[05:18] <oSoMoN> good morning desktoppers
[05:21] <jibel> good morning oSoMoN and all
[05:23] <oSoMoN> salut jibel
[05:23] <oSoMoN> good morning ricotz
[05:24] <didrocks> good morning
[05:26] <oSoMoN> salut didrocks
[05:28] <didrocks> salut oSoMoN, ça va ?
[05:29] <duflu> Morning oSoMoN, jibel, didrocks
[05:29] <oSoMoN> didrocks, nuit courte et agitée, mais ça va, et toi?
[05:29] <oSoMoN> hey duflu, good afternoon
[05:31] <didrocks> hey duflu ;)
[05:31] <didrocks> oSoMoN: ça va ici, juste fatigué :)
[06:02] <tsimonq2> o/
[06:03] <tsimonq2> I'm publishing the Ubuntu Weekly Newsletter at the moment, I haven't seen the Desktop Team update in a few weeks... am I missing it or is it just not being published? :)
[06:05] <oSoMoN> tsimonq2, the last newsletter was sent on August 11th
[06:06] <oSoMoN> https://lists.ubuntu.com/archives/ubuntu-desktop/2017-August/005117.html
[06:07] <tsimonq2> oSoMoN: Ah ok, but that's not within the past week (15th to 21st is what this issue covers) ;)
[06:07] <oSoMoN> tsimonq2, no indeed, it's not very recent news
[06:07] <didrocks> tsimonq2: I guess grab Will once he's around
[06:08] <didrocks> (but wish him a happy birthday first :p)
[06:08] <tsimonq2> oSoMoN, didrocks: ok
[06:08] <tsimonq2> shhh, we gotta surprise him, didrocks ;)
[06:08] <didrocks> :-)
[06:32] <willcooke> morning all
[06:34] <oSoMoN> good morning willcooke
[06:34] <oSoMoN> happy birthday!
[06:34] <willcooke> :) ha, thanks oSoMoN!
[06:34] <didrocks> hey willcooke, happy birthday!
[06:34] <willcooke> morning didrocks, thanks!
[06:35] <didrocks> to celebrate this, upload one package! :)
[06:35] <willcooke> hehe
[06:36] <oSoMoN> one upload a day, keeps the doctor away?
[06:36] <didrocks> ahah ;)
[06:36] <didrocks> waow
[06:36] <didrocks> refreshing the seed, I saw:
[06:36] <didrocks> Removed gdm3 from desktop [s390x]
[06:37] <didrocks> Removed gnome-shell from desktop [s390x]
[06:37] <didrocks> Removed ubuntu-session from desktop [s390x]
[06:37] <didrocks> like "WHAT"?
[06:37] <didrocks> and then, I noticed [s390x]
[06:37] <willcooke> :DD
[06:37] <didrocks> knowing that I did those removals in the archive a couple of days ago :)
[06:37] <didrocks> phew
[06:37] <didrocks> so… the other change is:
[06:37] <didrocks> Added gnome-shell-extension-appindicator to desktop
[06:37] <didrocks> \o/
[06:37] <willcooke> yay!
[06:54] <koza> duflu hey and https://code.launchpad.net/~kzapalowicz/bluez/+git/bluez/+merge/329308
[06:54] <duflu> koza, OK
[07:00] <tsimonq2> willcooke: Happy birthday! :D
[07:00] <willcooke> koza, duflu - I can't do the BT meeting today I'm afraid.  And seb128 is off as well.
[07:00] <willcooke> tsimonq2, thanks!!
[07:00] <duflu> willcooke, yeah I cancelled too
[07:01] <duflu> ... FYI koza
[07:01] <tsimonq2> oSoMoN, didrocks: https://wiki.ubuntu.com/UbuntuWeeklyNewsletter/Issue517 ;)
[07:02]  * didrocks likes avoiding "shocks" ;)
[07:02] <didrocks> thanks for the mention tsimonq2!
[07:03] <tsimonq2> didrocks: You're welcome, keep up the great work ;)
[07:03] <tsimonq2> (*everyone*, that is ;) )
[07:04] <koza> duflu, willcooke, right then I will cancel as well :-) Only one thing I had was the update but once duflu acks and top-approves my MP I'll merge it and send the src package for someone to upload. so duflu no pressure but much depends now on you ;-)
[07:05] <willcooke> thanks koza
[07:05] <duflu> koza, well, I cancelled because I am away this afternoon. We'll see
[07:19] <didrocks> willcooke: got ownership of the appindicator extension repo, transferring now to ubuntu org
[07:19] <didrocks> (funny that it's needed to transfer to me first so that I can retransfer ;))
[07:20] <willcooke> didrocks, great stuff!  Saw the emails with the upstream, great work
[07:22] <didrocks> and https://github.com/ubuntu/gnome-shell-extension-appindicator :
[07:22] <didrocks> we got nice exchanges on the bug report as well, they rock :)
[07:23]  * didrocks adds the ubuntu branch to it now
[07:23] <didrocks> and https://github.com/ubuntu/gnome-shell-extension-appindicator/tree/ubuntu :)
[07:23] <didrocks> let's publish the blog post thus
[07:26] <didrocks> and published/tweeted/g+ssed
[07:27] <willcooke> reshared
[07:30] <didrocks> hum, linux unplugged from yesterday is entitled "forks done right"
[07:30] <didrocks> I think this is related to the blog posts, will need to listen to it while running :p
[07:31] <didrocks> in the shownotes "Ubuntu just showed every other distro how to roll a Dock"
[07:31] <didrocks> funny :)
[07:31] <willcooke> listening now
[07:31] <didrocks> (linking to phoronix though :p)
[07:35] <willcooke> didrocks, looks like it starts around 49m
[07:35] <willcooke> "What happens when an adult makes a Linux distribution"
[07:35]  * tsimonq2 sleeps, o/ all
[07:35] <willcooke> night tsimonq2
[07:36] <didrocks> willcooke: thanks! Will definitively listen to it :)
[07:36] <didrocks> good night tsimonq2
[07:39] <jibel> didrocks, both "ubuntu" and "ubuntu xorg" log me in x11. is it known? It's an upgraded machine not a fresh installation
[07:39] <didrocks> jibel: no, so it means we still have the session bug in gdm :/
[07:39] <didrocks> jibel: can you get from a vanilla machine some reproduceable steps?
[07:39] <didrocks> that will help debugging
[07:40] <jibel> didrocks, bug # ? or you want a new bug?
[07:40] <didrocks> but clearly, gdm is mixing up session
[07:40] <jibel> didrocks, I'll try an upgrade
[07:40] <didrocks> hum, let's reopen the old one, one sc
[07:40] <didrocks> sec*
[07:40] <didrocks> jibel: upgrade not selecting the right session (or ubiquity not setting the right session after installation) is probably a different bug
[07:40] <didrocks> jibel: we are talking here about: you select "ubuntu" and got xorg, (so ubuntu-xorg), right?
[07:41] <didrocks> https://bugs.launchpad.net/bugs/1704050 this is for the second one, you can retitle ofc ^
[07:42] <jibel> didrocks, yes. I have a choice between "ubuntu" and "ubuntu on xorg" and get ubuntu-x11 in both cases
[07:43] <didrocks> jibel: ok, so if you can get that from a vm, with reproduceable steps, that would be great
[07:43] <duflu> koza, looking now (I hope it's easier than fixing meson problems :)
[07:43] <didrocks> jibel: and let's get another bug report for upgrade (if there is any issue) and installation (where you mentioned there is one)
[07:43] <didrocks> jibel: making sense?
