[03:22] jamesh, if you're looking for a snapd branch to write see https://forum.snapcraft.io/t/getting-translated-snap-metadata-via-snapd/1242 [03:23] If we pass through the HTTP Accept-Language header from the client to the store then translations should be possible... [03:23] (once the store has an appropriate infrastruture) [03:27] There's also the local snap case, which I guess would be language keyed fields in the YAML [03:29] robert_ancell: okay. I guess I should add that to the desktop priorities thread (and update it in general) [03:30] and thanks for merging the snapd-glib branch [03:30] np, thanks for making it! [03:30] I'm just writing a thread about the "verified developer" feature. Not sure if that's in the priorities either [03:48] So, on reboot, my gnome-shell segfaults. This seems suboptimal. [04:29] And a fresh user session doesn't segfault. How "exciting". [05:22] morning [05:26] Morning/good day jibel [05:26] and seb128 [05:28] Hi duflu [05:29] good morning [05:32] Morning didrocks [05:32] A very orange morning [05:32] duflu: ahah, updated to our theme? [05:32] duflu: ironically, despite having ran the theme for a week and half now, I did upgrade yesterday to -0ubuntu2 which rolled me back to no theme :p [05:33] so, I upgraded the theme to everyone's but me :p [05:33] didrocks, yes but I need a fresh machine... my desktop is still booting with the old Ubuntu Gnome splash [05:33] ah, you did install Ubuntu Gnome [05:33] we can change that if you are interested [05:34] it's an alternative on plymouth theming [05:34] didrocks, I was going to suggest the login screen be grey (with organge highlights) but haven't seen how it plays with the purple boot splash yet... [05:34] duflu: yeah, better to check with the real boot splash for continuing [05:35] orange on the grey didn't really work, we did try this [05:35] but I bet we'll change GDM/lock screen after a while [05:35] duflu: /usr/share/plymouth/themes/default.plymouth is what points you to your current plymouth theme [05:35] The lock screen is wallpaper you configure in Settings isn't it? [05:35] (and so /etc/alternatives/default.plymouth) [05:35] not really [05:35] you have 2 lock screens step, correct? [05:36] the first one (showing the date) is the wallpaper your configure [05:36] the second one (the aubergine background) is the same than GDM [05:36] and harcoded in Shell code and css [05:36] I forgot. Only recently disabled the password bit [05:36] https://didrocks.fr/images/artful-shell-transition/default-lock-screen.png [05:36] step 1 ^ [05:36] then, you start typing [05:37] https://didrocks.fr/images/artful-shell-transition/user-unlock.png [05:37] step 2 ^ [05:37] Yes, I skip step 2 [05:37] It's less annoying [05:37] Instead, I lock the door :) [05:40] didrocks, still the two bar transparencies are different. Do you know of any plan for that? [05:40] duflu: see my blog post, we kept that one for now because we don't know what's GNOME design team is going to do [05:40] duflu: they might remove the transparency [05:41] if they keep it, it won't be in the current state anyway [05:41] and we'll have to modify it [05:41] ideally, doing something than they do, if we can, like changing transparency based on top panel one [05:42] didrocks, that would also affect the colour (top is black, side is Ubuntu grey) [05:42] good morning desktopers [05:42] hey duflu [05:43] duflu: yeah, probably [05:44] Weird, the top bar flickers if I scrub the Terminal menus. What is the top bar trying to change to? [05:44] I don't see this [05:45] Doesn't matter. We have bigger issues [06:19] Morning duflu didrocks seb128 [06:22] hey flexiondotorg [06:38] Morning flexiondotorg [06:40] hey flexiondotorg [07:14] Does anyone know why $WAYLAND_DISPLAY is set in a Xorg login? (yes it really is Xorg) [07:35] duflu, about bug 1705157 if, in gdm, you select xorg -> Ubuntu -> xorg then login do you get an xorg session ? [07:35] bug 1705157 in gnome-session (Ubuntu) ""Ubuntu" and "Ubuntu on Xorg" randomly start the same session (maybe Wayland maybe Xorg)" [High,Confirmed] https://launchpad.net/bugs/1705157 [07:36] duflu, then if you select ubuntu -> xorg -> ubuntu do you get a wayland session? [07:36] jibel, I recall using that as a workaround in the past yes. But I haven't hit the bug myself for some time [07:37] anyone else hitting this bug could do this verification ^ ? [07:40] I actually thought it might be fixed now [07:42] it is not, it