=== alket_ is now known as alket [07:13] good morning [08:08] Good morning. [08:46] hiho lordievader [08:50] Hey soee, how are you doing? [08:51] lordievader: good, but its colad as hell here in office - 17° :/ and outside is only -4 [08:51] *cold [08:52] soee: That is too bad. Compile KDE and warm your hands at the cpu :P [08:52] Or grab a cup of coffee ;) [08:57] hehe no time for KDE atm, working with Laravel + AngularJS [09:01] http://xkcd.com/1172/ [09:25] Good Morning Folks [09:26] So I'm running Plasma5, and the desktop keeps locking up. The only things I can do are switch tty, or reboot. [09:27] I want to keep running Plasma5 in production so I can test it for 15.04 release, but I need away to be able to identify, and kick the service that is locking up [09:27] Rebooting every 30 minutes, is not going to work out. [09:28] Sick_Rimmit: Try alt+tab next time. It might just be the desktop effects freezing. [09:29] lordievader: OK, thank you, I will try that one [09:30] So far I've noticed it when Firefox is has loaded LinkedIn, and also just now when Rekonq tried to load forum.kde.org [09:32] Sick_Rimmit: If that ain't it, look at the system utilization. [09:32] Perhaps there is some memory leak and it is heavily swapping or something. [09:33] Ah OK, that's a good point I had been looking at Top to see, if I could see something running away with the system, but I didn't check Swap.. [09:34] Also If I can create some repeatable issues, where should we report these KDE ? [09:35] Sick_Rimmit: "vmstat 1" is a nice utility for that ;) [09:38] Oooh, not heard of that [09:38] * Sick_Rimmit ducks off to the man pages section [09:46] Sick_Rimmit: what I do is `killall -9 plasmashell && plasmashell &` [09:46] then: `plasmashell /dev/null &` [09:46] alt+tab and all gentler measures didn't work for me [09:47] but with up-arrow, this takes just a sec from yakuake [09:48] alt+shift+F12 to disable effects [09:56] valorie: did you investigate that with someone? just killing it aint not solving the underlying issue ;) [09:57] Riddell: that's the new thing you wrote? https://community.kde.org/Kubuntu/PPAs#Kubuntu_Continuous_Integration_.28CI.29 [09:57] sitter: yep [09:58] someone wants https downloads of images, is he hiding it from gchq? http://paste.kde.org/p9ijnvhag [09:59] that page is entirely filled with internal documentation :P [09:59] alas, I misled you into thinking the page should be on community, while in fact it probably should have been techbase ^^ [10:00] it's a Kubuntu project and we only have Kubuntu stiff on commuity for now [10:00] stuff [10:00] Riddell: community in general is internal [10:00] sitter: yes, I commented on the bug report [10:00] and supposedly the fix is in [10:00] * sitter got lodsa whining when he put phonon release notes there after having been told so by someone -.- [10:01] however, I've not gotten it in ppa/ next [10:01] what be the bug number? [10:01] ooo [10:01] gosh so long ago, let me search [10:02] Riddell: that person has a point probably [10:03] not so much about the download itself, more about the website though ... since once can't https the website a bad person could maninthemiddle the website thus providing a bogus ISO link and a bogus checksum [10:03] sitter: it is supposedly because of the repeat of "xgeWireToEvent: Unknown extension 148, this should never happen." [10:03] 8 billion times [10:04] blurg, searching on that gives 8 million results too.... [10:04] ah, log filling up perhaps, though I think that should be fixed [10:04] then again looking at my xsession-errors I see kwin warnings gone wild as well [10:04] ... [10:05] hasn't filled my HD yet [10:05] most days I think we should just compile everything in release mode [10:05] kbroulik is the one complaining the most about it [10:05] that would also make software crash less and force developers to figure out more userfriendly ways of data gathering xD [10:05] xgeWireToEvent? I know that! [10:05] it causes my xsession errors to eat all harddisk [10:05] and krunner to eat all ram \o/ [10:05] valorie: it wouldn't fill the hd [10:06] krunenr eating all memory seems more like it [10:06] anywho [10:06] all is vain! [10:06] let's build in release mode [10:06] muahhaha [10:07] seems a bug in krunner though, it doesnt detach when launching an application [10:07] so apparently it keeps track of all the output (and plashashell spams me with xgeWire shit) and eventually eats GBs of ram [10:07] hmmm, I'm not launching from