[07:43] <didrocks> really bad that you select a session in gdm and it's giving you another one…
[07:43] <didrocks> probably an offset "somewhere"
[07:44] <jibel> didrocks, yeah
[07:58] <duflu> koza, I think a bug report (titled "Please release...") and debdiff is now required, based on https://git.launchpad.net/~bluetooth/bluez/log/
[07:59] <koza> duflu, merged so you have pushed it to artful already, nice
[08:00] <duflu> koza, yes and updated the changelog again. Do you want me to do the bug/debdiff/release?
[08:01] <koza> duflu, if you have time I would be awesomely grateful
[08:02] <jibel> didrocks, I filed bug 1712287 yesterday for the installation case
[08:02] <jibel> trying the upgrade now
[08:02] <didrocks> jibel: full-upgrade?
[08:02] <didrocks> willcooke: mind adding that one to trello? ^
[08:02] <jibel> didrocks, from z
[08:02] <Laney> morning
[08:03] <didrocks> jibel: sweet :)
[08:03] <didrocks> jibel: I'm retargetting to gdm
[08:03] <jibel> didrocks, and the theme bug is bug 1712289
[08:03] <didrocks> could be ubiquity, I have to look at it goes
[08:03] <didrocks> willcooke: same, please add that one so that I don't forget about it ^
[08:03] <didrocks> jibel: thanks :)
[08:04] <didrocks> the second one isn't unexpected, I need to check some things though (and we need ubiquity out of proposed first)
[08:04] <didrocks> hey Laney!
[08:04] <didrocks> jibel: btw, it's not radiance but adwaita
[08:04] <jibel> jbicha, bug 1711625
[08:04] <willcooke> didrocks, done - but it found a load of other ubiquity ones too
[08:05] <jibel> jbicha, it is fixed, right?
[08:05] <duflu> koza, BTW I think the original patch output from 'git format-patch -1 5252296b725ef1599' would have been just as good.
[08:05] <didrocks> jibel: it's fixed AFAIK
[08:05] <didrocks> willcooke: hum, "great"? ;)
[08:05] <willcooke> didrocks, they're added to "Proposed"
[08:05] <didrocks> Trevinho: hey, can I publish your ubuntu-themes bilbetto changes?
[08:05] <didrocks> good, thanks!
[08:08] <koza> duflu, you mean in favor of what I have added to debian/patches? probably but since the format requires a few extra fields I have edited it. you know more info for future us
[08:09] <duflu> Yeah, doesn't matter. And no matter what format I use, the reviewer often complains. It depends on the project, and the mood of the reviewer
[08:14] <koza> right
[08:14] <koza> duflu, luckily for me you have ben in a good mood  ;-)
[08:14] <koza> *been
[08:15] <Laney> hey didrocks
[08:19] <Laney> laney@nightingale> systemctl --user list-timers                                                                                                                                                                                   ~
[08:19] <Laney> Failed to list units: Process org.freedesktop.systemd1 exited with status 1
[08:20] <Laney> O_O
[08:30] <Trevinho> didrocks: yeah, I had to rebuild it as there was a branch causing an issue
[08:30] <didrocks> Trevinho: so, publishing now?
[08:31] <Trevinho> didrocks: this still needs approval https://code.launchpad.net/~3v1n0/ubuntu-themes/better-destructive-action/+merge/329392
[08:32] <didrocks> Trevinho: ok, let me test it and come back later today
[08:34] <jibel> didrocks, after an upgrade from Z, I've 'Ubuntu on Xorg' and 'Unity'
[08:34] <jibel> I'll file another bug
[08:35] <didrocks> jibel: why?
[08:35] <didrocks> jibel: you don't have "Ubuntu"?
[08:35] <jibel> didrocks, no 'Ubuntu'
[08:35] <didrocks> waow, wasn't expecting that one
[08:35] <didrocks> you rebooted, correct?
[08:35] <jibel> didrocks, yes, it's the last step of a release upgrade
[08:36] <didrocks> hum, you do have /usr/share/wayland-sessions/ubuntu.desktop
[08:36] <didrocks> gdm is the default
[08:36] <didrocks> and /etc/gdm3/custom.conf have #WaylandEnable=false (commented)
[08:36] <didrocks> right?
[08:37] <popey> oSoMoN: did you confirm that the ordering of variables in the launcher is the problem with chromium snap?
[08:37] <jibel> gdm is the default, I've /u/s/w-s/ubuntu-desktop and #WaylandEnable=false
[08:37] <jibel> didrocks, ^
[08:37] <didrocks> jibel: thanks for checking, please log it then
[08:37]  * didrocks is puzzled
[08:37] <didrocks> can be that lightdm deselect the default session on logout/shutdown
[08:38] <didrocks> because it doesn't see the "ubuntu" session in wayland
[08:38] <didrocks> but then, why gdm doesn't pick it, weird…
[08:38] <Trevinho> didrocks: you add me as collaborator in https://github.com/ubuntu/gnome-shell-extension-appindicator so I can add branches there instead of adding it in personal
[08:39] <Trevinho> not that changes much, but..
[08:39] <didrocks> Trevinho: github ID?
[08:39] <Trevinho> 3v1n0
[08:39] <didrocks> Trevinho: done! :)
[08:39] <Trevinho> ta
[08:39] <didrocks> yw
[08:41] <jibel> didrocks, bug 1712504
[08:41] <didrocks> thx jibel
[08:42] <flexiondotorg> Morning desktopers
[08:42] <jibel> Laney, I've the ALT-Fn switches to VT bug again. eg ALT-F4 doesn't close the window but goes to VT4
[08:42]  * jibel goes to the vet, BIAB
[08:42] <popey> jibel: i have that bug again today too
[08:44] <Laney> jibel: ok, what did you do?
[08:44] <Laney> that might be something else entirely
[08:44] <Laney> does alt-left-right switch vt?
[08:44] <popey> yes
[08:45] <Laney> so what did you do?
[08:45] <Laney> the thing we were working on yesterday was an upgrade problem in the console-setup package
[08:45] <popey> did my updates
[08:46] <popey> it's not a heavily used laptop, use terminal to update and chrome to browse. I also often plug an external display in, and switch audio stuff between local and hdmi audio devices
[08:47] <Laney> did you get an update of console-setup?
[08:48] <popey> yes, looks like it, from dpkg.log
[08:49] <Laney> probs the same thing then
[08:49] <Laney> should be fixed soon
[08:50] <popey> yay
[08:50] <duflu> koza: I need to run off soon, but you can track http://reqorts.qa.ubuntu.com/reports/sponsoring/
[09:14] <oSoMoN> popey, nope I couldn't confirm, and I'm afk for the morning, will be back around 12:30 UTC
[09:28] <Mirv> hey. doko is asking ~ubuntu-desktop to subscribe to https://launchpad.net/ubuntu/+source/hfst-ospell - could you do that?
[09:28] <Mirv> similar to libvoikko. related to bug #1708428
[09:29] <Mirv> oh now, he says ~desktop-packages
[09:30] <Mirv> ..or ~desktop-bugs , which has more than seb128 and a couple of others
[09:31] <Mirv> as seb is not around, I'll go ahead and ask Laney instead to subscribe ~desktop-bugs to hfst-ospell bugs
[09:35] <koza> duflu, sure, thanks
[09:39] <flexiondotorg> didrocks We've been getting mixed reports about some snaps not working, at all, for some users.
[09:40] <flexiondotorg> I've been investing a few and so has popey, and we've potentially found a solution.
