[07:48] Riddell: I think we need to do some more advertising, no female ones have sold yet... [07:58] Good morning. === jackson is now known as Guest46812 [08:17] jussi: have you written to the kub. user list? [08:17] posted on the forums? [08:17] valorie: I havent. mind doing that? === Guest46812 is now known as Noskcaj [08:18] if you ask in a couple of days [08:18] :) [08:18] the workmen arrive in a few hours to demolish our front bath [08:21] * apachelogger has a really terrible headache today -.- [08:23] apachelogger: hair of the dog :P [08:23] *cough* [08:26] why it aint no hangover [08:26] I'd not be up at this hour if it was :P [08:41] haha [09:23] maco: https://wiki.ubuntu.com/TrustyTahr/ReleaseSchedule [09:26] Sput, ScottK: what's new in quassel 0.10? [09:28] a reluctance to declare it stable? [09:29] Riddell: how is that new? :P [09:30] apachelogger: https://github.com/quassel/quassel/blob/master/ChangeLog [09:30] cheers [09:30] (yes, I'm actually keeping this up to date mostly, these days) [09:34] ^^ [09:42] agateau: where is it best to install the kapidox files to? [10:00] anyone with kdeconnect around? [10:02] Riddell: you mean kapidox code, or the output of kapidox? [10:03] Riddell: kapidox code is a standard Python app, just follow Ubuntu way of installing Python stuff [10:03] kapidox output, I don't really know. /usr/share/doc/kf5? [10:09] apachelogger: yes [10:10] agateau: it's the code I was pondering, dunno if it should go in /usr/lib/python or in its own place [10:11] apachelogger: I do... [10:11] Riddell: isn't there a standard way of doing this on Ubuntu? [10:12] agateau: yes, I think my question is if it's a module (library) used by other things or standalone [10:12] I guess it's standalone [10:12] Yes, it's standalone [10:14] agateau: should the parts which the user is supposed to run be in /usr/bin? e.g. kgenapidox [10:14] Riddell: yes, I'd do this [10:15] soee, jussi: can one of you createa screenshot of something not too technical for https://wiki.ubuntu.com/TrustyTahr/Beta1/Kubuntu please [10:15] soee, jussi: a systemtray icon if it has one, or short of that the KCM [10:15] (ultimately with default theming and stuff please :)) [10:16] bah, default theme :/ [10:16] ^ :) [10:16] also it does not see my device :D [10:17] awww [10:17] poor device [10:17] soee: have you paried it? [10:18] jussi: yes i have it listed, but this settings are from older installation [10:18] actually, weird, seems buggy here als [10:18] android 4.3 [10:19] I just paired my note 2, sees the device in settings, can pair/unpair, ping, but it doesnt see it in the normal systray thing [10:19] 4.3 also [10:19] ahh now it came up [10:20] hmm [10:22] jussi how does it pair exactly ? [10:23] i have Device not reachable [10:25] soee: right click, kde connect settings, select device, request pair [10:26] jussi: yes yes ;) it ends with Device not reachable info message [10:26] soee: is the device on, kde connect started, maybe click refresh there? latest version of kdeconnect from store? [10:27] jussi yes :) [10:27] i hope 14.04 is not a problem here [10:27] soee: Im rtunning 14.04 [10:28] brb [10:31] Riddell: i386 symbols patch committed [10:31] Riddell: kde4support ^ [10:32] sgclark: groovy [10:35] ubiquity report sent [10:36] Riddell: did you upload plasma-nm with a po/ directory? [10:42] apachelogger: I hope I did [10:43] " * New upstream snapshot with translations" yay I did [10:43] \o/ [10:44] * apachelogger wonders about the size of the l10n QA card -.- [10:45] I really think we should get an xx language going somehow [10:47] apachelogger: I did suggests this to launchpad translations people back in the day but they were very skeptical for reasons I never worked out, and presumably there aren't any lp translations people now [10:48] jussi: doesn't work for me at all [10:48] Riddell: yeah, but even if we only cover software that is not piped through launchpad we'll have like 80% of the installation covered [10:49] (i.e. only things going through launchpad are our apps, libreoffice, and firefox) [10:49] true [10:49] although, I am not even sure about firefox [10:49] are you sure about libreoffice? [10:49] given the cadance of releases it likely doesn't get piped through launchpad