=== jalcine is now known as JackyAlcine_ === JackyAlcine_ is now known as jalcine [00:20] heya Riddell === jalcine is now known as JackyAlcine_ === emma is now known as em === webjadmin is now known as JackyAlcine_ === JackyAlcine_ is now known as webjadmin === webjadmin is now known as jalcine [02:21] if he was looking for the beta version of gimp, there isn't a ppa i would trust === jalcine is now known as JackyAlcine_ === JackyAlcine_ is now known as jalcine [02:52] rbelem: I think Riddell's gone for a few days. [03:42] i'm looking for distros to hand out at a booth, i noticed that kubuntu didn't offer the user the ability to reformat currently installed operating systems, is this being worked on? [03:47] ybit: What do you mean by "reformat the current operating system"? You can certainly install over an existing operating system and reformat. [03:50] reformat a partition which another OS currently resides in* [03:50] hmm... [03:50] i installed kubuntu [03:51] i then went to install kubuntu on top of that kubuntu and it didn't give me an option to reformat the current installation [03:51] that was my testing expereience [04:25] You can. [04:25] Although, unless the partition is messed up there's really no need to. [06:09] 4.8.1 on oneiric being installed currently [09:04] nice, KDE 4.8.1 is arriving in the repos :) [10:11] If I have a bug in icon only task manager, against which package do I file it? [10:53] jussi: bugs.kde.org :) [10:53] debfx: thats fail. :/ Whyever cant it be reported in K/Ubuntu? [10:55] jussi: if it's an upstream issue it will likely be ignored on the ubuntu bugtracker === rdieter is now known as rdieter_away [12:30] hmpf. after today's kubuntu upgrade, kmail and akonadi keep crashing somewhere in nepomuk/soprano... [13:05] Riddell, apachelogger: is someone working on the installer? I have time to work on it [13:06] bulldog98: No one is working on it, IIRC. [13:06] I think nixternal triaged a bunch of bugs, but that's as far as he got. I may have other plans. [13:10] 'Morning all [13:14] yofel: want me to help uploading 4.8.1? [13:21] I take that as a yes ;) [13:21] * debfx grabs kdepimlibs [13:22] debfx: yes, I'm at school and won't be able to upload anything for the next 2 hours [13:23] ok === jalcine_ is now known as jalcine === bulldog98_ is now known as bulldog98 [13:51] shadeslayer: would it be ok for you if I would apply for becoming Kubuntu dev the same time as you? [13:51] bulldog98: I don't see why :D [13:59] debfx: Are you uploading anything else? [13:59] * ScottK might do a few. [14:00] bulldog98: The fewer meetings the better. === jalcine is now known as JackyAlcine_ [14:03] bulldog98: no one working on it [14:04] apachelogger: so I’ll have a look at it :) [14:04] bulldog98: what you want to do? [14:04] apachelogger: I would like to port the most stuff to qml [14:05] uh [14:05] bulldog98: don't we have more important things todo :P [14:05] apachelogger: eg? [14:05] don't get me wrong, proting and cleanup and stuff needs to happen, just not when we are close to release :P [14:05] bug fixing [14:05] bug triage [14:05] QA [14:06] <--- qa coordinator :P [14:07] apachelogger: don’t we have something like a bug list somewhere on launchpad? === JackyAlcine_ is now known as jalcine [14:10] bulldog98: For the installer, there's (sorry, ugly LP URL follows): [14:10] https://bugs.launchpad.net/ubuntu/+source/ubiquity?field.searchtext=kde&search=Search&field.status%3Alist=NEW&field.status%3Alist=INCOMPLETE_WITH_RESPONSE&field.status%3Alist=INCOMPLETE_WITHOUT_RESPONSE&field.status%3Alist=CONFIRMED&field.status%3Alist=TRIAGED&field.status%3Alist=INPROGRESS&field.status%3Alist=FIXCOMMITTED&field.assignee=&field.bug_reporter=&field.omit_dupes=on&field.has_patch=&field.has_no_package= [14:11] * ScottK does kactivities. [14:14] I have done that but: Signer is not permitted to upload to the component 'main' ... [14:15] Oh. So we need the package set list updated. [14:15] I just uploaded it. [14:15] debfx: Any others you've done? [14:16] Or more particularly any others you know you can't do because they aren't in the packageset? [14:16] ScottK: nope, haven't done any others [14:16] so who broke my nvidia driver! :/ [14:16] micahg: What's the process now for fixing obvious omissions due to new pakcages from a packageset? [14:17] jussi: nvidia. By definition. [14:17] hehe [14:17] likely all that are new in 4.8 [14:17] who broke its integration with ubuntu more to the point ;) [14:18] ScottK: I'm pretty sure it's still "ping cjwatson" [14:18] debfx: I suspect so, but the DMB may have powerz now. Not sure. [14:21] even if they have the power I doubt anyone else would be willing to touch it [14:21] micahg is fearless. He'd do it. [14:24] OK. kactivities and kde-baseapps up. === jalcine_ is now known as JackyAlcine_ === JackyAlcine_ is now known as webjadmin === webjadmin is now known as JackyAlcine === JackyAlcine is now known as jalcine_ === jalcine_ is now known as jalcine [14:27] * ScottK does libkipi next. [14:32] OK. That's done. Time to pretend to work for awhile. === vivek is now known as Guest24242 [14:44] ScottK, is he sick? [14:45] I don't think so. [14:45] He just mentioned something about