[01:31] Anyone packaging up mgraesslin's kwin fix for 4.7? (see packagers) [02:30] any idea what caused this build failure? https://launchpadlibrarian.net/79514260/buildlog_ubuntu-lucid-ia64.qt4-x11_4%3A4.6.2-0ubuntu5.3_FAILEDTOBUILD.txt.gz [02:37] * micahg dreads the idea of respinning qt4-x11 for ia64 [03:28] micahg: No. I'd hit retry and see if it magically works. [03:34] ScottK: ok, it's only 12 hours, thanks [10:25] ScottK: quassel-core creates world-readable directories :( [10:29] and the generated ssl cert+key file is world-readable [10:29] also the logs [10:29] ScottK: urgh, some idiot broke phone line here and I dont have internet === Quintasan_ is now known as Quintasan [10:30] Quintasan: that’s always a nice feature :) [10:31] bulldog98: Care to elaborate how having no internet while runtime is apparently broken nice? [10:32] Quintasan: it was meant ironic [10:33] * bulldog98 handles cookies and tea to Quintasan [10:33] Oh, I see [10:33] Well, now the damn PC wont boot [10:33] ffs [10:33] * Quintasan need to take the pc to the shop to get it repaired or replaced [10:34] Quintasan: have you tried to recover grub? [10:35] bulldog98: It's not GRUB related [10:35] If it was GRUB then I would have fixed it by now [10:35] it's faulty hardware [10:35] Quintasan: that sucks [10:36] Quintasan: have you put new RAM into it? that was my problem at friday [10:36] I believe it's a shitty power supply but I dont have a spare one confirm it [10:37] bulldog98: I did but the problem was occuring even before I added memory [10:37] when I shutdown the PC it refuses to boot when I turn it on immediately after [10:37] it powers on, fans spin, hdd starts working but nothing happens [10:38] when I reboot it from os it reboots normally [10:38] Quintasan: have you placed the RAM tight enought, my PC refused to boot, cause I hadn’t put it in hard enought [10:38] thats not the case [10:38] hm [10:38] Im telling you I had this even before putting new memory [10:39] strange [10:42] i know. that's why Im going to shop tomorrow and tell him to fix at once or rma the parts [10:42] I dont care which part is at fault [10:49] apachelogger, ScottK: the quasselcore-makecert is currently broken (tries to write into the wrong dir). do you want to keep it? I think we shouldn't encourage creating self-signed ssl certs ... [10:50] on the other hand quassel doesn't seem to be able to verify certificates :/ [10:50] debfx: I think any cert is better than no cert at all [10:53] bulldog98: yes but that's hardly a justification for telling people to create self-signed certs [11:05] debfx: the point of the cert is not for it to be verified but for it to encrypt communication from core to client [11:05] much like an ssh cert [11:05] s/cert/key [11:05] kubotu: order coffee [11:05] * kubotu slides coffee with milk down the bar to apachelogger. [11:06] kubotu: order tea [11:06] * kubotu gives bulldog98 a nice hot cup of tea. [11:06] kubotu: order a plate of eggs in the basket [11:06] * kubotu slides a plate of eggs in the basket down the bar to apachelogger [11:07] kubotu: order pork with potatoes [11:07] * kubotu slides pork with potatoes down the bar to bulldog98 [11:07] eh? if you don't verify the cert somehow ssl is mostly useless [11:07] debfx: nonono, you verify the cert [11:08] debfx: I think the guys, that install quassel-core know how to check the fingerprint of an ssc cert [11:08] much like an ssh key [11:08] you can scp the cert to your pc and import it as knowen cert [11:08] once you have established that your server has a given cert you get encrypted connection plus man-in-the-middle would trigger a cert warning or something [11:10] sure, but if quassel had a proper certificate you wouldn't need to verify it at all [11:10] *verify manually [11:11] bulldog98: yes, but not all users install their own quassel core [11:11] debfx: then the guy that installed it has to send them a gpg encryped mail containing the ssl cert [11:12] on the other hand gaining a singed cert is a bunch of work [11:12] with no advantage [11:13] as one can see in the nice problems of ssl certs [11:13] namly google.com [11:16] bulldog98: quassel is supposed to accept certs signed by trusted CAs so that won't help you at all [11:17] debfx: one point for you :) [11:17] the advantage is of