[09:40] <flexiondotorg> It appears to be related to LD_LIBRARY_PATH "order".
[09:42] <flexiondotorg> In the desktop helpers the original LD_LIBRARY_PATH is prepended, but we've found that is some (perhaps all) cases appending the original LD_LIBRARY_PATH is resolving the issues.
[09:43] <flexiondotorg> I'd be interested to know if the desktop helpers prepend the existing LD_LIBRARY_PATH on purpose and for a good reason that I'm not aware of?
[10:02] <doko> what is the difference of the desktop-packages and desktop-bugs teams?
[10:02] <doko> as a bug subscriber?
[10:12] <Laney> dunno, I think we're using desktop-bugs though
[10:16] <Laney> Mirv: I added it, it's probably best to talk with the owning team before in future though
[10:20] <jamesh> the initial polkit support branch has been merged to snapd master
[10:27] <Laney> nice
[10:27] <Laney> what's next?
[10:27] <willcooke> jamesh, nice, congrats!
[10:29] <jamesh> Laney: polkit wise there's two avenues to improve things: (1) extend the support to package install/remove so gnome-software can function without a store login, and (2) get better integration with the snap command line tool (e.g. make sure interactive auth is only enabled if there is a terminal, and use the tty agent)
[10:30] <jamesh> (1) still seems to be a bit controversial, but should be easier to argue with the initial support landed.
[10:30] <Laney> seems the approach they like is 'provide a patch and we can argue on the PR'
[10:32] <jamesh> yep.  The patch for (1) is going to be a few lines of XML config and setting a field on a few more Command structs
[10:32] <jamesh> all the logic to support it exists now
[10:32] <Laney> will gnome-software need modifying?
[10:33] <Laney> or does it only ask you to login if snapd refuses the request?
[10:35] <GammaDraconis> Gnome Software is bad !
[10:35] <GammaDraconis> It is very incomplete
[10:35] <jamesh> looking at the code, it might not.  It looks like it is set up to authenticate on error
[10:35]  * Laney nods
[10:36] <Laney> I think I remember seeing that
[10:36] <GammaDraconis> It lacks a lot of software
[10:36] <jamesh> we'll probably want some snapd-glib changes to make use of this though: we can replace all the snapd-login-service code with direct communication with snapd
[10:36] <GammaDraconis> (in Gnome Software)
[10:38] <GammaDraconis> On the French forum of ubuntu, many users are dissatisfied with gnome software (cf with Didrocks, he is aware)
[10:39] <jamesh> GammaDraconis: is there particular types of packages it isn't showing that you think it should?
[10:41] <GammaDraconis> i don't understand very good your question
[10:42] <jamesh> GammaDraconis: you say gnome-software is incomplete.  I'm asking what you think is missing
[10:42] <jamesh> I'm trying to understand what we could do to improve things
[10:43] <GammaDraconis> I have no example on the moment (I do not use personnaly) but it is what is said on the forum.
[10:45] <GammaDraconis> There is not only that as a problem
[10:45] <GammaDraconis> gnome software is low
[10:45] <GammaDraconis> and bug
[10:45] <jamesh> It definitely takes a while to start the first time, yes.
[10:46] <GammaDraconis> For example with VLC : https://framapic.org/TjhhSMoFMCH7/tsynDWoaTsT5.jpg
[10:46] <Mirv> Laney: thank you! I didn't realize the owning team must be from a specific list (if there's one, the "owning" word at https://wiki.ubuntu.com/UbuntuMainInclusionRequirements could be linked or explained a bit more), so we initially thought ~giellatekno-hfst-voikko would be enough as the team to monitor the bugs
[10:46] <GammaDraconis> bug
[10:47] <Mirv> that said, it's been obvious that Canonical needs to stand behind the main packages so..
[10:49] <GammaDraconis> and i find that totem is a bad choice (for video) because it does not read much format without external codec. It would be better to put MPV or VLC instead
[10:51] <Laney> jamesh: Did you get my email last night? (I'm not sure because I sent it from my phone; maybe that doesn't work....) Wondering what the status of the Snap plugin for Builder is - I kind of lost track a bit :(
[10:51] <Laney> Basically we were wondering if we could/should try to get it into 17.10
[10:51] <jamesh> GammaDraconis: totem should have pretty similar format support with gstreamer1.0-libav installed (this is essentially the same ffmpeg codecs the other players use)
[10:53] <GammaDraconis> Why would you prefer totem than vlc or mpv?
[10:53] <jamesh> Laney: yeah.  I don't have much to report on gnome-builder yet.  To implement things the way upstream wants, I need some snapcraft extensions.  I've described what I'd need on the forums and got some feedback, but at this point I'll probabably need to look at doing the snapcraft bits myself
[10:53] <GammaDraconis> Most users know VLC because it is multiplatform (windows, mac, linux, android...)
[10:54] <jamesh> GammaDraconis: totem matches the overall style of the rest of the desktop, and is using the same media framework as other GNOME apps
[10:54] <GammaDraconis> Would be more convenient to use the best-known software and he is free/libre
[10:55] <jamesh> GammaDraconis: also, there are legal reasons why the libav/ffmpeg codecs aren't installed by default
[10:55] <GammaDraconis> Vlc and mpv fits well into gnome
[10:55] <jamesh> they are problematic in countries with software patents
[10:57] <GammaDraconis> As much propose vlc
[10:58] <GammaDraconis> And in addition vlc is a French software (and France is the best country as you know)
[10:59] <jamesh> GammaDraconis: if that's how you pick software, you'll be happy to know that totem's main author (Bastien Nocera) is also French
[11:00] <GammaDraconis> Good ! but i prefer vlc, it is more complete.
[11:01] <GammaDraconis> Mpv is not bad either
[11:01] <GammaDraconis> VLC > MPV > Totem
[11:02] <GammaDraconis> The complete passage to Gnome 3.26 is for when on Artful ?
[11:03] <GammaDraconis> Currently, gnome-shell is still in 3.24
[11:04] <jamesh> GammaDraconis: I haven't really looked into the current state of media playback on the desktop, but we definitely used to have a system where you'd get prompted to install codec packages when first trying to play an unhandled file.  That would be the preferred way to fix the problem of totem not handling enough formats
[11:05] <jamesh> there is a gstreamer1.0-packagekit package in universe, but I don't know how well it works
[11:40] <Laney> my ctrl-alt-l is gone
[11:42] <doko> Laney: could you change the bug subscriber of libvoikko as well? https://launchpad.net/~desktop-packages doesn't look like an official team ...
[11:44] <Laney> there's some debate about this
[11:44] <Laney> I think some reports were looking at one and not the other or something
[11:44] <jbicha> jibel: yes, the xdg-user-dirs bugs is supposed to be fixed now but I haven't done a new install recently to test it
[11:45] <jbicha> didrocks: I'd like to stop building the ubuntu-desktop binary package for s390x
[11:45] <jbicha> good morning
[11:46] <jibel> Laney, one more thing I just discovered about 1710637 is that when I switch back to gnome-shell the app eventually receives the event
[11:46] <jibel> jbicha, okay, I'll check with a new installation
[11:47] <didrocks> flexiondotorg: yeah, it's on purpose, you should use first the libraries you are embeeded in your snap rather than the ubuntu core provided one, as this is the snap philosophy, correct?
[11:49] <flexiondotorg> didrocks Yes, I agree that we should first use the libraries from the snap.
[11:49] <flexiondotorg> But I don't think that it what is happening.
[11:49] <flexiondotorg> Because the snap paths are appended to the existing LD_LIBRARY_PATH. So are in fact, not found first.