being away. [14:46] ScottK, i hope he is completely recovered from that accident [14:47] AIUI he's not fully recovered, but mending slowly and ~as expected. [14:47] apachelogger: as far as I can see we need to port ubiquity from PyQt to PySide to use Qml [14:48] ScottK, so, if you have some time, could you help me with the active pkgs? [14:49] rbelem: Possibly, but the time I have for Kubuntu today, I'm trying to help upload KDE 4.8.1. === Guest24242 is now known as pvivek [14:50] ScottK, oki :-) when you r free, could you ping me? [14:51] I probably won't remember. Once 4.8.1 is done, give me a shout. [14:51] ScottK, will do [14:51] :-) [14:56] pfff now chrome/chromium doesn't respect color scheme. it forces black text like libreoffice used to do [14:57] hmpf. I guess I should've disabled the beta ppa [15:00] * ScottK does libkdeedu [15:06] Now okular [15:13] I'll take kde-workspace and oxygen-icons [15:15] * ScottK updated the wiki page. [15:16] * ScottK does -runtime. [15:16] ScottK: have you seen the "ffe?" on the wiki? [15:17] apachelogger: pinga [15:17] Saw it. Approved it already. [15:17] * ScottK should make sure the bug is closed in changelog [15:18] :) another spalsh qml http://www.youtube.com/watch?v=2KZreVEHp5g [15:18] Yep. It's there. [15:19] debfx: One thing I just thought of though ... kde-runtime will need to go through binary New and if I upload it, I can't do that. Would you please review and upload? [15:19] * ScottK doesn't do -runtime. [15:20] yep, can do [15:21] Great. [15:22] * ScottK does libkexiv2 instead. [15:35] * debfx grabs kde-wallpapers === vivek is now known as Guest98591 [15:46] what's the deal with libkwineffectsabi2-gles? [15:47] why is it empty? [15:48] I assume we don't need it anymore since we dropped our double build hacks? [15:49] It's referred to all over the place. [15:51] where? (except kde-window-manager-gles) [15:55] Nevermind. It was kde-window-manager-gles I was looking at. [15:56] We do build kdeinit_kwin_gles.dir/effects.o [16:07] debfx: I think you're right. As nearly as I can tell from the build log of 4.8.0, whatever was in libkwineffectsabi2-gles is now part of kde-window-manager-gles. [16:15] ScottK, does kde-window-manager-gles give more graphics effects options for pci graphics cards like nvidia 7000 series for example ? [16:16] IIRC it's mostly designed for contrained environments where all of GL isn't available. [16:17] ok , hence the nouveau dependency [16:18] guys i have a problem with qml [16:18] /main.qml:2:2: module "QtMultimediaKit" is not installed [16:18] i installed libqtmultimediakit1 [16:18] it still breaks [16:19] * ScottK bets apachelogger knows the answer to that one. [16:20] ScottK: i get a nice core dumped [16:20] xD [16:22] * ScottK looks at marble [16:25] Peace-: sounds like you need libdeclarative-multimedia [16:27] debfx: i wil try [16:28] debfx: nice now it doesn't play anything but.. at least doesn't crash [16:28] debfx: :) thanks fixed [16:31] Hi! would be great if someone with upload privileges could have a look at the Zanshin package I pushed here: https://launchpad.net/~agateau/+archive/ppa/+packages and upload it [16:32] ok now even the splash theme has the audiio === jalcine is now known as JackyAlcine_ [16:34] agateau: It should probably wait until after more of 4.8.1 is uploaded so it doesn't get mixed up with that. Later today or tomorrow. [16:34] ScottK: no problem, there is no urgency [16:34] Peace-: Please make sure there's an easy way to disable said audio. [16:35] ScottK: well i am not working on official kubuntu stuff [16:35] :) [16:35] OK. [16:36] but i will === JackyAlcine_ is now known as jalcine === jjesse_ is now known as jjesse [17:33] guten abend from cebit [17:33] so 4.8.1 is released, yay [17:33] free beer on the KDE stall [17:33] Riddell: oh, having fun there? [17:33] yofel_: how did you get on with 4.8.1? [17:33] Tm_T: this place is huge! [17:33] Riddell: We mostly need to upload stuff. [17:34] Riddell: Also rbelem was looking for help on his active stuff. If you've got any time, it'd be lovely if you could help him since several of us can upload 4.8.1 stuff. [17:34] Riddell: is KDE booth having a good set of different distros and hooks, like plasma active? === vivek is now known as Guest65638 [17:35] I can't do much, I'm in the middle of a conference [17:35] * ScottK does libkdcraw. [17:36] Tm_T: one desktop with kubuntu, one spark [17:36] Riddell: OK. If you could give me a hint about where things are so we can help him out, that'd be good. === Guest65638 is now known as pvivek [17:36] ScottK: on active? [17:36] yes. [17:36] * ScottK thinks apachelogger is looking for something to do. [17:37] images should be being made and rbelem's latest stuff was uploaded on monday [17:37] and I think I stopped kubuntu-active installing plasma-desktop [17:37] so go test the images [17:37] maybe it magically works [17:37] OK. [17:37] He said he had more stufz. Not sure what. [17:38] just see what's new in ~kubuntu-active PPA since monday [17:39] Riddell: nice [17:42] and free currywerst :) [17:42] wurst? [17:42] yes [17:43] sounds delicious [17:43] no, it's wiener [17:50] guys I tested precise and I found this bug (also