=== maclin1 is now known as maclin
[05:27] <FourDollars> thunderbird-locale-zh-hant is lacking in http://people.canonical.com/~ubuntu-archive/seeds/ubuntu.bionic/desktop.minimal-remove.
[07:28] <didrocks> good morning
[07:28] <jibel> good morning
[07:30] <duflu> Morning didrocks, jibel
[07:30] <duflu> Ping tjaalton?
[07:32] <duflu> ping RAOF
[07:33] <didrocks> hey duflu, jibel
[07:34] <RAOF> Yo yo
[07:35] <duflu> RAOF: This fancy new glvnd stuff... do you know how to make it not ignore LD_LIBRARY_PATH (if I build a local mesa)?
[07:36] <RAOF> There's probably an environment variable you can use...
[07:36] <RAOF> I don't know off hand, though.
[07:36] <RAOF> grep for getenv 😀
[07:37] <duflu> RAOF: I've been searching the source for such things but it looks nontrivial and indirect
[07:37] <duflu> Relies on json text files and such
[07:37] <oSoMoN> good morning/afternoon desktoppers
[07:37] <duflu> Hi oSoMoN
[07:37] <oSoMoN> hey duflu
[07:37] <RAOF> EGL is driven by some json in /usr/share
[07:37] <tjaalton> duflu: I don't know either
[07:38] <duflu> tjaalton, OK. Yeah I can't bisect Mesa till I can convince the system to use my built Mesa :P
[07:38] <RAOF> No getenv calls at all?
[07:39] <RAOF> Odd
[07:39] <didrocks> morning oSoMoN
[07:39] <duflu> RAOF: Yes, but only for config dirs, that are expected to contain text files, which contain file names, not directories
[07:40]  * duflu stops complaining and starts guessing
[07:41] <RAOF> Does libglx communicate the right libgl to glvnd? I think that's the flow.
[07:43] <duflu> RAOF: glvnd provides the one and only /usr/lib/x86_64-linux-gnu/libGLX.so.0
[07:43] <duflu> And vendors provide files by different names
[07:46] <tjaalton> LD_PRELOAD doesn't work?
[07:47] <tjaalton> maybe harder to get gdm on it
[07:47] <duflu> tjaalton, yeah I vaguely recall gdm is a problem. Remind me why? Does it clobber LD_* ?
[07:48] <tjaalton> just guessing, since it should work for testing failing apps/games with 'LD_PRELOAD=libfoo app'
[07:48] <tjaalton> but in this case it's gdm and/or gnome-shell that would need that
[07:54] <tjaalton> or cut corners and only point i965_dri.so to the bisected build, don't think the other libs would care
[07:55] <duflu> tjaalton, using just the drivers last week yielded no commit without the bug. So I'm trying with full Mesa
[07:56] <tjaalton> ok
[07:57] <duflu> tjaalton, OK. Wedging the dir of a local Mesa build in here works: /etc/ld.so.conf.d/x86_64-linux-gnu.conf
[08:00] <oSoMoN> salut didrocks, ça va?
=== pstolowski|afk is now known as pstolowski
[08:15] <didrocks> oSoMoN: ça va bien, bon week-end, et toi ?
[08:16] <oSoMoN> didrocks, très bien, super week-end, la forme!
[08:39] <tsimonq2> Belated morning o/
[08:46] <oSoMoN> good morning tsimonq2
[08:52] <tsimonq2> o/ oSoMoN
[09:00] <willcooke> morning all
[09:02] <Laney> hey
[09:12] <didrocks> hey willcooke, Laney
[09:16] <willcooke> hi didrocks, how goes?
=== czajkows1i is now known as czajkowski
[09:18] <didrocks> willcooke: good, but getting colder again. Yourself?
[09:18] <czajkowski> morning
[09:19] <willcooke> didrocks, yeah cold :)  Kids & Wife had sickness bug over the weekend.  So far I've avoided it.  Washing my hands in bleach.
[09:19] <willcooke> hi czajkowski
[09:20] <didrocks> willcooke: urgh, good luck with avoiding getting cotaminated :p
[09:21] <didrocks> morning czajkowski!