course that you don't need to verify the cert manually [11:18] still it’s much overhead getting a ssl cert singed and most times you have to pay for that. (If you only have 6 users, you can tell them: that’s the fingerprint, import it and everything will work fine) [11:21] it takes about 30 seconds to create a cert signed by startssl and doesn't cost you anything [11:24] anyway that makecert script is completely broken [11:25] and has been since at least lucid [11:26] debfx: maybe we should just print some instructions howto create a singed cert? [11:26] apachelogger: bug 846027 is a bug which I said you last time here [11:26] Launchpad bug 846027 in kde-workspace (Ubuntu) "strange behavior of second panel" [Undecided,New] https://launchpad.net/bugs/846027 [11:30] ari-tczew: quite honestly I believe this to be a graphics driver issue [11:30] I know that JontheEchidna is also running a 2 panel setup, and he does not have this problem [11:30] anyhow [11:30] bulldog98: imho the ssl cert business is something server admins should know about (or be able to google), so creating an inital certificate on package install is good enough [11:31] ari-tczew: you should move the bug to bugs.kde.org, it does not sound particularly kubuntu specific IMHO [11:32] apachelogger: yofel and me also have 2 panels [11:32] apachelogger: look at this my report https://bugs.kde.org/show_bug.cgi?id=280739 [11:32] KDE bug 280739 in general "plasma crashes when I have second panel" [Crash,Unconfirmed] [11:32] plasma_applet_message_indicator [11:33] that BT suggests that it is not a bug in plasma but the message indicator [11:33] agateau_ ^^^^^ [11:35] ari-tczew: does it still crash with the latest message-indicator package? [11:35] the backtrace looks a lot like bug #631794 [11:35] Launchpad bug 631794 in plasma-widget-message-indicator (Ubuntu Oneiric) "widget crashes when starting Thunderbird or Evolution" [High,Fix released] https://launchpad.net/bugs/631794 [11:36] debfx: hmm not sure, because now I'm ingoring all crashes due to out of time. I'll make sure on next time boot. [11:36] debfx: I use evolution, so maybe this is the bug which you gave number [11:37] bulldog98, yofel_, JontheEchidna: what's your graphic card? [11:37] ari-tczew: nvidia [11:37] bulldog98: my too ... [11:37] ari-tczew: but I also had that running on intel [11:37] ari-tczew: nouveau or nvidia driver? [11:38] bulldog98: nvidia driver [11:38] * bulldog98 too [11:38] hm strange [11:39] ari-tczew: where are the 2 panels? At the bottom and the top? [11:39] hm I’ve got one left and one right [11:39] and that works [11:39] that is vastly different though :P [11:40] also mind that a dual screen setup could be messing with it [11:40] the plasma is not particularly dual screen friendly [11:41] apachelogger: hmm, I'll plug-off TV [11:41] bulldog98: yes bottom and top (up and down) [11:42] yofel and me don’t use that setting === yofel_ is now known as yofel [11:43] ari-tczew, bulldog98: I actually saw that bug yesterday after finally rebooting in a while [11:43] apachelogger: I thought aseigo was going to fix that [11:43] I don't get a crash though, just that one panel isn't where it's supposed to be [11:43] yofel: I haven’t got that with the packages of 8th [11:43] *shrug* [11:43] * apachelogger stopped caring about plasma 5 releases back [11:44] but I only have one monitor on my screen [11:44] bulldog98: not sure if I had all updates installed, I don't have much bandwidth here [11:44] sheytan: dude, I like need a quick n dirty mockup for muon on a mobile phone [11:44] will try later [11:45] ari-tczew: maybe you should talk to the guys in #plasma they know best how to handle with stuff like that [11:45] yofel: did you see wrong panel displaying like in my bug? [11:46] ari-tczew: yep, like an additional invisible panel beside it [11:47] nice! I'm not alone with that bug [11:47] yofel: what's your graphic card? [11:48] nvidia [11:48] ari-tczew: btw why don’t you have the panel completly at top, you can move the crashew to change that [11:48] bulldog98: *that* is the bug [11:48] yofel: ah ok [11:49] then I haven’t get the numbering of the stuff [11:49] yofel: me too nvidia [11:49] I set mine to the left, but when I login again it's a bit away from the screen edge like there's an invisible panel