is reproducible on a fresh install but I cannot find a pattern in which case it starts wayland and which case it starts xorg [07:48] jibel, sounds like it's about time someone just looked at the code, and searched for uninitialized variables etc :) [07:48] Or rather found out what/where "the code" is [07:51] yeah [08:01] morning all [08:02] hey willcooke [08:02] moin [08:03] hey Laney [08:03] hey didrocks!!!! [08:03] what is happening [08:04] apart from spamming of planet gnome [08:04] Laney: not really spamming, unsure how the planet importer work (or rather doesn't :p) [08:04] Laney: so had to repost at least :p [08:05] but I think next posts will be ok, crossing fingers [08:05] otherwise, I guess today is the g-c-c day :p [08:05] adding a new panel, blablabla [08:05] and you? [08:10] mmm [08:11] didrocks: I think the planet importer has an algoritm to try to find an old embarassing post [08:11] jbicha: clearly, hence my followup post :p [08:11] I got my UOA/GOA split post a few months ago :( [08:11] ahah, nice! [08:12] jbicha: especially as the source of the joke (the GNOME release team announcement on the 1st April) isn't even around anymore :p [08:12] I wouldn't add to a new planet without using a tag specific feed [08:12] * Laney is scared of this kind of thing [08:12] :) [08:12] think I spammed planet debian back then [08:12] Laney: it is a tag specific one [08:12] things related to GNOME was tagged as such at the time :) [08:12] then you got what you wanted! [08:13] well, 2011/2012 posts? :p [08:13] knowing that I have a bunch more recents on my feed? I don't think so :p [08:13] afranke asked me to add my blog to the planet for the ubuntu changes to GNOME Shell [08:14] the importer should have import none or just the last ones in the feed [08:14] oh it missed some? [08:14] yes [08:14] https://didrocks.fr/tags/gnome/ [08:14] all the "Day X" [08:15] fail [08:15] + feedback from GUADEC [08:15] yep [08:15] there is a bug apparently (said afranke) opened in 2010 about the first import doing something crazy [08:15] sounds like the importer picks its posts, indeed :) [08:15] yeh, that's what I was referring to [08:15] but I never actually *wanted* to import old posts [08:16] oooooooooooh I'm getting a deja vu [08:16] brain malfunction [08:17] ahah, tired already? [08:18] * didrocks just read Gitlab as "Glib" [08:18] should go to bed as well :) [08:19] One theory of déjà vu attributes the feeling of having previously seen or experienced something that is currently being seen or experienced to that of having dreamt about a similar situation or place and then forgetting about it until one seems to be mysteriously reminded of the situation or the place while awake [08:19] * Laney dreamed about planet.gnome.org [08:22] isn't that your brain and memory are asynced rather? [08:22] I thought it had something with the speed of memory and current time transmission [08:22] and so, you think you have already experienced it because it has been store quicker in your memory than you "living/feeling" it [08:25] doesn't seem to be settled, from wikipedia at least [08:26] ah, interesting [08:49] hey Laney [08:49] ahoy seb128 [08:49] are you well [08:49] ? [08:50] jibel, did you see my comment from yesterday? your steps on your bug are still not clear, is the second login giving you the same as the first or is that the third? [08:50] Laney, yes, bit of a crazy schedule yesterday/this morning but I should be back to more quiet/normal now which is nice [08:51] willcooke, wb! had good holidays? [08:51] hey seb128, great thanks. Could use another week though ;) [08:52] oh, hi willcooke! [08:52] yeah all that time by the pool must have been knackering [08:52] factor in the kids [08:52] :) [08:54] fly paper [08:54] :P [08:54] it'll be ok, just give it 16(?) years [08:55] :'( [08:55] doesn't seem to have made the psycho effect that Laney wanted [08:56] seb128, Laney - it's my AWS meeting again today, could you run the meeting? [08:56] I can do [08:56] merci [08:56] np [08:56] Laney sent the newsletter on friday so it's my turn to do something :-) [08:57] didrocks knows the score too [08:57] seb128, yes I saw. As I said earlier the pattern is not clear. But it seems that the session type is not set unless you go back and forth between 2 sessions. [08:58] Laney: I'll give you soon a Dr Friedlander badge :) [08:58] seb128, like u -> x -> u then login will start a wayland session, and x -> u -> x will start