[09:21] <duflu> Morning willcooke, Laney, czajkowski
=== muktupavels_ is now known as muktupavels
[09:44] <duflu> tjaalton, happy Monday: https://bugs.freedesktop.org/show_bug.cgi?id=105518#c10
[09:44] <ubot5`> Freedesktop bug 105518 in Drivers/DRI/i965 "Framebuffer corruption (briefly) when logging into Xorg sessions ("Xorg -background none") (Intel gen9 only)" [Normal,New]
[09:44] <duflu> (the bug isn't happy, the bisection is)
[09:45] <duflu> Tis indeed gen9 specific
[09:46] <tjaalton> duflu: cool, I'll ping ben
[10:13] <jibel> is there a reference somewhere that explains why ecryptfs has been demoted in 18.04?
[10:18] <xnox> jibel, well, ecryptfs-utils is still not demoted.
[10:21] <xnox> jibel, opened https://bugs.launchpad.net/ubuntu/+source/ecryptfs-utils/+bug/1756840 and dropping ecryptfs-utils from d-i requirements
[10:21] <ubot5`> Ubuntu bug 1756840 in ecryptfs-utils (Ubuntu) "Buggy, under-maintained, not fit for main anymore; alternatives exist" [Undecided,New]
[10:21] <Trevinho> Hi guys
[10:21] <jibel> xnox, yes but it will. It's just to answer to users asking why the option has been removed from Ubuntu.
[10:21] <tseliot> Laney: hey, do you know why gdm uses XWayland in bionic?
[10:21] <jibel> xnox, thanks
[10:22] <xnox> jibel, i was expecting "demotion" to already have happened, and some bug # already mentioned, but it was not =)
[10:22] <Laney> tseliot: not sure, some g-s-d thing maybe?
[10:24] <tseliot> Laney: maybe it defaults to Wayland if the system supports it. This kind of breaks systems with intel+nvidia
[10:25] <Laney> it does default to wayland
[10:25] <Laney> you asked about Xwayland though
[10:25] <tseliot> Laney: Xwayland is all I can see on my system
[10:25] <Laney> you won't "see" wayland in the same way
[10:25] <tseliot> gdm        922  1.0  1.4 566324 55136 tty1     Sl+  11:18   0:00 /usr/bin/Xwayland :1024 -rootless -terminate -accessx -core -listen 4 -listen 5 -displayfd 6
[10:26] <darkxst> hey desktopers!
[10:26] <Laney> right
[10:26] <Laney> but "what is using Xwayland" is quite a different question to "is the session a wayland session"
[10:26] <tseliot> Laney: how can I change that?
[10:26] <darkxst> gdm still runs wayland for gdm in bionic
[10:26] <darkxst> once you login you get Xorg
[10:26] <Laney> tseliot: there's an option in /etc/gdm3/custom.conf
[10:26] <Laney> WaylandEnable=false or something
[10:27] <darkxst> but I wouldnt suggest using that^ unless gdm doesnt start
[10:27] <Laney> I'm answering the questions that are asked.
[10:28] <tseliot> The thing is, gdm starts, but I don't see my username or the icons
[10:30] <tseliot> it doesn't seem to work
[10:30] <darkxst> tseliot, but otherwise stuff is rendered ok?
[10:32] <tseliot> darkxst: https://drive.google.com/file/d/1H47ilOtqzZkRIEWSObZaCrPgjywh2xsW/view?usp=sharing
[10:33] <darkxst> tseliot, that is just wrong
[10:33] <tseliot> things are fine once I log in
[10:33] <tseliot> yep
[10:33] <tseliot> and Xwayland is still there, BTW
[10:33] <darkxst> can you enable debug in /etc/gdm/custom.conf
[10:33] <darkxst> and grab logs
[10:33] <darkxst> yes Xwayland is still required for gdm login
[10:35] <tseliot> darkxst: is there a man page that explains how to do that?
[10:36] <tjaalton> just edit the file, under [debug]
[10:36] <tjaalton> "Enable=true"
[10:36] <darkxst> yeh its like the last line in that file
[10:37] <darkxst> then gdm will log to systemd journal
[10:37] <darkxst> next time you reboot
[10:37] <tseliot> oh, so the correct file is /etc/gdm3/custom.conf
[10:38] <darkxst> yep
[10:39] <tseliot> and yes, disabling Wayland seems to work now
[10:41] <tseliot> I think hybrid graphics + glvnd + the custo glx path got Xwayland confused
[10:42] <darkxst> what is custo?