present with qt 4.8 in oneriric): https://bugs.launchpad.net/ubuntu/+source/kde4libs/+bug/942102 [17:50] Launchpad bug 942102 in kde4libs (Ubuntu) "Notifications have no alert sounds" [Undecided,Confirmed] [17:51] also all panel bugs are still present :( [17:51] schnelle: What we really need is testing on precise with Qt 4.8 and KDE 4.8. [17:52] ScottK: the bug above happens on precise [17:53] also panel bugs are present on precise [17:53] for both qt need to be patched [17:53] OK. Maybe agateau would know about what needs doing. [17:53] * ScottK does kdepim-runtime [17:56] thanks Riddell, ScottK :-) [17:58] Riddell, today's image should have all new stuff? [18:40] Now kdepim [18:46] rbelem: It should be all whatever's in the official archive (not the PPA) [18:47] ScottK, nice :-) everything is there [18:53] * ScottK does pykde4 [18:53] ScottK: well, depends on which packageset and why [18:54] micahg: Adding kactivities to the Kubuntu packageset since it's a new KDE package for KDE 4.8. [18:54] Nothing else uses it. [18:55] ScottK: just e-mail cjwatson to add it to the generated packageset assuming it's seeded [18:55] or rather to rerun his scripts [18:55] OK. Thanks. [18:57] Done. === yofel_ is now known as yofel [20:10] ok i have released :D [20:11] http://kde-look.org/content/show.php?content=149385 [20:11] if anyone wants test it :D [20:19] Peace-: you don’t have a download path anywhere [20:20] bulldog98: ? http://kde-peace-settings.googlecode.com/files/nwd2.tar.gz [20:20] ups [20:20] sorry [20:22] bulldog98: http://kde-peace-settings.googlecode.com/files/nwdsound.tar.gz [20:22] bulldog98: thank you [20:23] i used the old link :/ [20:52] ScottK: I am still looking at plymouth and try to make sense of it [20:53] Ouch. [20:53] bulldog98: regarding bugs... I sent a mail about that back when feature freeze started, clearly no one bothered to read it :P [20:54] bulldog98: also about pyside ... booohoooo [20:54] Riddell is not gonna like that one [20:54] also pyside is not nokia maintained anymore from what I heared [20:54] apachelogger: I’ll put the link into the header === bulldog98 changed the topic of #kubuntu-devel to: Kubuntu: Friendly Computing | 4.8.1 https://wiki.kubuntu.org/Kubuntu/Ninjas/Packaging | Precise: Feature Freeze, time to squish beasties | Bug hunting: https://bugs.launchpad.net/~kubuntu-bugs/+packagebugs| http://status.ubuntu.com/ubuntu-precise/group/topic-precise-flavor-kubuntu.html TODOs! | http://www.kubuntu.org/news/12.04-lts-announce | Jussi's new baby's name is Elodi! :) [20:58] friends! what's with the long header? [20:58] also [20:59] help! friends! I got tricked by a guinea pig! [20:59] Nice. [20:59] They are reasonably tasty. [20:59] I tried one in Peru once. === jussi changed the topic of #kubuntu-devel to: Kubuntu: Friendly Computing | 4.8.1 https://wiki.kubuntu.org/Kubuntu/Ninjas/Packaging | Precise: Feature Freeze, time to squish beasties | Bug hunting: https://bugs.launchpad.net/~kubuntu-bugs/+packagebugs| http://status.ubuntu.com/ubuntu-precise/group/topic-precise-flavor-kubuntu.html TODOs! | http://www.kubuntu.org/news/12.04-lts-announce [21:00] Peace-: so you are now login into a community? :O [21:00] ScottK: you are a strange man [21:00] friends! what's with the almost equally long header? [21:01] someone tinyurl that shoot [21:01] * jussi grumbles once again at the nvidia driver [21:01] or something [21:01] apachelogger: topic is open, fixor it! [21:01] jussi: why wouldn't you know, fglrx is doing the old abi incompatible segfault dance :) [21:01] jussi: It's quite a common food locally there. When in Rome and all that. [21:01] I cannot even not watch no prn [21:02] apachelogger: eh-heh [21:02] jussi: rome? [21:02] Peace-: huh? [21:02] ScottK: rome is in peru? [21:02] oh you all fail [21:02] "when in rome do as the romans do" [21:03] ah :S [21:03] "When in Austria, go to Germany?" [21:03] substitute rome/romans for $place/$people [21:03] ScottK: xD [21:03] suite case point [21:03] oh [21:03] a suite case, picture that [21:04] it's how the doctor travels [21:05] Peace-: good sir, you really do not want to use oga but wav [21:06] apachelogger: ? [21:06] why wav? [21:07] cuz it is less slow [21:07] ah === bulldog98 changed the topic of #kubuntu-devel to: Kubuntu: Friendly Computing | 4.8.1 http://ur1.ca/8kalp | Precise: Feature Freeze, time to squish beasties | Bug hunting: https://bugs.launchpad.net/~kubuntu-bugs/+packagebugs| http://status.ubuntu.com/ubuntu-precise/group/topic-precise-flavor-kubuntu.html TODOs! | http://www.kubuntu.org/news/12.04-lts-announce [21:07] then again you are loading qtmmkit in a splash screen plus a bazillion pictures, one gets to question wherever it makes any difference === bulldog98 changed the topic of #kubuntu-devel to: Kubuntu: Friendly Computing | 4.8.1 http://ur1.ca/8kalp | Precise: Feature Freeze, time to squish beasties | Bug hunting: http://ur1.ca/8kam4| http://status.ubuntu.com/ubuntu-precise/group/topic-precise-flavor-kubuntu.html TODOs! | http://www.kubuntu.org/news/12.04-lts-announce [21:08] also I wanted to ask everyone something, but I forgot [21:08] friends! what's with the topic spam already? [21:08] also === bulldog98 changed the topic of #kubuntu-devel to: Kubuntu: Friendly Computing | 4.8.1 http://ur1.ca/8kalp | Precise: Feature Freeze, time to squish beasties | Bug hunting: http://ur1.ca/8kam4| http://ur1.ca/8kamo TODOs! | http://www.kubuntu.org/news/12.04-lts-announce [21:08] friends! did we ever decide which background to use for splashes? ariya or upstream ariya version for splashes? [21:09] for me the default stuff it's not "Hot" [21:10] that aint not no answer [21:10] :D [21:10] also why does digikam of all things use dh_apparmor? :O [21:11] great no one wants to answer me [21:11] Presumably because it has a profile. [21:11] no new splashes then :P [21:11] apachelogger: because it uses mysql??? [21:11] Oo [21:11] it uses mysql? [21:11] Oo [21:11] :O [21:12] -.