I can't find a way to run the inspector on it...
[15:08] <Trevinho> any tip?
[15:08] <larsu> Trevinho: terminal is weird because it doesn't use a GtkScrolledWindow
[15:09] <larsu> Trevinho: actually I *think* rishi worked on this in the 3.18 cycle
[15:09] <larsu> Trevinho: set GTK_DEBUG=interactive to always open the inspector
[15:09] <larsu> you need to do that on the server command though
[15:09] <Trevinho> larsu: it doesn't work in terminal... Also using the server
[15:10] <Trevinho> as the server only shows infos about the server itself, but not about the window it creates
[15:11] <larsu> even when you use the widget picker?
[15:11] <Trevinho> yep
[15:11] <Trevinho> larsu: if I change the theme (for example), from visual It doesn't apply to that window even
[15:11] <larsu> weird
[15:12] <Trevinho> larsu: yeah.. can you?
[15:12] <larsu> there's a compile time option to add a Help/Inspector menu item
[15:12] <larsu> let me check
[15:13] <larsu> ya same problem
[15:14] <larsu> also getting warnings about TerminalWindow not being on the correct screen
[15:14] <Laney> i think you have to build with --enable-debug then use a menu item
[15:14] <Trevinho> mh, i see
[15:15]  * larsu wonders why
[15:30] <Laney> MEETING TIME DING DING DING
[15:30] <willcooke> Woo
[15:30] <didrocks> DONG DONG DONG
[15:30] <andyrock> \o/
[15:30] <willcooke> #startmeeting Desktop Team Weekly Meeting - 2015-09-08
[15:30] <meetingology> Meeting started Tue Sep  8 15:30:28 2015 UTC.  The chair is willcooke. Information about MeetBot at http://wiki.ubuntu.com/meetingology.
[15:30] <meetingology> Available commands: action commands idea info link nick
[15:30] <willcooke> Roll call:  andyrock, attente, desrt,  dgadomski, didrocks, fjkong (out), happyaron (out), hikiko (hols), laney, larsu, qengho (out), seb128 (hols), sweet5hark, themuso (out), tkamppeter, trevino, robert_ancell (out)
[15:30] <Sweet5hark> .
[15:30] <happyaron> .
[15:31]  * happyaron please call me a bit later in the queue, writing atm
[15:31] <larsu> \o
[15:31] <attente> hi
[15:31] <dgadomski_> hey
[15:32] <willcooke> happyaron, ack
[15:32] <willcooke> Let's roll...
[15:32] <willcooke> #topic andyrock
[15:32] <andyrock> hey
[15:32] <andyrock> Updated branch lp:~azzar1/unity/gnome_session_manager_more_tests
[15:32] <andyrock> Updated branch lp:~azzar1/unity/sc_dbus_manager_tests
[15:33] <andyrock> Created MP for bug 1491555 using patch from external contributor
[15:33] <andyrock> Created MP for bug 1491913 using patch from external contributor
[15:33] <andyrock> Keep working on screen resize bug on compiz (damn you compiz!)
[15:33] <andyrock> basically that was my week
[15:33] <willcooke> thanks andyrock
[15:33] <willcooke> #topic attente
[15:34] <attente> investigated and wrote out a list of requirements we need in Mir to support hardware keyboard layouts
[15:34] <attente> refresh non-gnome-unity-ibus-support branch for u-c-c, thanks Laney!
[15:34] <attente> more work on proper popup menu placement in gtk-mir continued...
[15:34] <attente> todo: fix gtk apps not working on Ubuntu Touch devices due to not supporting the available pixel formats Mir can provide
[15:34] <attente> (eof)
[15:34] <willcooke> thanks attente
[15:35] <willcooke> #topic desrt
[15:37] <willcooke> let's come back to desrt
[15:38] <willcooke> #topic dgadomski_
[15:38] <dgadomski_> hey, I was informed that libosmesa6 from lts-mesa-{utopic,vivid}. Looks like it was removed because of bug #1424059. Luckily tjaalton is going to take a look at it.
[15:38] <dgadomski_> That's all in the desktop area this week. Thanks!
[15:38] <willcooke> thanks a lot dgadomski_
[15:38] <willcooke> #topic didrocks
[15:38] <didrocks> hey
[15:38] <didrocks> Ubuntu Make:
[15:38] <didrocks> - Released 15.09 which contains notably the Unity 3D support. Wrote blog post + social media
[15:38] <didrocks> - More reviews and merged android sdk support + exit code cleanups from 2 contributors (one being new \o/)
[15:38] <didrocks> - Fix some pep8 issues on trusty only for CI tests and other small tests broken by above.
