[06:08] Hi [06:57] Good morning [07:37] morning pitti [09:02] hey hey [09:09] hi Laney [09:11] hey hikiko [09:11] happy fake friday [09:13] hahahaha [09:13] happy fake friday :D [09:24] ♫ ♪ Hot cross buns, Hot cross buns ♪ ♫ [09:27] Laney, any thought on this one: https://bugs.launchpad.net/ubuntu/+source/libunity/+bug/1538471 [09:27] Launchpad bug 1538471 in libunity (Ubuntu) "scope-runner-dbus.py assert failure: *** Error in `/usr/bin/python3': free(): invalid next size (fast): 0x0000000001ca17b0 ***" [Critical,Confirmed] [09:27] see comment #25 from jamesh [09:27] willcooke: I saw that, but not straight away [09:28] nw, thx Laney [09:28] probably bisecting / looking at git would be helpful [09:29] ricotz: ^? any idea? [09:31] willcooke: ♫ ♪ Stop! Hammertime ♫ ♪ surely that the easiest way to get someone nailed to a hot cross? [09:31] ooof [09:31] :) === king-afk is now known as cking [09:52] it's cool running metacity [09:52] muktupavels: the patch helps, think we need to scale decorationsn too though so trying to port that other one [09:53] Laney: port from mutter or my wip branch in metacity? [09:53] you did it already? [09:54] ah yeah 57c3fb46993a730f01ba8619832dde119f53c863 [09:54] what is wip about it? [09:54] for 3.20, but I have made many changes so it wont apply to 3.18 [09:55] mmm, would like it on 3.18 for xenial [09:56] I have not tested it on HiDPI display, only by changing scaling-factor in dconf-editor... [09:57] I can check it [09:57] it needs GTK+ 3.20 [09:58] * Laney will backport the patches [09:58] Also in 3.20 metacity by default now use GTK+ theme. You will need change settings under org.gnome.metacity.theme. [10:11] flocculant: you about? [10:18] willcooke: new issue I just went to install from pendrive on hardware and I'm getting Failed to create swap space I assume because it is mounted [10:22] how the hell do I access a tty from the ubiquity session [10:32] davmor2, CTRL+ALT+F1 [10:36] jibel: that is the ubiquity session and F2-7 give you a black screen with a flashing cursor you can't type in [10:39] ctrl-alt-t ? [10:45] hikiko, hi! Any news on that upgrade crasher this morning? [10:46] oh willcooke [10:46] I was about to ask you on that [10:46] that was spooky hikiko :) [10:46] well so far: [10:46] andyrock said he was going to do some more playing last night [10:46] I fixed 1 post inst script of udev (andy suggested that the udev restart might be relevant) but it didnt fix it just a warning [10:46] but [10:47] I installed i386 [10:47] and upgraded successfully [10:47] so I think that what I told you yesterday about libc [10:47] might be the problem [10:48] (libc went multi arch, maybe at the time we get the crash the new libc is installed but the dynamic loader has the previous libc paths) [10:48] so, I am now trying to test libc post-inst scripts [10:48] the thing is [10:48] if we know the libc maintainer... [10:49] he might be able to test that [10:49] much faster than me [10:49] because I am not so familiar [10:49] and I google everything [10:49] Laney, any ideas who could advise? ^ [10:49] someone who knows about libc [10:49] also there might be something more simple than that like they forgot to run an ldconfig or so [10:50] about libc packaging [10:50] not libc [10:50] willcooke, jibel: http://people.canonical.com/~davmor2/vids/video20160324_104355766.mp4 [10:52] anyway, I am going to look at it today if we don't know who is the maintainer [10:52] ogra_: ubiquity session not live desktop session so that doesn't work [10:52] davmor2, lets see if cyphermox has any ideas when he comes online [10:52] * willcooke has a feeling that today is going to be a long day [10:53] willcooke: I knew that Monday [10:53] :) [10:56] davmor2, https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1552539 [10:56] Launchpad bug 1552539 in ubiquity (Ubuntu) "Ubiquity Erase Disk and Install Fails to create Swap Space" [Critical,Triaged] [10:56] pitti, any clues with that one? ^ [10:57] willcooke: infinity is the maintainer of glibc for us [11:00] thx Laney [11:01] hikiko, could you ping