an x session [08:58] jibel, did you try to stop the pulseaudio process to release the previous session to see if that makes a difference? [08:59] seb128, in the debug log when the bug happens the message "Setting session to type ..." is not displayed [08:59] seb128, no but it happens even if you never logged in ie first boot after a fresh installation the default session is Ubuntu but it starts an xorg session [09:00] i'll try to stop pulseaudio to check if it makes a difference [09:03] jibel, I can't confirm the new install case [09:03] wfm [09:04] I tried several installs [09:04] did you rule out your vm/video drivers not liking wayland? [09:07] seb128, yes [09:07] seb128, and I tried on hw [09:19] yeah. I did a fresh install on my new laptop over the weekend, and it seems to be stuck in an X session [09:20] It's run of the mill Intel graphics [09:21] shrug [09:21] I can't confirm here [09:21] anyway, there is really something fishy, tried in my vm yesterday to switch to the GNOME vanilla session, XDG_CURRENT_DESKTOP was coming from /usr/share/wayland-sessions/gnome.desktop but Exec= line was the one from /usr/share/wayland-sessions/ubuntu.desktop [09:21] would be nice if somebody who has the issue could at least report it upstream [09:21] so, even that it's a mix… (and ofc, I logout and log back in, no more issues…) [09:22] didrocks, there is at least https://bugzilla.gnome.org/show_bug.cgi?id=785918 [09:22] Gnome bug 785918 in general "After gnome-classic session, gnome session looks like gnome-classic" [Normal,Reopened] [09:22] which is confirmed [09:22] GNOME_SHELL_SESSION_MODE is carried over between sessions because of remaining pulseaudio process [09:22] which keeps the old dbus env active [09:22] seb128: ah, that was it! [09:22] I really miss lightdm :p [09:23] yeah, the env variable was what I meant by "GNOME_SHELL_SESSION_MODE" env set to ubuntu [09:30] logging out, picking "Ubuntu on X", logging out again, and picking Ubuntu seems to have switched me to the wayland session [09:30] as jibel found [09:35] jamesh, that seems normal behavious? [09:36] ubuntu is meant to be wayland [09:36] seb128: Prior to trying this, I'd left it as the default "Ubuntu" [09:37] seb128: this would lead to DESKTOP_SESSION=ubuntu-xorg [09:37] after switching to "Ubuntu on X" and back, it is actually starting a Wayland session [09:38] k [09:38] well dunno how you get in the buggy state [09:38] it'd be great if someone felt like debugging this... [09:38] if you have it again maybe poke to the gdm/accountsservices configs [09:38] it started in the buggy state :) [09:39] I can't get the buggy state here [09:39] works fine on new install [09:39] and on changes [09:52] I just installed the daily and "WiFi settings" isn't doing anything from the indicator. jibel do you have a new(ish) install which you can confirm that? [09:55] willcooke, not new(ish) with a wifi card [09:55] jibel, ack, np [09:56] hummm [09:56] willcooke, but I can confirm it does nothing on an upgraded system [09:56] If I click on VPN settings, then I get the panel up [09:57] and wired settings works [09:57] It isn't doing anything here either (fresh install from Saturday) [09:58] thanks jibel jamesh [09:58] willcooke, I get this error [09:58] sept. 05 11:57:50 herm gnome-shell[3240]: Settings panel for desktop file gnome-wifi-panel.desktop could not be loaded! [09:58] willcooke, you report it or do I? [09:59] jibel, if you would, thank you! I'll try and work out what's changed there [09:59] I suspect it's VPN related [10:02] I think gnome-wifi-panel.desktop is new [10:03] ah, so could be related to the new control center [10:03] https://git.gnome.org/browse/gnome-control-center/commit/?id=016efdfac8e06a6c223e8dd8ede990bb2d49b8bd [10:06] so, we've got a 3.24 release of gnome-control-center, but a 3.25 development version of gnome-shell [10:06] right, it's waiting for the patches to be rebased, didrocks is working on that [10:07] (and seb128 ;)) [10:08] jibel: FYI, you also have the display panel which doesn't work with new G-S/mutter [10:08] so we'll need to get the new version anyway [10:08] (just a warning so that you don't inspect more there) [10:10] didrocks, all right [10:10] bug 1715117 so we don't forget :) [10:10] bug 1715117 in gnome-control-center (Ubuntu) "Cannot launch wifi settings from wifi indicator "Settings panel for desktop file gnome-wifi-panel.desktop could not be loaded!"" [Undecided,New] https://launchpad.net/bugs/1715117 [10:10] thanks jamesh [10:10] and thanks jibel [10:22] not much point opening bugs about g-c-c/g-s mismatch imho [10:22] but anyway, if people feel better doing it feel free [10:23] we are going to update for sure so they don't really need tracking [10:26] jbicha, hey! I just installed the daily and I don't see the cap. portal config file. Is it being done via a setting instead? [10:27] willcooke, it should be in /usr/lib/NetworkManager/conf.d/20-connectivity-ubuntu.conf [10:27] ahhh [10:27] thanks seb128, I was looking in /etc/NetworkManager [10:27] yw [10:28] etc is for admin changes [10:28] systems default are in usr [10:28] why does n-m captialise its directories I wonder [10:28] yeah, unsure [10:28] it's the same for the tarball/git repo name though [10:28] so at least they are consistent [10:29] but yeah, it's a bit uncommon and weird [10:29] The best thing about standards is that there are so many different ones [10:31] yeah [10:32] changing location, be back online in 10/15min [10:52] Can't we adjust ubuntu-dock transparency to match exactly the top bar transparency? Or vice verse? I think that those two different transparencies looks odd. [10:53] amano: read my blog post, I explained why we didn't touch it yet [11:00] Can anyone running Xenial proposed tell me if they have the correct G-Software icon? i.e. the Ubuntu orange one? [11:01] * didrocks reboots on X as hacking on the Shell [11:05] willcooke, do you have ubuntu-software installed on that xenial-proposed? [11:06] if you opted in for some binaries you might have forgotten that one and it got uninstalled because of versions mismatches [11:06] seb128, ah, that's probably what it is [11:07] seb128, installed it, ran it, still have the GNOME icon, maybe I need to restart my session [11:07] The name is correct in the about box [11:07] The Hiri banner looks cool though [11:08] willcooke, yeah, you probably need to restart [11:16] willcooke, I've the orange case with a capital A on it [11:17] thx jibel [11:33] is the shell slow to start for everyone or just me? On a low spec laptop it takes more than a minute between the moment I press the login button and the moment I can use the shell [11:41] it takes me like 10-15s on a 6 years i5 laptop with rotational disk [11:45] takes ~2s here with a 6 years i7 laptop and ssd [11:46] i've a g-s crash maybe that explains it [11:46] https://errors.ubuntu.com/bucket/?id=/usr/bin/gnome-shell%3A11%3A__GI___libc_free%3Ag_free%3Ameta_wayland_finalize%3Ameta_finalize%3Ameta_run [11:46] oh probably [11:46] * jibel deletes the crash file and tries again [11:50] better, no crash but still over 30s [11:59] nobody on the french forum reported this [12:00] so, at least, not that widespread [12:00] (and I can testify they are testing the changes) [12:00] rotational disk? [12:00] (as part of the low spec) [12:00] we are opening some more files than barebone shell, but it's like 15 of them… [12:03] Hello all [12:05] https://bugs.launchpad.net/ubuntu/+source/gnome-orca/+bug/1714688 [12:05] At this time blind people couldn't use Ubuntu because Orca is broken. For me it is a RC bug because a blind person without orca can't use his computer, it is similar if the screen becomes black. [12:05] willcooke: Do you think someone from the desktop team could take a look to this bug? I've investigated last Saturday the origin of the issue and I've find a way to figure out the issue so it's more or less a matter of uploading so. [12:05] Ubuntu bug 1714688 in gnome-orca (Ubuntu) "orca fails to start" [Undecided,Confirmed] [12:06] I just try to help you a little bit in this area to ensure no regression for the blind Ubuntu user-base. [12:08] alexarnaud, why didn't you keep using bug #1704847? [12:08] bug 1704847 in gnome-orca (Ubuntu) "orca crashes on startup" [High,Invalid] https://launchpad.net/bugs/1704847 [12:08] seb128: it's the request of jbicha [12:08] alexarnaud, if you have changes you want to get sponsored the process is described on https://wiki.ubuntu.com/SponsorshipProcess [12:09] so you were talking to jbicha about it? [12:09] seb128: I've added him on CC of the bug indeed. [12:09] seb128: it's just a one-line change. [12:10] one line or more the sponsoring process is still the same [12:11] alexarnaud, that bug has no diff nor explanation of by autoreconf is needed and what the issue is === jlnr_ is now known as jlnr [12:18] So let me time to take a look more deeply on the