=== sergiusens_ is now known as sergiusens [06:39] Hi seb128 [06:39] hey duflu, how are you? [06:39] good morning desktopers [06:39] seb128, alright.. you? [07:07] good morning desktoppers [07:10] duflu, I'm good, not fully awake yet, working on it though! [07:10] lut oSoMoN [07:13] salut seb128 [07:13] good afternoon duflu [07:16] Good morning oSoMoN [07:39] good morning [07:39] Hi didrocks [07:46] hey duflu [08:41] anyone knows what's up with autopkgtests? the queues are suspiciously empty, and they're neither running nor queued for packages recently built in bionic-proposed [08:42] like https://launchpad.net/ubuntu/+source/chromium-browser/64.0.3282.167-0ubuntu1 or https://launchpad.net/ubuntu/+source/libreoffice/1:6.0.1-0ubuntu1 [08:42] jibel, ^ [08:42] oSoMoN, hi, no idea [08:42] Laney, ^ [08:46] oSoMoN: libreoffice needs to make it through the new queue first. Chromium just finished building on armhf so its autopkgtests should be triggered soon [08:48] right, the "(New)" suffix should have been a hint [08:48] thanks jbicha [08:49] I upgraded from 16.04 (Unity) to 18.04 (Unity) and lost CTRL+ALT+T for terminal, is this expected? [08:49] I'd never seen the queues so empty [08:51] jbicha: hey! Do you mind commenting/helping on https://github.com/Ubuntu/gtk-communitheme/issues/135#issuecomment-365672257? [08:54] it's a symbolic icon I believe [08:56] morning [08:58] jbicha: mind answering to them? [08:58] hey willcooke [09:02] hi [09:03] hey Laney [09:03] thanks jbicha [09:11] Morning willcooke [09:11] moin didrocks [09:11] And morning Laney [09:11] moin duflu! [09:11] how goes? [09:13] didrocks: can I interest you in doing some binary NEW processing for bionic? specifically libreoffice and libgweather [09:13] jbicha: if I new libgweather, I won't review the MIR [09:13] so your pick [09:13] Goes well [09:13] I can have a look to libreoffice [09:14] I don't want to NEW and MIR the same package, that's lacking a double check [09:14] libgweather is already in main [09:14] ah, it's only a bin NEW [09:14] can do in a while [09:15] cool, thanks [09:16] binnew is usually processed pretty quickly, but since I wanted to do multiple transitions without intertangling them I've been a bit impatient :) [10:18] jbicha: libgweather NEWed [10:35] Good morning seb128! Is there a way to see a generated .pot file when building in a PPA? Asking because of this: [10:35] https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/1707898/comments/22 [10:35] Ubuntu bug 1707898 in systemd (Ubuntu) "systemd translations are not synced with upstream" [Medium,In progress] [13:13] o/ desktop, I'm facing a situation where a GCE instance running Ubuntu server doesn't shutdown (when pressing the stop webconsole button) when the nvidia_drm is in use by Xorg and managed by lightdm. The only thing we see in the log at the moment of the stop action is "Power key Preseed" and then nothing mre happens.We have notice that lightdm handle the keypresses. Any idea why the shutdown doesn't complete when lightdm is started ? Is it [13:13] because it is awaiting for user input within the xorg session maybe to complete ? [13:19] ninja: error: unknown target 'gweather-3.0-pot', did you mean 'libgweather-3.0-pot'? [13:19] yeah I guess I did mean that 😳 [13:22] jbicha: I noticed you've been working on packaging pipewire for Debian. How confident are you in what you've got so far? [13:23] (just noticed that latest xdg-desktop-portal depends on it) [13:24] jamesh: not confident at all, that's why it got stuck a few months ago [13:24] ah [13:24] the packaging is at https://salsa.debian.org/gnome-team/pipewire [13:25] I got stuck in figuring out how we want to split it into different packages [13:25] ah. looks like I can disable it [13:25] so we can put that off for another release or so [13:26] jamesh: are you aware that xdg-desktop-portal 0.10 is already in bionic? [13:27] jbicha: I was building from source to test out a not-yet-landed feature [13:27] rather than making the support conditional on whether pipewire is available, you need to use --disable-pipewire [13:27] kenvandine: good morning [13:29] jamesh: I just filed LP: #1749672 but you're going to need xdg-desktop-portal-gtk too, right? [13:29] Launchpad bug 1749672 in xdg-desktop-portal (Ubuntu) "[MIR] xdg-desktop-portal" [Undecided,New] https://launchpad.net/bugs/1749672 [13:30] jbicha: yeah. xdg-desktop-portal without xdg-desktop-portal-gtk doesn't do much [13:30] (or xdg-desktop-portal-kde, but that's not so interesting to the main Ubuntu flavour) [13:36] jamesh, yeah i disabled pipewire [13:37] jbicha, oh you filed the MIR, you are awesome! [13:39] jbicha: at this point, it mainly gets interesting once we have the snap support working (which is what I'm still working on) [13:39] kenvandine: do you want to file the -gtk MIR then? [13:42] and smcv already updated xdg-desktop-portal in debian... i did that yesterday because i didn't think he'd get to it quite this quickly :) [13:42] jbicha, i can [13:42] jamesh, how's that snap support coming? [13:42] kenvandine: so um, there's an issue with your Flatpak update in bionic [13:43] what's that? [13:43] smcv was surprised you packaged 0.11 since 0.10 is supposed to be an LTS series [13:43] hmmm [13:43] but 0.11.1 was the version that split out the document portal [13:43] that would conflict [13:44] we figured that was it and that you were unaware that Alex appears to have backported that to 10.4 that he released yesterday [13:44] https://blogs.gnome.org/alexl/2018/02/14/moving-a-portal/ [13:45] smcv is usually pretty prompt at updating flatpak and xdg-desktop-portal(-gtk) [13:45] ugh [13:46] i didn't see 0.10.4 :( [13:46] Alex also backports quite a lot to his LTS series (at least he did for 0.8) [13:47] 0.10.4 doesn't say anything about remove the permission store [13:47] but i guess it must [13:47] is "--disable-document-portal" not sufficient? (I haven't looked) [13:47] i haven't looked [13:48] but in 0.11.1 release it specifically says it removed both the document portal and permission store [13:50] so we might end up wanting to use a "really" version number here but you can ask Alex what he thinks [13:50] jbicha, indeed it looks like disable-document-portal does suppress build the permission store [13:51] ugh... i hate really versions :) [13:51] jbicha, do you know if smcv is actively working on xdg-desktop-portal-gtk now? [13:52] he probably will be soon. You can find him in #debian-gnome on OFTC [13:54] kenvandine: I've got most of my test setup for the new xdg-desktop-portal set up. Just trying to get my portal-test snap rebuilt (partly bitrot, partyl that it doesn't build on straight xenial) [13:54] anyway, it's really cool that Snap & Flatpak are working together on some common standards & tech [13:54] jbicha, i might end up distro patching the snap support patch in our xdg-desktop-portal package [13:54] for the short term [13:54] until it's in a released version [13:55] jbicha, just so we can get a head start testing it as feature freeze is quickly approaching [13:55] jbicha, it is really cool [13:55] kenvandine: smcv likes collaborating on these packages (even though hardly anyone has shown up to help him yet), so he may be willing to cherry-pick the patch in Debian too [13:55] jbicha, that would be great [13:57] on the snap side, niemeyer was against maintaining long term Ubuntu-specific patches, so it is great we've got upstream support [13:57] grrr, the flatpak 0.11.1 release notes really should have mentioned something about the stable series. i didn't know he uses the even/odd versioning [14:00] kenvandine, hey [14:00] kenvandine: it was mentioned way back in the 0.10.0 release notes. I think he just assumed everybody knew :( [14:00] kenvandine, but you should know better than updating random projects to new series without wondering first how they are handled :) [14:01] seb128, indeed :) [14:01] it works fine though :) [14:01] i actually installed some flatpaks yesterday :) [14:01] the version number is confusing, 0.11 isn't obviously a new series compared to 0.10 :( [14:01] yeah [14:03] I believe Debian & Ubuntu were the only distros who stayed on the previous 0.8 "LTS" series while everyone else used 0.9 stuff [14:39] Laney, do you know where one can view the results of autopkgtests triggered against a ppa? [14:42] signal flatpak works, while signal-desktop .deb from signal.org segfaults :P [14:42] on launch [14:44] seb128: they are hidden at a URL like [14:44] https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-bionic-jbicha-temp20171026/ [14:44] for PPA jbicha/temp20171026 for bionic [14:44] jbicha, how do you find out the url? [14:44] I kicked some yesterday [14:44] they were on the main page so I know that worked [14:45] but I would like to see the results/logs today [14:45] but can't figure out how [14:45] just use that URL but replace the part after autopkgtest- for your ppa name [14:45] or the digits don't change? [14:45] how weird [14:46] once you have the page load, it should show your several names, just add the name to the end of the URL like [14:46] bionic/amd64/f/flatpak/20171029_131527_d6930@/log.gz [14:46] becomes [14:46] https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-bionic-jbicha-temp20171026/bionic/amd64/f/flatpak/20171029_131527_d6930@/log.gz [14:47] thx [14:47] shrug [14:47] that notify-osd/dconf/armhf issue on bionic is over me [14:47] it seems indeed to fail with the proposed version of dconf [14:48] but the tests/same script doesn't fail with bionic-proposed and the new dconf on a porter box [14:48] I tried to rebuild old dconf to see if that's a rebuild with the current toolchain creating the issue [14:48] but tests are working [14:49] seb128: the disturbing part to me about that notify-osd/armhf issue is that basically the only change was: [14:49] https://git.gnome.org/browse/dconf/commit/?id=701d19d1 [14:50] yeah, I though maybe it was not an new issue since https://autopkgtest.ubuntu.com/packages/notify-osd/artful/armhf doesn't have dconf entries [14:50] if it were up to me, I'd like to just ignore that test failure 🤷 [14:50] but seems it is :/ [14:51] at the same time there is no reason we need the new version now [14:51] it is a bit of a problem since we are trying to rename d-conf source package to dconf [14:52] that doesn't really have to be done for bionic, but … [14:53] you could upload a rename of .0 [14:53] like we delete was is in proposed [14:53] upload that [14:53] and sync back the new version then [14:54] is there a way to show multiple times (timezones) in GNOME? [14:54] mdeslaur, install gnome-clocks [14:54] then add locations in there [14:54] then they should show up in the panel calendar dropdown [14:55] awesome, thanks seb [14:58] yw [14:59] seb128: I could do the opposite and just revert the new dconf commit with a patch and see if that passes autopkgtest [15:00] hello all [15:00] jibel, you around? [15:09] seb128: btw it's documented at https://wiki.ubuntu.com/ProposedMigration/#Testing_against_a_PPA [15:10] gsilvapt, Hey [15:10] Laney, DOH, that's the page I used to figure out the url to do the querie, I should have read more :p [15:10] Laney, thx [15:11] jibel, how are you doing? I have some questions regarding autopkg so when you have a couple of minutes you could spend with me, that would be great :) [15:12] gsilvapt, i'm good, thanks! sure, just ask your questions. [15:12] and I'll try to answer the best I can [15:13] jbicha, what doesn't make much sense to me is that it works on a porter box :/ [15:13] The thing regarding adding snaps functionality to autopkg. So I want to add a new argument to the command line options like --snap and --snap-channel so it knows it is a snap, it needs to get that snap from the store [15:13] I found a couple of arguments under autopkg_args.py but I'm not sure how to work with that. I can see some options allow to simply write the bash commands but there are some arguments I have no idea what they represent and where I can read more about it [15:13] jibel, ^ [15:16] gsilvapt, I can understand why you'd need a 'channel' argument, but why a snap? you know it is a snap because the packaging format is snap. I don't think you need to explicitly mention it. [15:16] You think? Hum. I thought that would be better so autopkg already knows how to handle instead of having to search the package directory to ensure whether it is a snap or not [15:17] gsilvapt, also it'd be useful to run from a local snap too. So the dev doesn't have to upload to the store to test his package and it simplifies the workflow [15:17] So, if --snap command is called, a couple of different methods are called to take care of things. [15:17] jibel, you can do that without calling any channel command, no? [15:34] jbicha, oh, I debdiffed dconf versions, that's interesting :) [15:36] jbicha, [15:36] -/* dconf.c generated by valac 0.30.0, the Vala compiler [15:36] +/* dconf.c generated by valac 0.37.1, the Vala compiler [15:36] jbicha, between .0 and .1, there are lot of code changes in the generated .c due to the change of vala version [15:37] jbicha, so could be that vala 0.37.1 generated buggy code on armhf [15:38] thanks, let's do a rebuild then since ricotz says newer vala fixes several bugs [15:39] he suggested rebuilding packages but I believe bionic will be getting an archive rebuild for spectre mitigation so I thought we'd wait for that [15:41] Hey everyone, the minimal install option in the desktop installer looks really cool. I was wondering if the list of removed packages will be documented somewhere? Eg in the release notes or on a file on the disk after install? As it would be nice to know which packages I may need to reinstall to get X feature working, Eg names like vino, remmina, baobab are not as obvious as thunderbird, rhythmbox. [15:44] mdeslaur, but you might need to log out and back in [15:44] willcooke: yeah, that's what I figured [15:45] took me