[11:50] <flexiondotorg> This is an example of what we have in the helpers now:
[11:50] <flexiondotorg> export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$RUNTIME/usr/lib/$ARCH/mesa
[11:50] <flexiondotorg> Which I think should be:
[11:50] <flexiondotorg> export LD_LIBRARY_PATH=$RUNTIME/usr/lib/$ARCH/mesa:$LD_LIBRARY_PATH
[11:52] <didrocks> flexiondotorg: appended, not prepended?
[11:52] <didrocks> hum
[11:52] <didrocks> no
[11:52] <didrocks> ecoh $LD_LIBRARY_PATH
[11:52] <didrocks> echo*
[11:52] <flexiondotorg> https://github.com/ubuntu/snapcraft-desktop-helpers/blob/master/common/desktop-exports#L23
[11:53] <didrocks> yes, but that's not all the story
[11:53] <flexiondotorg> I suspected ;-)
[11:53] <didrocks> first, snapcraft prepend the $SNAP path to LD_LIBRARY
[11:53] <flexiondotorg> Which is why I'm asking.
[11:53] <didrocks> then, the launcher is called
[11:53] <didrocks> that's what it's appended
[11:53] <flexiondotorg> Curious.
[11:53] <didrocks> maybe in some of those snaps, snapcraft didn't prepend the correct paths for them?
[11:53] <didrocks> look at their .wrapper script
[11:54] <didrocks> (the one generated by snapcraft)
[11:54] <didrocks> this is why I don't prepend to it
[11:54] <didrocks> or the RUNTIME will be before the $SNAP one
[11:54] <didrocks> (RUNTIME being platform snap)
[11:54] <didrocks> making sense? (at least in theory?)
[11:54] <flexiondotorg> The snaps I'm looking at are all using the desktop helpers. To avoid adding any other variable.
[11:55] <flexiondotorg> Some are classic some are strict.
[11:55] <flexiondotorg> But adding a new launcher that prepends the $SNAP LD_LIBRARY_PATH is correcting the issue.
[11:56] <didrocks> hum
[11:56] <didrocks> the desktop helpers isn't design to work on classic
[11:56] <didrocks> look at their .wrapper script
[11:56] <didrocks> maybe they are "classic snap"
[11:56] <didrocks> and so snapcraft doesn't add anything to the .wrapper script
[11:56] <didrocks> and this is why
[11:56] <didrocks> did you read my ubuntu-make blog post serie about classic snap?
[11:57] <didrocks> I mentioned this :)
[11:57] <flexiondotorg> didrocks Yeah, needed for classic Electron applications. Which have ship there own launcher as well.
[11:57] <flexiondotorg> I have read lots from you and z_yga about classic.
[11:57] <flexiondotorg> We can for the moment ignore the classic snaps.
[11:57] <flexiondotorg> Since the issue is present in strict snaps.
[11:59] <flexiondotorg> I'll gather LD_LIBRARY_PATH from all the affected snaps and see if that helps identify anything.
[11:59] <flexiondotorg> For example, has snapcraft regressed? I'll investigate.
[11:59] <flexiondotorg> didrocks Thanks for the info.
[11:59] <didrocks> flexiondotorg: yeah, I bet it's classic snap only
[12:00] <didrocks> flexiondotorg: yw! really look at this .wrapper script
[12:00] <didrocks> I guess it will be a good start
[12:01] <flexiondotorg> I'll ignore classic snaps for now and just focus on understanding why certain strict snaps are not working as expected.
[12:02]  * ogra_ notes that 'especially* in classic snaps you want the $SNAP/lib path prepended to LD_LIBRARY_PATH since they can ship their own libs and newer versions 
[12:06] <didrocks> I wonder if there is a way via an env variable to detect classic snaps
[12:07] <didrocks> and that way, have the desktop launcher doing the $SNAP prepending
[12:13] <didrocks> jbicha: we should do that, indeed
[12:16] <jibel> didrocks, the theme is incorrect in the live session. Same bug than ubiquity-dm or different one?
[12:20] <didrocks> jibel: echo $XDG_CURRENT_DESKTOP?
[12:21] <jibel> didrocks, ubuntu:GNOME
[12:21] <jbicha> didrocks: oh we already don't publish ubuntu-desktop/s390x, if we want ./update to stop trying to do desktop/s390x, maybe we need to split ubuntu-desktop into a different source package?
[12:21] <didrocks> jibel: I'm puzzled puzzled puzzled
[12:22] <didrocks> jibel: gsettings get org.gnome.desktop.interface gtk-theme
[12:22] <jibel> didrocks, I don't have yesterday's image, but Im pretty sure it was correct
[12:22] <didrocks> jibel: try a gsetting reset just in case
[12:22] <didrocks> jibel: hum, ubiquity wasn't, right?
[12:22] <jibel> didrocks, u-dm was not but the live session was
[12:22] <jibel> today both are broken
[12:22] <jbicha> maybe that ubuntu-desktop split is something to try earlier in the cycle, I don't want to break things like that this week
[12:23] <jibel> didrocks, but I'll double check
[12:23] <didrocks> jibel: I was expecting u-dm to be broken, but not the live…
[12:23] <didrocks> (due to this env variable not triggering glib to read the correct keys ^)
[12:23] <didrocks> jibel: mind doing this gsettings get?
[12:23] <jibel> didrocks, rgsettings returns Adwaita
[12:24] <didrocks> jibel: /usr/share/glib-2.0/schemas/10_ubuntu-settings.gschema.override is installed, correct?
[12:26] <jibel> didrocks, it is
[12:26] <didrocks> jibel: can you try reverting the key?
[12:26] <didrocks> and get again ?
[12:26] <didrocks> resetting*
[12:26] <jibel> didrocks, I filed bug 1712557, mind adding all the question there and I'll continue my review
[12:27] <didrocks> jibel: thanks! I'm really really puzzled on that one, will be great to have the manifest version diff between the 2 isos (if you can retain them)
[12:27] <jibel> okay, I'll do that
[12:27] <didrocks> jibel: unsure I'll have time before the next couple of weeks looking at it with the fit and finish sprint, but I'm worried now :p
[12:27] <didrocks> especially about the live session, I don't have a firm explanation
[12:27] <didrocks> and my gnome boxes is crashing :(
[12:27] <jibel> and also confirm which build is the last good build
[12:28] <jibel> use plain qemu :)
[12:28] <didrocks> jibel: ubiquity-dm is a no-brainer, it's the env variable missing, what I was expecting
[12:29] <didrocks> but I need to look and ensure it's not going to bring the dock, and want the correct theme for G-S once we tweak
[12:29] <didrocks> the live session… hum
[12:29] <didrocks> especially with the "worked yesterday, didn't today"
[12:29] <didrocks> qemu-system-x86_64: -cdrom artful-desktop-amd64.iso: Could not get temporary filename: Permission denied
[12:29] <didrocks> nice!
[12:31] <Laney> bet gnome-boxes is https://git.gnome.org/browse/gnome-boxes/commit/?id=06099b1b43a2607eecbde07c698d98c12f2e8534
[12:31] <Laney> new version out 3 minutes ago!
[12:31] <didrocks> Laney: you are so on the edge!
[12:31] <didrocks> yeah, sounds like a good candidate
[12:32] <Laney> well tested updates ;-)
[12:32] <didrocks> yep… release AND updates :)
[12:32] <didrocks> ok, doing it
[12:33] <didrocks> jbicha: FYI, please test the package before uploading (at least, that the apps launch) ;)
[12:33] <Laney> the update?
