[00:05] I don't know of it's a good idea, but the rebranding could actually be an opportunity for Kubuntu to match Ubuntu colors, which I always thought made some sense [00:05] an aubergine tinted Air could look good [00:08] it will look good getting replaced right away :p === jonathan__ is now known as jjesse === jonathan_ is now known as jjesse [02:13] * NCommander builds a full kdebindings on ARM and has yet to get a segfault ... [02:13] and I'm at 81% ... [02:16] NCommander: What did you change? [02:17] Or are you just playing roulette? [02:18] persia: the annoying bit? Nothing. [02:18] persia: this happened before when I tried to debug it and then I pounded retry and it blew up again, then I forgot about it === gatlinAFK is now known as gatlin [03:27] Anyone happen to know which package does the compositing for kwin desktop effects? A user in #ubuntu-devel claims that glxinfo is required, which slipped out of main, which, if true, probably indicates a missing recommends (unless this was intentional). [03:29] Sarvatt: I just asked your question a moment ago. [03:30] * genii waits for NCommander's ARM to blow up extravangtly and spectacularly [03:30] genii: not yet :-) [03:30] Cool === gatlin is now known as gatlinAFK [04:08] persia: what do you mean "does the compositing"? kwin does its own compositing if i understand it correctly [04:09] yuriy: Does kwin expect to find glxinfo? [04:09] ah is that the question. i don't see why it would, but i have no idea [04:11] hmm nepomuk is using over a gig of ram, plus almost 500M for dbus-daemon :-\ [04:13] Yeah. If something *does* want glxinfo, it ought be on the CD. [04:18] KWin definitely does not need glxinfo to do compositing [04:19] Sarvatt: With that, please file a bug about your issue. [04:19] JontheEchidna: Thanks. [04:19] * JontheEchidna giggles wobbly windows sans-glxinfo ;-) [04:19] *jiggles [04:22] oh? will do then, I reproduced it 3 seperate boots where it wouldn't let me activate gl compositing until I installed mesa-utils and then it worked properly [04:35] yeah I grabbed the kdebase-workspace source and it shouldn't need mesa-utils, on the plus side I see I need to submit some patches to to kde to extend these hardcoded gl vendor strings to work for nouveau and gallium in general === gatlinAFK is now known as gatlin [06:42] ScottK: we spoke yesterday, but did you ever think of something for a novice with C++ experience to do? [06:42] I like Kubuntu, this is to be a labor of love [06:43] gatlin: JontheEchidna or apachelogger would be better people to ask. [06:43] alright, thank you very much === gatlin is now known as gatlinAFK [09:58] hey Riddell, good morning. I've just seen bug 180994, and I wanted to ask you: is there not a locale for 'sco'? Looking at http://sourceware.org/git/?p=glibc.git;a=tree;f=localedata/locales;hb=HEAD, it seems there isn't. Do you know what the difference between Scots and Scots Gaeilc is (and which code corresponds to each one, if they are different)? [09:58] Launchpad bug 180994 in langpack-locales "Scots cannot be selected" [Undecided,New] https://launchpad.net/bugs/180994 [09:58] My concern is that we seem to have a Scots team in Launchpad, and we're letting people translate into Scots, while these translations won't make it to the distro if there isn't a locale [10:32] debfx: Scots and Gaelic are entirely different languages, Scots is Germanic while Gaelic is Celtic [10:33] Scots is sco, not sure what Gaelic is [10:33] I don't think there's a locale for Scots [10:39] dpm: ^ [10:40] thanks Riddell, debfx [10:42] oh aye, wrong d [10:43] NCommander: how did kdebindings do? [10:44] Riddell: tuxdavis took over for me, and might have found the cause of the crash [10:47] hmm, another seg fault :( http://launchpadlibrarian.net/40195659/buildlog_ubuntu-lucid-armel.kdeedu_4%3A4.4.1-0ubuntu1_FAILEDTOBUILD.txt.gz [10:51] NCommander: well if you or tuxdavies want a new challenge, there it is [10:54] tuxdavis: ah, just the person for a new challenge http://launchpadlibrarian.net/40195659/buildlog_ubuntu-lucid-armel.kdeedu_4%3A4.4.1-0ubuntu1_FAILEDTOBUILD.txt.gz [10:55] can anyone here confirm that setting the date/time via systemsettings is NOT possible in 9.04 and kde 4.4.1? or the opposite? [10:56] i.e., there is no dialog asking for authentication, nothing. just