[08:01] morning! [08:01] morning! [08:01] hey Laney, willcooke [08:03] hey Laney, how are you? [08:03] * pitti waves to willcooke [08:04] bonjour didrocks, comment ça va ? je ne toi vois pas ce matin [08:04] hey pitti, doing good thanks! [08:04] you? [08:04] didrocks: s/toi/te/ (je crois) [08:04] good morning pitti, oh, I forgot to say hi here :) [08:10] good morning desktopers [08:10] hey Laney willcooke didrocks pitti [08:10] re seb128 [08:10] bonjour seb128 ! [08:24] Laney: hi, thanks for the review. I've just made the last few changes you recommended. How shall I proceed? Shall I merge the branches or upload? (I think last time somebody did the former) [08:24] tseliot: Can you use the train? [08:25] Laney: is that a trick question? :P What is it? [08:25] CI train [08:26] Laney: I've never used that [08:31] tseliot: ok, one second [09:06] tseliot: can you add a commit message to the MP please? [09:06] Laney: sure. I'll do it for both requests [09:07] thanks [09:07] hi Laney! [09:08] hey larsu! [09:08] what's up? [09:08] tired :) [09:08] you? [09:08] Laney: done [09:08] feeling a bit like I'm juggling things [09:08] climbed last night, felt like ages since I went before that [09:08] \m/ STRONG \m/ [09:13] tseliot: it's going to turn up in https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-009/+packages (wily) and https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-028/+packages (trusty) soon [09:13] please test a bit and then I can hit the button to upload [09:14] Laney: ok, thanks [09:17] seb128, bug 1488431 [09:17] bug 1488431 in mir (Ubuntu) "Clients fail to create surfaces when using Mir server in compositor mode" [Critical,New] https://launchpad.net/bugs/1488431 [09:17] robert_ancell, trying [09:18] alan_g, can you confirm if the above bug matches what we discussed yesterday? [09:19] ChrisTownsend, could you also confirm if the above bug is working fine on your system? [09:32] Laney, do we have a branch for gedit 3.16 packaging somwhere? [09:33] * larsu needs to add some patches [09:33] larsu: I've never worked on that [09:33] oh, so I guess not, then [09:33] Unless you did before [09:33] Ubuntu GNOME probably have something [09:33] I didn't - maybe darkxst? [09:35] larsu, Laney, we might be able to just get the new version with csd and no menubar/toolbar, as long as we push the popdown menu to the hud [09:35] larsu, Laney there should be a gedit 3.16 on gnome3-staging, but it doesnt have the UI patches required [09:35] seb128, awesome! [09:35] darkxst, no branch? [09:35] larsu, can we push the menu content to the hud easily? [09:35] larsu, no branches for the ppa [09:36] been trying to set that up for the last year ;( but its a lot of work no one want to do ;( [09:36] seb128: oh? [09:36] seb128, dunno about the state of the hud to be honest. We definitely can if we expose it as a menu on the bus, but then it would also appear in the menu bar [09:37] you mean the hud would become the primary interface? [09:37] Laney, yeah, I was going to bring news from the meeting we just had on the other channel before the discuss just went weird [09:37] sorry :( [09:37] no worry [09:38] * seb128 needs some coffee [09:38] brb [09:38] darkxst, ok, I'll work on that, then [09:38] seb128, enjoy! [09:38] larsu, thanks ;-) [09:38] seb128, grab some cookies as well (the ginger ones) [09:38] * larsu misses those [09:38] short bread ftw! [09:45] * larsu tries to open the hud [09:45] seb128: can you summarise this meeting? [09:46] oh wow, somebody put it on Shift+Alt+? ... [09:46] you probably [09:47] never would I tamper with my system in such a way! [09:47] ha [09:49] larsu, ok thanks, I'm sick as tonight, should probably try and eat something then sleep [09:50] Laney, meeting is basically "design is giving a principle +1 on using upstream UI with csd/no menubar, as long as hud keeps working/showing the menu items since that's a feature a part of our userbase like and give us good vibes" [09:50] larsu, 3.16 packages are probably still in vivid pocket also [09:50] seb128: hoooooooly shit! [09:51] Laney, willc needs to demo what gedit 3.16 looks like to uper instances for validations [09:51] but in principle john thinks