[12:33] <didrocks> Laney: you are doing it?
[12:33] <Laney> nope
[12:33] <didrocks> so yeah, I'm going to do the new version update
[12:33] <didrocks> juts because I want it first!!! :)
[12:34] <Laney> go for it
[12:34] <jbicha> didrocks: thanks
[12:34] <didrocks> thanks for noticing it Laney!
[12:34] <Laney> probably take some minutes for the tarball to appear
[12:34] <didrocks> yeah
[12:34] <didrocks> can wait, quite happy to not have to debug that one TBH
[12:35] <Laney> wonder why my VM has started telling me no schemas are installed
[12:35] <didrocks> especially when they are obviously there!
[12:36] <Laney> XDG_DATA_DIRS is wrong
[12:36] <Laney> =/var/lib/snapd/desktop ¬_¬
[12:38] <didrocks> oh, might be why it doesn't find the theme on the live session
[12:38] <Laney> probably some hack I did
[12:38] <didrocks> I guess the theme is the last of problem with that env variable :)
[12:39] <didrocks> I'm sure ogra can live with that env variable :)
[12:39] <didrocks> jibel: FYI ^
[12:39] <Laney> the desktop doesn't even start
[12:39] <Laney> probably not this
[12:41] <ogra_> didrocks, i didnt set it ... not my fault :P
[12:45] <jibel> didrocks, I've XDG_DATA_DIR=/usr/share/ubuntu-xorg:/usr/local/share/:/usr/share/:/var/lib/snapd/desktop
[12:45] <didrocks> ogra_: https://www.youtube.com/watch?v=yNY6ZstdUdY
[12:45] <didrocks> jibel: bah, can't even blame Laney, snif ;)
[12:45] <jibel> didrocks, /usd/share/ubuntu-xorg doesn't exist
[12:45] <didrocks> yeah
[12:45] <ogra_> didrocks, sure :P
[12:45] <jibel> /usr/*
[12:45] <didrocks> jibel: but that shouldn't be an issue
[12:46] <didrocks> ok, something to investigate thus
[12:46] <didrocks> getting the manifest version diff would be grand! (we should autogenerate them if we don't already)
[12:46] <ogra_> veery iso has a manifest
[12:46] <ogra_> *every
[12:46] <ogra_> check cdimage
[12:47] <ogra_> (oh you want a prepared diff ...)
[12:47] <didrocks> yep
[12:48] <ogra_> yeah, ursinha had some code that was never applied for that
[12:48] <ogra_> (back in the phone days)
[12:48] <didrocks> unreleased code (so, it's not free, release it!)
[12:48] <ogra_> heh
[13:04] <jbicha> didrocks: I think we can get gnome-shell 3.25.90 in today, is that ok with you?
[13:08] <didrocks> jbicha: that would be perfect!
[13:09] <didrocks> jbicha: just be aware that upstream is probably going to revert the transparency panel
[13:09] <didrocks> so let's not make too much fuzz about it :)
[13:09]  * didrocks F5 and still didn't get any gnome-boxes on ftp://ftp.gnome.org/pub/GNOME/core/3.25/3.25.91/sources/
[13:09] <didrocks> sad
[13:10] <didrocks> jbicha: do you know if g-s is going to migrate today out of proposed?
[13:10] <didrocks> I wonder if I should just pick the gjs stack to hack tomorrow morning
[13:10] <didrocks> (as I'll be in the train to London)
[13:11] <didrocks> the css diff itself is easy anyway, not big, just wondering if I tackle some other things…
[13:12] <jbicha> didrocks: I think gjs will migrate in an hour or so when its autopkgtests complete, and then I'll upload mutter and gnome-shell (mutter's soname was bumped so it needs easy NEW processing)
[13:12] <didrocks> jbicha: ok, sounds like doable if it's in an hour ;)
[13:12] <jbicha> so I think gnome-shell will migrate today but I don't know if excuses will come up with some excuse
[13:14] <willcooke> jbicha, re: captive portal - did you have anything prepared to upload for enabling it?  Would like to get that switched on before FF if poss.
[13:14] <jbicha> could you remove gnome-session's s390x binaries again? (I'm hoping we won't have that issue any more once gjs migrates?)
[13:14] <jbicha> willcooke: it's on my to do list for today
[13:15] <didrocks> jbicha: ah, because it was built on -proposed
[13:15] <willcooke> jbicha, super, thanks!
[13:17] <didrocks> jbicha: 6 packages successfully removed.
[13:17] <didrocks> (removed them from -proposed)
[13:26] <GammaDraconis> The top bar of Gnome 3.26 will remain translucent in 17.10 and 18.04 i hope ?
[13:26] <GammaDraconis> Translucent is better than black
[13:28] <jbicha> btw, I'll be uploading gnome-shell/mutter 3.25.90 because with 3.25.91 I had this problem: https://bugzilla.gnome.org/786660
[13:30] <didrocks> GammaDraconis: upstream is going to revert it as said above ^
[13:30] <didrocks> a lot of design issues with it until they get proper blur support in the shell
[13:31] <didrocks> jbicha: small bug (kidding) :p
[13:41] <GammaDraconis> good bye, i go.
[14:08] <jbicha> Laney: could we ignore gjs/armhf/artful autopkgtest failures?
[14:08] <Laney> maybe
[14:09] <Laney> depends what's going on
[14:10] <jbicha> it's related to https://bugzilla.gnome.org/786578 and my excuse is that Firefox hasn't built on Ubuntu armhf for multiple Firefox releases
[14:10] <Laney> looks rather like 1.4.90 regressed doesn't it
[14:10] <jbicha> that was when we switched from mozjs38 to mozjs52
[14:11] <Laney> 49*
[14:11] <jbicha> and Firefox stopped building on armhf after FF 50
[14:12] <jbicha> so it's likely that mozjs52/armhf doesn't work :(
[14:16] <jbicha> trello says we decided to "stop building 32bit ISOs for 18.04", maybe stop armhf Desktop iso's now?
[14:16] <King_InuYasha> didrocks: hey
[14:17] <King_InuYasha> I saw you took over the appindicator extension in the ubuntu org
[14:17] <King_InuYasha> you don't plan on imposing the Canonical CLA on it, do you?
[14:18] <didrocks> King_InuYasha: no, it's a project lead by the community and we only have a branch in it, why?
[14:18] <didrocks> (same rule than for dash to dock)
[14:18] <King_InuYasha> didrocks: I'm paranoid about these things :)
[14:18] <King_InuYasha> and most people expect stuff like that to happen from Canonical controlled repos
[14:19] <King_InuYasha> so you should probably make that clear somewhere
[14:19] <didrocks> King_InuYasha: there are a lot of project in the ubuntu org that don't have CLA
[14:19] <didrocks> like ubuntu-make for instance
[14:19] <didrocks> so not an exception
[14:20] <jbicha> King_InuYasha: see https://www.canonical.com/projects/directory
[14:30] <Laney> jbicha: just running `gjs' crashes
[14:30] <Laney> I think skipping the tests might be a little bit dishonest here
[14:33] <didrocks> 1https://launchpad.net/ubuntu/+source/gnome-boxes/3.25.91-0ubuntu1/+build/13286768/+files/buildlog_ubuntu-artful-i386.gnome-boxes_3.25.91-0ubuntu1_BUILDING.txt.gz
[14:33] <didrocks> nice
[14:33] <didrocks> but it built here!