now [11:49] once you adjust the screen edge it moves back where it's supposed to be [11:49] yofel: and I hadn’t that when I logged in this moring [11:49] ok, installing updates again, will take a while [11:50] ari-tczew: do you have latest packages installed? [11:50] yofel: could you comment on my bug reported? [11:50] bulldog98: which package? [11:50] ari-tczew: all? [11:51] ari-tczew: later once I know if the updates help or not [11:51] there was an x update yesterday or the day before yesterday, maybe that fixed it? [11:53] bulldog98: x* packages are up-todated, kdegraphics has got something to update (new upstream release 4.7.1), I'll update later my system because I need to go out in a while [11:54] hm [12:01] ScottK: workspace compiles, doesn’t it? [12:05] * bulldog98 takes over fixing that [12:13] yofel: it could also be a qt regression [12:13] since I don’t have 4.7.4 installed by now [12:14] seems I have [13:09] shadeslayer: any news about 2.4.3 packages for Natty? [13:10] Also there still is no announcement for Oneiric and the packages are available since quite some time [13:20] apachelogger: does that looks like a bug fix for you? http://paste.ubuntu.com/686855/ [13:21] where did you get that from? [13:21] it looks rather bogus [13:22] the length has nothing to do with the width really... one could wordwrap and whatnot [13:23] apachelogger: I wrote it myself [13:24] what are you trying to solve? [13:24] apachelogger: problem is German translation is cut of [13:24] in shutdown dialog [13:24] screenshot [13:25] also what class does KSMPushButton derive from? [13:25] apachelogger: can you do a screenshot, while logging out? [13:25] ksnapshot with time set to 3 seconds or so [13:25] trigger logout dialog [13:25] wait for ksnapshot to take a shot [13:25] cancle logout [13:25] apachelogger: that guy again with his broken Kubuntu: http://forum.kde.org/viewtopic.php?f=20&t=96851&sid=e9737065e8f68f47c773187d9094b7f2&p=203673#p203614 [13:26] I moved that part of the thread as it has no relation to Amarok whatsoever [13:26] also, what in the word "revoked" didn't he understand? [13:26] he did not do what I told him [13:27] namely wipe all of qt apps and reinstall kubuntu-desktop [13:27] since he does not do what I tell him to, I really cannot help him [13:27] that guy is rather confused I might say... but, but! someone should make him a minion [13:28] I feel he would make a great minion, he certainly doesnt give up easily [13:28] apachelogger: can I batpaste screenshots? [13:29] no [13:29] there is a plasmoid [13:29] you can drag right form ksnapshot [13:30] apachelogger: pastebin? [13:31] *shrug* [13:31] some paste [13:31] there are of course more paste plasmoids than clock plasmoids [13:31] of course just like with the clock plasmoids they all seem rather pointless to me [13:31] *shrug* [13:33] apachelogger: btw did you manage to get the unicorn in the analog clock to vomit? [13:33] no [13:42] JontheEchidna: http://i.imgur.com/ZwntD.png === tazz_ is now known as tazz [14:04] hey all, is there a way to do an apt-file search on the kubuntu backports PPA ? [14:05] in particular kwineffects.h (and all other related kwin headers) which were in kdebase-workspace-dev appear to have vanished [14:05] sure [14:05] (they're back in project-neon though) [14:05] add the the ppas deb-src line to your sources.list :P [14:06] it's there. [14:06] "deb-src http://ppa.launchpad.net/kubuntu-ppa/backports/ubuntu/ natty main " [14:07] and on apt-file update I get: [14:07] http://ppa.launchpad.net/kubuntu-ppa/backports/ubuntu/dists/natty/Contents-amd64.gz [14:07] Ignoring source without Contents File: [14:08] on a quick check there doesn't appear to be a Contents.gz file on any ppa. [14:14] maybe indeed it does not work [14:14] #launchpad will know alllll the details [14:14] also they are at fault if it does not work :P [14:20] evening [14:48] great, we throw away all upstream quassel desktop file translations [14:49] apachelogger: evaluating if we should get rid of that desktop file stripping would be a good uds session [14:50] please add [14:51] I was thinking about it [14:51] but [14:51] we are having l10n sessions for decades and nothing ever changes [14:52] debfx: actually the stripping has advantages... like when you edit a desktop file