either [10:49] Riddell: no, we'd not care eitherway though :P [10:49] only thing missing is our stuff [10:50] and that can likely be checked by other means, alas a bit more messy [10:51] woudl require a) having a script to get all templtes a package creates, then querying launchpad api whether those templates are available for translation b) another script that autogenerates xx at build time and stores it in a special path so that the mangler doesn't eat it when ripping out translations [10:51] or b) could be done outside the archvie package build I guess somehow [10:51] not impossible at any rate, plus b) we'd want for all package builds actually [10:51] soee: and the device has wifi on? [10:52] and is connected etc [10:52] since nothing ships with the test language [10:52] we'll have to create it from scratch [10:55] jussi: ah wifi was on but not connected :/ [10:55] works now :) [10:57] Riddell: kfileaudiopreview ready [10:57] https://trello.com/c/epuzP9Vy [11:05] "l10n review #2 on 14.04 Deadlines & Quality Control is due tomorrow at 12:00 pm " [11:05] much scared [11:06] jussi, soee: so, do I get a screenshot? :P [11:07] i would have to create new account with default settings :) i can do that wafter work [11:07] groovy [11:07] Morning [11:07] ahoy manchicken [11:08] I'm giving kubuntu-debug-installer the hour that I (hopefully) have before the kids wake up. [11:10] manchicken: :) [11:10] Poop. I can replicate the problem of the sources being duplicated. [11:12] better than not being able to replicate a problem [11:13] Now the question is whether that's QApt or kubuntu-debug-installer. I'm worried that it is the former, but I'm pretty sure that my automated tests cover that. [11:14] automated tests I had to disable :) [11:14] Which worked for me though. [11:14] I'm going to play with that really quick. [11:14] only if qapt is already installed [11:14] if it's not the tests don't work [11:15] Changes I think you guys made to QApt were disabling the tests (due to header file name use?) and then turning off some of the dependencies in the CMakeList.txt that were duplicated from the top-level. [11:15] Am I missing anything? [11:16] only this patch manchicken http://starsky.19inch.net/~jr/tmp/kubuntu_disable-tests.diff [11:16] Okay, cool. [11:17] If I sent you a patch, would you be able to test it quickly? [11:17] manchicken: ok [11:17] I was thinking about it, I think that replacing the include_directories in the tests/CMakeLists.txt with one that didn't have hard-coded relative paths would resolve the issue. [11:18] http://pastebin.ubuntu.com/6999233/ [11:18] (I'm enjoying Qt Creator these days) [11:19] manchicken: is that to the top level CMakeLists.txt ? [11:19] * Riddell does hate the filename CMakeLists.txt [11:19] tests [11:19] I also am not a fan, especially since there's so many "CMake*" artifacts left around. [11:19] s/there's/there're/ [11:19] manchicken meant: "I also am not a fan, especially since there're so many "CMake*" artifacts left around." [11:20] I think that'll resolve the build issue, because I'm guessing your build directory isn't going to match what my setup was. [11:22] Ooh, still broken. Now it breaks my build. [11:30] Poop. make[2]: *** No rule to make target `src/qapt_automoc.cpp', needed by `src/CMakeFiles/qapt.dir/qapt_automoc.cpp.o'. Stop. [11:30] make[1]: *** [src/CMakeFiles/qapt.dir/all] Error 2 [11:34] ' Morning folks [11:36] * manchicken blames apachelogger... [11:36] wut [11:37] manchicken: that makes no sense [11:37] I think you broke QApt's build: with 2c0397e4d31b14c2a6889c8774ac86469a160cf8 [11:38] Riddell: http://paste.ubuntu.com/6999296/ [11:38] It's not properly generating workerdbus stuff anymore. [11:38] manchicken: wipe your build dir please [11:38] I have many times :) [11:39] wtf [11:40] manchicken: make clean; cmake .. -DCMAKE_VERBOSE_MAKEFILE=TRUE && make -j1 [11:40] paste output of all of that please [11:40] I cannot reproduce the moc failure [11:40] in fact I think qapt_automoc.cpp should not be generated at all [11:41] -automoc4_add_executable(qaptworker2 [11:41] +add_executable(qaptworker2 [11:41] ^ because of this change [11:41] I'm totally open to it being my fault. [11:41] manchicken: are you sure you don't have a