=== jono is now known as Guest14995 [05:32] mgraesslin, ping [05:33] tvoss: pong [05:33] mgraesslin, good morning (under the assumption you are in Germany) :) [05:33] yes :-) good morning [05:33] * tvoss loves the brain melting temperatures early in the morning right now [05:34] mgraesslin, so I was looking through the kwin source code and wonder if you could help me in getting a few questions answered [05:34] I can try [05:35] mgraesslin, great :) here we go: So for integrating with an EGL/GL platform, OpenGLBackend is the class to implement? Do I understand that correctly? [05:35] yes [05:35] but that's just for the OpenGL compositor obviously [05:36] the XRender one is currently getting adjusted [05:37] mgraesslin, yup :) so one question (out of curiosity): is there a reason you do not use the q platform abstraction layer to "hide" those GL bits? [05:37] when we started the OpenGL part, Qt's OpenGL was not up to where it is now [05:38] maybe that changed with Qt 5 - I haven't looked into it enough [05:38] but Qt 4 doesn't provide what we need [05:39] mgraesslin, ah okay, that makes perfect sense then [05:39] mgraesslin, so for interfacing with GL it's OpenGLBackend. I was wondering where you bridge over to the input subsystem. Or do you read directly off evdev? [05:40] quite simple: that code is not there yet [05:40] the egl_wayland_backend does some nasty hacks [05:40] but in general the code still needs to be written [05:41] mgraesslin, ah okay :) I would think the same holds true for app mgmt/focus/window placement strategies? [05:42] no that all exists, just only for the X11 backend [05:42] mgraesslin, ah, that is cool :) can you point me to the code? [05:42] mgraesslin, that is, for the existing x11 backend [05:42] well that's everything :-) [05:43] tvoss: best explained in http://community.kde.org/KWin/Hacking [05:43] mgraesslin, got it, thx [05:44] tvoss: at the moment I highly recommend to not try to integrate KWin with Mir - it's a bad point in time [05:44] we are about to switch to Qt 5 and KF 5 [05:44] so the code will break badly [05:44] and the code will change [05:44] we don't mind moving large code parts around [05:45] mgraesslin, hmm, that would be great as we have a working qpa implementation for Mir. Do you have some ballpark estimate for the qt5 switch? [05:46] when master opens up again - in a few weeks [05:46] we plan to have a development preview release this year [05:47] for Kubuntu the earliest release on Qt 5 can be 14.10, maybe 15.04 [05:47] 14.04 will come 4.11 just like 13.10 [05:48] mgraesslin, okay, thanks, that helps me. I might start playing around with the kwin code as is, though, totally being aware that I'm standing on shaky ground :) [05:49] mgraesslin, would you be the right guy to talk to for XMir, too? [05:49] tvoss: I have no interest in anything with Mir, it's a downstream problem from my point of view [05:49] and it's a downstream I'm not running, so I cannot help anything there [05:50] mgraesslin, okay, fair enough [05:50] tvoss: in fact I don't think you will find any upstream developer who is able to help there - no KDE workspace core developer is to my knowledge using Kubuntu [05:52] mgraesslin, okay, however, my questions were answered :) thanks for that [05:52] you're welcome [06:12] good morning [07:56] Damn it, okular still is crapped. [07:56] I REALLY HATE SYMBOLS... [08:05] Re-uploaded... [08:05] smartboyhw: make sure you supply pkgkde-symbolshelper with both the build logs [08:05] shadeslayer, ? [08:06] both = both i386 and amd64 [08:06] wait [08:06] smartboyhw: how are you fixing the symbols file? [08:06] by hand? [08:06] shadeslayer, hand yeah. [08:07] ... [08:07] shadeslayer, that's why I'm stupid:P [08:07] shadeslayer, btw how do you do it? [08:07] pkgkde-symbolshelper [08:07] * smartboyhw