[10:43] <tseliot> *custom
[10:43] <tseliot> the intel driver must be using nvidia's libglx.so
[10:43] <tseliot> for some reason
[10:50] <darkxst> tseliot, I dont have a hybrid machine with nvidia, but maybe it is getting confused
[10:51] <darkxst> libglvnd has had its fair share of fallout
[10:51] <darkxst> but gdm should in theory failback
[10:53] <tseliot> in this case both intel and nvidia support wayland
[10:53] <tseliot> except they won't work together
[10:56] <darkxst> gdm fallback is based on failure, not advertised support? afaik
[10:57] <tseliot> yes, I suspect that there is no failure there. Then gnome-shell crashes
[10:58] <tseliot> darkxst: http://paste.ubuntu.com/p/jNrQXpSmty/
[11:01] <darkxst> tseliot, please file a bug, and I will take a look tomorrow
[11:02] <tseliot> darkxst: against gdm?
[11:02] <tseliot> or gnome-shell?
[11:03] <darkxst> It probably doesnt matter, it could be either
[11:03] <darkxst> subscribe me to your bug
[11:06] <darkxst> its late here, so I have to go
[11:07] <tseliot> ok
[11:41] <oSoMoN> good morning kenvandine! got this for you: https://code.launchpad.net/~osomon/evince/snap-fix-cbr/+merge/341634
[12:08] <ahasenack> seb128
[12:09] <ahasenack> does anybody know if he is on holidays?
[12:12] <didrocks> he is until Friday
[12:25] <ahasenack> ok, thanks
[13:13] <kenvandine> oSoMoN, thx!
[13:14] <Laney> jbicha: did you want to merge g-c-c with debian or otherwise hold off?
[13:15] <Laney> I'm merging Gunnar's region stuff now
[13:16] <jbicha> you can merge if you like. I just uploaded g-c-c to bionic (without the scrollbar fix) a few minutes ago :)
[13:17] <Laney> whyyyyyyyyyyyyyyyyyy
[13:18] <Laney> I don't really fancy merging though
[13:18] <Laney> this makes the bzr stuff really annoying :(
[13:18] <Laney> urgh
[13:18] <jbicha> bzr isn't my fault … ;)
[13:19] <jbicha> Seb's away this week though… lol
[13:21] <jbicha> I'm fine with us not merging this time
[13:22] <jbicha> btw, https://code.launchpad.net/~ubuntu-desktop/ubuntu/+source/gnome-desktop/+git/gnome-desktop is on git too
[13:22] <jbicha> I published that after Budapest before we said we weren't really going to switch all our pkgs to git right now
[13:24] <Laney> I don't mind, but some people might ignore it
[13:25] <jbicha> that's fine, that particular bzr branch was mostly ignored too
[13:25] <jbicha> https://code.launchpad.net/~ubuntu-desktop/gnome-desktop/ubuntu
[13:30] <jbicha> Laney: do we need a FFe for LP: #1752579 to sync brlaser from Buster?
[13:30] <ubot5`> Launchpad bug 1752579 in brlaser (Ubuntu) "Needs sponsoring: Upload brlaser 4" [High,New] https://launchpad.net/bugs/1752579
[13:31] <Laney> jbicha: what are the changes?
[13:32] <jbicha> switch to cmake, add duplex printing support. I'm just reading from https://github.com/pdewacht/brlaser/releases/tag/v4 ;)
[13:34] <jbicha> version 3 was July 2014
[13:35] <jbicha> the duplex thing looks pretty minimal, so the switch to cmake is the big one. Better support for more printers.
[13:36] <jbicha> I don't have a printer so I'm not actually running this myself
[13:38] <Laney> I don't really know what this is, but sounds like it to me - want to bounce it back to Till and ask him to fill in some FFe details including having done some testing?
[13:38] <jbicha> sure
[13:38] <Laney> cheers
[13:39] <Laney> binary debdiff would be good to check, but maybe the sponsor can just do that
[13:49] <jbicha> I closed LP: #1699216 😢
[13:49] <ubot5`> Launchpad bug 1699216 in gnome-initial-setup (Ubuntu) "Encrypted home support" [Wishlist,Won't fix] https://launchpad.net/bugs/1699216
[13:52] <oSoMoN> kenvandine, please comment on https://forum.snapcraft.io/t/evince-snap-does-not-open-comic-book-file-cbr-while-deb-does-so/4523 when you publish an update to the evince snap
[13:52] <kenvandine> oSoMoN, will do
[13:52] <oSoMoN> thanks
[13:56] <kenvandine> oSoMoN, got a link to a cbr file i can test with?