[15:38] <didrocks> - Bug management
[15:38] <didrocks> Developer experience:
[15:38] <didrocks> - Add transitions and episod on it for test app
[15:38] <didrocks> - various warnings and code cleanups on test app + opened last bugs
[15:38] <didrocks> - installed snappy on my machine and a raspberry pi 2 to evaluate Go developer experience (+ running first examples)
[15:38] <didrocks> Misc:
[15:38] <didrocks> - reviewed unity3d upstream deb about "bad packaging" complains: the issue is mostly warnings about /opt "bad quality" for sotfware center due to this path. Nothing we can fix as this is the best practice for such out of repo components.
[15:39] <didrocks> - doing some small snappy patch
[15:39] <didrocks> - patch pilot and some archive admin duties
[15:39] <didrocks> .
[15:39] <didrocks> (and thanks tjaalton! no xorg crash for 5h now! ;)
[15:39] <willcooke> thanks didrocks
[15:39] <larsu> didrocks: new record?
[15:39] <willcooke> and thanks tjaalton
[15:39] <didrocks> larsu: yep :p
[15:39] <willcooke> #topic FJKong
[15:39] <willcooke> update code of pinyin search
[15:39] <willcooke> * support advance search feature
[15:39] <willcooke> * make scaning more faster
[15:39] <willcooke> * support run as daemon
[15:39] <willcooke> * remove duplicate result of searching
[15:39] <willcooke> attend CCN meeting at Changsha, show demo of pinyin searching
[15:39] <willcooke> #topic larsu
[15:39] <willcooke> err
[15:40] <willcooke> I meant laney, but larsu - you go
[15:40] <larsu> sure
[15:40] <willcooke> keep Laney on his toes
[15:40] <Laney> this is like final destination
[15:40]  * Laney has cheated his turn
[15:40] <larsu> - finish up notify-osd work (long standing blurry-on-hidpi bug)
[15:40] <larsu> - finish up Laney's indicator-datetime test fixes (sorry again for putting you through this) and cleaned up those tests whlie I was at it (some weren't even running...)
[15:40] <larsu> - sift through timezonemap bugs and start working on disabling geonames.ubuntu.com for now due to bugs in returning cities for partial names
[15:40] <larsu> - help out with dbus-test-runner test issues on ppc

[15:40] <willcooke> thanks larsu
[15:40] <Laney> can you get someone to review datetime?
[15:41] <larsu> Laney: you do it!
[15:41] <Laney> not sure about that :(
[15:41] <willcooke> larsu, do you think the OSB changes are doable for Thursday?
[15:41] <larsu> willcooke: I think we should merge as is
[15:42] <larsu> unless someone objects
[15:42] <larsu> ok time for objections is over. Glad you all like them so much!
[15:42] <willcooke> larsu, wfm
[15:42] <willcooke> thanks
[15:42] <willcooke> #topic Laney
[15:42]  * larsu grins
[15:42] <Laney> • Things what Lars did for me which are about dbus
[15:42] <Laney> ∘ I looked more at i-datetime but my fix was bad so larsu took over
[15:42] <Laney> ∘ I noticed that dbus-test-runner was failing a test on ppc64el and larsu is working on a fix
[15:42] <Laney> • Tried to find out about beta testing the font, not really with much success but at least there's a sort of working source package now.
[15:42] <Laney> • Packaged gstreamer 1.6 RC
[15:42] <Laney> ∘ -bad failed to build on armhf, test build is in the works now
[15:42] <Laney> ∘ had to fix libav to build, uploaded to Debian, synced
[15:42] <Laney> • Uploaded a new gtk (3.14.15) to a silo for the overlay ppa which Mir people want
[15:42] <Laney> • Updated file-roller, glib, e-d-s, evo, dbus (+ Debian as appropriate)
[15:42] <Laney> • Smooth some autopkgtests so things can migrate
[15:42] <Laney> • Got some training/instruction on the armhf/ppc64el autopkgtest infrastructure - can do light admin there now
[15:43] <Laney> • Some release team stuff like FFe reviews
[15:43] <Laney> • Uploaded a few things through the train (u-c-c, ubuntu-themes)
[15:43] <Laney> • Patch piloted
[15:43] <Laney> • Helped upload things to move from libav (back) to ffmpeg
[15:43] <Laney> ⁂
[15:43] <larsu> Laney: s/my fix/tiny part of my fix/ - thanks for your work on that!
[15:43] <willcooke> thanks a lot Laney
[15:44] <willcooke> #topic qengho
[15:44] <willcooke> - releasing chromium 45.0.2454.85 with GPU rendering re-enabled.
[15:44] <willcooke> - Cr test failures. Not Cr, not Sikuli. Jython grammar bug exposed by openjdk-8-jdk update?
[15:44] <willcooke> - A little more Cr/ozone/mir.