does not work. [10:57] This is the same bug I found in kdebindings, and I believe NCommander found a similar one in another package, the issue resides in the toolchain and not kde [10:59] tuxdavis: did your inline fix work as a workaround? [11:00] no it still segfaulted [11:01] basically I first recompiled everything in debug mode, as it was the smokegen part of kdebindings that was segfaulting [11:01] Then I re-ran make with VERBOSE=1 to find the exact command, then ran it through gdb [11:01] When it crashes you can tell gdb to print two values [11:02] value1 = "print m_session->token_stream->tokens[0].kind" (should print just fine) [11:02] value2 = "print m_session->token_stream->kind(0)" (crashes with segfault) [11:03] the method for "kind" is "int kind(std::size_t i) {return tokens[i].kind;}" [11:04] Since a very similar bug is being found on the same architecture, we've successfully narrowed it down to a bug in the gcc/g++ toolchain [11:08] * Riddell hugs tuxdavis [11:15] Torch: works ok in lucid [11:17] Riddell: i'm suspecting sth is either broken after the update to kde 4.4 or on my machine regarding policykit on karmic. hence the question. [11:17] yes it'll be a policykit issue [11:17] "set date and time automatically" is broken on lucid however :( [11:17] Riddell: i have no authentication manager on the session bus. i think there should be one, right? [11:17] (my understanding of policykit is very limited still) [11:21] Torch: I'm not sure, policykit 0.9 works differently from 1.0, one has it on the system bus the other doesn't [11:22] Riddell: yeah, i just noticed lucid has the newer one. works for me too in my vm on lucid. [11:23] Riddell: do you happen to know which package on karmic has the auth agent for kde and/or where it's installed? [11:25] (i just noticed i was writing 9.04 in my introductory question... should have been 9.10) [11:27] kdelibs and kdebase-workspace both build-dep on polkit === Nightrose2 is now known as Nightrose === Tonio__ is now known as Tonio_ [12:17] shtylman_: seen bug 532533 ? seems nasty [12:17] Launchpad bug 532533 in ubiquity "Kubuntu ubiquity crashed after selecting the language" [Undecided,Confirmed] https://launchpad.net/bugs/532533 [12:24] * Tonio_ is still woried about knetworkmanager... [12:25] anyone tried to right click on the icon ? crap window with empty lines [12:25] the same than with left click, but broken [12:25] am I the only one concerned ? [12:25] Tonio_: that's agateau's fault, he's working on solutions [12:25] Riddell: oh great then :) [12:26] Tonio_: one of which was to fix the plasma applet, we need testers for that [12:26] Riddell: I've contacted upstream about the vpn connections issues [12:26] Riddell: I can test... I just built a fresh svn snapshot, the plasma applet is completly broken [12:26] Riddell: where's agateau work ? [12:26] in svn :( [12:27] Riddell: but yeah, let knm in the background and use the plasma applet would be nice [12:27] Riddell: ah :'( [12:27] he's also working on a workaround for the systray applet I believe [12:27] Riddell: well fresh svn is probably more broken than the current package reguarding to the plasma applet... [12:27] I hope one day kde will have a decent network manager gui :) [12:28] Riddell: it's been kde isue forever as far as I can remember... [12:28] seems we're not that far right now anyway :) [12:28] yes, it's painful [12:28] Riddell: I'm also concerned about a hudge issue with oxygen window theme [12:29] what's that? [12:29] Riddell: I made several tests and I can confirm it reduced the global plasma/kde FPS from 50% at least.... [12:29] Riddell: it's not very visible with a powerfull network card, but with an old chip it's a nightmare [12:29] Riddell: I benched all window themes, oxygen is the guilty [12:30] Riddell: I know changing is out of purpose, but in my case for example [12:30] I have an intel 4500 HD whose driver is know for medium performances... [12:30] this is an issue for LTSP? [12:30] the all effects get slow and not smooth at all as soon as I go with the oxygen windeco [12:31] Riddell: sure it is [12:31] Riddell: but also it is an issue with modern laptops, most of them have the same chip than me [12:31] Tonio_: what's broken about the nm plasmoid? [12:31] mgraesslin may know more about window decorations [12:31] sebas: it won't display my connections [12:31] lemme show you [12:31] sebas: and "I !!! good to see you" [12:32] sebas: so first issue with current svn is I don't have an icon in the systray,but an empty space... [12:32] Tonio_: kquitapp knetworkmanager && sleep 3 && qdbus org.kde.kded /kded loadModule networkmanagement [12:32] Tonio_: yes we have a small performance issue in Oxygen window decoration due to the animations [12:32] you need to load the kded module [12:33] mgraesslin: yeah :( [12:33] mgraesslin: I don't get how can a window decoration theme cause this issue anyway... [12:33] Tonio_: unfortunately it's driver dependent, that's why we didn't notice in beta cycle [12:33] mgraesslin: any other one works like a charm here [12:33] Tonio_: http://vizzzion.org/blog/2010/02/tokamak-iv-network-management/ and http://vizzzion.org/blog/2010/02/tokamak-finished/ for background and progress [12:33] it's the active/inactive animation [12:33] mgraesslin: yep that's true, with nvidia you won't get any trouble as I said [12:33] sebas: looking [12:34] in Aurorae it's even worse [12:34] mgraesslin: I personnaly don't use arora but rekonq so I can't tell about that [12:34] Tonio_: Aurorae is a windowdeco :-) [12:34] mgraesslin: but if all modern intel chips are concerned, which seems to be the case, that's a hudge problem for an LTS release... [12:35] mgraesslin: oh !! ;) [12:35] Tonio_: it's possible to disable the animations [12:35] that should improve the performance [12:35] it's unfortunately a hidden option [12:36] sebas: indeed that's better with your command :) [12:36] sebas: still no icon in the systray anyway, but that's really better :) [12:36] sebas: well I just lack a background on the icon it seems [12:37] sebas: very close to something very nice :) [12:38] sebas: one thing I noticed also is that it will let me create a vpn connection when I don't have any network-manager vpn plugin installed [12:39] sebas: is that to be reported as a bug, cause it'll cause a crash, or already in the todo list ? === Tonio__ is now known as Tonio_ [12:46] what's different? [12:48] you get a load of ugly stipes down it I'm assuming it's to do with vm and screen size [12:49] sebas: so knetworkmanager or plasmoid for lucid? [12:49] davmor2: got a screenshot? [12:49] plasmoid IMO [12:49] sebas: oh really, that's not what I expected [12:49] we're || that close now :) [12:49] :) [12:49] Riddell: I'll grab you one [12:49] that's good to read [12:49] Actually, I think it's a good time to start testing, I just put in the last bigger UI changes last night [12:50] reminds me ... I still have uncommitted code === Tonio__ is now known as Tonio_ [12:50] sebas: I'll perform many tests and will give you feedback [12:50] Tonio_: I haven't tested VPN yet, that's next on my list [12:50] sebas: good point is that I can test vpn pptp and openvpn too [12:51] Tonio_: are you able to make us a package to put in kubuntu experimental PPA? [12:51] sebas: from my experience, it works with pptp but crashes, and fails with openvon [12:51] Riddell: I have the package ready [12:51] Riddell: also do we want qdbusfornm ? [12:52] I think we do yes, we have it currently [12:52] gatlinAFK: things to do ... that entirely depends on if you want to do something in c++ only [12:52] Tonio_: able to throw your package into ppa:kubuntu-ppa/experimental ? [12:52] yeah but for some reas the patch won't apply, look like topgit is broken, misses apply paatches directive [12:52] Tonio_: even better, if you can give me backtraces, I can probably fix that [12:52] Riddell: I'll fix the package for the patch then and will push it [12:52] sebas: sure [12:52] I need to set up a VPN here actually [12:52] Tonio_: also, make sure to always grab the latest from svn [12:53] Riddell: no promiss to do that today anyway, I have a HUDGE work waiting for me on a broken active directory (thanks MS) [12:53] sebas: yep [12:53] cool :) [12:53] Riddell: as soon as sebas commits I'll try to get the package to work [12:53] just committed :> [12:53] and upload [12:53] sebas: perfect [12:57] Riddell: any idea how can I test debian patches with topgit ? [12:57] Riddell: http://www.davmor2.co.uk/kub-ins.png [12:57] aka like fakeroot debian/rules patches or something ? [12:57] * Tonio_ is tired of the 1 millions way to get