lingering diff in the worker cmakelists? [11:42] Yeah [11:42] I get nothing on git diff or git diff --cached [11:42] peculiar [11:42] well, get me that log please [11:43] * apachelogger gets some coffee meanwhile [11:44] sgclark: it's a plugin so it's ok for it to act weird as it's not a real library so ignore that [11:46] ok [11:46] http://pastebin.ubuntu.com/6999327/ [11:47] Full disclosure: I have a rather basic level of what I hesitate to call "understanding" of CMake building still. [11:49] manchicken: asking on #kde-devel often helps with cmake foo [11:49] That was weird... #kubuntu-devel just disappeared from my "All Chats" list. [11:49] there's probably also a #cmake [11:49] Riddell: I struggle with this, though... I have spent vastly more time struggling with build issues in the last several months than I have on code. [11:49] That makes me sad. [11:50] manchicken: for some reason your build uses the automoc4 binary rather than the builtin automoc [11:50] :O [11:50] apachelogger: It could be that I have qt5 installed and I keep having to tell the thing to pick up qt4 [11:51] that should not matter [11:51] manchicken: export QT_SELECT=4 [11:51] then wipe your build dir and try again [11:52] manchicken: also what's the output of grep -ri automoc /home/manchicken/Documents/Devel/kde/libqapt [11:54] apachelogger: I sent you the screenshot of the Qt Creator dialog that seems to be weird about Qt setups. [11:54] mh [11:55] manchicken: I tend to run cmake manually outside qtcreator and then open the project in qtcreator, that way one knows what is going on ^^ [11:56] apachelogger: http://pastebin.ubuntu.com/6999360/ [11:56] i.e. mkdir build && cd build && cmake .. -Detcetcetc -> qtcreator -> open cmakelists.txt -> point build dir to mine -> let qtcreator run cmake without arguments [11:56] as suspected there is nothing calling automoc4 :S [11:57] * apachelogger creates a saucy chroot [11:58] apachelogger: Here's my cmake output: http://pastebin.ubuntu.com/6999371/ [11:58] I just added the QT_SELECT -D, seems like it doesn't like how I did that. [11:59] manchicken: export QT_SELECT [11:59] it's an environment variable for qtchooser, which in turn is the thing that actually backs /usr/bin/qmake /usr/bin/moc etc [12:00] Yeah, for some reason I thought it would take -D as well as off of the environment.. [12:01] Updated cmake output: http://pastebin.ubuntu.com/6999387/ [12:01] Arg, baby is up. [12:02] apachelogger: https://launchpad.net/~rohangarg/+archive/nightly/+files/libkpeople_0.2.1-0ubuntu1.dsc [12:02] upload pzl [12:02] manchicken: I'll have a look in a chroot [12:02] *plz [12:02] it all looks fine, except it isn't [12:02] Okay, Dacia has me covered for now, I probably only have about 15min left. [12:03] shadeslayer: are you going to pick up a card to drop kpeople dep? [12:03] apachelogger: Updated build output: http://pastebin.ubuntu.com/6999401/ [12:03] shadeslayer: also you might ask cj nicely to add libkepeople to the package set if not ;) [12:04] yeah, libkpeople needs adding to supported [12:04] I [12:04] I can do that myself, then next time meta package is uploaded, it'll get added to the packageset [12:04] manchicken: ls -Rla in your build dir please [12:05] Uploading libkpeople_0.2.1-0ubuntu1_source.changes: done. [12:05] Successfully uploaded packages. [12:06] thx [12:07] http://pastebin.ubuntu.com/6999421/ [12:07] the amount of deps qapt has -.- [12:08] manchicken: is that output from qtc? [12:09] (I cannot reproduce the issue in a saucy chroot either) [12:09] That's output that I did `ls -Rla > ls.txt` from [12:09] You mean the build dir [12:10] manchicken: the make output I meant, sorry [12:10] That's from qtc running `make -n` [12:10] (or make all -n, I guess) [12:11] apachelogger: re kpeople deps, sure, write a card to "Fix all broken deps" since kde-workspace has one too [12:11] so a meta card would be nice [12:11] manchicken: can you please try to build it entirely outside of qtc [12:12] Trying that now [12:12] https://bugs.launchpad.net/ubuntu/+source/kde-workspace/+bug/1197261 [12:12] Launchpad bug 1197261 in kde-workspace (Ubuntu) "Can't install kde-style-skulpture on KDE 4.11 Beta 2 - conflict with kde-window-manager package" [Undecided,Confirmed] [12:13] ohohoh [12:13] manchicken: it's the -n [12:13] -n, --just-print, --dry-run, --recon [12:13] Print the commands that would be executed, but do not execute them. [12:13] huh? [12:13] :O [12:14] manchicken: I think your qtc build rules might be misconfigured ^^ [12:14] I'm not sure why that ended up in there... I think I was trying to walk it through troubleshooting. [12:15] I've had damn near no end to build problems, so anything's possible. [12:15] I'm building it outside qtc now. [12:15] http://i.imgur.com/DBHVlhQ.png [12:15] It's not dead yet. [12:15] qtc should have the -n on your build page [12:15] in the ardditional arguments line I guess [12:17] oh nvidia drivers update, hope it will work with prime [12:19] shadeslayer: regarding the bug, I think the breaks is there becasue MG changed API but didn't want to bump soversion or something like that [12:19] It does [12:20] manchicken: drop it and all should be fine again [12:21] apachelogger: right, skulpture hasn't been updated in a loooonnggg time and I would argue that it's bordering on unsupported software [12:21] REMOVE IT [12:21] * apachelogger drops some pain killers [12:21] debian doesn't have it as well [12:21] kubotu: order painkillers [12:21] * kubotu slides painkillers down the bar to apachelogger [12:21] apachelogger: will do [12:21] * shadeslayer orders some wine for apachelogger and leaves for lunch [12:22] how can i list guggested packages for some package ? [12:22] apachelogger: I killed it, I got another build error, I'm cleaning up again and trying again. [12:22] shadeslayer: :O [12:23] So much of this seems like it's me cleaning up very rough trial-and-error builds from before. [12:24] Can I still blame you, though? [12:24] Successful qtc build. [12:24] manchicken: you can, I will just disprove your claim :P [12:25] I'm okay with that. [12:25] I'm going to make a suggestion for the tests/CMakeLists.txt, you tell me if you think it'll fix the build issue on the tests. [12:26] http://pastebin.ubuntu.com/6999487/ [12:26] I had ../src in there before. [12:28] manchicken: it's at least better cmake code ^^ [12:28] Okay, I'm going to commit that to master. [12:29] but AFAIR you simply need to change your includes to fix the build [12:29] Is this a generated file: src/worker/org.kubuntu.qaptworker2.policy ? [12:29] Naw, you said I was using camel-cased includes, but I'm not. [12:29] instead of you'd use and make sure the folder containign it is in the includes_dirs [12:29] manchicken: oh, I misremembered then [12:29] I'm pretty sure that it's because the includes were hard-coded to ../src [12:30] That makes the most sense... [12:30] entirely possible I did not look at it in detail [12:30] as for the policy file, yes that is generated from a desktop file [12:30] Okay, git keeps complaining that it's deleted. [12:30] huh [12:30] oh no [12:30] manchicken: it's not generated, but it should be ^^ [12:31] (i.e. in kdeware it would be geneated because there's no other way to get it translated, however qapt does not care about translations :S) [12:31] Yeah, so it shouldn't even be in the repo, yeah? [12:31] it should [12:32] you need the policy file for polkit [12:32] ultimately however it should be replaced by the desktop file I was talking about (alas, that sort of requires a hard build dependency on kde) [12:33] manchicken: not sure why it would be delted tho [12:33] seems a tad odd [12:33] Okay, I don't know how it ended up killed. [12:35] Okay, that tests/CMakeLists.txt has been pushed. [12:35] I'm checking now to see if something (further) is wonky in my dev which could have killed that file. [12:35] My cleanup scripts don't do anything with that file, so it wasn't one of those. [12:36] w00t! My tests build and pass once again. [12:38] Okay, I don't feel like I accomplished what I wanted to (getting kubuntu-debug-installer dusted off), but I did accomplish something. [12:38] :) [12:38] Hopefully the tests will build for someone other than me. Do you think you'll get to those today? [12:38] (or anybody else) [12:38] I know you're a busy guy [12:38] ah dang, I could have tried that in my chroot, already closed it tho -.