[15:44] <willcooke> #topic Sweet5hark
[15:44] <Sweet5hark> not too much:
[15:44] <Sweet5hark> - testbuilds of 5.0.x on wily armhf/powepc -- all good
[15:44] <Sweet5hark> - FFE LibreOffice 5.0 for wily
[15:44] <Sweet5hark> - finding a sponsor for it (thanks cyphermox)
[15:44] <Sweet5hark> - nlpsolver and writer2latex build fixes for LibreOffice 5.0
[15:44] <Sweet5hark> - conference preperation: slides and foo
[15:44] <Sweet5hark> - addding a staging area for unstable tests upstream ("make -k stagingcheck")
[15:44] <Sweet5hark> - bugzilla grinding
[15:44] <Sweet5hark> EOF
[15:44] <willcooke> thanks Sweet5hark
[15:44] <Laney> qengho: you might profit from giving sikuli an autopkgtest
[15:45] <willcooke> #topic TheMuso
[15:45] <willcooke> * More work on Orca's gsettings backend which when implemented, will help the accessibility profile implementation.
[15:45] <willcooke> * Discovered that Unity in Wily has a regression with accessibility with the launcher and quicklists. Quicklists are not speaking with Orca, and once opened, the launcher itself goes silent. I started debugging, will report when I have more.
[15:45] <willcooke> #topic robert_ancell
[15:45] <willcooke> Off sick so only two days to catch up on email and look at XMir issues. Released LightDM 1.16.0.
[15:45] <willcooke> #topic happyaron
[15:46] <willcooke> desrt, are you back?
[15:46] <happyaron> unfortunately quite short ones..
[15:46] <happyaron> 1. Meeting at Changsha with NUDT
[15:47] <happyaron> 2. Packaging of pinyin search, written by fjkong
[15:47] <happyaron> EOF
[15:47] <willcooke> thanks happyaron
[15:48] <willcooke> #topic desrt
[15:49] <willcooke> hrm
[15:49] <willcooke> he ded
[15:49] <willcooke> #topic Any Other Business
[15:50] <willcooke> Sprint is go.  See email for more details.
[15:50] <Laney> prague plz
[15:50]  * larsu seconds prague
[15:50] <Trevinho> willcooke: you forgot me? :)
[15:50] <Laney> hahaha
[15:50] <willcooke> and till
[15:50]  * happyaron really hopes Toronto, but no go...
[15:50]  * didrocks +1 prague as well
[15:50]  * larsu hugs Trevinho 
[15:50]  * Laney shuts the door on Trevinho's cage
[15:50] <Trevinho> :°(
[15:50] <willcooke> Trevinho:  Went to China.  EOF
[15:50] <larsu> Laney: let me out first!
[15:51]  * Trevinho is ready to paste :P
[15:51]  * larsu was still hugging
[15:51] <willcooke> Trevinho, two secs...
[15:51] <willcooke> any more votes for sprint locations?
[15:51] <Trevinho> willcooke: I can be more verbose (but that is just rigth) :D
[15:51] <larsu> willcooke: 10 more for prague
[15:51] <didrocks> we did budapest twice, prague one, we need to be even! :)
 prague
[15:51] <Trevinho> willcooke: well prague never saw it, so i'd go there
[15:51] <Trevinho> :D
[15:51] <larsu> willcooke: not sure that's how voting works
[15:51]  * happyaron send everyone to China
[15:51] <didrocks> I'm sure seb would +1 prague as well :p
 prague
[15:51]  * Laney coughs
[15:51] <Trevinho> but I love Budapest as well...
[15:51] <didrocks> ahah
[15:52] <happyaron> Laney: lol
[15:52] <Sweet5hark> prague  <3
[15:52] <Trevinho> lol
[15:52] <Trevinho> So... Prague will be! (?)
[15:52] <attente> abstain
[15:52]  * larsu believes it when he sees it
[15:52] <Laney> like the existence of toronto
[15:52] <willcooke> oki good
[15:52] <larsu> attente: THINK OF THE BEER
[15:52] <larsu> wait, budapest is equally good for that
[15:52] <attente> larsu: alcohol is ruining my life :P
[15:53] <larsu> attente: maybe we should do it in disneyland, then?
[15:53] <willcooke> let's loop back to Trevinho quick...