- [21:12] :@ [21:12] LOL [21:12] apachelogger: at least you can make it use mysql [21:12] I think default is sqlite [21:12] on windows I have this many user space apps using mysql: 0 [21:12] * ScottK doesn't think anything actually "uses" mysql so much as it endures it. [21:13] on kubuntu I this many user space apps using mysql right now: 2 [21:13] mind my sequel [21:14] dup dup dup [21:14] so, really [21:14] what wallpaper to use? [21:14] for waht? [21:14] (FWIW: the upstream default splash wall is more suitable for low color depths) [21:14] splashes [21:15] i would use dark splash wallapaper [21:15] walla walla xD [21:16] yeah and I would use a picture of nakkid bulldog98, no luck for both of us as neither is an option... -.- [21:16] I could totally roll a dice [21:16] better yet, I'll spam the list [21:16] \o/ [21:16] spam it LOL [21:17] what I like most about systemsettings is the affordance the icons provide [21:17] desktop effects having a desktop and all that [21:18] totally makes me not click it when I want to get desktop settings [21:18] <3 [21:18] nixternal: will I ever get an answer on my mail btw? :P [21:21] sqlite > mysql for a light single user db. [21:26] oh oh [21:26] http://aplg.kollide.net/tmp/aubertruck-png.png [21:26] what I found [21:27] s/found/made/ [21:27] tazz: You did something wrong... Try s/you/me/ or tell me "help sed" [21:27] :p [21:28] no [21:28] that was jon the tacko [21:28] who is once again not here [21:28] kubotu: seen JontheEchidna [21:28] JontheEchidna was last seen 1 day, 18 hours, 21 minutes and 18 seconds ago, quitting IRC (Quit: Konversation terminated!) [21:28] using quassel so he can avoid me [21:28] eh [21:28] konversation [21:29] <-- using quassel because he is in the credits apparently [21:30] He was the litigator for Konversation in the last great Quassel versus Konversation Kubuntu Council case, so no surprise there. [21:31] in the last IRC client war you mean [21:32] +1 for http://aplg.kollide.net/tmp/snapshot020.png [21:32] he probably argued for konversation because he is not in the credits of quassel [21:32] * apachelogger only got told today by a fellow student ^^ [21:33] apachelogger: Yes. [21:34] we should rename ariya to vertigo [21:34] way easier to spell [21:35] oh my tmp folder is filled with bug screenshots [21:35] cool [21:35] http://aplg.kollide.net/tmp/snapshot9.png [21:35] bulldog98: fix that [21:37] * apachelogger pokes fabo into fixing Qt as per mailing list [21:37] apachelogger: I’ll have a look at it [21:38] yofel: u here? [21:38] bulldog98: I am being lazy, could you check some packaging issue for me? [21:39] Peace-: btw, you can have multiple qml files ;) [21:40] apachelogger: i am plannin to rename stuff and put them on one page [21:40] with multiple links :D [21:41] apachelogger: you know the fist walla walla is better for me [21:41] first [21:41] cuz it is made for splashes :P [21:41] no clue whoever made the ariya wallpaper the background [21:41] or how [21:42] and since I am being lazy and bulldog98 doesn't wanna help we might never find out [21:42] apachelogger: helping with what? [21:42] find out why the thing in ksplash is ariya [21:42] oh [21:42] maybe I know already [21:42] sec [21:43] ok, I don't [21:43] bulldog98: something is astray there [21:43] all the stuff in /usr/share/kde4/apps/ksplash/Themes/Default is ariya [21:44] apachelogger: which arch, since on my amd64 it works on the i386 not [21:44] even though it should be that center gradient thing [21:44] amd64 [21:44] that doesn't matter for share files anyway :P [21:45] * apachelogger needs to clean out his bzr dir for it is a right mess [21:45] kde-workspace-data, kde-workspace-data-extras: /usr/share/kde4/apps/ksplash/Themes/Default [21:46] apachelogger: with ubiquity-slideshow I don’t see where I would need to change something [21:46] dantti: did I tell you about print-mgr ppa btw? === jalcine is now known as JackyAlcine [21:46] cuz it is a right mess too === JackyAlcine is now known as jalcine [21:47] bulldog98: you'd need to make the text have a background, which looks equally crappy [21:47] unless you find a way to make the text flow around the pic [21:48] ohohoh [21:48] bulldog98: u on .1? [21:48] that would explain why you have the upstream background for splash [21:48] they might have changed it for .1 === apachelogger changed the topic of #kubuntu-devel to: Kubuntu: Friendly Computing | 4.8.1 http://ur1.ca/8kalp | Precise: Feature