didn't know that existed:P [08:07] smartboyhw: please go through http://pkg-kde.alioth.debian.org/symbolfiles.html [08:08] shadeslayer, hmm interesting. [08:10] After okular, it would be kwallet. [08:10] The damning .install file error;P [08:14] On this pace, we can get 4.10.80 out when 4.10.81 releases:P [08:14] (Kidding) [08:15] not really kidding. I think we've been doing that for the last 2 releases ^^ [08:17] yofel, yeah!:) [08:20] * shadeslayer is trying to do everything as fast as possible [08:21] Currently existing problems are for korundum (code error), kwallet (I will fix it), rocs (cmake grantlee error), kajongg (weird build-dep error on python-twisted), kdegraphics-mobipocket (due to okular) [08:21] kdegraphics-stringi-analyzers failed (code again) [08:21] s/kdegraphics/kdenetwork/ [08:21] smartboyhw meant: "kdenetwork-stringi-analyzers failed (code again)" [08:21] well cmake rather [08:21] it's missing a check for boost [08:22] (and the package is missing the dep) [08:22] kdewebdev (.install files problem) [08:22] I think we can be done by this weekend [08:22] hopefully [08:22] klickety (again .install files problem) [08:23] Fortunately I can join in this mess at the right time:) [08:25] yofel: kajongg is still fcksd on i386 [08:25] Seems like kiten is ok now [08:26] someone else look into kajongg, that's some python issue in the archive [08:26] ^^ [08:27] * shadeslayer won't be available on the weekend [08:27] erm [08:27] s/on the weekend/in the evening/ [08:27] shadeslayer: You did something wrong... Try s/you/me/ or tell me "help sed" [08:27] help, I now have the exact error as the ORIGINAL okular i386 plasma-active-dev issue (what? the build in ninja is ok while here not) [08:27] pastebin them build logs [08:28] shadeslayer, erm it is just this... [08:28] your chroot does have saucy-proposed enabled? [08:28] yofel, erm no. [08:28] you'll need that ;) [08:28] 1. that's my actual computer:p [08:28] yofel, meh. [08:29] that's why we have chroots... [08:29] if you don't want to use pbuilder a simple debootstrap chroot works too [08:29] or a VM, but that's overkill [08:29] ^^ [08:29] lxc [08:29] ftw [08:29] true ^^ [08:30] Heh [08:35] huh odd [08:35] there is a applets/kdeobservatory/src/ [08:35] however [08:35] there is no kdeobservatory widget that's built [08:51] oh [08:51] hmm [08:51] Qwt5-Qt4_FOUND [09:02] shadeslayer, I'm supposed to include http://paste.kde.org/777638/ in symbols too? [09:04] that kind of looks odd but yes [09:04] just use pkgkde-symbolshelper [09:09] Hmm, it really takes a long time to set up pbuilder i386... [09:12] Maybe I should upload this thing first into Ninjas and see how symbols fail in i386.... [09:13] Uploaded ~ppa5. [09:13] Hopefully I don't need a ~ppa10. [09:18] that's what I usually do ^^ [09:19] yofel, yeah. I hope not this time..... [09:19] and ppa10 is fine, with the build priority ninjas has you'll still be done today [09:19] yofel, yeah, that's the important point:P [09:19] Thank you private PPAs:P [09:24] "To get the mighty Kubuntu pbuilder-hooks (written by mighty Harald Sitter aka apachelogger), run:" [09:24] lol [09:25] apachelogger, :O [09:25] * smartboyhw feels terror terror terror terror terror [09:28] yofel, shadeslayer dh_sameversiondep: cannot continue because the reference package libokularcore2abi1 could not be found in debian/control or dpkg status [09:28] The strange error [09:28] Since the symbols have been solved now it seems. [09:28] look through the control file [09:29] there's some sameversiondep definition in a Depends: field somewhere [09:29] yofel, ok. [09:29] yofel, yeah got it. [09:29] * smartboyhw builds ~ppa6. [09:33] Uploaded. [09:33] Hope that this build finally works. === yofel changed the topic of #kubuntu-devel to: Kubuntu - Friendly Computing | https://trello.com/kubuntu | https://notes.kde.org/p/kubuntu-ninjas 4.10.80 in progress | 4.10.4 SRU yofel WIP [09:38] ErrorMessage: não pode copiar dados extráidos para './usr/lib/kde4/libexec/ksendbugmail' para '/usr/lib/kde4/libexec/ksendbugmail.dpkg-new': fim de ficheiro ou stream inesperado [09:38] if only I knew what it means [09:39] yofel: good to see you SRUd finally :P [09:39] I would guess it failed to write a file [09:39] yofel: /opt/project-neon5/share/pkg-project-neon5/0/default-settings.mk:3: *** missing target pattern. Stop. [09:39] yofel: file stream got interrupted unexpected I suppose [09:39] yofel, \o/ [09:39] We can get 4.10.4! [09:40] locale on terminal is le fufufufufufufufufu [09:40] apachelogger: why are you using default-settings directly? Either use project-neon5.mk, or write your own %: target [09:40] yofel: include /opt/project-neon5/share/pkg-project-neon5/0/project-neon5.mk [09:40] see lp:~neon/project-neon5/qtbase [09:41] brrrrr [09:41] looking [09:41] yofel: don't seem to find default-settings [09:42] ah, I am just bad at copynpaste [09:42] yay kde-workspace compiled [09:42] Riddell, yep: [09:42] :) [09:43] yofel: NEON_PROJECT_NAME: := project-neon5 [09:43] apachelogger: hm? [09:43] excess : [09:43] eeeek [09:44] fixed [09:44] works without the colon it seems \o/ [09:45] :=) [09:45] :P [09:46] hi Oli [09:46] apachelogger: I triggered a build and need to be off for lunch, bbiab [09:47] xcb.cpp:46:27: fatal error: xcb/xcb_icccm.h: No such file or directory [09:47] :( [09:48] !? [09:48] :O [09:49] Yes Okular BUILDS!!!!!!!!!!!! [09:49] * smartboyhw goes off cheering:P [09:49] still publishing but at least it builds without issues. [09:53] !find randr.h [09:53] File randr.h found in e17-dev, libghc-x11-dev, libghc-x11-doc, libgnome-desktop-3-dev, libgnome-desktop-dev, libxcb-randr0-dev, libxrandr-dev, x11proto-randr-dev, xserver-xorg [10:02] Okular successfully uploaded. [10:23] kwallet done, now kdepim-runtime. [10:23] you're on a roll [10:24] Riddell: maybe even send a email about analitza ABI breakage [10:24] Riddell: commit 3c1f0b4ea784eeed754a68e611a1f6f3aa4f9cac fixes it [10:25] Riddell, heh. [10:26] shadeslayer: groovy [10:32] Riddell, shadeslayer I saw kgoldrunner in the kubuntu-ninjas notes, but I can't see that package in our PPA. Did it disappear or just it wasn't built? [10:33] it wasn't uploaded [10:33] yofel, for what reasons? [10:33] as that was the first batch, bzr-buildpackage-ppa failed [10:33] for whatever reason [10:34] (usually it's patches) [10:34] Grrrrrrr :P [10:49] JontheEchidna: https://bugs.launchpad.net/ubuntu/+source/kde4libs/+bug/1183615 you totally should close this bug because I forgot again when a file stream error occurs :P [10:49] Launchpad bug 1183615 in kde4libs (Ubuntu) "package kdelibs5-plugins (not installed) failed to install/upgrade: não pode copiar dados extráidos para './usr/lib/kde4/libexec/ksendbugmail' para '/usr/lib/kde4/libexec/ksendbugmail.dpkg-new': fim de ficheiro ou stream inesperado" [Undecided,New] [11:03] kdepim-runtime uploaded! [11:16] Hey all [11:16] smartboyhw: awooga! [11:17] Riddell, :P === mydogsnameisrudy is now known as monkeyjuice [11:21] Riddell, any thoughts on http://paste.kde.org/777746/ ? [11:27] yofel: I can build source package \o/ [11:27] \o/ [11:28] \o/ [11:28] yofel, apachelogger thoughts on http://paste.kde.org/777746/ ? [11:29] not sure, but smokegen is broken too, fix that first [11:29] considering there's _SMOKE stuff missing that might be the reason [11:30] yeah, fixing smoke before fixing stuff that uses smoke is a good idea generally :P [11:30] ah hm, smokegen is only that stupid lib symlink [11:31] gawd [11:31] kmix is so broken it's just not funny anymore [11:32] oh fun, kdenetwork-strigi-analyzers has 0 copyright files [11:32] of course [11:32] why am I not