it should be ok [09:51] I guess show the appmenu up top? [09:51] Laney, larsu, we also need to fix that fact that csd windows under unity7 have solid square corners [09:52] Laney, no, we can do without menubar [09:52] if that's what the upstream design of the app is [09:52] and save us maintainance work [09:52] appmenu is the single item [09:52] the app name [09:52] oh [09:52] I guess then yes [09:52] currently we get (e.g.) "Disks Disks" [09:52] where the second one is the menu [09:52] right [09:52] try XDG_CURRENT_DESKTOP=GNOME gnome-disks [09:53] should fix that I think [09:53] good point [09:53] the app-menu would normally fallback to a button in the CSD's [09:53] I think we could show it probably [09:53] maybe not in the LIM case? [09:53] something to think about [09:53] (for non-shell) [09:53] add a "menu here" triangle if there is one [09:53] and get unity to show it under the title [09:54] or something [09:54] seb128, there's a compiz issue in there :/ [09:54] anyway, man, this is cool in principle [09:54] I'll investigate again, but from what I remember switching to argb windows broke shadows or something [09:54] but ... I'm hesitant to ask .... what about LIM? [09:55] I think just make it do ShellShowsAppMenu false [09:55] larsu, well, LIM only applies when there is a menubar [09:55] no menubar, nothing to include in the decoration [09:55] those apps would be like apps without menus [09:55] seb128, many apps export a menubar even with csd [09:55] we patched them upstream [09:55] it's a gtk feature, even [09:55] and on a different note, are there really still framebuffers that only support 4bpp (for plymouth) or is that just ancient cruft? [09:56] seb128, which also gives us perfect hud integration [09:56] i.e 16 colours [09:56] larsu, hum, unsure what to do with those, no easy reply [09:56] hide it with lim on :) [09:56] seb128, ok, thanks [09:56] Laney, or show the gear menu? [09:56] does it mean hud becomes the only way to access some items? [09:56] that's not good if that's the case [09:57] larsu: Isn't that ShellShowsAppMenu false? [09:58] Laney, yes, and ShellShowsMenubar. But LIM needs those set for other apps [09:58] seb128, no, hud only for things that are also in the menu or the ui, I think [09:59] good [09:59] I forgot the details maybe [10:20] larsu app-menu [10:20] ? [10:20] uses the same format as all other menu ui [10:20] xml in most cases [10:21] not sure what your point is [10:22] I don't see why LIM wouldnt work with the app-menu's [10:22] oh! No, it works fine with app menus, even right now [10:23] the problem is that LIM draws menus in the title bars, which don't exist for CSD apps [10:24] larsu, ok, don't worry about me, I am too sick to really even think atm, bloody food poisoning or so :( [10:24] darkxst, no worries :) Hope you get better soon! [10:30] jgdx, re bug #1488087, which type would you like to see that translated to? QList> ? [10:30] bug 1488087 in gsettings-qt "Lack of support for signature “a(ss)”" [Undecided,New] https://launchpad.net/bugs/1488087 [10:31] second ass patch? [10:31] * darkxst out for the night, hoping a solid sleep will help a bit [10:31] ya :/ [10:31] robert_ancell: you could also try unity-system-compositor-spinner [10:32] larsu: welcome back to work! *thud* === hikiko is now known as hikiko|bbl [10:32] haha [10:32] thanks :) [10:33] larsu, :) [10:33] but the average client -o -f should be the same.. [10:33] larsu, works for me [10:33] jgdx, not sure if that is idiomatic qt, but seems like the closest match to me [10:33] do they have a tuple type? [10:34] tuple seems to be qtpair [10:34] seems weird to only make this work for pairs [10:34] ya, QPair [10:34] how do you mean? [10:34] pair is their tuple, no? [10:34] a(ss) good a(sss) bad? [10:34] asss is a typo [10:35] but you have a good point [10:35] QList then? That also seems weird. [10:35] but then, next up will be a(si) [10:36] maybe it's best to do what qtdbus does [10:36] larsu, could pass it as a variant [10:36] jgdx, the question what to put in there still remains :) === pstolowski is now known as pstolowski|lunch [10:37] they make you give a struct IIRC [10:38] ah, and you have to register that with