- [12:40] Okay, I'm going to do the whole getting ready to go to work thing, hopefully folks have better luck with that stuff today. [12:40] Later. [12:43] /tmp/libqapt/tests/transactionerrorhandlingtest.cpp:26:31: fatal error: LibQApt/Transaction: No such file or directory [12:43] #include [12:43] manchicken: I told you it was a camel :P [12:45] Riddell: libqapt master should now build tests as expected [12:45] good enough if we pick it for the next release though IMO [12:45] for/through [12:46] I am not sure the tests will work right now anyway [12:46] because I don't think they abstract the worker, so it requires a complete polkit/dbus setup I think [13:05] apachelogger: awooga [13:10] Riddell: wierd error after I made symbols: http://paste.ubuntu.com/6999655/ [13:15] sgclark: what's the error? [13:15] there's multiple libraries in that package so they need their symbols added too [13:15] which you might be able to do with just batchpatch [13:16] ok [13:16] but I suspect needs the symbols file moved out the way, a new symbols file made, then copy and paste them together. ug [13:27] test cases almost done [13:27] * apachelogger sighs a bit [13:34] cool pipelight runs amazon oinstant video inside ubuntu B-) made my day [13:35] btw hello folks [14:02] Sadness: http://blog.hostilefork.com/qtcreator-debugger-no-locals-ubuntu/ <-- totally having this problem. [14:07] !testers | beta 1 candidates testing [14:07] beta 1 candidates testing: Help is needed in #kubuntu-devel. Please ping Riddell, yofel, soee, Tm_T, shadeslayer, BluesKaj, James147, smartboyhw, Quintasan, lordievader, shrini, tester56, parad1se, mamarley for information [14:07] http://iso.qa.ubuntu.com/qatracker/milestones/312/builds tests needed [14:08] jussi: you did not throw shirt ads on kubuntu's g+ page it appears [14:10] ugh... libqapt has issues with saving for sure. [14:10] QQ [14:12] manchicken: the writing seems rather straight forward, are you sure that it's not the actual sourcelist management classes that screw up? [14:12] Digging deeper in my tests I'm finding some test cases I missed. [14:12] I've gotta go to work now, but I'll dig deeper tonight. [14:13] Riddell: patching did not work, tried copy and paste and I am only making things worse :( Do you mind taking a look at it? [14:13] I may need someone who knows how to better troubleshoot dbus stuff to help me with this, it looks like I'm having issues with that. [14:13] sgclark: yeah can do [14:13] It could be, but the fun part is that some places I see writing duplicates, other parts I see it writing nothing. [14:14] I'll have a better look at this when I'm not on five hours of sleep about to trudge into the office. [14:15] I'll catch you guys later. [14:15] thanks manchicken [14:15] appreciate all your effort [14:16] No problem, it's interesting :) [14:19] ScottK: apachelogger https://bugs.launchpad.net/ubuntu/+source/kde-style-skulpture/+bug/1285173 [14:19] Launchpad bug 1285173 in kde-style-skulpture (Ubuntu) "Please remove kde-style-skulpture from Ubuntu" [Undecided,Triaged] [14:19] sgclark: yeah can go [14:19] sgclark: yeah can do [14:19] * Riddell feels left out of shadeslayer's ping list [14:19] heh [14:19] Riddell: ^^ [14:20] * shadeslayer hugs Riddell [14:20] Riddell: will be ready for you shortly [14:20] shadeslayer: if you quote policies at least follow the policy :P [14:20] "Before taking any actions, at least one attempt should be made to assess whether upstream really is dead. The preferred way to do this is sending a mail to the biggest copyright holder or a respective mailing list. If no information is provided regarding the state of upstream within 14 days, the software is to be considered unmaintained and general viability rules apply. " [14:21] or propose a change to the policy :P [14:21] that step is there to cover our own behind, so we can say "yo, we did our best" when someone complains ^^ [14:22] oic [14:22] * shadeslayer will email author [14:22] apachelogger: Riddell https://jenkins.qa.ubuntu.com/search/?q=upgrade-kubuntu [14:23] yofel: ^^] [14:23] yay [14:23] shadeslayer: can we get that going for ppas? ^^ [14:23] apachelogger: see that page again [14:23] upgrade-kubuntu-precise-trusty-desktop-backports-i386 [14:24] I mean, ppa snapshot 1 -> ppa live; ppa snapshot 2 -> ppa live [14:24] alas, that might be too