infinity when he's on later? [11:02] sure willcooke :) thank you! [11:03] * ksamak yawns [11:04] * ksamak finished some more modifs to the zoom algo on compiz [11:13] davmor2, we should start collecting a list of all of these issues. Shall we just a LP tag? Maybe you already have a process in place? [11:13] jibel, ^ [11:17] willcooke: I'll be adding it to the iso tracker in a second just caught up on something else for a second [11:18] willcooke, all the issues should be linked on the iso tracker [11:18] what davmor2 said [11:18] ta :) [11:19] willcooke: http://iso.qa.ubuntu.com/qatracker/milestones/358/builds tag to look out for is iso-testing [11:27] muktupavels: if you have time, could you check out wip/iainl/gnome-3-18-window-scaling please? [11:28] I'm applying it twice somewhere, the titlebar text looks to be 4× [11:28] and the button spacing too, but maybe the same problem [11:29] Laney: ok, I will check [11:30] muktupavels: feel free to push fixes to it [11:30] thanks! [11:39] Sooooo [11:39] quick recap [11:40] hikiko has a theory that ldconfig isn't being called [11:40] 2 theories :) [11:40] and it looks like a change along these lines went in yesterday: [11:40] https://launchpad.net/ubuntu/+source/glibc [11:40] * Revert dropping the ldconfig wrapper, xenial still has a lot of [11:40] packages that don't ship a trigger but instead call in postinst. [11:40] * Add more belocs options to locale-gen to appease our installers. [11:40] Published on 2016-03-23 [11:41] oh, yeah, hikiko you're right it says release in main [11:41] let me re-upgrade [11:41] if this doesn't fix it [11:41] we have to check if they've updated the paths to be multi-arch everywhere [11:42] ack [11:42] to see if the dynamic loader looks at the right place [11:47] willcooke: on a plus side netboot just passed [11:47] nice! [11:47] ship it [11:58] hello ubus [11:59] how desrt [11:59] howdy [12:00] I know I'm a bad typist, but I think the batteries might be running out in this keyboard [12:00] you use wireless? [12:00] * desrt makes plans to set up a sniffing device [12:00] *shame* [12:00] yeah [12:01] it's ok though, the batteries are nearly flat so it has a range of about 3 inches at the moment ;) [12:03] willcooke: buy the logitech solar keyboard no batteries running out there :) [12:04] * willcooke makes a note that davmor2 doesn't do work once its dark [12:04] as I always expected [12:04] lol [12:05] willcooke: I have spots and currently I'm actually using the xps13 laptop keyboard cause you know it's attached to the device and stuff [12:24] willcooke: unity8 session I just get a black screen once I get passed the lightdm greeter page [12:24] davmor2, which gfx card? [12:24] willcooke: intel but on prime setup [12:25] could be prime drivers causing it [12:25] willcooke: however no nvidia bits installed [12:25] oh [12:25] still could be related [12:26] also, sometimes on first go around it doesn't start [12:26] willcooke: that was going to be my test install unity8 session then try and install the nvidia binary to test the additional drivers [12:26] try restarting lightdm [12:26] only Intel is supported atm, AIUI [12:26] willcooke: rebooted same thing [12:26] so might not be much point there [12:27] willcooke: oh wait it's here no [12:27] +w [12:27] yeah [12:27] it does that [12:27] that's why its in universe [12:27] :) [12:28] davmor2, I would suggest you put U8 session at the bottom of the list. It will be very good to know whats broken, but right now all I really care about is the default session [12:28] until that works, everything can wait [12:29] tea, brb [12:29] drinking tea, not eating tea [12:29] * desrt wonders why there are not nfc(ish) keyboards [12:30] for a long time i used one of these things: https://s-media-cache-ak0.pinimg.com/736x/c9/bc/bb/c9bcbb39b9b1bfe3b52c170d12350767.jpg [12:30] there's some serious potential there [12:31] security issues would be vastly reduced and i'm pretty sure it could be done without batteries [12:32] Laney: pushed one commit, should be better, but... Ambiance theme still does not