their type system? [10:38] yeah [10:38] with the streaming operators [10:38] but how do they know where to put the values? [10:38] ah. right. [10:39] otherwise you can nest QPairs [10:40] which is a bit grim [10:41] I think I'm going to special case this to QList> [10:41] it's ugly, but unlikely that we'll see a(sss) [10:41] haha [10:41] I understand that [10:41] and when we do, we might want to think about doing something more elaborate [10:41] * larsu wishes QVariant were more like GVariant [10:53] qvariant doesn't seem to know about qpait [10:53] *qpair [10:56] ah, because it would need to know about all type combinations [10:57] if I declare my own metatype, will that be possible to use from within qml? [10:57] * larsu guesses no and goes for QList instead [11:04] robert_ancell: I thought we were discussing nested mir sessions with a client? [11:05] larsu: http://doc.qt.io/qt-5/qmetatype.html#Q_DECLARE_METATYPE says QPair is automatically registered [11:05] Laney, the compiler told me otherwise... [11:05] The system-compositor only expects fullscreen surfaces (associated with displays) [11:05] :-o [11:18] * Laney is on phone to Dell again [11:18] :-/ [11:18] computer broken? [11:21] jgdx, Laney, branch is up - reviews appreciated [11:23] not broken, but there is a loose screw inside and the speakers don't work [11:23] probably since the last engineer came to replace the keyboard [11:24] larsu: link? [11:24] never mind [11:27] Laney, ugh, annoying! [11:28] shows I don't use the speakers very much [11:28] this was 3-4 weeks ago [11:28] larsu: I feel like you could support any number of 's' now [11:28] would this be too annoying? [11:29] not really, but it would make all of this less readable [11:30] and I really think this is an exception [11:30] and noone needs anything but this for a while [11:30] haha [11:30] I guess this is about input sources? [11:30] you hope :) [11:30] I don't have an example [11:30] just trying to avoid future work [11:30] pretty sure I lobbied for a{sv} when desrt and attente decided this, but there was some reason that didn't work [11:31] this was discussed a lot [11:31] i think attente even sent a patch to stefw [11:31] but it never got reviewed/merged [11:31] desrt, making it an a{sv}? [11:31] aa{sv} iirc [11:32] wasn't there some problem about ordering? [11:32] oh wow, that wouldn't have helped us in this case [11:32] actually [11:32] aa{ss} i think we agreed on [11:32] * larsu just added a(ss) support to gsettings-qt [11:32] since 'v' is weird here [11:32] larsu, correct [11:32] stef was concerned about storing to a keyfile [11:32] I just bet that someone will come with a(sss) or a(ssssssssssssssssssssssssSSsssSSsssSSsssSSsssSSssssSSsssSSssSSSsssSSSSssss) in future and you'll need to generalise it anyway [11:32] but if you don't care about having to do it then, then whatevs [11:33] what's the size == 2 check for? [11:33] Laney, ha, there's actually *one* instance: org.gnome.yelp.gschema.xml: [11:33] more worried about phone people adding new random keys [11:33] Laney, not crashing. In all other cases we don't crash but return default values (variant.toXXX() returns a default-constructed thing) [11:34] and I'm accessing tuple[0] and [1] below [11:34] com.ubuntu.notifications.hub uses a(ss) as well [11:34] man, I think we might have to go all the way at some point [11:35] I bet you will [11:35] then you will need to re-load this into your head [11:35] * desrt swears loudly [11:35] Laney, no, I mean do what qtdbus does and expose every type [11:36] inotify is evil. [11:36] desrt, burnt fingers? [11:36] my new perfect approach has a serious problem :( [11:36] i make the assumption that dirs cannot be hard linked, and that's true [11:36] s/perfect// [11:36] but they can be bind-mounted... [11:36] in which case two separate paths can give the same wd, and indeed, do... [11:37] which is potentially a bug [11:37] since one of those two paths could be unmounted... [11:37] ...which, of course, results in no event being sent [11:38] grrrrrr [11:38] sorry :( [11:38] hrm. inotify seems buggy vs. bind mounts. [11:39] indeed, it seems that there is absolutely no way to detect when they appear/disappear [11:39] larsu, for convergence we need yelp bookmarks btw plz implenement [11:40] larsu, thank you for the MP. Appreciate it! :) [11:40] hah [11:40] you're welcome [11:40] great. bug filed about this upstream since march 2013 with no reply. === MacSlow is now known as MacSlow|lunch [11:41] desrt, fix it! [11:41] * larsu hides [11:41] long ago i promised myself never to hack on the kernel..... :/ [11:43] good times: open nautilus to /mnt, mount -o bind / /mnt, refresh nautilus, umount /mnt, refresh nautilus [11:44] manually refreshing works here [11:45] of course inotify doesn't [11:45] but to be honest, I didn't really expect it to [11:45] that's only because you have low expectations already :) [11:45] lol [11:45] to me it seems obvious that ^R should be completely redundant in nautilus [11:46] ie: you only need to use this if there is a bug or misconfiguration somewhere along the way [11:46] the one exception i may make (and even i consider this suboptimal) is network filesystems [11:49] I agree that this should be the ideal [11:49] not saying I don't want this to work, just that I didn't expect it to [11:49] :) [11:49] well [11:49] i guess i can always watch /proc/mounts and respond to that [11:50] so it's not 100% hopeless [11:51] in fact, both systemd and udisks tell me about this on the system bus, so that's even better [11:52] * desrt goes back to the drawing board for a moment [11:54] Laney, thanks for the review. Who can take care of landing this? [11:55] larsu: I guess jgdx with whatever wants to use it [11:55] or kenvandine [11:56] cool [11:56] * larsu is not a fan of this process [11:56] * Laney screams [11:57] dear dspam, please learn to learn [11:57] getting identical spams and it's not catching them after being trained [11:58] Laney: are they being marked as spam> [11:58] no [11:59] I've had it like 10 times now [12:00] uh oh [12:12] le sigh. Wifi. [12:16] willcooke, office wifi is pretty good, no? [12:16] well, it depends who you ask. [12:16] haha [12:16] Most people seem to be ok [12:16] but I'm getting kicked off every 10 mins or so [12:17] *This* is why you don't skimp on the wifi card spec'd in your laptop [12:17] i.e. Broadcom [12:17] vs Intel [12:17] ya... === pstolowski|lunch is now known as pstolowski [12:17] I looked at swapping it out [12:17] find the cable and guard it jealously [12:17] but I have to dismantle the entire laptop [12:17] which I can't be arsed to do [12:18] didn't you have an x220? [12:18] yeah, but I got the cheapo wifi [12:18] * larsu didn't know you could get them with shitty chips [12:18] https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1485659 [12:18] Ubuntu bug 1485659 in indicator-datetime (Ubuntu) "Date/time indicator doesn't update after changing time zone" [Low,Confirmed] [12:18] iz glib boog [12:19] Laney, it is, but indicator-datetime should not be using a file monitor for that [12:20] desrt is working on it :) [12:20] on what? [12:20] file monitoring [12:20] I would guess this code existed before timedated [12:21] I'm not so sure [12:21] it was rewritten once or thrice [12:22] probably directly ported though [12:22] * Laney is going to kill the file monitor [12:22] please do [12:23] hm, can I make xchat-gnome not show join/part messages? Very annoying and I won't have my irssi setup for another week or two [12:23] * larsu could run that locally [12:23] is polari any good? (and packaged?) [12:23] irccloud.com? [12:24] and I give them my password? [12:26] well, yeah, and all your conversations too [12:33] Laney, good catch that tz change to be a glib issue [12:33] I wouldn't have known where to start debugging it [12:33] I looked at the code and saw GFileMonitor :P [12:35] :-) [12:35] desssrrtttt === hikiko|bbl is now known as hikiko [12:59] robert_ancell: Can we try setting MIR_CLIENT_PLATFORM_LIB to /usr/lib/x86_64-linux-gnu/mir/client-platform/mesa.so.3? [13:05] did we ever find a solution for the desktop file renaming thing? [13:36] Laney: the packages seems work well here [14:10] tseliot: ok cool, will upload shortly [14:10] larsu: don't think so, someone needed to own the problem [14:10] Laney: thanks a lot! [14:11] * Laney lunch [14:14] Noskcaj, unsure