[15:53] <willcooke> #topic Trevinho
[15:53] <Trevinho> budapest is awesome for everything... But well, not that Prague is bad :)
[15:53] <Laney> haha yes, get this meeting back in order
[15:53] <Trevinho> * Not full week because of travel
[15:53] <Trevinho> * Back from China, got some involvement in unity7 from Kylin guys (longer report soon)
[15:53] <Trevinho> * Setup a new meeting (every two weeks) in #ubuntukylin-devel for syncing
[15:53] <Trevinho> * Prepared a document with some development hints for unity7 (raw version http://is.gd/6nrYBA)
[15:53] <Trevinho> * Initial review of the unity scope pinyin search
[15:53] <Trevinho> * Prepared new landing for unity7 and compiz branches
[15:53] <Trevinho> * Testing new landing, new scrollbars needs some fixes (terminal)
[15:54] <Trevinho> ***
[15:54] <willcooke> thanks Trevinho
[15:54] <Laney> I think you need more than theme fixes for the terminal
[15:54] <willcooke> #topic tkamppeter
[15:54] <Laney> see also webkit (e.g. devhelp)
[15:54] <larsu> Trevinho, Laney: I think rishi fixed the terminal scrollar problems. Is it too late to take .18?
[15:54] <tkamppeter> - cups-filters: Released 1.0.75 with a crasher fix in texttopdf
[15:54] <tkamppeter> - cups: Packaged 2.1.0 final release
[15:54] <tkamppeter> - ippusbxd: investigations on a bug
[15:54] <tkamppeter> - Bugs.
[15:54] <willcooke> thanks tkamppeter
[15:54] <Trevinho> Laney: yeah I think so.. But maybe not... At least it seems it works a little better
[15:54] <Laney> yes
[15:54] <willcooke> #topic AOB
[15:54] <Trevinho> Laney: come on.. :P
[15:55] <Trevinho> Mh, wherever we go (prague or budapest), I've no direct flights from florence :°(
[15:55] <larsu> I'm glad meetingology repeats the topic all the time. I constantly forget where the home of the desktop team is
[15:55] <larsu> Trevinho: florence has an airport? :P
[15:55]  * Trevinho hits at his darts target with local politicians faces
[15:56] <larsu> haha
[15:56] <didrocks> no direct for me to budapest, so 1h40 vs 4h30 in average :/
[15:56] <didrocks> on*
[15:56] <Laney> get the train
[15:56] <larsu> not even with shitty airlines?
[15:56] <larsu> there's easyjet to both from berlin
[15:56] <didrocks> larsu: 4h30 is with shitty airlines
[15:56] <willcooke> ok, gonna wrap the meeting, we can continue discussions though.
[15:56] <larsu> ah, got it
[15:56] <willcooke> #endmeeting
[15:56] <meetingology> Meeting ended Tue Sep  8 15:56:56 2015 UTC.
[15:56] <meetingology> Minutes:        http://ubottu.com/meetingology/logs/ubuntu-desktop/2015/ubuntu-desktop.2015-09-08-15.30.moin.txt
[15:56] <willcooke> Thanks all
[15:57] <didrocks> thanks!
[15:57] <Trevinho> larsu: mh, well... if you want to call it so. The landing is like going in roller coaster, due to very short strip
[15:57] <larsu> :)
[15:57] <Trevinho> flights lands with brakes on... I broke my tablet because of that
[15:58] <willcooke> BTW - really good Hangout on Air with the Mycroft CTO
[15:58] <larsu> seriously? That's some mad stuff
[15:58] <larsu> willcooke: with dholbach?
[15:58] <willcooke> larsu, yeah
[15:58] <larsu> ya he said something about that
[15:59] <Sweet5hark> prague is 6:30 even with train for me (changing once in berlin)
[16:00] <larsu> I might take the train as well
[16:00] <larsu> it's a beautiful ride
[16:04] <Sweet5hark> larsu: yeah. and in october chances are its neither too hot or too cold for deutsche bahn.
[16:04] <larsu> hehe
[18:34] <willcooke> g'night
[18:36] <gQuigs> isn't Ubuntu  Make the reboot of Ubuntu Quickly?  (https://wiki.ubuntu.com/ubuntu-make  | https://wiki.ubuntu.com/Quickly)
[18:36] <gQuigs> time to retire Quickly/
[18:38] <jcastro> no
[18:38] <jcastro> quickly was a quick way to write apps
[18:39] <jcastro> umake is a way to grab IDEs and development packages directly from upstream
[18:42] <gQuigs> jcastro: so quickly is still something we should recommend to developers?
[18:43] <jcastro> I'm not sure what the status of quickly is
[18:44] <gQuigs> hmm.. maybe it's more like the Ubuntu SDK replaced Quickly
[18:48] <gQuigs> yea, last trunk commit was 2012-11-01, will email desktop team to see if there is anything I'm missing
[22:21] <larsu> Laney: morning. Do you have a bug about that test failure? Turns out it is a small fix: https://code.launchpad.net/~larsu/dbus-test-runner/fix-test-race
[22:21] <larsu> I'll merge propose this in the morning