sponsorship process. I could only do that on my spare time. [12:19] alexarnaud, looks like jbicha has been responding to you so maybe he's wanting to upload the change you suggest [12:20] seb128: where? I've never seen the answer. [12:21] alexarnaud, you said he's the one who asked you to open a new bug? [12:25] yes, you're right, here https://bugs.launchpad.net/ubuntu/+source/gnome-orca/+bug/1704847/comments/11 [12:25] Ubuntu bug 1704847 in gnome-orca (Ubuntu) "orca crashes on startup" [High,Invalid] [12:27] I've discovered cdbs last Saturday so I don't really understand what is the purpose of the autoconf.mk but the fact is it was included in previous release and it works. [12:27] As I understand it takes care of new files. [12:29] right [12:30] jibel, thanks for sending that session bug upstream, I still find your descriptions confusing though, I hope upstream gets less confused [12:30] jibel, like the test case have ubuntu=xorg and the workaround ubuntu=wayland [12:31] I think it mixes descriptions from the time were xorg was default with now where wayland is default? [12:31] anyway let's see if they ask for details and which ones [12:49] alexarnaud: https://launchpad.net/ubuntu/+source/gnome-orca/3.25.4-0ubuntu2 thank you for reporting and figuring out the issue! :) [12:55] you're welcome.:) [13:05] Laney: do you think LP: #1699216 needs a FFe? And do you think you'll have time to review the accountsservice patch? (I remember you briefly looked at an earlier version of it) [13:05] Launchpad bug 1699216 in gnome-initial-setup (Ubuntu) "Encrypted home support" [Wishlist,Confirmed] https://launchpad.net/bugs/1699216 [13:15] jbicha, thanks for that orca upload [13:15] jbicha, also on the firefox bug, did you remove the s390x from the title/description on purpose? [13:17] seb128: yes, sorry for not commenting about that: firefox/s390x was already removed in yakkety or zesty [13:18] oh ok [13:24] jbicha: I dunno, it's probably fine as long as it doesn't affect any other codepaths, and I don't know if I'll have time to review. [13:24] I did ask around if some people would but they didn't reply to me. [13:26] I can review the other packages but I felt uncomfortable with my knowledge of how libraries work to know if the accountsservice patch was ok [13:29] I'm not extremely happy with having to patch many things downstream for this feature FWIW [13:30] but unless someone wants to work on an upstreamable way to make users with encrypted home I don't see a way around it [13:44] there was talk about encrypted home eventually using native ext4 encryption which should be upstreamable [13:54] IRC logs aren't updated yet... what's the blocker for encrypted home? [13:55] what blocker? [13:56] err.. what's "but unless someone wants to work on an upstreamable way to make users with encrypted home I don't see a way around it" talking about? [13:57] the patches use an ubuntu only flag to adduser [14:03] hey desktopers, is there a chance to go through another poppler transition to update to 0.59.0? [14:03] https://launchpad.net/~ricotz/+archive/ubuntu/staging/+sourcepub/8235372/+listing-archive-extra [14:04] gQuigs: the blocker is someone reviewing the accountsservice patch from LP: #1699216 and checking that it won't cause us maintainability problems [14:04] Launchpad bug 1699216 in gnome-initial-setup (Ubuntu) "Encrypted home support" [Wishlist,Confirmed] https://launchpad.net/bugs/1699216 [14:09] ricotz, you need a ffe -> release team [14:10] why would we want to do that? [14:10] ricotz: [14:10] you should provide a strong argument otherwise I don't think so [14:17] ricotz: does it fix the jpeg regression? - https://bugs.launchpad.net/ubuntu/+source/poppler/+bug/1714596 [14:17] Ubuntu bug 1714596 in poppler (Ubuntu) "PDF images are blank" [Undecided,Confirmed] [14:17] seb128, I don't find a better way to describe the bug than its title. I've 2 laptop exhibiting this bug, 1 always starts xorg the other always wayland ... [14:17] * gQuigs testing [14:17] and now upstream wants a proof that it's an upstream bug, that's helpful [14:18] jibel, right, don't worry, the title/issue is understandable, it's the steps/testcases which are a bit confusing but I guess it's because nobody found a reliable way to hit the issue [14:18] bah [14:18] would be good to reproduce it in like fedora with multiple sessions maybe? :p [14:19] the triager is annoying [14:19] let me reopen [14:24] jibel: Is this issue related