work good, at least with 3.20 it looks better. :( [12:34] Laney: for gnome-3-18 branch use theme: prefix not libmetacity: [12:35] muktupavels: ok, I copied and pasted it :P [12:43] muktupavels: title text is still too big, button size looks right though [12:44] Laney: I know, there is also problem with left, right border [12:44] how is the text scaled? [12:47] meta_frame_style_apply_scale? [12:49] Oh, wait. It is not used with metacity theme... [12:51] Laney: pushed fix for font size. [12:51] divide by the scale in meta_theme_get_title_scale ? [12:51] yes [12:51] :) [12:51] * Laney tries that [13:00] davmor2: generally not during day till lunchtime ish ... [13:04] flocculant: no worries there was already a bug about it so it is fine now :) [13:05] muktupavels: nice! [13:05] I assume it was some sort of upgrade thing :p [13:05] muktupavels: do you want this pushed to 3.18? [13:05] flocculant: nope there is a bug in swap formatting [13:06] oh right [13:06] willcooke: bug 1546388 [13:06] bug 1546388 in unity-control-center (Ubuntu) "unity-control-center crashed with SIGSEGV in _cogl_check_extension()" [Medium,Confirmed] https://launchpad.net/bugs/1546388 [13:06] Laney: probably, but it is not ready yet... [13:06] willcooke: just been bitten by that on the real hardware install [13:06] Does right/left border looks good to you? [13:06] good in what way? [13:07] I don't see any particular problem there [13:07] I think it's upscaling low resolution images for the window controls though [13:08] they look blurry [13:08] davmor2, does it work if you relaunch it? Is it broken all the time? [13:08] willcooke: all the time [13:08] davmor2: fresh install? [13:09] can you show screenshot with how it looks for you? [13:09] Laney: yeap amd64 bit on my Lenovo ideapad [13:09] k, talk to tjaalton [13:09] muktupavels: okay [13:09] Laney: crash triggered took me to that bug [13:13] muktupavels: http://people.canonical.com/~laney/weird-things/metacity.png <- there's a 1px line below the title bar [13:28] Laney: there is also 1px line at left sidze of title bar... [13:28] * Laney stares [13:29] that is hard to see [13:31] cyphermox: what's a way to install a package to get it used for only-ubiquity mode? [13:31] seems vt switching doesn't work [13:31] will a usb drive work if I break=bottom?!?! [13:31] yeah, it should [13:32] or you could just download the package [13:32] hm, maybe there is no network though ;) [13:32] no network [13:32] heh [13:43] how long can dd-ing /dev/zero to a stick actually take? [13:44] * Laney checks he didn't type /dev/sda [13:44] a bit late :P [13:44] :D [13:44] TELL MY WIFE I LOVE HER [13:44] a huge and slow stick tend to take its tme [13:45] it's only 4gb [13:45] Laney, really depends on the blocksize you picked [13:45] probably just slow and crap [13:45] 8M [13:45] that will surely taske 10-15min [13:46] been like 25 [13:46] whatevs, it's still flashing [13:46] the activity light thing on it [13:47] for me 4GB with bs=64M usually takes 3min on my Sd card reader [13:47] max 5 if i do other busy IO stuff [13:47] if i turn the blocksize down it takes significantly longer [13:48] I picked 32M the other way when writing an ISO and my system went crazy [13:48] scared now :) [13:48] well, depends on your ram and IO throughput [13:48] (hardware thing) [13:48] yeah thought it would be good enough [13:48] new-ish xps 13! [13:49] most likely your usb-stick has a problem then [13:50] as long as it can get a metacity deb into this ubiquity session [13:50] then it can explode into flames for all I care :) [13:50] Laney, btw is there way to update the thunderbolt firmware with linux on a xps 13? [13:50] dunno [13:51] hmm, bios itself is, but this intel thunderbolt thing :\ [13:51] I haven't used it [13:51] did update the bios a few times though, touchpad woes on the initial firmware [13:52] I see, installing windows for that seems overkill [13:53] you can just put the exe on a stick and do it direct [13:53] Trevinho: remind me of the gtk branches to upload please? [13:54] Laney: I wanted to, but I didn't want