using menuedit it will be copied to your home, meaning you'll not get translation updates and stuff [14:58] apachelogger: imho that's a small advantage considering how hacky and broken it is [14:59] just saying [14:59] * bulldog98 is in favour of that [15:01] btw if up update an iso, will the installation (not done yet) also include the latest packages? [15:01] s/iso/live system/ [15:01] bulldog98 meant: "btw if up update an live system, will the installation (not done yet) also include the latest packages?" [15:05] apachelogger: sexy [15:06] oioi, I am a married man [15:12] lol, was talking about your screenshot [15:12] but grats on the marriage [15:13] apachelogger: http://imgur.com/V6uHQ [15:14] JontheEchidna: thx ^^ [15:15] bulldog98: this is a non-trivial problem and the patch you presented only works around it [15:16] apachelogger: that’s exactly what it was supposed to do :) [15:16] apachelogger: how to fix that properly? [15:16] bulldog98: fix the ksmpushbutton [15:17] bulldog98: I suppose that the ksmpushbutton is a regular qpushbutton with a pixmap [15:17] something seems to go wrong with the layouting of the text and the pixmap [15:17] so first you'll have to find the logic for layouting [15:18] which probably either is a qlayout or manual layout calculation (probably in the paintEvent()) [15:18] * apachelogger points out that this dialog might actually be best implemented in qml + plasma [15:18] apachelogger: you are married? to who? (or is it just phonon :P?) [15:18] as it is themed to the point of not looking like qwidgets anyway [15:19] you could achieve the very same result by using qml and at the same time resolve the layout problem [15:19] as that button would be a Rectangle { MouseArea{} Text {} Image {} } [15:19] in case you want to actually redo the thing properly ;) [15:19] apachelogger: point me to a good qml tutorial and I maybe able to do that [15:20] apachelogger: but for 4.7.x I suggest that workaround [15:20] no, the workaround is bogus and might fail for various languages [15:20] the layout code needs to be fixed for 4.7u [15:20] as for qml tutorials -> qt documentation on qt quick [15:20] bulldog98: p.drawText(10, 0, width() - (m_smallButton ? 16 : 32) - 8, height(), [15:20] Qt::AlignVCenter | Qt::AlignLeft | Qt::TextWordWrap | Qt::TextShowMnemonic, m_text); [15:21] in KSMPushButton::paintEvent( [15:21] so it does in fact calculate layouting by itself [15:21] and that is probably where it goes wrong [15:21] or the sizehint is just wrong ^^ [15:21] who'd know ^^ [15:22] apachelogger: I guess it gets width() only form the "&Cancel" and not from the translated stuff [15:22] apachelogger: file? [15:22] also se ::init() [15:22] see even [15:22] there it does some calculation on minimumsize [15:23] supposedly the calculation is wrong [15:23] though it is using qfontmetrics, so technically it should be good [15:23] *should* [15:24] you could probably just add some 4px additional padding to the final minimum width as workaround [15:24] though finding out why the width calculation is off by a couple of pixels would of course be better [15:25] bulldog98: FWIW... the size calculation is done on m_text, which is the translated text as per i18n() [15:38] apachelogger: as I read the code there should be a space of 4pix between text and pix [15:39] probably [15:39] see, in qml that would again be very simple :P [15:39] Image { anchors.left: text.right; anchors.leftMargin: 4; } [15:41] apachelogger: what do I have to do to convert that stuff to qml? [15:42] where do you have your dragon qml, I want to learn form that [15:42] kde:scratch/sitter/dragon [15:43] I am not sure it is something you would want to learn from for starters :P [15:49] JontheEchidna: kde:scratch/sitter/muon-mobile [15:53] bulldog98: It did compile, but I think we definitely want to add mgraesslin's kwin fix that he mailed packagers about. [15:54] ScottK: that’s what I’m up to do [15:54] Great. [15:57] ScottK: when I finished it should I request a merge for it? [15:58] Yes. You might also check the kde/4.7 branch in KDE's git for anything else that is worth a cherrypick. We'll release with 4.7.1, so we'll only get further bug fixes we manually cherrypick. [16:03] apachelogger: bash: git://anongit.kde.org/scratch/sitter/muon-mobile.git: No such