[13:57] <kenvandine> so far all i've found point to 404 errors :/
[14:10] <oSoMoN> Laney, I could use your packaging expertise to detangle a dependency problem:
[14:10] <oSoMoN> the server team had intended to demote tomcat8 from main, but libreoffice depends on it via libreoffice-sdbc-hsqldb -> libhsqldb1.8.0-java -> libservlet3.1-java (source:tomcat8)
[14:10] <oSoMoN> libreoffice-sdbc-hsqldb itself is a recommends of libreoffice-base-drivers, which is pulled in by libreoffice-base (which is not installed by default)
[14:11] <oSoMoN> I can't see an obvious way out
[14:26] <Laney> oSoMoN: you have to break it somewhere
[14:27] <Laney> I guess put base in universe or break the recommends from -drivers
[14:29] <kenvandine> if we don't ship base by default, should be fine to send it to universe
[14:29] <Laney> dunno, we might want to support it for some reason
[14:30] <Laney> like we used to have evolution in main for some commercial support contracts even though that wasn't installed by default
[14:30] <oSoMoN> is there a looser relationship than recommends, that would allow a package in main to optionally pull in another one in universe?
[14:30] <Laney> only suggests which is more informational than anything else
[14:31] <Laney> we could contact the STS guys and ask if there's any reason on their side to keep it in main
[14:31] <Laney> that's what happened for evo
[14:50] <oSoMoN> Laney, thanks, I'll do that
[14:51] <oSoMoN> it looks like libreoffice-base doesn't have any rdepends in main, so technically that shouldn't be hard
[14:53] <jbicha> btw, libreoffice-base (& libreoffice-dev) is the only actual package that holds Java in main, see my comment on LP: #1735482
[14:53] <ubot5`> Launchpad bug 1735482 in openjdk-9 (Ubuntu) "OpenJDK JRE should be ported to GTK3" [Undecided,New] https://launchpad.net/bugs/1735482
[14:58] <oSoMoN> jbicha, ack, thanks
[14:59] <ackk> hi, does anyone have any suggestion about https://bugs.launchpad.net/ubuntu/+source/hplip/+bug/1731417 ?
[14:59] <ubot5`> Ubuntu bug 1731417 in hplip (Ubuntu) "Installed network printer removed automatically when turned off" [Undecided,Confirmed]
[15:05] <GunnarHj> Hi Laney, I have refreshed the language stuff MP, so if those changes are what you want to go in, it ought to be simple to merge now.
[15:08] <Laney> hi GunnarHj, thanks, I've got it under control
[15:13] <GunnarHj> jbicha: Did my email reply yesterday make sense?
[15:45] <amano> Trevinjo, the regression in your nautilus crash fixes seems to be fixed now in master: https://gitlab.gnome.org/GNOME/nautilus/merge_requests/153
[15:47] <amano> Trevinho, the regression in your nautilus crash fixes seems to be fixed now in master: https://gitlab.gnome.org/GNOME/nautilus/merge_requests/153
[15:47] <Trevinho> amano: yeah, I noticed and backported to 3.26 right now
[15:48] <Trevinho> amano: actually my fix wasn't causing the crash initially... Then I just did what upstream asked and that... lead to a crash xD
[15:48] <amano> 😄
[15:49] <Trevinho> Anyway 3.26 branch has all the changes now
[15:57] <flexiondotorg> willcooke: https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-390/+bug/1752739
[15:57] <ubot5`> Ubuntu bug 1752739 in nvidia-graphics-drivers-390 (Ubuntu) "PRIME Synchronization doesn't work with linux-kernel 4.15." [High,Fix released]
[16:01] <willcooke> flexiondotorg, ta
[16:02] <flexiondotorg> willcooke: There is another bug, I'm husting for it now.
[16:02] <flexiondotorg> *hunting
[16:02] <willcooke> flexiondotorg, that one is fixed it says, but you think it still breaks snaps?
[16:03] <flexiondotorg> https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-390/+bug/1756226
[16:03] <ubot5`> Ubuntu bug 1756226 in nvidia-graphics-drivers-390 (Ubuntu) "nvidia-driver-390 fails to start GUI" [Undecided,Confirmed]
[16:17] <willcooke> tseliot, hey, have you seen that 2nd bug flexiondotorg linked to yet? ^
[16:22] <jbicha> GunnarHj: I understand your answer as "maybe, but it would be nice to have dh_translations handle it automatically"
[16:22] <GunnarHj> jbicha: Sounds like a fair summary.
