=== tedg__ is now known as tedg_ === dobey_ is now known as dobey === maclin1 is now known as maclin === happyaro1 is now known as happyaron [08:59] Hey willcooke, hey folks. [08:59] hey TheMuso === faenil_ is now known as faenil === faenil is now known as faenil_ [09:02] muhahaha [09:04] hey TheMuso willcooke Laney [09:06] Morning all [09:06] Trevinho, bad news [09:07] your fix doesn't fix the segfault :/ or it's another one gnome session related [09:07] if I restart u7 40 times :p [09:07] there was 1 caused by a tex missing in decorations, I've fixed that [09:08] and then I got again the gnome session related [09:08] hey seb128 hey davmor2 hey hikiko hey TheMuso hey willcooke!!!!! [09:09] hi Laney seb128 davmor2 TheMuso willcooke flexiondotorg etc [09:09] I am sure I forgot to hi-light someone [09:09] :-) [09:09] Morning hikiko and desktopers [09:09] Hello TheMuso [09:09] Morning Laney how grim is it up north [09:10] davmor2: pretty grim actually [09:10] there's been this mist hanging around for like 4 days now [09:11] mist sounds better than -20C [09:11] Laney: yeah we got that today [09:11] ogra_: nice! [09:11] well, not anymore but this night was *cold* [09:12] although, not sure I've ever actually experienced -20 [09:12] we have that every few years for a day or two [09:12] Laney: it's cold you don't want to [09:12] mmm no, Trevinho it seems to be a dbus lambda, different :) [09:18] heh I've experienced -20!!! once, in Greece :p I kno, you 'll never believe me though :p (but we have some regions that get that low temperatures for a few days every few winters :P) [09:20] hey TheMuso willcooke Laney seb128 davmor2 hikiko flexiondotorg ogra_ hikiko [09:21] hi happyaron :) [09:22] and ogra_ [09:22] :) [09:24] happyaron o/ [09:25] :) === bigon_ is now known as bigon [10:18] hmm, hmm [10:27] good morning Sweet5hark [10:29] Sweet5hark, i have a clash for the DF advisory board =( [10:41] hikiko, Trevinho, seb128 - are you guys free for a quick hangout now? [10:41] want to get this low gfx thing sorted asap [10:41] yes [10:42] xnox: huh? [10:44] document foundation advisory board [10:45] * Sweet5hark was distracted by a build system bikeshed on #libreoffice. [10:46] xnox: ah: clash in the sense of a scheduling conflict? [10:46] * Sweet5hark is utterly underprepped for that call too. [10:49] yeah [11:01] in other news, my local 5.3.0.1 build finished somehow ... [11:02] dpkg that is. [11:03] while 5.3.0.2 snap looks more curious: success on amd64 \o/ .... [11:03] ... but fail on armhf and i386. [11:05] both with different test fails :/ [11:06] eh, retry. the windows administraitors secret weapon. [11:17] * flexiondotorg nips out for early lunch === hikiko is now known as hikiko|ln [12:41] Sweet5hark, hi, please let me know if there is a deb-source-package up === maclin1 is now known as maclin [12:51] willcooke, sorry was at lunch [12:51] seb128, np, i think its ok [12:51] did you have it? I don't think I'm key in that discussion [12:51] ok = in hand [12:51] k [12:51] good! === jdstrand_ is now known as jdstrand [13:35] willcooke: sorry, missed that... [13:35] Trevinho, np. I think hikiko|ln is all set [13:35] hikiko|ln: did you get any valid stack trace? === hikiko|ln is now known as hikiko [13:36] sorry I was for lunch [13:37] Trevinho, this method doesn't work not even for ccsm (see mail) [13:37] but I discussed with willcooke and I am going to try to fix/extend libcompizconfig to do what I want [13:37] then if I fail we could do a hangout and see next step [13:38] check in a guest env yourself [13:38] hikiko: the main problem I see with the ini files, is that we're switching to something that loads settings from some .ini configuration, instead of gsetttings [13:39] and I'm not sure then ccsm how will act, and what profile will really use [13:39] as writing on Default.ini means making ccsm to use it, instead of the unity(-lowgfx) one [13:40] So... To me, if that change doesn't work, it will on reload... Which we might still trigger using upstart/systemd in case... [13:41] i kno [13:41] I changed that [13:41] it still doesnt work [13:41] ccsSetBackend fixes that [13:42] I ve found something else worth investigating now [13:42] basically I am looking at the code that loads the Default.ini [13:42] hikiko: the thing is that if that setProfile doesn't work, it would be worth understanding why it's that... [13:42] that's where I have to start i thing [13:42] yup [13:43] and why only when Default.ini is changed [13:43] the change takes place instantly [13:43] Personally I didn't think it was a problem to even say users to logout to get the session properly set, but... Indeed it's better if that's not the case [13:43] and the other times it doesnt [13:43] this is what I am investigating [13:43] hikiko: I think it could be caused by the gsettings backend making something weird [13:43] yeah [13:44] as in that case it's using the .ini one I guess [13:44] and.... We