surprised... [11:32] they moved to git [11:32] the whole audio setup is a mess ...a nice comprehensive audio suite would be nice [11:33] Eh:P [11:33] and after some 7 modules having moved to git still no one remembers to copy the flipping copying files [11:33] BluesKaj: eh? [11:33] anyway, need to fix cmake first [11:33] yofel, wait a minute: Didn't smokegen compiled successfully? [11:34] apachelogger, kmix is just a small part of the problem [11:34] smartboyhw: yeah, it's just list-missing [11:34] * apachelogger gets a headache [11:34] BluesKaj: it is? [11:34] yofel, ah. [11:34] try to retry what failed [11:34] or maybe you need to sequentially retry smoke too, if something was chained in the chain [11:34] yofel: someone should revise them qtbase build deps, I have a feeling half of those are not needed :O [11:34] *changed [11:34] aw c'mon apachelogger , you know what I mean [11:35] BluesKaj: I really don't [11:35] apachelogger: does this look sensible? http://paste.kde.org/777758 [11:35] OK yofel I will fix smokegen. [11:35] smartboyhw: just add the dropped not-installed file back [11:35] yofel, OK. [11:35] Then? [11:35] yofel: no, see strigi above? :P [11:35] Just that? [11:36] apachelogger: I copied that from kdepim... [11:36] apachelogger, I'm oin no mood to rant about alsa's negelect , pulseaudio , pavucontrol , gstreamer etc tetc [11:36] then kdepim is silly too [11:36] yofel, ? [11:36] yofel: that command will fail immediately if it is not found [11:36] yofel: macro_optional_find_package OTOH will only fail at the very end [11:37] ok, so macro_optional_find_package(Boost REQUIRED) ? [11:37] or how does that work? [11:37] macro_log_feature(STRIGI_FOUND "Strigi" "Metadata extraction" "" FALSE "" "Needed to get metadata for .torrent files") <- you'll simply want a TRUE in the middle there, marking it as required [11:37] aaaah [11:37] ok, next try [11:37] BluesKaj: then don't? :P [11:37] \I won't [11:38] from where I am standing there is a comprehensive audio suite which is phonon which is based on pulseaudio :P [11:39] prefer to run without pulse , but if i want web audio then I need it , alsa used to do the job nicely until pulse got shoved down our throats [11:40] -rw-r--r-- 1 root root 43M Jun 19 13:38 project-neon5-qtbase_0.0+git20130619-9~saucy0.tar.xz [11:40] ooomph [11:40] that could take a while [11:40] BluesKaj: it didn't [11:40] apachelogger: paste.kde.org/777764 [11:41] the amarok forum actually has like 30000 cases of proof how alsa did not work [11:41] and the phonon bugzilla has about 10 ongoing reasons of why alsa still does not get the job done nicely [11:41] yofel: looking good [11:41] yofel, can't understand: What not-installed file? [11:41] thanks [11:42] odd, I didn't have any problems with alsa ,and of course I don't use amarok [11:42] Ah found it now. [11:43] BluesKaj: yeah, that's neither objective nor solid evidence of anything, so I'd very much prefer if you didn't pull the multimedia stack down [11:43] as a matter of fact it looks the same on every platform that has been around for more than 5 years [11:44] on windows 7 you have some six supported means to play audio [11:45] apachelogger, like I said earlier it would be nice iif there was an audio suite that just worked out of the box , a comprehensive integrated one [11:45] yofel: I am uploading to ppa:neon/kf5, we'll see what comes of it [11:45] though I doubt it will build [11:45] yay [11:45] or at least master did not build with gcc 4.8 [11:45] it's there, that's progress :P [11:45] yofel, so I just add back the not-installed and upload it again? ..... [11:45] though I got told dev should [11:45] so [11:45] * smartboyhw does not understand what he should do now.... [11:45] maybe we are lucky ^^ [11:46] smartboyhw: yeah, that'll be enough [11:46] yofel, don't tell me that