krunner much now that most things can be launched from the taskbar [10:08] and for instance kpat has to be launched in yakuake [10:08] krunner doesn't do it, since the paths are messed up [10:08] kbroulik: the output shouldn't be there to begin with ^^ [10:09] valorie: that should be fixed [10:09] cool! [10:09] there's a fix but oc kubuntu doesnt ship it :) [10:09] debian broke the paths [10:09] re the xgewire thing [10:09] I reported the bugs [10:09] and found the workarounds [10:09] sitter: only reply to my fam question is a non-concluside one about it maybe being problematic http://mail.kde.org/pipermail/kde-frameworks-devel/2014-November/020611.html [10:09] so when things are fixed, great [10:10] http://patchwork.freedesktop.org/patch/22124/ [10:10] kbroulik: my conservative pov is that a patch consitutes a micro fork, so if upstream had an interest in not having people suffer from this upstream would release an update :P [10:10] and who packages updates? right. downstream. [10:10] Riddell: yeah, I dunno, apparently dfaure is MIA actually [10:11] for example I need to run a random CI ppa on Kubuntu to get a recent Qt version since the 5.3.0 shipped in kubuntu has broken touch support [10:11] 5.3.2 has fixed that but the patch did not get backported by kubuntu .. [10:11] kbroulik: my point is unless there is a patch release there is nothing to package [10:11] downstreams often patch software themselves, so .. [10:12] I know [10:12] which is why I keep on saying that people shoudl stop calling themselves downstream :P [10:12] and upstream should stop calling downstream downstream [10:12] it creates a neat mental separation between my product and your product [10:12] when in fact it is the same product [10:12] we're all one big fractious family [10:13] this ^ [10:13] and somebody should totally give Ivan some good weed or something [10:13] so it's everyone's fault that many uses have to suffer from this \o/ [10:13] * valorie is out of ideas [10:14] valorie: that still not resolved? [10:15] my last plea went unanswered, I believe [10:15] it was Thanksgiving then my hubby's bday plus GCi begins in a few hours.... [10:16] * sitter calls for a fork [10:17] who will step up to maintain? [10:17] that is the only thing we're waiting for, I guess [10:17] Ivan seems immovable as a stone [10:18] just have someone setup a CI job to merge ivan's thing into the fork thing :P [10:18] right, self-maintaining software! [10:19] like the perpetual motion machine [10:19] * valorie checks in #kubuntu-ci to see how that's going [10:19] hmm, SoK blog request from someone who seems to have achieved compiling gcompris with unity screenshots http://saikrishna17394.github.io/feed.xml [10:20] bleah [10:20] also, they are late [10:20] valorie: there isn't much to maintain ... the patch proposed by MK is 2 lines, that would be the entirety of the fork (well, and having to release things ^^) [10:20] it is no longer November [10:21] whatever, I'm weary and going to sleep [10:21] valorie: not heard anything from my guy doing planet kde, I'll send him another e-mail to remind about the blog post [10:21] sleep well valorie [10:21] have a great day [10:26] I really hate the subdiring on debian git [10:26] sitter: what would you prefer? [10:26] flat structure like on kde [10:26] before I can clone something I have to lookup where the thing lives, so very annoying [10:28] and konversation is a full source, brrrrrrrrrrr [10:29] yep [10:30] I also don't like full source branches but diane I think said not to worry about updating the upstream source [10:31] for CI I think there mustn't be upstream source [10:33] oh gosh that would be a nightmare [10:33] even worse than UDD [10:36] well, exactly the same as UDD really :P [10:52] mh, somehow my konversation toolbar has no tools :'< === greyback_ is now known as greyback === tazz_ is now known as tazz [11:29] lordievader: You're a very clever fellow. It would appear that you were correct, I just another desktop freeze, and Alt + Tab got me out of it. I'm a happy boy [11:29] Now I just need to trace the cause, and I'll have me a juicy bug report ;-) [11:29] Sick_Rimmit: :D [11:30] I believe it is a know bug, but if not, let me know I'll confirm the bug ;) [11:30] Sick_Rimmit: alt+shift+F12 help too? [11:30] I'm also suffering this problem recently [11:31] Ah, yes that turns Compositor Off / On IIRC [11:32] Ah there seems to be something going on with the Window Manager [11:44] d_ed: 192.168.1.2 [11:44] d_ed: SOCKS [11:44] d_ed: port 8080 [11:59] - serverGroup->setName("freenode"); [11:59] + serverGroup->setName("Ubuntu IRC"); [11:59] why ever are we doing that Oo [12:00] who knows, drop it [12:03] qtcurve has packagin issue btw [12:03] I think the files in /usr/share/ are shared between both versions, but only packaged in the 4 version and there is no dep tie between the two [12:04] Riddell: are you doing konversation beta upload? kubuntu_unstable should be good for branching to archive [12:04] shadeslayer: is there a particular reason the kf5 dh module doesn't use list-missing? and didn't we already talk about list-missing being provided as module? [12:05] eitherway... without list-missing we have a reliability problem for CI projects that use the module rather than the mk include [12:10] sitter: I could. but so could you since you're already looking at it no? [12:11] Riddell: second set of eyes ;) [12:11] there's 300 things that need CI set up -.- [12:11] ubottu: bug 1048606 [12:12] bug 902762 in system-config-printer (Ubuntu Quantal) "duplicate for #1048606 scp-dbus-service.py crashed with ImportError in __main__: No module named asyncconn" [High,Fix released] https://launchpad.net/bugs/902762 [12:13] * sitter scratches head [12:13] I hate patches [12:14] From #kubuntu: < alvin> The amd64 zsync file on http://files.kde.org/snapshots/kubuntu-plasma5/ is wrong, it has the wrong filename [12:16] sitter: vale vale [12:16] shadeslayer: so when is phonon getting fixed? [12:17] lordievader: what does the filename matter? === kbroulik is now known as kbroulik-lunch [12:17] zsync won't be able to find the iso file to download? [12:17] "failed to retrieve from utopic-desktop-amd64.iso" oh yes [12:18] Riddell: I'm just proxying ;) [12:18] shadeslayer, sitterL you have access to that place to remove those .zsync files? [12:18] if the ISO is renamed after the lb run, you'll have to re-generate the zsync file [12:27] Riddell: gone [12:27] thanks [12:28] sitter: debian bug 771220 [12:28] Debian bug 771220 in libphonon4qt5-dev "libphonon4qt5-dev cannot be used in Debian jessie" [Important,Open] http://bugs.debian.org/771220 [12:28] I can't do much more than that without being a Debian developer :p [12:29] fix it in kubuntu :P [12:29] actually phonon probably should get add to ci [12:29] oh ah ok [12:29] mh [12:29] shadeslayer: thats not the problem I meant [12:30] what I meant was that experimental is needed to use the cmake config but -dev doesn't depend on experimental [12:30] shadeslayer: You could make an actual debdiff and attach it to the bug. [12:30] sitter: ah that too [12:30] You might even commit that to the Qt-KDE git repo. [12:30] ScottK: I have a workaround for the bug in the git repo [12:30] but then sitter committed a fix upstream, so I should probably fix that [12:31] s/fix/use/ [12:31] shadeslayer meant: "but then sitter committed a use upstream, so I should probably fix that" [12:31] ... [12:31] kubotu: scru u [12:31] Yeah. [12:31] Can't get uploaded until the current upload migrates in two days anyway. [12:32] shadeslayer: oh and in kubuntu we even have 2 dev packages :O [12:32] ScottK: roger [13:17] 'Morning folks [13:17] hiho BluesKaj_ [13:18] hey soee [13:42] hi :D [13:42] i get this [13:42] sudo add-apt-repository ppa:neon/kf5 [13:42] Cannot add PPA: 'ppa:neon/kf5'. [13:42] on 14.04 so :) [13:42] i wish to install again plasma 5 to send some feedbacks [13:42] https://blogs.kde.org/2014/11/20/kubuntu-ci-replacement-project-neon [13:42] helps ? [13:43] oh sitter thank you [13:43] sitter: but just isos? [13:45] you can add ppas and install P5 [13:45] plasma-next ppa + ci-* [13:45] oor wait you talking about 14.04 ? [13:48] soee: yes [13:48] i doubt you can use it oin 14.04 [13:48] Riddell: ^ [13:48] :( [13:49] guys i just bought a ssd new , removed the old one with windows 8 [13:49] when i use kubuntu i feel at home [13:49] windows 8.1= nightmare xD [13:49] sorry no daily packages for 14.04 any more, you need 14.10 (or vivid) [13:49] i have no idea, never used win 8 [13:49] wifi doesn't work bluetooth idem [13:50] and it's even ugly no sense system [13:50] soee: i hope you do not even try it it's a mess even with drivers [13:50] yeah I had to dust off my win8 for my SoK project, yuck... [13:50] Peace-: i do not need any windows since 1-2 years [13:51] soee: i needed to flash my android phone [13:51] Kubuntu is all i need [13:55] * Sick_Rimmit appears in a puff of dust, bits and bytes [13:58] Hmmm... I think the locking issue on Plasma can be replicated by actively with the Right Click Window manager menu, If you open, select a sub menu then click back to the Window this seems to cause a desktop lock. [13:58] Is anyone else able to replicate this ? [13:58] Sick_Rimmit: that e-mail you answered, I like how his biggest problem with vivid is that he doesn't like the busy cursor :) [13:59] Riddell: Yes, I read it all as a complement too.. [13:59] Riddell: Plus, I think it's good to respond in a postive manner, makes people feel welcome === kbroulik-lunch is now known as kbroulik === alket_ is now known as alket [14:11] Whoops, that was me closing Quassel without configuring Hide to SysTray.. [14:14] :) [14:14] Sick_Rimmit: I don't think I can recreate your replication of locking but can you describe it again? [14:15] Sure [14:18] Right click on the Title bar of a Window. Then point to more actions in the menu, a sub menu appears, now click the mouse back on the title bar, or window to return focus, you should find that it's locked. Then use Atl+Tab to get it back. [14:19] Sick_Rimmit: nope not in my case [14:19] Hmm, ah Ok [14:20] It appears that I can get control back by clicking on another window and then back again, which is different behaviour to the whole desktop locking up [14:20] I suspect they're separate issues.. [14:20] meh [14:20] So I'm no further forward yet then... [14:21] No worries, I'm hammering Plasma 5 as my daily production desktop, I'll get to the bottom of it [14:21] disable one compositing plugin at a time? [14:22] Yeah could be worth a play, [14:23] lordievader: suggested memory leak, but on the last two lockups, I checked that and it still had plenty of FreeRAM, [14:23] I get the feeling is signal related... [14:23] But, I'll take a look at the xorg logs next time, see if there is anything in there [14:28] sitter: you renamed konsole .deb to konsole5 in kf5? is that to keep the kde4 konsole? why would you want to keep the binary? [14:28] Riddell: don't carry that into vivid [14:29] it's utopic compat [14:29] since we can't do a proper restructured migration what with konsole(4) not being CI'd [14:29] for vivid you probably only want to adopt the build-deps and the mk include increement from rules [14:29] oh cool [14:31] sitter: whole desktop lockup or single app ? [14:31] Sick_Rimmit: ^ [14:37] This time, I got whole desktop [14:38] Using Libreoffice to select records from a spreadsheet, clicked and hold on 1 row, then scroll through to select more, push mouse way of screen to spee up select, and LOCK, whole desktop, Mouse icon is a hand. [14:38] Alt+Tab resolved [14:40] Sick_Rimmit: I also can't recreate although I'm not sure I'm following right [14:49] It doesn't do it every time, from LibreOffice [14:49] sgclark: ping [14:49] bukai: pong [14:49] sgclark: your site is fine now? [14:50] i nean the server [14:50] bukai: ahh yeah it is up for the moment [14:51] sgclark: Ah! great! [14:51] bukai: ty === Blizzzek is now known as Blizzz [14:57] Da dah!! [14:57] freeze?# [14:58] OK, In LibreOffice I can repeat this, [14:58] Yep [14:59] I have a sheet with 3000 records in it, I click on say the row at 750, and drag the mouse down to select a load of them. Then pusht the mouse up to unselect, and go past to begin selecting those below 750, and it locks up [15:00] In this case Alt_Tab failed to get me out, but I think I know what happened. [15:00] I switched tty's to go take a look at mem usage, and when I switch back, sddm had locked the screen. [15:01] But something else had keyboard focus so I couldn't type my password in. [15:01] I pressed Alt+Tab and the screen lock vanished (That's probably a bug), and put me back to the desktop no password required. [15:02] But I was still locked out. The junk I had tried typing whilst the screenlock was showing, had been captured in Libreoffice as a cell entry! [15:02] But when I tried to edit, and enter anything I still had no keyboard. [15:03] Alt+Tab showed me Window selections, but I could not get the keyboard to connect / focus into any of them [15:06] Ah blow it, I can't get that to repeat now either... [15:07] I'm none the wiser, sorry, I thought I had something there for a moment [15:07] it's possible it's not very repeatable [15:08] which is an annoying bug [15:09] Agreed, for sure it's real problem, because I'm getting 5, or 6 lock ups each day, ( So far Friday, and Today ) [15:09] I think I'll have to go log diving on the