a while :) [15:47] jbicha, rebuild doesn't help, but the package is building from the generated .c no? what we would need is to generated the .vala->.c no? [15:48] hmm, we can force it to rebuild from the vala sources [15:49] or somebody needs to do a new upstream release [15:56] 0.27.1 was converted to meson which doesn't have the problem of generated source files in the tarball [16:07] jbicha, update, maybe that fixes it :) [16:08] ok [16:08] jbicha, in any case I spent half a day on that, I need to go back to other work, sorry I didn't figure it out [16:09] but yeah I suggest we try regenerating the .c with a newer vala [16:09] though I still don't understand why it works on a porter box then [16:09] maybe overriding the autopkg is right :) [16:09] Laney would probably disapprove of that though [16:13] I just tried the test using autopkgtest-build-lxd here and it fails like on autopkgtest.u.c [16:13] so that might be a way to go poke at it if you wanted to [16:15] Laney, k, thanks, I guess I just need to make some disk space to be able to do that, it just felt more tedious than using the porter box [16:15] but would have been the right choice [16:16] now I wasted half a day trying to get porter chroots to reproduce the issue and doing ppa rebuild tries [16:19] :( [16:20] it might be easier at this point to rebuild with a newer vala first [16:20] just to see if that resolves it [16:38] * oSoMoN leaves early to visit a makerspace, have a good evening everyone! [18:00] night all. Long weekend of decorating for me, back Tuesday [18:00] ping me on Telegram etcs [20:00] long enough [20:01] I'll be on later tomorrow as I've got an appointment [20:01] so long! [20:08] Laney, have a nice evening [20:15] seb128: regenerating with new vala didn't seem to help :( [20:15] :-( [20:17] I guess it needs proper debugging [20:17] I'm off tomorrow but I can try to have a look next week [20:18] or maybe Laney wants to help, he said he has a setup to reproduce, he could maybe at least get a bt [20:24] jbicha, oh, you didn't include the change from https://code.launchpad.net/~vanvugt/gtk/fix-1698270 in the gtk upload :( [20:25] seb128: I squeezed it into 3.22.28 upstream :) https://gitlab.gnome.org/GNOME/gtk/commits/gtk-3-22 [20:27] oh nice, I see now that it's listed in the Debian changelog entry :) [20:27] great, duflu is going to be happy [20:27] on that note calling it a day (and a week, I'm off tomorrow), have a nice evening/friday/w.e desktopers! [20:41] robert_ancell, popey was testing my gnome-software snap on his kde neon install [20:41] and gets an empty catalog [20:41] https://paste.ubuntu.com/p/8mMXBQJmWb/ [20:42] is the console output [20:42] robert_ancell, it works fine for me in gnome on bionic, 17.10 and debian as well as 16.04 with unity [20:42] seems odd to me that it could have anything to do with the desktop session [21:00] robert_ancell: lemme know if there's anything I can do to help debug this :) [21:01] popey, maybe gnome-software has some if KDE then > /dev/null :) [21:01] kenvandine, popey empty-catalogue means not apps at all show? [21:01] robert_ancell, "No Application Data found" [21:02] is all his screenshot showed [21:02] kenvandine, do you have the screenshot handy? [21:02] https://usercontent.irccloud-cdn.com/file/7qNraUNo/Screenshot_20180215_210219.png [21:02] ^ [21:03] i just forwarded it via telegram [21:03] but popey was faster :) [21:03] :D [21:03] 14:46:21:0081 GsPluginPackageKit ignoring /usr/share/metainfo/org.gnome.Software.Plugin.Snap.metainfo.xml as does not exist [21:04] interesting paths [21:05] popey, does running 'appstreamcli search test' from the command line show anything? [21:05] actually, I'll just spin up a VM - popey, what image are you using for Neon? [21:06] https://pastebin.com/1JNZ9XWN [21:06] eek [21:06] so there's something provided... [21:06] pastebinit has a bug, i think IS changed our pastebin implementation from serial numbers to unique random characters [21:06] oh, it did work [21:06] ignore me [21:06] kenvandine, I guess the snap includes appstream-glib? I wonder if there's different paths on KDE Neon [21:07] perhaps [21:07] a strace might show something [21:07] yes, it's bundled in the snap [21:07] popey, user edition from https://neon.kde.org/download ? [21:07] i think they made the iso harder to find [21:08] for some reason [21:08] I'm using user lts I THINK [21:08] great marketing [21:08] https://files.kde.org/neon/images/neon-userltsedition/current/ [21:08] robert_ancell, gnome-software snap is in the edge channel now [21:08] lemme find the iso i used to confirm [21:09] kenvandine, nie [21:09] nice [21:09] I am the