[16:23] <jbicha> we can just wait for people to complain since it seems like a headache now to add extra rules for handling .desktop translation to all the meson-using packages in main
[16:25] <GunnarHj> jbicha: Yeah.. But I'd recommend to think through which packages include patches which creates Ubuntu specific .desktop files. Maybe there are not so many of those, though.
[16:25] <jbicha> gnome-control-center & gnome-software
[16:26] <GunnarHj> jbicha: And gnome-session. But that and g-c-c we handled. What about gnome-software?
[16:27] <jbicha> is "Ubuntu Software" correctly translated for you?
[16:29] <GunnarHj> jbicha: I'm not on bionic now, bug last I checked that string didn't exist in LP, which made me conclude that we have switched to call it just "Software", and which made me change the docs accordingly... Please don't say that I misunderstood.
[16:30] <jbicha> I guess you can add gnome-software to your to-do list ;) it's supposed to be "Ubuntu Software"
[16:31] <jbicha> there is another bug where it doesn't use the Ubuntu branding in the "Ubuntu on Wayland" session
[16:32] <GunnarHj> jbicha: Shit! Ok, I'll look at gnome-software. Which is the proper branch to suggest a change for?
[16:33] <jbicha> the string isn't translatable (it's a bit of a hack in debian/rules). Only the English version was changed to "Ubuntu Software", all other languages are using the upstream "Software" name which is wrong
[16:33] <GunnarHj> jbicha: What?? Can't we just decide that "Software" is right? :)
[16:34] <jbicha> if you're submitting an actual patch, it's https://gitlab.gnome.org/Ubuntu/gnome-software
[16:34] <jbicha> but packaging is at https://code.launchpad.net/~ubuntu-desktop/gnome-software/ubuntu
[16:34] <jbicha> GunnarHj: its name is "Ubuntu Software" on Ubuntu sessions, that's not my decision :)
[16:35] <GunnarHj> jbicha: Right, I know who made the decision...
[16:35] <jbicha> lol, ok
[16:42] <alan_g> tjaalton, do you know if Mesa will still have the obsolete, broken "Mir EGL" patch when bionic ships?
[16:45] <tseliot> willcooke: looking
[16:47] <tseliot> willcooke: yes, I spoke with the original reporter via email. I haven't tested the multiple nvidia GPUs test case yet
[16:47] <tseliot> flexiondotorg: are you using KMS for nvidia?
[16:49] <flexiondotorg> I had the nvidia-driver-390 packages installed. I didn't tweak the configuration, so it was using whatever is default. They are all currently purged, so I can test what you need.
[16:51] <tseliot> flexiondotorg: any chance you can get me the journalctl output after reproducing the problem?
[17:05] <flexiondotorg> OK
[17:10] <didrocks> andyrock: before I leave, if you address Brian's comment on https://code.launchpad.net/~azzar1/software-properties/add-canonical-livepatch/+merge/341427 (he seems to have commented), I'm happy to give another review tomorrow
[17:10] <andyrock> didrocks: kk I'll do that later today
[17:11] <didrocks> thanks :)
[17:19] <flexiondotorg> tseliot: While I preapre the test, one other thing. the nvidia-390-drivers in the graphics/ppa currently is uninstallable.
[17:19] <tjaalton> alan_g: will not
[17:22] <tjaalton> alan_g: i've dropped them all from git
[17:22] <tjaalton> just waiting for a new upstream rc or release..
[17:24] <alan_g> tjaalton, thanks. (Our "auto detect" when building Mir is confused ATM because we have the patched headers, but it doesn't actually work.)
[17:26] <alan_g> Was just wondering if it will sort itself, or we need to be cleverer.
=== pstolowski is now known as pstolowski|afk
[18:03] <flexiondotorg> tseliot: I've attached some log to https://pad.lv/1756226
[18:04] <ubot5`> Launchpad bug 1756226 in nvidia-graphics-drivers-390 (Ubuntu) "nvidia-driver-390 fails to start GUI" [Undecided,Confirmed]
[19:28] <willcooke> hoy robert_ancell
[19:29] <robert_ancell> hello
[20:38] <willcooke> Right, successfully signed the Eclipse CLA despite their impossible captchas
[20:38] <willcooke> and that'll do
[20:38] <willcooke> night all