of that's a parole or gstreamer issue but that started with your most recent update, bug #1488494 [14:14] bug 1488494 in parole (Ubuntu) "/usr/bin/parole:11:handler_unref_R:g_signal_handler_disconnect:parole_gst_shutdown:parole_player_quit:parole_player_stopped" [Undecided,New] https://launchpad.net/bugs/1488494 [14:14] Laney, I think seb128 renamed the gnome-terminal one, right? [14:14] I didn't rename anything [14:15] haha [14:15] desrt, what about that desktop name map in glib? Still up for it? [14:31] Laney, +1 on using the tz property, but we should also look at the GFileMonitor issue [15:00] seb128: ya, I think "we" are already ;-) [15:01] Laney, is there a bug open about it? [15:02] upstream [15:02] good [15:02] as look as we keep an eye on it and makes sure he doesn't fall through the cracks ;-) [15:04] pretty sure this is related to what desrt is doing right now [15:04] cool [15:29] tseliot: someone did it already? [15:30] Laney: did what? [15:30] uploaded u-s-d [15:31] *cough* [15:31] * Laney looks around [15:31] mumble grumble [15:31] wifi [15:31] #startmeeting Desktop Team Weekly Meeting - 2015-08-25 [15:31] Meeting started Tue Aug 25 15:31:46 2015 UTC. The chair is willcooke. Information about MeetBot at http://wiki.ubuntu.com/meetingology. [15:31] Available commands: action commands idea info link nick === meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-08-25 | Current topic: [15:31] Hello everyone! [15:32] Roll call: andyrock (out), attente, desrt, dgadomski, didrocks, fjkong, happyaron (out), hikiko, laney, larsu, qengho, seb128, sweet5hark, themuso (out), tkamppeter, trevino, robert_ancell [15:32] hey [15:32] Hi hi [15:32] hello [15:32] Hello [15:32] o/ [15:32] * Sweet5hark touching down after vacation [15:32] meeting time! [15:32] o/ [15:32] * larsu missed that [15:32] hey [15:32] hi [15:32] larsu: we don't believe you :) [15:33] hai [15:33] I think seb128 is in a meeting [15:33] hey (in a meeting yes, trying to follow up/post my update still) [15:33] Trevinho, but it's true. ish. [15:33] okay, let's go [15:33] :) [15:33] seb128, you can't schedule *a* meeting over *the* meeting [15:33] #topic andyrock === meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-08-25 | Current topic: andyrock [15:34] larsu: i'm guessing someone else scheduled it :) [15:34] * Investigate on the way compiz sets GL_PROJECTION matrix to properly [15:34] fix bug #1485727 and others compiz glitches. [15:34] * Code reviews [15:34] desrt, should be forbidden [15:34] bug 1485727 in unity (Ubuntu) "Corrupted unity desktop after resizing vbox/vmware window." [High,In progress] https://launchpad.net/bugs/1485727 [15:34] #topic attente === meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-08-25 | Current topic: attente [15:34] uh oh [15:35] attente is also in london, no? [15:35] yes [15:35] (or en route?) [15:35] give him the hard stare [15:35] a kick under the table [15:35] proposed a branch for packaging libertine containers in a click package [15:36] attempt at debugging https://bugs.launchpad.net/ubuntu/+source/xorg-server/+bug/1481330, no luck [15:36] Ubuntu bug 1481330 in xorg-server (Ubuntu) "Xmir crashes when client closes display" [Critical,Confirmed] [15:36] bisect it [15:36] uh oh [15:36] D: [15:36] poor attente [15:36] d:\ gets a bad rap these days [15:36] i remember i used to have some good files there [15:37] looking again at correct menu positioning in gtk-mir [15:37] (eof) [15:37] thanks attente [15:37] #topic desrt === meetingology changed the topic of #ubuntu-desktop to: Home of the Desktop Team, https://wiki.ubuntu.com/DesktopTeam | For help or questions, try #ubuntu | Breaking your systems: GTK 3.16 - file bugs with tag "gtk316" | Desktop Team Weekly Meeting - 2015-08-25 | Current topic: desrt [15:37] - attended debconf. seems they are trying to make it into a meeting-place type of conference and i met lots of interesting glib users there (tails, guile, others) [15:37] - hard not to notice all of the inotify bugs lately -- been working on a proper solution for that [15:38] . [15:38] desrt, guile uses glib? interesting... [15:38] the maintainer is doing something weird with searching [15:38]