next one.. [15:09] gdb is your friend [15:10] well maybe, you can run gdb but I'm not sure how you'd tell is when to give you a backtrace [15:10] since it unblocks on alt-tab [15:10] run gdb in a console and wait for that console to freeze then blindly type bt ? [15:19] I can give it a go.. [15:19] obviously test it unfrozen first to make sure it works === rdieter_work is now known as rdieter [15:56] qca2 needs uploading again with updated symbols files. [15:58] for the obscure architectures [15:59] * Riddell adds to todo [16:00] Riddell: just compiled/installed sddm from source, rm'ed the init.d script, et voila: sddm starts fine using systemd [16:03] hm, why would libsoprano-dev now be needed to build Amarok? We didn't add any new dependencies... [16:03] kfunk: utopic or vivid? does it start with upstart? [16:04] Mamarok: are you compiling by hand or with kubuntu packaging? [16:04] utopic [16:04] kfunk: note that we also have the legacy sysv init script [16:04] also [16:04] kfunk: /etc/init.d is for sysv init scripts [16:04] so you still have /etc/init/sddm [16:04] probably [16:04] yeah. I *think* I had to remove that, but let me try again [16:06] Riddell: I do by hand, this wqas mentionned by a usyer who follows my blog to do so [16:06] I already have libsoprano-dev, probably pulled in by something else, but I don't remember having installed it myself knowingly [16:07] what I do recommend is to get in the dependencies by running build-dep [16:07] but in this particular case the guy has always followed my blog, and this was never dragged in for him up until now [16:08] (talking about Utopic with backports enabled) [16:09] Mamarok: I'm pretty sure soprano has always been in build-dep, what's changed is that nepomuk has been removed [16:09] soprano probably can be too [16:09] hm, so maybe I was not aware of that because I reinstalled my system completely for Utopic [16:10] but: we do not have "actice" baloo support in current Amarok, it used to have nepomuk, which is now disabled [16:11] no there isn't any baloo build-dep [16:39] * ScottK discovers https://launchpad.net/~ubuntu-cdimage/+livefs/ubuntu/vivid/kubuntu [16:40] build logs all moved [16:40] some cjwatson magic I expect [16:43] Yes. [16:51] ScottK: oh heh, quite old === toscalix_ is now known as toscalix [17:14] Riddell: your really working on all those? [17:15] sgclark: bash for loops for the win [17:15] shrug ok [17:17] On my 15.04, any particular repositories/ppa that should be used? ( like ninjas etc ) [17:18] not currently [17:18] genii: did you install from the daily image? [17:19] Riddell: No, I had to use debootstrap. [17:20] ah [17:20] we are packaging KDE Applications 14.12 in ninjas so you can try that but goodness knows if it works or not [17:21] OK === greyback is now known as greyback|afk [17:54] Riddell: Hi Riddell. This bug is still not fixed https://bugs.launchpad.net/ubuntu/+source/kde-runtime/+bug/1322968 [17:54] Launchpad bug 1322968 in kde-runtime (Ubuntu) "Changing country leads to invalid locale" [Undecided,Confirmed] [17:55] 14.04 being LTS release, I think it is important [17:56] sitter: ↑ ? [17:57] Riddell: sitter said "Invalid country language combinations lead to invalid locales." and closed the bug and then you reopened it: "Reopening as we shouldn't allow it to set incorrect combinations if those give errors" [17:58] just to remind you :) [17:58] mm, he improved it somewhat in utopic [17:58] but I think it's too fragile a change to backport to 14.04 [18:00] many people are affected, there is one more bugreport about the same bug but I cannot find it atm [18:02] sitter: feelings on an SRU? [18:07] Riddell: so is libsoprano-dev a direct dependency of Amarok now or an indirect one due to the reshuffling of dependencies when moving from Nepomuk to Baloo? [18:08] I am pretty sure we never depended on it, it never showed up in our README file [18:31] Mamarok: doesn't seem to be either [18:32] Mamarok: looking at the build log it says successfully found "* Soprano - Soprano libraries required by Nepomuk Collection" but since it also doesn't find nepomuk I think we can just remove it as a build-depend [18:33] I think so, too [18:39] Riddell: did you see my comments about konsole on friday? [18:40] Man. This sddm faces thing is annoying. Doesn't automatically pull in users' ~/.face files ( which is apparently related to org.freedesktop.Account ). Looks like the current workaround is to make a custom /usr/local/sbin/adduser.local