file or directory [16:04] its broken [16:04] omg [16:04] *shrug* [16:04] JontheEchidna: maybe anongit takes a bit [16:04] if nobody works on it, I propose to backport the kwin's patch into kde-workspace-4.7.1 (see kde-packagers) [16:06] [lp:~rekonq/rekonq/rekonq-ubuntu] Jonathan Kolberg * 5 * (11 files in 3 dirs) Update for rekonq daily build [16:08] bambee: bulldog98 is working on that [16:08] ok [16:08] :) [16:08] np [16:09] bambee: I’m currently doing a test build [16:09] awesome! [16:09] should I push it to ninjas then, so we can test that [16:10] is the tarball still in ninjas? So can I upload only the other stuff? [16:10] you could even blog about that, it's an important fix... [16:12] bambee: problem is I’m not in planet ubuntu [16:12] oh , right :\ [16:13] bambee: but maybe that will change on tuesday [16:13] :) [16:17] bulldog98: Not ninjas. There's nothing private about it. Ninjas should only be used for pre-release tarballs that aren't public yet. [16:17] ScottK: so experimental? [16:17] Sounds reasonable. [16:19] apachelogger: do I have to wrap all the stuff ksmserver can do into QObjects to do qml? [16:19] depends on how you do it [16:19] but generally that would be an approach [16:19] apachelogger: does stuff like i18n("&Cancel") works in qml? [16:20] no [16:20] you'll need magic for that [16:21] apachelogger: then we can’t use qml for ksmserver, since we would drop features [16:21] I did not say you cannot do it [16:21] I said you need to pull magic to make it possible [16:21] that said, the plasma foo might actually have something already [16:22] apachelogger: but kdeclarative will break as soon as Qt 5.0 is out [16:23] wah? [16:23] what you be talking about? [16:27] apachelogger: they said it at ds, that libkdeclarative (the plasma foo) will only work until 4.8 later it won’t be possible the way they do it atm [16:27] ah [16:27] the plasma strikes again [16:28] high quality code at work [16:28] bulldog98: write your own then [16:28] apachelogger: problem is that every foo will break with 5.0, since they change the way Qml works [16:29] bollocks [16:29] there will be qt quick 2 [16:29] qt quick 1.x stays the way it is [16:29] anything you do in qt quick 1 will work in qt 5 [16:29] ah ok [16:47] ScottK: do you have some time to review my quassel packaging changes? [16:48] [lp:~rekonq/rekonq/rekonq-ubuntu] Jonathan Kolberg * 6 * patches/ (fix_history_ordering.diff dont_set_app_as_parent.diff series) Updated the patches for the new version [16:51] micahg: do you think bug #846922 warrants security uploads? [16:51] Launchpad bug 846922 in quassel (Ubuntu) "quassel-core creates world-readable directories" [Undecided,New] https://launchpad.net/bugs/846922 [17:12] debfx: looking [17:13] * bulldog98 testbuilded the fix, I will now upload it as 0ubuntu2~ppa1 into experimental [17:14] Mamarok: iirc i did see natty packages in backports [17:14] debfx: indeed, I think so [17:14] and by backports i mean the kubuntu backports ppa [17:19] shadeslayer: nice, thank you, now it should be announce as well [17:20] Mamarok: yeah, ryanakca_ was working on that iirc, i've lost the credentials to login to the site and can't make a announcement myself === jussi01_ is now known as jussi [18:22] kde-workspace with the fix is up at experimental [18:26] someone needs to test it [18:29] ScottK: qt4-x11 seems to have built on ia64, gnash is test building, time to look at kdesdk [18:29] * micahg just has to figure out how to use the porter box now [18:44] bulldog98: next time, add a better description "Added a fix for kwin" is too ambiguous. For example "Backport revision e142a1a1 from kwin's repository. It fixes an important bug which improves performance greatly" [18:44] bambee: I never was very creative at writing texts [18:44] :) [18:44] :) [18:53] ScottK: I actually have to wait for access for the porter box, so it'll be sometime this week for kdesdk === ejat- is now known as ejat [20:06] apachelogger: desktop file translations are even more broken than I thought. the assumption that gettext domain = desktop_ isn't true anymore. [20:06] e.g. for kate it's desktop_kdebase_kate [20:09] maybe we should drop it already this cycle and rebuild the packages [20:34] micahg: Thanks for looking into it. If you get access, you might see if you can