to bother too :) [13:54] I want to be reminded now [13:54] Laney: https://code.launchpad.net/~3v1n0/gtk/unity-maximized-headerbar-buttons-hide/+merge/288552 [13:54] need to get everything done today, on holiday for 8 days after tonight [13:54] thx [13:54] does that make this toolbar-mode thing work? [13:55] Laney: yeah, I heard that [13:55] Laney: yes :) [13:55] ok [13:55] Laney: it also needs a theme [13:55] I have the theme [13:55] Laney: it's already in ci-train [13:55] Laney, the thunderbolt firmware is not part of the bios afaik [13:55] silo 067 [13:55] I mean, a theme change [13:55] try it [13:56] Laney: it's in https://requests.ci-train.ubuntu.com/#/ticket/1173 [13:56] nope [13:56] Laney: I'm speaking of the ubuntu-themes change [13:57] https://requests.ci-train.ubuntu.com/#/ticket/1182 [13:57] Laney, oh, btw: pkill -USR1 ^dd$ [13:57] (from a second terminal) [13:57] what does that do? [13:57] will give you the current status [13:57] huh [13:57] good tip! [14:01] how could someone check if a mirror is updated (apt repo)? [14:03] willcooke: would you like to review https://code.launchpad.net/~3v1n0/ubuntu-themes/toolbar-mode-headerbar-gradient/+merge/288701 (here's how things would look http://imgur.com/a/VETka)? [14:05] Trevinho, sure, one mo === Drac0 is now known as Guest8381 [14:18] Trevinho: it makes csd apps get a white border when not maximised? [14:19] Laney: white border? No... [14:21] Trevinho, Laney same - white [14:22] that's only using that theme? [14:22] as it shouldn't change anything when not using the gtk patch [14:22] probably a different one in the landing [14:24] yeah [14:24] rebuilding it without one of the mps [14:25] Trevinho, also - what am I looking for, the gradient should be light -> dark in which direction? [14:25] willcooke: in which widget or state you mean? [14:26] Trevinho, when a csd headbar window is maximized [14:26] you need the gtk patch? [14:27] oh [14:28] this one: https://bugs.launchpad.net/ubuntu/+source/gedit/+bug/1515810 ? [14:28] Launchpad bug 1515810 in gtk+3.0 (Ubuntu) "Doesn't merge into unity panel when maximized" [High,In progress] [14:28] well, the gtk patch is needed to get that effect. But without the patch you shouldn't get any breakage [14:29] oki, well on my fairly standard test install I'm getting white corners [14:29] willcooke: when the window is maximized (rule .maximized .titlebar.header-bar.toolbar-mode:not(:backdrop)), the bg-image should use a gradient tat goes from dark to lighter [14:30] Maybe I'm running out of date versions then [14:30] I'll upgrade again [14:31] willcooke: the needed gtk is not landed yet... it's what Laney has to do... So maybe we can wait his package and then you can try it later [14:31] works for me :) [14:31] does it fix white corners too? [14:31] Laney: corners are fixed for everything... BUt, gnome-calendar has an issue that is specific to that as well [14:31] I am using calendar to test [14:32] Ok, so calendar is the only one broken I know [14:32] but it's something related ot its theming [14:32] we discussed about that with seb few days ago. Basically if you set background-color: transparent to the gtkwindow is fine [14:32] for some reason it picks a different bg color [14:33] so use gtk3-demo or gtk3-widget-factory to test [14:33] or any other headerbar app [14:33] I have a calendar theme fix for highlighting the current day, I could put that theme fix in there as well [14:33] (with this change, we could probably even avoided to patch gedit and others :P) [14:33] willcooke: would be nice [14:34] Trevinho, I will wait for the gtk fixes to land, and then I'll take a look [14:34] but I've to figure out what's the best way to do [14:34] I'm proposing to add a calendar specific css file [14:34] if that makes it any easier? [14:36] https://code.launchpad.net/~willcooke/ubuntu-themes/ubuntu-themes-calendar-focus [14:39] willcooke: the workaround for calendar corners is [14:39] GcalWindow { background-color: transparent; } [14:39] but I'm not happy with that, as I'd prefer to fix the root cause... [14:39] for