Freeze -> Bug hunting: http://ur1.ca/8kam4| http://ur1.ca/8kamo TODOs! | http://www.kubuntu.org/news/12.04-lts-announce [21:49] apachelogger: yeah I’m on .1 but also on my eeepc, where the problem still exsists [21:49] apachelogger: nah [21:49] apachelogger: but why you don't change the default wallpaper [21:49] dantti: kubuntu experimental [21:49] apachelogger: I'm stuggling with a mem leak right now :P [21:49] dantti: also see my wicked mail on kubuntu-devel [21:49] apachelogger: it is making my plasma desktop eat all my ram... but hopefully it's the last one.. [21:50] dantti: basically you just need to tell me a commit that works well and I can roll an update for all supported releases in less than a second (not including build time which is like 3000 hours ;)) [21:50] apachelogger: I'm not subscribed to k-dev [21:50] poor soul [21:50] apachelogger: right now I'd say no commit works well until I kill that leak [21:50] dantti: https://lists.ubuntu.com/archives/kubuntu-devel/2012-February/005888.html [21:51] otherwise it will eat your ram after a few hours [21:51] https://code.launchpad.net/~kubuntu-ppa/+archive/experimental [21:51] once you are fine with moving to to backports or beta-backports we can also easily do that [21:51] rbelem: also I still do not know where your copyright script is ;) [21:51] apachelogger: kde-workspace-data-extras is faulty [21:52] * bulldog98 is gona have a look at kde-workspace [21:52] but how [21:52] ./kde-workspace-data-extras.install:usr/share/kde4/apps/ksplash/Themes/Default/600x400/background.png [21:52] it just installs whatever is there [21:52] apachelogger: and that is the wrong image [21:52] but how? [21:53] surely we do not patch images [21:53] grep doesn't list any such business anyway [21:53] apachelogger: have we checked upstream, if the images are right? [21:53] apachelogger: btw do you know why ksplashqml is not on kubuntu 11.10 ? [21:53] http://websvn.kde.org/trunk/KDE/kde-base-artwork/ksplashx-themes/default/ [21:53] unless dirk rolled a wrong tar that should be the background [21:54] Peace-: it is in some extra package [21:54] someone is writtin here that i am an asshole because kslashqml doesn't exist on k11.10 :D [21:55] oh [21:55] it is not in 4.7! [21:55] funny [21:55] so not in stock 11.10 [21:55] you'd need to upgrade to .8 [21:55] i am on 12.04 :D [21:55] well that has .8 [21:56] a kid has downloaded my theme ... [21:56] and it's angry xD [21:56] *he is [21:56] anyway => bed [21:56] cya [21:57] just when I wanted to praise some work [21:57] oh well [22:01] apachelogger: tarball is ok [22:01] then I do not get it [22:01] so we seem to somehow copy something from horos to default [22:01] oh [22:02] no [22:02] tarball does not seem ok [22:02] /tmp/kde-workspace-4.8.1/ksplash/ksplashx/themes/default/1920x1200 is ariya [22:02] bulldog98: please mail kde-release and kde-artists [22:03] this is not healthy [22:03] apachelogger: ok [22:04] apachelogger: for me it’s that stripe thing [22:06] apachelogger: I did apt-get source kde-workspace and all under kde-workspace-4.8.1/ksplash/ksplashx/themes/default/ is what it is supposed to be [22:06] http://www.golem.de/news/freier-desktop-zeitplan-fuer-kde-sc-4-9-steht-fest-1203-90168.html lol, now I really need to make dragon3 happening :O [22:06] bulldog98: I did pull-lp-source [22:06] bulldog98: well yes [22:06] bulldog98: the stripe thing is ariya [22:06] which is wrong [22:06] it should be the center gradient [22:06] http://websvn.kde.org/trunk/KDE/kde-base-artwork/ksplashx-themes/default/ [22:07] maybe ask someone in #oxygen [22:09] dantti: did you valgrind the leak? [22:10] apachelogger: yup [22:10] when I fix the leak it does not talk to cups correctly [22:10] oh? [22:10] whats the leak? [22:11] codewise [22:11] kcupsconnection.cpp:190 [22:11] apachelogger: if it’s announced you have to hurry up :P [22:11] a string dup that is not deleted [22:12] (if I delete it still fails( [22:13] * I pushed a fix but them things stoped working :P [22:13] *then [22:13] bulldog98: that's not particularly trivial [22:13] bulldog98: while you are over in #oxygen ask nuno nicely wherever he has time to come up with a UI design for dragon3 [22:14] that'd would at least only leave the technical problem in phonon with format negotiation [22:14] apachelogger: wait aren’t you the developer of dragon3? [22:14] case in point [22:14] *developer* [22:14] not UI artist [22:15] apachelogger: but still you should ask him [22:15] I am too shy [22:15] also waitign for quickgit to time out ;) [22:17] dantti: it is funny how the dir is called libqcups and has Kclasses ;) [22:18] https://projects.kde.org/projects/playground/base/print-manager/repository/revisions/master/entry/libqcups/KCupsConnection.cpp#L190 ? [22:18] apachelogger: even better in the package kde-workspace-data-extra the pics are all the same and the right pics [22:18] eh? [22:19] bulldog98: that doesn't make sense [22:19] all pics in the source tar are ariya [22:19] and not the gradient [22:19] apachelogger: Ii renamed the classes [22:20] apachelogger: the names confuse me tell me what I’m supposed to see on the pics [22:20] so where is the code :P [22:20] bulldog98: the one from