figure why the latest qt4-x11 upload on oneiric failed on powerpc (I already retried it once and it hung at the same point). [20:35] ScottK: I'd prefer to leave that for didrocks if he'll take it [20:35] Understandably ... [20:41] debfx: it makes me lol all over [20:41] debfx: quite honestly to me that sounds a lot like we need to act, fast, as the release is drawing close [20:42] so I propose you bring this up at the meeting on tuesday and we'll have the council make a call [20:48] apachelogger: If it's broken, it sounds like someone who understands the problem ought to just write kubuntu-devel and explain the options. KC is only needed if we can't agree. [20:49] because discussion on kubuntu-devel is always very conclusive, right? :P [20:50] but yeah, anything will be fine... just thought since we have a meeting anyway we might as well run it through there [20:57] In any case I'd like a discussion in advance so people have a chance to think it over. [20:58] like there is much to think :P [20:58] debfx: please toss a mail on kubuntu-devel [20:58] if no decision was reached by tuesday we'll make one at the meeting based on the discussion until then or something like that [21:39] debfx: I'd appreciate it if you'd run your ISO size diff script again on our alternates relative to beta 1. I made some changes yesterday and they didn't help nearly as much as I'd anticpated. [21:46] apachelogger: mail sent [21:46] time to debug video players \o/ [21:48] ScottK: the script won't pick up your changes if the image hasn't been rebuilt since then [21:48] It has. [21:48] The current one reflects the changes. [21:48] ah ok, then it should already be up-to-date [21:49] it runs every couple of hours [21:54] Thanks. [21:55] What's the URL again? [21:56] ScottK: http://felix.fobos.de/kubuntu/kubuntu-oneiric-cd-alt-amd64-diff.htm [21:56] Thanks. [21:56] firefox-locale-ja [21:56] uffta [21:57] also I think we should throw some locales out [21:57] de, fr, pl, ru and ja surely consumes a lotta space [21:57] and pt it seems [21:57] and it [21:57] and es [21:57] zomg [21:59] apachelogger: the langpacks pull the firefox locales in now since they're built from the firefox source, this should be fixed either for or after the LTS [21:59] * apachelogger would rather have it fixed for right now :P [22:00] but I think simply moving some languages off the seed would be a good enough thing to do for now [22:00] the language packs are a bit broken [22:01] language-pack-kde-de depends on (and replaces) language-pack-kde-de-base [22:01] ah, nothing's like old news :P [22:01] apachelogger: That's true, but we also got llvm2.9 and gtk3.0. [22:01] ScottK: we need gtk3 [22:01] I'd like to see if we can get rid of those first. [22:02] apachelogger: Why? [22:02] so I can make fun of kubuntu when I am socializing with opensuse people [22:02] Yeah, well I got it fixed once before. [22:02] ScottK: didn't you try to give it the boot last time we were oversized? [22:02] right [22:02] looks like the translations have been moved from -base to the non-base package [22:02] I did and I succeeded. [22:03] but -base is still pulled in (and contains old translations) [22:03] it is like a zombie, no matter how often you kill it... it keeps coming back [22:03] * apachelogger shivers in fear [22:03] so I guess currently translations take up twice as much space [22:03] that is unfortunate [22:05] all:libllvm2.9 | llvm-2.9 | libgl1-mesa-dri | Ubuntu Developers Sounds like fixing that would help too. [22:09] what is there to fix? [22:10] mesa uses llvm for software rendering [22:10] apachelogger: well, WRT the firefox-locales, it's the same as before, just as a separate binary [22:11] debfx: No, I meant the translations. Sorry. [22:11] ah [22:11] I guess I'm not sure why we didn't have llvm2.9 there before. [22:14] ldm (seeded in ship) pulls in compiz [22:14] that doesn't seem right [22:15] No. It doesn't. [22:17] no? [22:18] ldm is on the cd and depends on compiz-core [22:19] At the very least that can move to the dvd I think. [22:19] IIRC it's there for some probably not very relevant ltsp related reason. [22:19] It's possible it could be removed entirely. [22:19] * ScottK needs to go out, so see you later. [22:20] So far the seeds are really confusing me. [23:48] * rbelem pokes ScottK