to make a symlink from /usr/share/apps/sddm/faces/username.face.icon to their picture in home dir. [18:49] d_ed: ^^ [18:54] yo! [18:55] yofel: nope, I know it needs to be made co-installable wit kde4 kpart, what else? [18:57] Riddell: yeah, that's on the todo. I was more complaining about the weird maintainer, wrong Vcs and watch URLs as the package is marked "done", but really isn't... [18:58] probably I need to tidy them all up for that [18:58] ah, the changelog is missing too [18:58] yofel: but what's wrong with the maintainer? [18:58] Debian/Kubuntu/Ubuntu Qt/KDE Maintainers [18:59] oh weird [18:59] anyway, if you're going to look through them at the end then ok [19:00] yofel: what changelog is missing? [19:00] Riddell: er... you wrote " * New upstream RC release" but reverted essentially all kpart splitting stuff that harald did in 4:4.14.2-0ubuntu1 [19:00] that's... kind of wrong... [19:01] or well, that's at least not a single change [19:01] but nvm, I was confused because the package was marked as done and looked rather incomplete [19:04] well it doesn't need to have the kpart splitting stuff surely? [19:04] thats only for kde4 [19:04] right, the changelog still needs to say that it was removed [19:05] oh right [19:05] shadeslayer: why did you highlight me? [19:06] d_ed: sddm thingamagjig above [19:08] genii: David says that it should be reading ~/.face [19:08] genii: also, maybe file a bug [19:09] .face.icon [19:17] d_ed: Yes, I have 3 users each with a ~/.face, it's not pulling them in ( just showing all as /usr/share/apps/sddm/faces/default.face.icon instead) . [19:17] I'll file a bug against it [19:18] ~/.face.icon [19:18] we have ~/.face though, not .face.icon [19:18] who is "we" [19:18] ubuntu at the very least, I don't know if that even has a specification [19:18] I'll rename and see if that works. [19:20] * Riddell has vauge memories of accountsservice doing this [19:28] Renaming to .face.icon works. Interestingly 14.04/14.10 lightdm also shows thumbnail if it's named that instead of just .face === rdieter is now known as rdieter_work [20:28] Hehe, heated discussion in #ubuntu-offtopic whether KDE is supposed to be an OSS version of CDE originally [20:32] Wikipedia says KDE was kreated by Matthias Ettrich because he was dissatisfied with the lack of consistency of the DEs available at the time. [20:37] * mamarley wonders why anyone would get into a flamewar over that anyway. [20:37] Does it really even matter? [20:42] Probably not :-) I think they were fighting more over who was right or not. [20:47] * Sick_Rimmit perhaps if they're board they'd be interested in some bug triaging ? [21:04] genii: sort of like CDE but good [21:05] Well, I definitely enjoy it. [21:10] IIRC Matthias Ettrich's initial posting about KDE didn't mention "like CDE". [21:12] mamarley: you would if you ever had to use CDE with Netscape Navigator for unix with xeyes with solaris-terminal with whatever the old SunOS office suite was === mamarley_ is now known as mamarley === Pici is now known as Guest74903 === DarthCodus__ is now known as DarthCodus === Mamarok_ is now known as Mamarok === _Groo__ is now known as _Groo_ [21:39] shadeslayer: now I'm at a point where I can't start sddm either way (nor via sysv init sript, nor via systemd, nor via plain invocation) :D [21:39] I'm giving up [21:39] * kfunk defeated [21:40] shadeslayer: one thing makes me curious: [21:40] sddm [21:40] 4: set "$0" "$@"; INIT_D_SCRIPT_SOURCED=true . /lib/init/init-d-script [21:40] that's the init.d script. it tries to source a file I don't even have. where does that come from? [21:50] kfunk: I'd git blame the sddm packaging to see where it comes from [21:53] shadeslayer: sorry, I can't take no more sddm today. I'll need to recover first ;) [21:53] it's very frustrating to debug your login manager, tbh === rdieter_laptop is now known as rdieter [22:43] kfunk: the really difficult bit is getting it to work with the live cd pre-desktop screen [23:16] Riddell: I think I even suggested a SRU [23:17] the code isn't more fragile or anything it simply enforces en_GB and en_US as a language to dictate locale, which I still do not agree with but since people apparently want to have this and the GUI doesn't really offer affordance that the combination is wrong it's the best I can offer [23:19] this sort of misconfiguration won't be possible with kf5 anyway as except for the actual LANGUAGE key you cannot diverge from established libc locales anymore