[14:33] <didrocks>  sbuild-build-depends-core-dummy : Depends: build-essential but it is not going to be installed
[14:33] <didrocks> ohoh
[14:34] <didrocks> I guess there is a bigger issue around ;)
[15:14] <jibel> didrocks, this is really weird, same image, tried several times and cannot reproduce the wrong theme in the live session ...
[15:16] <didrocks> jibel: hum… ok, that's weird
[15:18] <jibel> grr, I closed my entire session with alt+f4 :/
[15:19] <jibel> well x11 crashed
[15:33] <oSoMoN> wow, using gnome-tweak-tool to switch to static workspaces and decrease their number reliably crashes my session
[15:33] <oSoMoN> known bug?
[15:46] <jbicha> Laney: what is your suggestion then? should I re-upload gjs without building armhf binaries?
[15:48] <Laney> jbicha: I'm building mozjs so that I can get that debugging output
[15:48] <Laney> but I don't think it's right to knowingly let a completely busted package into the release, do you?
[15:48] <jlnr> This morning the hot corner for "Activities" disappeared. Is there any way in the Ubuntu session to get it back? I know there's an extension to disable the corner, but is there one to *enable* it?
[15:52] <jbicha> Laney: I believe gnome-shell 3.25.90 can still work with gjs 1.48/mozjs38 if you want to remove gjs, gnome-shell, polari, gnome-sushi from artful-proposed to try to do the gnome-shell transition first
[15:53] <jbicha> I mean, if we find an AA to do those removals
[15:54] <Laney> maybe it's okay to remove gjs/armhf, I don't know, what do you think?
[15:55] <jbicha> gnome-shell might not run on zesty/armhf but I don't have the hardware to know for sure
[15:56] <jbicha> gjs is used by more stuff than gnome-shell but if you can't run our default desktop, I'm not sure those other apps are that important?
[15:58] <jbicha> so I'd vote to drop gjs/armhf, it can be restored later this cycle if it ends up working, but it's holding things up now
[16:32] <oSoMoN> if that's of interest to anyone, bug #1712616 is the bug I filed for my session crash when using gnome-tweak-tool to decrease the number of static workspaces
[16:39] <popey> oSoMoN: any luck with chromium snap today?
[16:41] <oSoMoN> popey, haven't looked at it myself, but flexiondotorg was looking at that LD_LIBRARY_PATH ordering issue
[16:45] <popey> oSoMoN: i dont think he got any further than asking if we needed to do something with the desktop launchers, and asked didrocks if there was some reason they were ordered the way they were
[16:46] <oSoMoN> popey, ok, so I don't have a config where the chromium snap fails to launch, the issue is nvidia-specific, right?
[16:46] <popey> oSoMoN: it fails on nvidia with the binary driver, I don't know where else it fails, but it certainly fails there
[16:47] <oSoMoN> popey, and you confirmed that prepending the paths instead of appending them fixes the launch issue, right?
[16:48] <flexiondotorg> I can give Chromium a go.
[16:48] <popey> oSoMoN: I haven't forked the entire desktop launcher, just added some lines to your chromium launcher
[16:49] <popey> I am concerned that we have a completely broken high profile snap in stable channel. it doesn't look very good.
[16:49] <oSoMoN> popey, if I produce a snap with the modified desktop launcher, can you try it?
[16:49] <popey> i can
[16:49] <oSoMoN> ok, let's give this a go
[16:49] <oSoMoN> sorry I didn't get to it earlier, was/am busy fighting other chromium build issues
[16:50] <popey> Sorry to hear that, it must be quite a monster!
[16:51] <oSoMoN> let's say trying to build the dev branch which increasingly makes use of c++14 constructs on trusty which has g++ 4.9 presents interesting challenges :)
[16:52] <popey> lulz :)
[16:54] <willcooke> right, calling it a day.  Off to London tomorrow so ping on Telegram if you need anything
[16:54] <willcooke> night all
[17:12] <oSoMoN> popey, I’ve got a test snap but trying to figure out how to share it efficiently with you, as my upload link is slooow
[17:13] <amano> G-C-C 3.25.91 is tagged now on Git
[17:18] <amano> The tarball seems to be on ftp as well
[17:19] <amano>  - Several small UI improvements and layout adjustments - Various crash fixes - Translation updates
[17:21] <popey> oSoMoN: it should do delta uploads if you push it to candidate :)
[17:21] <popey> oSoMoN: or push the yaml etc to a junk branch and have launchpad build it for you :)
[17:21] <popey> (that's my favorite way to build snaps)
[17:24] <oSoMoN> popey, yeah, but I went the route of unsquashfs the original snap, modify it and resquash it
[17:24] <oSoMoN> I’ll try the delta thing, that sounds like it could work
[17:25] <popey> yeah, with a minor update it should
[17:25] <oSoMoN> Received 401: '{"error_list": [{"message": "Authorization Required", "code": "macaroon-permission-required"}]}'
[17:26] <oSoMoN> this is annoying
[17:27] <oSoMoN> popey, and even if that worked, I would need to request a manual review before I can publish it to the candidate channel, so not a good option
[17:27] <oSoMoN> I mean, not faster than my upload link anyway
[17:27] <popey> do you not have an exception for it?
[17:28] <oSoMoN> not yet, jdstrand needs to update the review tools for it
[17:28] <popey> ah
[17:28]  * jdstrand notes he did that. waiting on a store pull
[17:29] <popey> actually, if you upload it to the store, I can get it from there, whether it's approved or not ;)
[17:29] <popey> magic
[17:29] <oSoMoN> that’s if the push works…
[17:29] <oSoMoN> dunno why I’m getting a 401
[17:29] <popey> try snapcraft logout and snapcraft login
[17:29] <popey> snapcraft whoami will tell you who you're logged in as
[17:30] <popey> those macaroons timeout though
[17:30] <oSoMoN> ok, pushing from a different place works, this is in progress
[17:30]  * oSoMoN notes that this is not doing a delta upload, but since I'm pushing from a machine in ze fast cloud, that will do
[17:31] <popey> \o/ the cloud
[17:31] <oSoMoN> the magical cloud
[17:32] <oSoMoN> popey, https://dashboard.snapcraft.io/dev/snaps/8166/rev/21/
[17:32] <popey> whee
[17:32] <popey> blimey, 233MB
[17:32] <popey> chunky
[17:33] <oSoMoN> at 90KB/s, that would have taken the most part of an hour to get it to you :/
[17:33] <popey> ouch, your connection that bad?
[17:36] <oSoMoN> I’m on a 3G router and download speed is very decent, but upload is terrible
[17:37] <popey> oSoMoN: that snap launches on my nvidia laptop
[17:37]  * popey copies it to the intel laptop
[17:37] <oSoMoN> been hassling my ISP for 3 over weeks now so they come and add the extra 20 meters of fiber to my house, but no luck so far
[17:39] <oSoMoN> popey, for info, that’s the change I made to the snap: http://paste.ubuntu.com/25377793/
[17:42] <popey> oSoMoN: same snap works on my intel laptop too
[17:43] <oSoMoN> popey, that seems to confirm the ordering issue, WDYT?
[17:43] <popey> looks that way :)
[17:43] <popey> we need to get the desktop helpers fixed
[17:44]  * oSoMoN just killed an oversized mosquito full of blood - my blood :/
[17:44] <kenvandine> popey, have you seen a snap that works with the thumbnailer?
[17:44] <kenvandine> oSoMoN, revenge :)
[17:44] <popey> oSoMoN: you have all the fun
[17:44] <popey> kenvandine: not sure what you mean...?