websvn [22:20] which is the thing with the center gradient [22:20] ariya is the stripe thing [22:21] (FWIW the gradient is the base of ariya ;)) [22:21] remove the stripes and you get the gradient ;) [22:21] apachelogger: ok [22:21] apachelogger: a commit before that [22:21] apachelogger: even better on my eeepc I have horos pics at that locations [22:22] apachelogger: yes that are the stripe things [22:22] dantti: https://projects.kde.org/projects/playground/base/print-manager/repository/revisions/af3015ba428473bcdf609c5811ceb5f0369adda7/diff/libqcups/KCupsConnection.cpp [22:22] ? [22:22] oh nvm [22:22] I am being silly [22:25] dantti: btw you might want to read up on QStringBuilder [22:26] apachelogger: right [22:26] httpAssembleURIf(HTTP_URI_CODING_ALL, uri, sizeof(uri), "ipp", "utf-8", "localhost",.... [22:27] oh [22:27] I guess that utf-8 there is defining the encoding of the path? [22:27] cause my immediate guess would have been that destination.toUtf8() might screw it over [22:29] bulldog98: will nuno make me a design? :P [22:30] dantti: request = ippNewDefaultRequest(values.take("printer-name").toString(), isClass, operation); [22:30] apachelogger: I think so, but this works for all the rest of the code the real problem is that I need qstringlist to const char ** [22:30] toString().toUtf8()? [22:30] maybe [22:30] apachelogger: he hasn’t answered yet [22:30] apachelogger: that is not the leak [22:31] look at the loop where I copy the values [22:31] dantti: no, I mean in your fix for th eleak you do not explicitly convert to utf8 [22:31] is it still leaking with the latest commit? [22:31] or just not working? [22:31] not working [22:31] no leak [22:32] then try a toutf8 on that line from above [22:34] ohm [22:34] dantti: attr->values[i].string.text = qstrdup(list.at(i).toUtf8()); [22:34] actually [22:34] this does not make sense [22:34] actually the entire QStringList case does not make sense WRT attry [22:34] s/attry/attr/g [22:34] apachelogger meant: "actually the entire QStringList case does not make sense WRT attr" [22:34] that actually works, but qstrdup, needs cleaning... [22:35] ohm [22:35] dantti: you never do anything with that attr? [22:35] maybe I am not understanding the code though [22:36] IPP_TAG_COPY is said to copy the const char **, but this code also does not working if I delete http://privatepaste.com/e42e6a498e [22:37] apachelogger: I only need the attr to set up it's internal values [22:37] apachelogger: you get help form nuno, but I must warn you he uses mplayer [22:37] ah, so it is held elsewhere? [22:37] apachelogger: if I pass a const char ** to ippAddStrings then I can ignore the attr [22:38] apachelogger: yes it adds strings to ipp resquest [22:38] apachelogger: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=369083 [22:38] curious [22:38] Debian bug 369083 in libgnomecups1.0-1 "gnome-cups-manager: gnome-cups-icon bloated ?" [Normal,Fixed] [22:38] is ippAddStrings one of your functions? [22:38] the gnome-cups-had the same issue, but they managed to fix it but I'm trying to do the same and it's not working :P [22:39] apachelogger: no [22:39] it's a cups function [22:39] you should do a wrapper around it IMHO [22:39] to which you can pass a QString [22:39] just a thought on making the code better :) [22:39] doesn't really help us [22:40] apachelogger: well I use it just a bit... the problem is not with string also it is with string lists :P [22:41] this is wicked really [22:41] ok [22:41] so [22:41] attr->values[i].string.text = qstrdup(list.at(i).toUtf8()); [22:41] leaks because cups has refcounting internally and refuses to delete the string [22:42] only stuff add via ippAddStrings makes an internal copy if that copy flag is set [22:42] thus it is part of the refcount and deleted properly [22:42] working aroudn that function and direclty inserting strings then of course causes a leak [22:42] so that's that [22:43] doesn't explain why it is not working though :) [22:43] it's weird... [22:45] dantti: did you try that missing toutf8? [22:46] apachelogger: i think so.. [22:46] still not working? [22:46] nah [22:47] apachelogger: btw values[item] = list.at(item).toUtf8().data(); [22:47] see I was using the toUtf8() [22:48] what is values? [22:48] and what is list? [22:48] typeweise [22:48] typewise even [22:48] https://projects.kde.org/projects/playground/base/print-manager/repository/revisions/master/entry/libqcups/KCupsConnection.cpp#L190 [22:48] apachelogger: char *values[siz] [22:48] ohm [22:48] list better be around for the time values is used [22:49] otherwise the memory will become invalid [22:49] that should be fine [22:49] bulldog98: eh? you can use QMl with PyQt [22:50] but the toutf8 was elsewhere [22:50] Riddell: yes that’s possible [22:50] dantti: request = ippNewDefaultRequest(values.take("printer-name").toString(), isClass, operation); [22:50] that line is missing it [22:50] s/yes/what is that/ [22:50] bulldog98 meant: "Riddell: what is that that’s possible" [22:50] apachelogger: ah ippNewDefaultRequest is internal... [22:53] dantti: well, you were passing it a char* previously [22:53] other than that there is no difference that would make things not working [22:53] not