can erase the red light on the page:O [11:46] smartboyhw: it's just red because list-missing isn't empty [11:46] BluesKaj: yeah, we have that. [11:46] yofel, ah alright. [11:46] and not-installed will hide that file [11:46] if you choose to not use the solutions we provide then that's not the solutions' problem TBH [11:48] Uploaded. [11:48] apachelogger, it's difficult for new users to understand the audio setup , it's rather complicated to them and discouraging at times ..as it is for us trying to help [11:55] how so? you order devices by priority, there is nothing more one needs to do (or should need to anyway) [11:57] what ? [11:59] anyway , bbl [12:00] BluesKaj: http://wstaw.org/m/2013/06/19/plasma-desktopdI2287.png [12:01] https://launchpad.net/~neon/+archive/kf5/+build/4727447 [12:01] meh, one time I try to fix something in git and the server craps out [12:01] oh, yeah git servers are migrating right now [12:01] yeah, I just read the horror story in -sysadmin ^^ [12:02] oh wow, qtbase compiles \o/ [12:02] but for how long ^^ [12:03] that's already more than the first editions of project-neon-qt did :P [12:03] lol [12:03] shadeslayer: I may needs a server soonishy === greyback is now known as greyback|lunch [12:21] yofel: https://launchpadlibrarian.net/142832008/buildlog_ubuntu-saucy-i386.project-neon5-qtbase_0.0%2Bgit20130619-9~saucy0_FAILEDTOBUILD.txt.gz [12:21] that's the gcc problem [12:22] no clue what to do :( [12:26] or maybe I am still on master [12:26] git shallow clones are weird [12:26] * apachelogger reclones [12:29] smartboyhw: sorry missed it, how did you get on with your link error? [12:30] smartboyhw: looks like it's missing a -lsmoke_akonadi somewhere [12:31] shadeslayer: I still need kubuntu_krfb_link_to_pthread.diff in krfb [12:47] Riddell, OK. [12:47] smartboyhw: complain upstream is probably the best thing, rdale would be the chap [12:48] Riddell, heh:P [12:48] he seems to be on irc [12:49] Riddell, pinged him. [12:50] Wow, now only 4 packages build-fail!:O [12:51] Am I wrong or analitza has some --list-missing problems? [12:51] Uh, it's the symbols:( [12:51] * smartboyhw does not want to touch symbols today ... [12:52] But someone has to be the guinea pig, let me be it:P [12:52] Laters though, I need to take dinner. [13:16] hmm, I can't run kppp cos I'm not in dialout [13:18] i see only 3 packages failed to build for saucy? [13:20] soee, yep. But Riddell you haven't yet split all packages right? [13:21] just finishing kdenetwork [13:22] Riddell, great:) [13:22] Riddell, what's the response of korundum? [13:23] smartboyhw: did you see those two patches? [13:23] Riddell, no didn't. Now working on analitza. [13:25] libkwinglutils.so.1abi2 libkwinglutils1abi1 #MINVER# [13:25] * yofel thought he already fixed that o.O === greyback|lunch is now known as greyback [13:27] yofel: uploading new qtbase [13:27] this time apparently really from dev branch ^^ [13:30] Riddell, you going to fix korundum or me? [13:30] smartboyhw: I'm still on splits [13:30] Riddell, OK I shall do it then. [13:43] Guys, we have prob with krfb https://i142838148.restricted.launchpadlibrarian.net/142838148/buildlog_ubuntu-saucy-i386.krfb_4%3A4.10.80-0ubuntu1~ubuntu13.10~ppa2_FAILEDTOBUILD.txt.gz?token=87b02fb2551e0decb4c4509968692674 (cmake error?) [13:44] analitza uploaded BTW. [13:44] smartboyhw: thanks, build deps need fixing, will look [13:45] And can someone tell me how adding not-installed to smokegen works, I added it and it still is in red light... [13:47] Hmm sounds like an old problem came back on analitza..... [13:47] libsmokekate.......... [13:48] !? [13:50] what does analitza have to do with smoke? [13:51] yofel, sorry korundum:P [13:51] And actually I just found that korundum failed to build in armhf for 4.10.4 ..... [13:52] smartboyhw: and it