[17:45] <kenvandine> thumbnails get writter to ~/.cache/thumbnails/
[17:45] <kenvandine> which the snaps can't access
[17:45] <oSoMoN> popey, do you reckon we should go with this manual override in the meantime?
[17:45] <popey> it certainly works better for me
[17:45] <kenvandine> popey, just wondering if anyone has worked around that
[17:45] <popey> oSoMoN: worth a forum thread perhaps?
[17:46] <kenvandine> even in devmode it fails to access the thumbnails, but snappy-debug doesn't recommend anything for that
[17:46] <popey> kenvandine: not heard anything about that
[17:46] <oSoMoN> popey, definitely, I’ll start a thread now
[17:46] <popey> oSoMoN: awesome!Thank you!
[17:46]  * kenvandine is working on an eog snap :)
[17:47] <popey> kenvandine: you tried any of these snaps you're making on nvidia? do any of them need GL?
[17:47] <popey> (see above conversation) - desktop helpers seem to break on nvidia
[17:47] <kenvandine> nope :/
[17:47] <kenvandine> my nvidia desktop is still packed in a box from when i had to move out of my office to get the floors done
[17:48] <kenvandine> i guess i have a good reason to get that setup again
[17:48] <kenvandine> i thought the desktop helpers had the mesa paths set
[17:48] <popey> yeah, they do
[17:49] <popey> but it seems the ordering of LD_LIBRARY_PATH matters
[17:49] <popey> see the paste above
[17:49] <kenvandine> ugh
[17:49] <popey> the desktop helpers seem to put things in the 'wrong' order
[17:49] <kenvandine> ok, we need to fix that in the desktop helpers
[17:49] <popey> ya
[17:49] <popey> who 'owns' them?
[17:49]  * kenvandine can do that
[17:49] <popey> \o/
[17:49]  * popey runs up to kenvandine and gives him a big wet sloppy kiss
[17:49] <kenvandine> popey, can you file an issue on github?
[17:49] <popey> ya, where?
[17:49] <kenvandine> yuck
[17:49] <popey> haha :)
[17:49] <kenvandine> https://github.com/ubuntu/snapcraft-desktop-helpers
[17:50] <popey> ok, doing now
[17:53] <kenvandine> popey, thanks, i want the issue # to tag in the commit
[17:54] <popey> #71 kenvandine
[17:56] <oSoMoN> popey, https://forum.snapcraft.io/t/gl-applications-using-desktop-helpers-dont-work-on-nvidia/1825
[17:56] <popey> might wanna link to the github issue too?
[17:57] <oSoMoN> kenvandine, popey: didrocks was talking about the ordering earlier today, and apparently there are good reasons for it to be the way it is, so this probably requires careful thinking before we just go ahead and change it
[18:02] <kenvandine> popey, https://github.com/ubuntu/snapcraft-desktop-helpers/pull/72
[18:03] <kenvandine> oSoMoN, ah...
[18:03] <kenvandine> well let's see what he says on the PR
[18:06] <oSoMoN> kenvandine, popey : I’ll test building a snap with that change
[18:09] <ricotz> oSoMoN, hey, seeing this dbaccess_RowSetClones test problem of libreoffice with amd64 is not a good sign :(
[18:10] <jbicha> kenvandine: I'm not sure your commit is right: https://git.gnome.org/browse/simple-scan/commit/?id=e11adfdf3b1ad
[18:11] <jbicha> would that make the Ubuntu packages much larger for things like gnome-user-docs/gnome-getting-started-docs ?
[18:16] <oSoMoN> ricotz, no it's not, indeed. Where is this happening?
[18:17] <ricotz> oSoMoN, with 5.2.6~rc1 on zesty/amd64 with 4.4.0-92.115
[18:18] <ricotz> (a ppa build)
[18:19] <ricotz> only zesty -- xenial and trusty succeeded
[18:21] <oSoMoN> huh
[18:22] <oSoMoN> ricotz, have you already retried it?
[18:22] <ricotz> oSoMoN, I grabbed the log and retried it
[18:24] <oSoMoN> good, let's see what happens
[18:25] <ricotz> still not a good sign
[18:26] <ricotz> oSoMoN, are you aware of any progress on that kernel bug?
[18:27] <oSoMoN> popey, does https://dashboard.snapcraft.io/dev/snaps/8166/rev/22/ work for you the same as 21 ? (it’s using kenvandine's patch for the desktop helpers)
[18:28] <oSoMoN> ricotz, there doesn't seem to be much activity on that bug unfortunately
[18:34] <popey> oSoMoN: will test
[18:34] <oSoMoN> thanks
[18:40] <kenvandine> jbicha, is it actually making it larger?
[18:40] <kenvandine> i doubt it made much of a difference for simple-scan
[18:40] <kenvandine> but gnome-user-docs might
[18:41] <kenvandine> jbicha, it was leaving dangling links which breaks the snap
[18:41] <popey> oSoMoN: chromium snap 22 launches on both intel and nvidia, but on intel chrome://gpu shows hardware acceleration, on nvidia its doing software
[18:41] <popey> need to look into that
[18:42] <oSoMoN> popey, did 21 show actual HW acceleration on nvidia?
[18:42] <popey> re-testing, i think not
[18:43] <popey> two separate issues here.
[18:43] <popey> yeah, 21 software rendering too
[18:45] <oSoMoN> ok, so in fact the "fix" is not really doing what it should, it allows the snap to launch, but it's not finding the gl drivers
[18:45] <popey> yeah
[18:45] <popey> gets past one issue to another
[18:46] <oSoMoN> popey, wanna comment on the forum thread, or want me to do it?
[18:46] <popey> feel free :)
[18:46] <oSoMoN> ok
[19:24] <kenvandine> i guess the issue i'm having with thumbnails and eog is really path related, not confinement
[19:24] <kenvandine> the thumbnails get generated in $HOME/.cache/thumbnails
[19:25] <kenvandine> but $HOME is different for the snap
[20:28] <jackpot51> Any thoughts about this issue: https://bugs.launchpad.net/ubuntu/+source/nautilus/+bug/1437502 ?
[21:16] <robert_ancell> jamesh, success!
[21:16] <robert_ancell> (phase 1)
[21:28] <rbasak> My capslock seems to have inverted on Artful. But not for the GNOME shell (Alt-F2 and Super key) and not for the lock screen (different VT?). Firefox and gnome-terminal are inverted, however. Wayland related problem?
[21:29] <rbasak> I'm having to keep capslock on, except when talking to the GNOME shell and the lock screen. Frustrating! I'm sure rebooting will fix it, but is there anything someone would like me to try to figure out where the bug might be first?
[21:29] <rbasak> It does seem to persist after multiple suspend/resume and screen lock cycles.
[21:33] <gQuigs> rbasak: curious,  how about other apps, gedit, etc?  are all apps inverted?
[21:35] <rbasak> gQuigs: gedit is inverted
[21:36] <rbasak> Also gvim
[21:36] <rbasak> libreoffice writer is also inverted
[21:36] <rbasak> I can't think of any other classes of apps to try
[21:38] <rbasak> Aha!
[21:38] <rbasak> I can turn this behaviour on and off by demand.
[21:38] <rbasak> I wondered what a text login VT would do.
[21:39] <rbasak> It seems that my lock screen (gdm?) is on vt1, vt2 has a text login running, and vt3 is my desktop session.
[21:39] <rbasak> I don't know if that's normal.
[21:39] <rbasak> But if I switch to vt2, then toggle the caps lock, the light doesn't change state. However the username input does change state.
[21:39] <rbasak> If I toggle once and switch back to vt3, the inversion is gone and everything is normal.