programmer who says nie [21:09] :) [21:09] NI! [21:09] Also, a shrubbery. [21:09] nice shrubbery! [21:09] NaN [21:09] :-D [21:09] :D [21:10] the iso is on my phone and the file manager only shows me the first few characters (yay android) [21:10] neon-userlts... [21:10] oh... a linux iso on your phone [21:10] that seems odd :-p [21:10] (I downloaded the ISO over 3g in the USA at the gate while waiting to get on a plane) [21:10] It was quite an adventure [21:11] insane... there i said it... popey is insane [21:11] https://twitter.com/m_wimpress/status/960120216616423424 [21:11] oh my, those bottles of ketchup freak me out [21:11] hah [21:14] I'd never tried as many distributions as I have now we work on snaps. [21:15] Neat huh? [21:16] indeed [21:28] bing. Downloaded. [21:48] jbicha: not sure if gvfs made it out of -proposed yet, but I grabbed it there - all good here on the mtp front for me :) [22:01] kenvandine, my gnome-software in kde neon just crashes on startup :/ [22:01] robert_ancell, that's probable better than what popey was seeing [22:02] kenvandine, it seems to be crashing in the profiling code [22:02] hits an assertion [22:03] i just tested it on solus, it crashed at the end of installing quadrapassel [22:03] which seemed to be related to not having the platform snap installed [22:03] i installed gnome-3-26-1604 and now it all works :) [22:04] oh, I don't have that [22:04] that didn't effect startup though [22:04] just installed a snap that needed it and something crashed [22:04] not sure why [22:05] does snapd auto install the default providers now? [22:05] don't know [22:05] that could have been the trigger there [22:05] i know that was planned [22:05] anyway... kdeneon is the only real problem one so far [22:06] could that have anything to do with being in a VM? [22:06] maybe graphics related ? [22:06] kenvandine the crash I'm getting? [22:06] yeah [22:06] It shows the UI for a split second, but it's just failing to make a profiling object for some reason [22:06] maybe you aren't getting as far as popey because it chokes on graphics [22:06] i am on intel [22:07] yeah, and robert_ancell is in a VM [22:07] should be fine though... i run this in virtualbox [22:07] oh... i gotta go get the kiddo... bbl [22:09] bye [22:12] it seems to be refining two apps with a NULL id and hitting an assertion on that. Not sure what's going on there... [22:27] robert_ancell: hi, why do you need a transitional package for snapd-login-service? Can't we just have the package be auto-removable? [22:27] or in other words, is there anything outside of Ubuntu that should be depending on or recommending snapd-login-service? [22:28] jbicha, there's nothing that I know of that would have depended on it [22:30] jbicha, I just thought that's what had to happen. You probably know this stuff better than me. [22:31] I think we might want to do some kind of maintscript to remove the /etc/dbus-1/ file, but I'm not very good at maintscripts [22:32] robert_ancell: have you ever done maintscripts? [22:37] jbicha, I think I might have and found them to be terrible... [22:37] lol [22:37] oh damn. We have an /etc file. [22:37] yeah [22:37] so I have this bad idea of adding a maintscript to libsnapd-glib to remove that conf file [22:38] I think that would mean there has to be a transition package - because you can't really put the maintscript into another package [22:38] you might be right [22:39] I'm going to need some other opinions on how to handle this since I've never dealt with this situation beore [22:39] before [22:40] I used to just ignore conf files but it's nice to clean up after oneself :) [22:45] that'd be nice :p [22:46] not that things you do in here affect me much ;) [23:38] Hm, I'm seeing bug reports come in that unity is uninstallable on xenial [23:39] eh? [23:40] tsimonq2: and upgrades are busted [23:40] popey: Since bug numbers are cool, bugs similar to bug 1749840 [23:40] bug 1749840 in compiz (Ubuntu) "apt-get install ubuntu-desktop doesn't work" [Undecided,New] https://launchpad.net/bugs/1749840 [23:41] tsimonq2: I understand it was being discussed in #ubuntu-release but I don't have logs, and this doesn't look like it's updated yet to include the converation https://irclogs.ubuntu.com/2018/02/15/%23ubuntu-release.html [23:42] sarnold: Join #ubuntu-release ;) [23:43] So it looks like phased-updates was set to 0% [23:44] Anyone got a bug number I can mark new incoming bugs as dups of? [23:45] tsimonq2: and we just have to hope that folks have scripted apt-get upgrade rather than dist-upgrade ... [23:45] sarnold: Right... [23:45] Looks like Steve's on it though, so that's Good News.