don't want it [13:44] I am using gsettings [13:44] not ini [13:44] noticed that too [13:45] that's a good point but not the only problem [14:10] i love gsettings users! [14:10] er.. i mean.. good morning! [14:10] happy thursday [14:16] hey desrt, happy thursday to you as well === JanC_ is now known as JanC [16:10] seb128: barry is having a look to the fix for update manager [16:11] \o/ [16:18] davmor2, great, thanks for poking him about it [17:29] Laney: due to the ongoing complaints of people about AppStream downloading too much, I will disable icon-downloads in the AppStream package, and have them enabled by GUI tools explicitly [17:30] that way, the icons are only downloaded if they are really needed [17:30] which means that when you switch to AppStream 0.10.6, you'll also need to ship a small apt.conf snippet with GNOME Software [17:31] hi ximion [17:31] ongoing by who? [17:32] Laney: bug reporters, people on IRC, ... [17:32] riiiiiiiiight [17:32] this will not stop people complaining, but will make them complain at the right place ;-) [17:32] well I wouldn't do this in a stable series if I were you [17:33] also, the real reason why I do this is for server people and others using AppStream to find firmware on non-GUI systems. If there is no GUI, there is no point for downloading icons [17:33] I also hope this change will make it easier later to enable/disable HiDPI icons [17:34] the impact on Debian should be zero, since we will upload a new GS for Stretch in time [17:35] you should do it with a Breaks on broken SCs imho [17:36] those opinions aside, seems fair enough to me [17:36] Laney, attente, ximion, did you see any report/discussions about fd leaks in g-s? [17:36] no [17:37] hum, k [17:37] bug #1648534 is the most reported issue on e.u.c [17:37] bug 1648534 in gnome-software (Ubuntu) "/usr/bin/gnome-software:5:g_wakeup_new:g_main_context_new:g_dbus_connection_send_message_with_reply_sync:g_dbus_connection_call_sync_internal:g_dbus_connection_call_sync" [High,New] https://launchpad.net/bugs/1648534 [17:37] Laney: jup, I am thinking about that, but I might accidentially break Ubuntu's gnome-software that way [17:37] don't follow upstream bugs though, so there could be one there [17:37] but doesn't seems to do it for robert_ancell or me [17:37] haven't seen any either [17:38] there is over 1.8M reports [17:38] https://errors.ubuntu.com/problem/70c23d0f4e2be24b26672427d4218dc8f0823597 [17:38] wonder what's going on [17:43] why are almost all the incoming requests for an old version? [17:47] good question [17:47] could be a liveCD issue [17:47] is that the version on the iso? [17:48] it is [17:48] but if you click on the reports some of them are from installed systems [17:48] users don't do updates... [17:48] is it being stopped by phased updates? [17:49] * Laney tries to remember how to check [17:49] Laney, http://people.canonical.com/~ubuntu-archive/phased-updates.html [17:49] http://people.canonical.com/~ubuntu-archive/phased-updates.html [17:49] hmm nope [17:49] it's not on there so I guess not [17:49] wtf [17:49] I would just say that users don't do updates [17:50] we have issues with update-manager not auto-opening often in xenial [17:50] meh [17:50] well the numbers look good for it being better [17:50] but hard to know [17:50] and the gnome-software/notify-osd non interactive notifications are probably not very sucessful on getting them to open the upgrade tools [17:50] yeah :-/ [17:53] hope someone's looking at those update problems ;-) [17:55] we fixed some, unsure what's the situation today [18:06] * Laney posted a quick summary on the bug, maybe robert_ancell can confirm that if he tries [18:06] nighty night [18:07] night Laney & desktopers! === faenil_ is now known as faenil [18:47] night all [22:50] cyphermox: merge proposal for Humanity to add @2x icon support: https://code.launchpad.net/~dmj726/humanity/hidpi-2x/+merge/315166 [22:50] Fixes a couple hidpi bugs: [22:50] https://bugs.launchpad.net/ubuntu/+source/humanity-icon-theme/+bug/1657863 [22:50] https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1622686 [22:50] Ubuntu bug 1657863 in humanity-icon-theme (Ubuntu) "Icons are too big or the wrong icon is shown on hidpi screens" [Undecided,Confirmed] [22:50] Ubuntu bug 1622686 in ubiquity (Ubuntu) "double header in 16.10" [Medium,In progress] [23:38] dmj_s76: we should get input from the desktop team, because this won't only affect the installer. [23:39] certainly [23:40] This will affect how icons look on hidpi systems (making them look the same as on lowdpi systems) [23:42] Right now the Humanity icons look noticeably different on hidpi systems. [23:43] Icon size issues, icons in buttons using different metaphors on a hidpi vs lowdpi screen, icons generally having much thinner strokes on hidpi than low. [23:44] Adding @2x support fixes those issues while keeping lines and curves nice and sharp on hidpi screens.