[21:40] <gQuigs> yea, it is gdm doesn't have the desktop session be the same always
[21:40] <rbasak> I can reproduce the issue by switching to vt2, pressing the caps lock, and then switching back to vt3.
[21:40] <rbasak> And again to turn the behaviour off.
[21:40] <rbasak> Can anyone else reproduce please?
[21:46] <gQuigs> I can't reproduce, but all my machines are running lightdm still - did find the logout issue that I believe made us choose gdm though
[21:53] <rbasak> up 2 days,  6:22
[21:54] <rbasak> Last updated 2017-08-21  16:26:57
[21:54] <rbasak> (FTR)
[21:54] <rbasak> Thank you for checking.
[21:57] <gQuigs> interesting (maybe related) but - https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1250186
[21:57] <gQuigs> g2run
[22:31] <sunweaver> xnox: jbicha: some time ago jbicha thought about handing over upstream maintenance of libdbusmenu over to me.
[22:31] <sunweaver> Is that still an option? At the moment the debian/watch file of the Debian package points to the orig tarball in the Ubuntu pool.
[22:31] <sunweaver> Which is sort of suboptimal.
[22:32] <ochosi> Laney, jbicha: hey guys! i quickly wanted to point you to an upstream gtk+ regression that already bit arch and debian and is going to affect us in 17.10. i know this may not be problematic for the gnome-shell, but indicator-sound and some other implementations that rely on a similar widget (GtkRange/Scale inside a GtkMenuItem) will break in rather awful ways. that includes some xfce panel plugins that cannot be fixed (at least not withou
[22:32] <sunweaver> It would be good to have an active upstream with release tarballs or tags.
[22:32] <ochosi> Laney, jbicha: and without further ado, here's the bugreport: https://bugzilla.gnome.org/show_bug.cgi?id=786029
[22:32] <sarnold> ochosi: irc has line length limits; you were cut off at "at least not withou"
[22:33] <ochosi> rly? that's strange. worked fine here
[22:33] <ochosi> but thanks for the note, i'll re-paste the rest
[22:33] <sarnold> ochosi: yes, many clients don't know anything about it or how to handle it :(
[22:33] <sunweaver> xnox: jbicha: also Laney: would it make sense to move libdbusmenu over into the context of Ayatana Indicators?
[22:34] <ochosi> Laney, jbicha: continuing from above (as i was cut off): "(at least not without removing quite crucial functionality). unfortunately the response upstream was not as positive as i had hoped (apart from garnacho who provided a  working patch for the regression he caused).
[22:34] <ochosi> Laney, jbicha: what are your thoughts on this? on the one hand i've never been a fan of carrying "random" patches downstream, otoh not having the patch will make at least  xubuntu a little broken in prominent places... would you mind to weigh in on the upstream bug? (sorry that i just presume you're "on our side" since the indicator-sound widget was born in  ubuntu-desktop land...)
[22:34] <sunweaver> xnox: I will also try to clean-up the current Ayatana Indicator packages upstream next week and then upload the full stack to Debian unstable EOnW or week after.
[22:34] <ochosi> sunweaver: if you're taking care of ayatana indicators then the bugreport i linked may also be of interest to you
[22:35]  * xnox doesn't have time to talk about this right now. there is a bucket and a half of things broken in artful at the moment, and we are supposed to freeze tomorrow =)
[22:35] <sunweaver> Quick question: would there be any volunteer to de-phablet the indicator-datetime and/or indicator-sound?
[22:35] <sunweaver> ochosi: yes, noted.
[22:36] <jbicha> sunweaver: could you email/ping willcooke and L_aney about turning over maintenance of indicator stuff?
[22:36] <xnox> we did do some dephablet, there is merge proposal to make url-launcher not depend on ubuntu-ui-toolkit which thus makes most indicators _only_ depends on the c++ url-launcher code
[22:36] <sunweaver> ochosi: you maybe in exchange find this interesting, if not yet spotted: https://sunweavers.net/blog/node/60
[22:36] <xnox> but not on the custom ubuntu-ui-toolkit
[22:36] <xnox> that is good for now.
[22:37] <sunweaver> xnox: ack. So, url-launcher would actually have to be made available in Debian, right?
[22:38] <sunweaver> Is url-launcher maintained as a proper upstream project on LP?
[22:38] <sunweaver> https://launchpad.net/url-launcher -> Launch microsoft internet explorer URL files with the default browser.
[22:38] <sunweaver> url-dispatcher, we are talking about, right?
[22:39] <jbicha> ochosi: please open a LP bug for the gtk issue and ping us again next week
[22:39] <ochosi> sunweaver: oh, indeed, that's interesting. so one thing that depends on the indicator stack in xfce/xubuntu is this: https://git.xfce.org/panel-plugins/xfce4-indicator-plugin/
[22:39] <sunweaver> jbicha: ack.
[22:39] <sunweaver> will do that on Friday, tomorrow I have a day off.
[22:40] <jbicha> ok
[22:40] <sunweaver> ochosi: yes, I know. I will work on MATE integration first, then XFCE.
[22:40] <ochosi> awesome!
[22:40] <sunweaver> on the Debian side first...
[22:40] <ochosi> sure
[22:41] <sunweaver> jbicha: can you propagate the fix for the slider bug to the Debian BTS, too?
[22:42] <jbicha> sunweaver: does the bug affect you?
[22:43] <xnox> sunweaver, also yunit?ubiports? they still care about both touch stack and indicators....
[22:43] <jbicha> if so, maybe you 2 should ping flexion about it :)
[22:43] <sunweaver> me personally, not yet, I guess. But once indicator-sound has been ported to ayatana-indicator-sound, I fear it might.
[22:43] <xnox> sunweaver, i was somehow hoping that maybe the code can be kept as a single base, and have touch bits still available and/or maybe compile time optional.
[22:44] <ochosi> sunweaver: fyi, the affected widgets in xfce use the exact same widget (it was borrowed from indicator-sound)
[22:44] <xnox> unless we need a clean break / fork and have "x11"-classic indicators and qml-indicators code
[22:44] <sunweaver> xnox: ack about compile-time optional.
[22:44] <xnox> cause maintaining 3 sets of, mostly, common code doesn't sound engineering wise the most optimal way forward
[22:45]  * sunweaver nods.
[22:45] <sunweaver> the idea is to reduce effort and widen the user base...
[22:46] <sunweaver> I will compile-time disable url-dispatcher code in the indicators that I'll port in the near future.
[22:46] <sunweaver> also, I must check how much I have removed already in the projects already available via Debian.
[22:47]  * sunweaver lacks brain capacity...
[22:49] <sunweaver> xnox: jbicha: I am also think about re-adding support for the com.canonical.Unity DBus interface.
[22:49] <ochosi> jbicha: here's the downstream report: https://bugs.launchpad.net/gtk/+bug/1712701
[22:49] <sunweaver> currently, it is supported under a different (more generic) name (which might be nonsense to have): org.ayatana.Desktop
[22:55] <jbicha> ochosi: you said "Debian" but it's not been reported in Debian's bug tracker yet, right?
[22:55] <ochosi> it has been
[22:56] <ochosi> but i don't have the report id handy
[22:56] <ochosi> i can try to find it though if that helps
[22:57] <ochosi> oh, there it is
[22:57] <ochosi> i'll add it to the launchpad report
[22:58] <ochosi> jbicha: added the debian report to the launchpad report
[22:59] <jbicha> darkxst: btw, I hesitated uploading mutter 3.25.91 to the Staging PPA because of https://bugzilla.gnome.org/786660