seems you didn't add the patch to analitza [13:52] yofel, you mean in bzr or? [13:53] smartboyhw: well, I don't see it in bzr [13:53] yofel, wait what patch? [13:53] I only changed symbols. [13:53] Patch was for korundum... [13:53] smartboyhw: you didn't read the pad... [13:53] and your symbol changes are something which isn't allowed [13:54] yofel, ah? (for second sentence) [13:54] - _ZNK8Analitza12PlotsFactory11requestPlotERKNS_10ExpressionENS_9DimensionE@Base 4:4.9.80 [13:54] - _ZNK8Analitza11PlotBuilder6createERK6QColorRK7QStringPNS_9VariablesE@Base 4:4.9.80 [13:54] those are missing PUBLIC symbols [13:54] that must never happen without an SOVERSION change [13:55] * smartboyhw is too confused with symbols.... [13:55] the soversion change did happen didn't it? [13:55] Let me add patch first .... [13:55] that's why you need to cherry-pick 3c1f0b4ea784eeed754a68e611a1f6f3aa4f9cac [13:55] Riddell: right ^ [13:56] Sorry guys... [13:56] smartboyhw: well, we helped you this time around because this is a complicated matter. But you really need to investigate every single MISSING symbol [13:58] smartboyhw: reason: if an application uses the library and uses those methods, it will crash with a symbol lookup error unless it's at least rebuilt (or fixed if the API changed too) [13:59] yofel, OK. [14:00] Re-uploaded. [14:00] smartboyhw: btw, this is a good read on the topic: http://www.netfort.gr.jp/~dancer/column/libpkg-guide/libpkg-guide.html [14:00] you should've at least read the library packaging part of the debian policy though [14:00] Whoas.... [14:02] Anyways, analitza done, now can someone tell me why that smokekate returned to haunt korundum!?:O [14:04] Tm_T: you admin kopete-devel mailing list? could you let through my posts? [14:04] * smartboyhw cries.... [14:06] smartboyhw: btw. if you add header-less patches, please at least mention the patch filename and commit hash that was used if you add upstream patches. [14:06] We don't require DEP-3 or anything like that, but at least this is good practice [14:06] yofel, ok...... [14:07] So I have to re-upload again? [14:07] ... [14:07] * smartboyhw is already getting smoked of that smokekate thing... [14:07] where's that anyway? I only see akonadi stuff in kroundum [14:07] l [14:08] yofel, it just shows up. [14:08] http://paste.kde.org/777926/ [14:09] what o.O? [14:09] yofel, and there IS a kate folder in modules... [14:09] yeah, but we don't build the kate bindings... [14:09] too unstable I think [14:10] Grrrrr......... [14:10] It haunted us after 3 cycles:O [14:10] -- Skip SmokeKDE bindings: Kate [14:10] I agree.... [14:11] as for analitza: yes, you'll have to upload again [14:11] but wait, the package build will fail anyway [14:11] We aren't going to ship smokeKDE right? [14:12] huh? ofc. we ship it [14:12] Ah yep. [14:12] yofel, I mean kate. [14:12] smokekate. [14:12] * smartboyhw goes to fix analitza again:) [14:12] no, that not, we ship: [14:12] -- Build SmokeKDE bindings: Akonadi;Attica;KDECore;KDEUi;KFile;KHTML;KIO;KNewStuff2;KNewStuff3;KParts;KTextEditor;KUtils;Nepomuk;Okular;Plasma;Solid;Soprano [14:13] yofel, so I have to patch korundum to not make kate? [14:13] !? [14:14] meh, let me try to build it myself. Otherwise I'm just guessing [14:20] * Riddell cheers as he gets kopete working with no missing symbols [14:22] smartboyhw: I don't get your kate error here [14:22] yofel, :O [14:22] maybe you have some kate headers lying around? [14:22] (the cmake part for kate is weird...) [14:22] yofel, check your modules folder... [14:22] but the akonadi issue is ${SMOKE_AKONADI_LIBRARY} being empt... [14:22] :O [14:23] smartboyhw: modules folder? [14:23] I know that kate is there, but [14:23] -- Skip KorundumModules bindings: Kate [14:23] so I'm fine [14:23] yofel, hmm it doesn't skip for me weird. [14:23]