complicated for the present tech [14:24] ah nope, don't think so [14:24] Riddell: khtml is ready for you [14:25] Riddell: you mentioned 2 more packages yesterday? [14:26] shadeslayer: actually that might be unsuitable anyawy. ... how about ppa -> other ppa [14:26] namely give staging a greater purpose [14:27] apachelogger: possibly could be done, talk to jibel about that? [14:27] in #ubuntu-devel === Blizzzek is now known as Blizzz [14:28] awww, no, I am scared now :'< [14:28] actually, here's the bigger picture [14:29] someone may have suggest more vigorous application backports to LTS releases [14:30] to get that as cheap as possible and as scalable as possible one would need massive amounts of automation and in particular automated QA (biggest piece obviously being a) dependency resolution working and b) no file conflicts ever happening) [14:33] other than the QA it's actually a somewhat viable idea right now... you could maintain a whitelist of backport candidates, then you have an application compare versions between devel and lts and if there's a newer in devel it uses a dedicated ppa (i.e. one ppa per application) to attempt an automated backport, if the build succeeds a bunch of QA is run and if that succeeds the backport is moved to kubuntu-backports [14:36] sgclark: kactivities-kf5 I put the start of packaging in bzr [14:36] it compiles but everything else needs going over [14:36] sgclark: and plasma-framework is the other new one [14:37] Riddell: ok [15:47] ooh munich going with kolab http://www.theregister.co.uk/2014/02/26/munich_goes_kolab/ [15:48] jmux: does that mean we'll get kolab packages in kubuntu? [15:49] oh so they finally announced it [15:49] yay [15:50] Riddell: At least for Precise :) === jmux_ is now known as jmux [15:55] jmux: cool! [15:57] jmux: do you plan on making packages available publically ? [15:58] Ok - actually I'm not sure how everything will work out. I'm not involved in the project, as I'm not responsible for mail and web browser. [15:58] oh [15:59] Actually the company was faster to their announcement, then I got informed of the status of the tender. [16:02] Ss there are only two thing I'm actually sure of: we'll get Kolab infrastructure and it will run with KDE 4.12 Precise packages. [16:06] jmux: I was told about this 3 weeks ago :P [16:08] shadeslayer is quite the well informed person [16:08] shadeslayer: do you work for the NSA? [16:08] shadeslayer: Well the officiall (very good hidden) announcement was on 13.1.2014 (http://ausschreibungen-deutschland.de/149525_MigMak_2014_Muenchen) [16:10] Then we have to wait two weeks for judicial objections. [16:14] At the end there is a lot of administrative law involved. [16:16] Riddell: kactivities-kf5 ready, note that I had to patch to get it to build [16:17] apachelogger: you should know better than that... shadeslayer just hacked the NSA... :P [16:21] Riddell: plasma-framework states not a branch, I start from scratch or was it uploaded as something else? [16:23] sgclark: start from scratch [16:23] ok [16:25] what jussi said is true :P [16:25] <- master hacker [16:25] xD [16:48] lordievader: able to do any beta testing? [16:50] Riddell: I did yesterday some, but it seems the image was respun. [16:50] lordievader: a never ending task this game [16:50] respon? (past tense of respin anyhow) [16:51] Riddell: Maybe I have time tomorrow to do more testing. [16:58] pretty sure those are tested by jenkins ^^ [16:58] and can be marked as done [16:59] where's the jenkins test? [16:59] does it test the UI elements? [16:59] https://jenkins.qa.ubuntu.com/search/?q=upgrade-kubuntu [16:59] does it make a nice upgrade page for users? :) [17:00] does it check for a running desktop at the end? [17:00] mmm .. dunno [17:00] * Riddell reboots to upgrade test [17:00] * shadeslayer is running the lvm + encrypted disk test [17:27] apachelogger: do you have a page with langpack statistics [17:27] i.e. how much stuff is translated [17:57] apachelogger: pingggg [17:58] apachelogger: http://imgur.com/Ihw5ui2 < ENOCPU? [18:04] yeah, so, solid is still broken [18:04] hooray! [18:04] shadeslayer: and what sort of statistics are you looking for? [18:04] statistics of our stuff? [18:07] apachelogger: KDE stuff [18:07]