to yours? https://bugs.launchpad.net/ubuntu/+source/poppler/+bug/1714596 [14:24] Ubuntu bug 1714596 in poppler (Ubuntu) "PDF images are blank" [Undecided,Confirmed] [14:24] Ah sorry this one https://bugs.launchpad.net/ubuntu/+source/gdm/+bug/1714876 [14:24] Ubuntu bug 1714876 in gdm (Ubuntu) "GDM doesn't show xorg session at login (Artful)" [Undecided,Incomplete] [14:33] how do you think we should fix bug #1712289? [14:33] bug 1712289 in ubiquity (Ubuntu) "Wrong theme in ubiquity-dm and 'Install Ubuntu' " [High,Confirmed] https://launchpad.net/bugs/1712289 [14:33] it's because the XDG_CURRENT_DESKTOP isn't set there I think and now we have per-session overrides [14:34] presumably in bin/ubiquity-dm somehow [14:34] seb128, Laney, yeah, mostly enhancements and new features, so not really a bugfix release [14:35] ricotz: nope, it doesn't fix the jpeg regression - it seems like all jpegs won't display now... [14:35] gQuigs, hmm, I don't think so if is it related to the unmaintained jpx part [14:35] openjpeg is what is needed as the bug already suggests [14:35] Laney: yeah, now that previous ubiquity is out of proposed, someone can tackle this. I may do it after the g-c-c thing. It's a one-liner fix via XDG_CURRENT_DESKTOP IMHO + creating a new mode for the Shell [14:36] or you will have unmatching theme in the Shell [14:36] and if you set it to "ubuntu", you will have the dock, which you don't want :p [14:37] ricotz: well it used to work with a lot more PDFs in poppler in 17.04 and now doesn't [14:38] didrocks: ah right, there's already a mode for ubiquity, presumably that just needs the theme [14:38] Laney: hum, I didn't see it? I guess we are using the gdm one otherwise? (which might be enough) [14:38] it comes from ubiquity itself [14:39] oh [14:39] interesting [14:39] bin/ubiquity-dm: wm_cmd = ['gnome-shell', '--sm-disable', '--mode=ubiquity'] [14:39] yep [14:39] so just need to add the theme to it [14:39] Laney: I can have a look either by EOW or next one [14:39] sounds easy anyway, just unsure how to test it easily [14:39] * gQuigs wonders if most users open PDFs in browser these days... renders fine in Firefox.. [14:40] (I hope that ubiquity won't stays forever again in proposed) [14:41] bah [14:41] it's hard to hack ubiquity-dm [14:41] I wonder if a persistent usb key isn't the best :p [14:44] Laney: I think we need https://code.launchpad.net/~jbicha/ubiquity/adapt-to-gsd325/+merge/329902 [14:45] sure, but that's not the same bug [14:45] oh [14:46] hmm [14:46] maybe I can use a shared mount thingy in virt-manager to get at my home directory [14:46] probably won't work from busybox will it [14:47] Laney: what do you try to do? [14:47] copy ubiquity-dm back in [14:48] ah, yeah, from casper_bottom, you can't do much apart from trying to hack directly [14:48] which is doable for tests at least [14:49] it's really hard to insert lines there [14:49] especially in python [14:49] yeah, appending is easy with a cat >>, but editing… [14:50] * Laney optimistically tries to install vim /o\ [14:50] huh, it seems to work [14:50] when breaking at casper_bottom? [14:50] yeah [14:50] then chroot into it [14:50] waow :) [14:51] sounds like it's going to be easy then [14:51] do you want to export the XDG_CURRENT_DESKTOP unconditionnally for G-S? [14:51] probs [14:51] so I guess in wm_cmd = ['gnome-shell'… [14:51] https://bazaar.launchpad.net/~laney/ubiquity/1712289/revision/6552 [14:51] and ofc, the other change is just http://paste.ubuntu.com/25472549/ [14:51] !!! [14:52] yep! [14:52] as gnome-shell or ubiquity is supposively launching the other processes [14:52] it should be good to load the correct gtk overrides [14:52] (theme) [14:55] I think we need a dbus-update-activation-environment too [14:56] oh correct [14:58] noooooooooooooooOOOOOOOOOOOooooooooooooooope not enough [15:01] how's this override thing supposed to work? [15:04] just using the mode is enough [15:04] it's either the env var or the --mode option [15:04] you don't have the css taken into account? [15:05] or you mean the per desktop name override? [15:05] (gsettings) [15:05] you need the gtk theme too [15:05] yes [15:05] for the latter, it's simply gsettings (glib) reading the correct matching key [15:05] can you look if the ubiquity-gtk process has the env variable? [15:06] it does [15:07] but gsettings get gives Ambiance [15:07] erm [15:07] Adwaita [15:08] gsettings get in a terminal which has the env variable? [15:08]