some reason it doesn't pick some of the default values [14:40] oki, well let me know when you decide, I can drop it in there easily [14:42] I uploaded gtk, won't go in until after beta though [14:43] Laney: thanks [14:49] willcooke: I can't find anything better than that right now :( [14:50] in a whole 10 minutes! [14:53] If there isnt a better alternative I will push a new version of the theme with that fix in, but not MP it [14:53] well, I'll push it now, so I dont forget [14:53] and then if we need it, it's there [14:55] Trevinho, sitting here if we need it: http://bazaar.launchpad.net/~willcooke/ubuntu-themes/ubuntu-themes-calendar-focus/revision/469 [14:56] it won't harm in case.. I mean, if we find a better fix, they won't clash [14:56] Thanks [15:01] Trevinho: It's because it sets a background-color on the window [15:01] Laney: the app itself? [15:01] it has some css [15:01] ok, i was guessing that [15:02] no need to guess, we have the source :) [15:02] .views { [15:02] /* background color of every view */ [15:02] background-color: @theme_base_color; [15:02] background-image: none; [15:02] } [15:02] Laney: hehe, sure... I just didn't check that before [15:03] Laney: so, what you'd say.. Is it better to fix that css not to include the window or we keep the workaround in themes? [15:04] Laney: I'm just wondering why this won't affect adwaita though [15:05] mh, it might be some other parameter [15:06] It's really a compiz bug that it is drawing there no? [15:06] Oh, no adwaita is not fine as well, I think [15:06] Laney: no... It's the app that paints there === francisco is now known as Guest26205 [15:07] App has a rgba surface, it does whatever it wants on that area. Compiz only draws a shadows underneath it. [15:07] It's correct on gnome-shell [15:07] Laney: using adwaita or ambiance? [15:07] Adwaita [15:08] ah, ok... For Adwaita is a different thing then... But it shouldn't be anything related to compiz [15:09] Well it's wrong with both Adwaita and Ambiance under compiz [15:09] and right with Adwaita under shell [15:09] so... [15:10] It's a different thing... With different themes also other apps are wrong with Adwaita.. [15:10] under unity I mean [15:21] Laney: also in gtk3-demo... Increasing the border radious you can see that there are some tiny corners on top. Which I don't really know who is drawing them: http://i.imgur.com/WybfxIa.png But it's not compiz, for sure [15:22] the fact that the gtk magifier shows them as well, ensures that: http://i.imgur.com/J5HvZ9D.png [15:31] infinity, [15:31] here? [15:36] hikiko: ? [16:22] Anybody who knows who maintains this site: [16:22] https://apps.ubuntu.com/cat/applications/synaptic/ [16:22] Asking because we link to that page (and a few other pages on the same site) from the web version of the desktop guide, and there seems to be an update need (versions, USC -> the new thing). [16:26] Trevinho: umm, if not compiz then who? [16:26] Laney: it's gtk the for some reason paints that... I wasn't able to figure out where and why [16:26] #gtk+ to the rescue! [16:27] it indeed does that only in compiz, probably because of our patch,... But, I'm not able to track it down [16:29] willcooke, cyphermox: so guys given the 5 critical bugs and the control center opening issues and you still can't upgrade, what is the plan going forward? [16:29] infinity: ^ [16:32] oh and mokutils issue from additional hardware installs [16:35] davmor2, let's move this to -release === ochosi_ is now known as ochosi [17:09] Laney: can you test your branch with cherry-picking 3932dca006d845aa8a85d29d38dcf3f100959e32? [17:10] muktupavels: you think this is going to fix the borders? [17:11] Laney: it improves decorations for me... [17:11] let me try it [18:17] Laney: should I push your branch to gnome-3-18? I have updated commit message with commits from master branch. [18:19] muktupavels: It didn't seem to fix the border for me anyway [18:20] still 1px? [18:20] looks the same [18:21] it fixed problem that I was seeing. [18:22] I'm running out of time so I'll upload it to xenial anyway [18:22] way way better than it was before [18:23] Laney: ok then I will push your branch now to gnome-3-18 [18:24] Laney: do you need your wip branch or can it be deleted? [18:27] muktupavels: if you push it there's no need to keep it any more [18:28] Laney: ok, then I will delete it. [18:36] Laney, thank you very much for the pygobject [18:50] willcooke: np, also fixed the hidpi one [18:50] Laney, \m/ [18:50] I'm going to have to come back tomorrow morning to upload gstreamer and fix appstream [18:50] thanks a million laney [18:51] umm [18:51] willcooke: I have left two things hanging :/ [18:51] maybe you want to find someone to look at them next week [18:52] I can certainly try [18:52] 1. glib2.0 hangs on s390x in its testsuite on the archive builders but I *can* *not* reproduce it [18:52] that's a blocker to get the stable version in [18:52] xnox is away too I'm afraid [18:53] I synced the stable into unapproved [18:53] so we'll get one more build on the freak chance it started to work [18:53] but I doubt it [18:53] :) [18:53] maybe someone on foundations could look [18:53] kk [18:53] and the second one is that I totally failed to sponsor any appstream icons [18:53] Laney, is there a bug for it? [18:53] no [18:53] np [18:54] lemme open one [18:54] kthx [18:54] I reckon it's not glib's problem [18:54] it builds on debian/s390x [18:54] probably a kernel thing [18:54] weird [18:56] Laney: If it builds on debian but not ubuntu, kernel's one option, the other is -fPIE -pie [18:56] Laney: Our s390x toolchain defaults to pieness, Debian's doesn't. [18:57] infinity: I'm brain dumping due to going away for a week [18:59] infinity: https://bugs.launchpad.net/ubuntu/+source/glib2.0/+bug/1561698 [18:59] Launchpad bug 1561698 in glib2.0 (Ubuntu) "Build hangs on s390x" [Undecided,New] [19:04] Laney: the new appstream-generator generates HTML info pages too now - might be just a few weeks until I can try it in a production-like scenario [19:05] and I should merge some of the pull-requests for the dep11-generator [19:05] the "suite" thing looks sane for the dep11-gen, for asgen I hope it won't be necessary to implement the same thing [19:06] ximion: nice, not here now though ;-) [19:06] I'm going to merge them all into my branch and run a fork I think [19:07] Laney: sounds reasonable, for the shorter term [19:07] nod [19:08] long-term, I will make asgen the default at Debian, because it really is a bad idea to have two implementations of the same thing, especially when all the parsers and data generators in libappstream outperform the Python stuff [19:09] Laney: for the gettext-foo in dep11gen, I hope I can implement that in the new code easily (it's heavily Ubuntu-specific, so kind of low priority...) [19:10] asgen will have a way to enable and disable certain featuresets depending on the needs of the project using it, so you can e.g. disable processing of .desktop files w/o metainfo files [19:10] low priority for you maybe :) [19:11] priority isn't zero though, I care for Ubuntu ^^ [19:11] with the feature profiles in asgen, implementing distro-specific stuff should become really easy and less controversial [19:13] Laney: if I make it to London, we can talk about this more - maybe I'll have a fully working implementation by then, even - only screenshots stuff and result-compression is needed to get feature-parity :) [19:13] oh, and I need to clean up the code a little, so creating a .deb package will be asier [19:14] *easier [19:14] ximion: yeah, will be good to see it in action [19:14] ok, I'm off for the evening [19:15] back quuuuuuuuuuickly tomorrow to do a couple of things, but otherwise [19:15] tty in a week and a bit [19:15] have fun and a nice free time :) [19:15] I'll continue to nag hughsie and mhall119 meanwhile ^^ [19:16] good luck! [19:16] thanks! :) [19:40] desktoppers - anyone who can, please help with i386 ISO tests: http://iso.qa.ubuntu.com/qatracker/milestones/358/builds/115349/testcases === Guest31325 is now known as fredp === fredp is now known as Guest55861 === davmor2 is now known as davmor2_HOLS