in that class anyway [22:54] bulldog98: you're hoping to use QML in Ubiquity? [22:54] yes [22:54] Riddell: overdue that is [22:55] what we do there is quite the abomination ;) [22:55] bulldog98: to do what? the slideshow? [22:55] Riddell: yes that was the main purpus [22:56] dantti: https://projects.kde.org/projects/playground/base/print-manager/repository/revisions/db4d72570b457292bed1b4048e726b2704cff25f/diff/libqcups/KCupsRequest.cpp [22:56] there you stop setting m_retArguments [22:57] https://projects.kde.org/projects/playground/base/print-manager/repository/revisions/db4d72570b457292bed1b4048e726b2704cff25f/diff/printmanager-dataengine/PrintManagerEngine.cpp [22:57] and there you turned the last if [22:57] eh [22:57] else [22:57] else if [22:58] bulldog98: of course there are probably lots of bugs that need fixed before 12.04 first :) [22:59] apachelogger: that just make the Request class not setting returned args since jobs() is defined then... [22:59] bulldog98: the main issue with having a bling slideshow is having an artist to make it bling, do you think you can get sheytan or the like to make bling for it? (or maybe he already has?) [22:59] it's just the string stuff I'm actually thouching... [22:59] dantti: but then it must be the missing toUtf8 :P [22:59] https://projects.kde.org/projects/playground/base/print-manager/repository/diff/libqcups/KCupsConnection.cpp?rev=db4d72570b457292bed1b4048e726b2704cff25f&rev_to=af3015ba428473bcdf609c5811ceb5f0369adda7 [23:00] also you dup'd name [23:00] Riddell: I first have to experiment a bit with ubiquity and qml [23:00] dantti: I'd replace that newdefaultrequest argument with qstrdup(values.take("printer-name").toString().toUtf8() [23:00] see if that works === rdieter_away is now known as rdieter_laptop [23:00] oh [23:00] blimey [23:01] cupsDoFileRequest also needs a dup [23:01] or the COPY flag thing (if there is one) [23:02] resource.toUtf8() also smells funy unless that function is internal ;) [23:04] bulldog98: the current ubiquity design was done by I think nuno or someone from Oxygen team and implemented by shtylman using QWidget stylesheets [23:05] talking about oxygen [23:05] Riddell: sheytan wants to bling it [23:05] Riddell: he just needs a dev [23:05] bulldog98: what was the conclusion on splash background business? [23:06] apachelogger: nuno had no plan what the default is and I’m still waiting for feedback [23:06] fun [23:07] DAFT PUNK! [23:07] now, where was I [23:07] apachelogger: but the leak is on the string list coping... [23:07] apachelogger: sheytan wants to bling ubiquity? [23:07] it works if that leaks... [23:07] Riddell: yes [23:08] I am being confused here [23:09] dantti: db4d7257 does not work, right? [23:09] apachelogger: if sheytan wants to bling ubiquity all we need is a keen coder to code Qt Quick bling into Ubiquity, I wonder if we have anyone like that [23:09] like cups does not want to talk to you? [23:09] Riddell: bulldog98 apparently [23:09] Riddell: ubiquity needs major refactoring anyway [23:09] the code is really terrible to read (probably mostly since we do ugly thigns to qwidgets to get nice theming) [23:10] gosh, bulldog98 would be interested you say? that is fortunate [23:10] quite :) [23:10] apachelogger: those where other leaks... [23:10] dantti: you are confusing me ;) [23:11] if db4d7257 is not working then I'd say first order of business is making it work again [23:11] and attr->values[i].string.text = qstrdup(list.at(i).toUtf8()); leaks [23:11] as explained in the debian bug report [23:11] so either you need to use the ippAddStrings function somehow and set that COPY flag [23:12] or you need to do 'overlay refcounting' (that is, you must somehow track which requests were processed) so that you can delete the qstring at the right time [23:13] e.g. qmap where Request is the request all QStrings in QStringList are associated with, once Request is processed (e.g. you get a reply) you delete all QStrings of that list and remove the entry [23:14] that's how I understand the issue anyway [23:14] i prefer the copy flag :P [23:14] well yes, that is the way to go [23:14] the qmap approach is only a workaround ;) [23:14] I'll try http://snippets.dzone.com/posts/show/8283 [23:15] ohm [23:15] ah [23:15] eh [23:15] dantti: I don't get it [23:15] dantti: you simply need to convert a qstringlist into a char** to use the copy flag? [23:18] apachelogger: yes [23:19] apachelogger: internally the ippAddStrings() would copy them... [23:20] right [23:20] you should not need to memcpy then [23:22] dantti: toUtf8 returns a QBA, which should deep-copy from QString [23:22] apachelogger: no I would use qstrdup.. [23:22] dantti: on what? [23:23] values[item] = qstrdup(list.at(item).toUtf8()); [23:23] that does not make sense [23:23] toUtf8 deep-copies from QString already [23:23] so you doing qstrdup made that line do *two* string copies there [23:25] dantti: sec [23:25] apachelogger: well if I don't qstrdup it does not work :P [23:25] because toUtf8 yields a QBA :P [23:31] apachelogger: this doesn't work eithervalues[item] = list.at(item).toUtf8().data(); [23:31] dantti: http://paste.kde.org/435056/ [23:32] something like that anyway [23:32] as long as you hold the QBA the char** stay valid [23:32] dantti: what you pasted cannot work because the memory of data goes kaput once that line is done executing [23:32] as the QBA gets deleted [23:33] so you need to hold the QBAs of *all* values in your char** for as long as that char** needs to stay around [23:33] apachelogger: I know to UTf8 yields a QBA but using data() will return a char * [23:33] yes [23:33] of the memory inside the QBA [23:33] it won't do a copy [23:33] which is what my function uses [23:34] there is one string copy per value, which is the one from QString->QByteArray [23:34] then we simply hold those QBAs around to keep the char** data valid [23:35] actually my function could throw a qbaList->reserve(list.size()); somewhere in at the beginning [23:35] then the qbaList doesn't constantly need to resize either (with many entries) [23:37] also I forgot null termination ^^ [23:38] dantti: http://paste.kde.org/435062/ [23:38] that should be about right now [23:40] http://paste.kde.org/435068/ [23:40] called something like that [23:41] apachelogger: so, if I do toUtf8().data() it will delete the string because it will go out of scope right? [23:41] yes [23:41] just use my function already :P [23:41] and don't forget to add my copyright :P [23:41] paste it then :P [23:41] see pastes from above [23:42] bulldog98: oh it probably also needs ported to python 3 next cycle [23:42] dantti: might need some fixing though ... I did not actually compile it :) [23:42] although so does our pykde packages [23:43] Riddell: I’ll look what I’m able to do [23:43] dantti: the idea should be visible though, we simply use the constData (i.e. internal data of the QBA) in the const char **, we ensure the QBAs stay around by putting them in a list, once that list is deleted/goes out of scope the data is deleted and all we need to do is remove the pointer itself [23:44] I doubt you'll get much more performant than that... unless you manually conver the 16bit QChar* of QString in which case you'd not need the QBAList but about 300 more lines of code ;) [23:44] s/conver/convert/g [23:44] apachelogger meant: "I doubt you'll get much more performant than that... unless you manually convert the 16bit QChar* of QString in which case you'd not need the QBAList but about 300 more lines of code ;)" [23:45] apachelogger: and it crashes :P [23:45] the function might need pimping [23:45] dantti: please paste your code [23:45] apachelogger: I changed constData() to data() [23:46] why? [23:46] apachelogger: since it was not building [23:46] Oo [23:46] that cups function wants char**? [23:46] wtf [23:46] >> /home/daniel/code/print-manager/libqcups/KCupsConnection.cpp:96:32: erro: invalid conversion from ‘const char*’ to ‘char*’ [-fpermissive] [23:46] well, I dunno what line that is ^^ [23:46] dantti: please paste the entire file eitherway [23:46] no your qstrs.dfsdfs() wants [23:47] apachelogger: http://privatepaste.com/caf6542fe3 [23:47] am I off by one? [23:47] of course I am [23:47] line 96 fails to build [23:47] ptr[list.size() + 1] = 0; [23:47] ptr[list.size()] = 0; [23:48] way too late for char** magic really :) [23:48] ah [23:48] my bad [23:48] char **ptr = new char*[list.size() + 1]; [23:48] should be [23:48] const char **ptr = new char*[list.size() + 1]; [23:48] then it should be working [23:48] I think [23:49] though it doesn't matter whether you use const char and constData or char and data, the function implicitly makes it const char** on return anyway [23:49] what probably crashed it was the off-by-one [23:50] I was also newing the values again :P [23:50] http://paste.kde.org/435080/ [23:51] dantti: newing the values? [23:51] too much new :P [23:51] values = new [23:51] oh ^^ [23:51] qStringListToCharPtrPtr really shoudl creturn const char ** [23:51] otherwise it is misleading [23:52] apachelogger: http://privatepaste.com/6377a5b428 [23:52] not working either.... [23:52] unless the ipp thing requires char ** [23:52] in which case it is rubbish [23:52] dantti: what's the error? [23:52] oh, crash? [23:52] if so, where? [23:52] apachelogger: there are no errors, it simple doesn't the data [23:53] for (int i = 0; i < list.size(); ++i) { [23:53] kDebug() << values[i]; [23:53] } [23:54] oh [23:54] / values[list.size()] = '\0'; [23:54] it needs a \0 termination? [23:54] it = cups [23:54] * dantti things it's not dunno [23:55] lol [23:55] curious, try the kdebug then [23:55] or qdebug, whatever makes sense [23:55] I think cups is not copying the strings... [23:55] see if there is sane values in values [23:56] oh [23:56] The fix is simple here: when [23:56] its last argument is not NULL and the tag IPP_TAG_COPY is present, the [23:56] well that is the case [23:56] dantti: check if the values are valid please [23:56] so we can rule out the new function is failing [23:57] oh [23:57] (IPP_TAG_URI | IPP_TAG_COPY) [23:57] that should be & not | [23:57] eh [23:57] no [23:57] man [23:57] I really should go to bed ^^ [23:58] apachelogger: http://privatepaste.com/a43e5d3f9c [23:58] the array is strange.... [23:58] why? [23:59] apachelogger: hmm with const it works http://privatepaste.com/f99e349030 [23:59] told ya [23:59] const ftw :P