patches to apply... and hates quilt [12:58] there are only too? :P [12:59] apachelogger: hum, not my opinion ;) [12:59] rules apply-patches or something and manually [12:59] aahhhh [12:59] my pet is super big on the screen [13:00] fakeroot debian/rules apply-patches doesn't work [13:00] then it was some other name [13:00] dh: Unknown sequence apply-patches [13:00] apachelogger: I read the tg2quilt.mk file and couldn't find anything [13:00] Tonio_: just do it manually via quilt? [13:00] apachelogger: and I can confirm the paatch won't apply during the build [13:01] apachelogger: shame on me but I don't know how to do so with debhelper 7 :) [13:01] apachelogger: any code example ? [13:01] there's no quilt.mk anymore... [13:01] ln -s debian/patches; quilt push -a [13:01] that has onthing to do with dh or cdbs :P [13:01] man quilt [13:02] oh, with above commands you shouldnt forget to rm patches again before building the package [13:02] apachelogger: I promisses myself I wouldnt't learn quilt [13:02] otherwise dpkg-source will start crying ^^ [13:02] apachelogger: have overengeneering apps :) [13:03] quilt isn't :P [13:03] you just expect too little of a patch system [13:03] which makes you shiver in light of quilt's mightyness ^^ [13:04] http://imagebin.ca/view/h7bNrx.html [13:05] Tonio_: I've no idea what topgit is [13:06] a patch queue manager [13:06] davmor2: wow, trippy [13:07] I'm assuming not right though [13:07] davmor2: don't know what causes that I'm afraid [13:07] the artwork hasn't changed since karmic there [13:07] I'll try on hw some point soon [13:08] Riddell: http://repo.or.cz/w/topgit.git?a=blob;f=README [13:08] Riddell: AFAIK topgit is like the superior implementatin of quilt, because it uses git and thus is capable of forming more relations than a rather crude quilt patch stack is able to offer [13:09] which probably makes sense if you do maintain your source in git ^^ [13:09] Riddell: and what's with the vmware/vbox logo in the bottom right of the desktop ;) [13:10] davmor2, Riddell: that problem looks graphics driver related [13:10] apachelogger: then how can I use it to apply patches ? [13:10] I have seen similar issues in alpha versions of karmic too [13:10] in vbox of course :) [13:10] apachelogger: if I build the currentl knm package it fails to apply... [13:10] apachelogger: yeah I'm assuming it's because it's on kvm [13:11] apachelogger: and I don't know how that works and have no time to learn about that ;) [13:11] Tonio_: quilt push applies a patch [13:11] quilt pop unapplies a patch [13:11] apachelogger: yeah but I don't whant to change the existing packaging :) [13:11] that is the beautiy of quilt [13:11] it is like a version control in itself [13:12] say you have one patch in your stack [13:12] so you run quilt push [13:12] that applies this very patch [13:12] but at the same time it does all sorts of tricky things so that you can go back [13:12] so once you run quilt pop [13:12] it will restore the original source again [13:12] building up on that you could run quilt push -a; quilt pop -a [13:13] to just for fun apply all patches and then unapply them again ^^ [13:14] apachelogger: yeah I know quilt is super powerfull, but well... I have enough in my mind with bzr, svn and git :) [13:14] I do not see the problem then :P [13:14] apachelogger: that said, the current knm package is supposed to apply patches according to rules, but it doesn't work [13:14] could you eventually give a little hepl so that I understand and learn a bit about it ? [13:14] Tonio_: does it not apply them or does it not apply them successfully? [13:14] not apply [13:15] apachelogger: I have this in the rules : [13:15] ifneq (,$(shell which tg)) [13:15] QUILT_PATCH_DIR := debian/patches [13:15] include /usr/share/topgit/tg2quilt.mk [13:15] endif [13:15] and it should work I have the binaries there... [13:15] but they won't apply in pbuilder... [13:15] is that a cdbs mk? [13:16] also why do you use topgit to begin with? [13:16] apachelogger: cause this is how the package is made :) [13:16] oh my, exam on introduction lecture to computer science in t-1h [13:17] apachelogger: I don't want to diff from debian on that point [13:17] ahhh [13:17] * apachelogger thinks that he can imagine who introduced that in the debian package ^^ === txwikinger2 is now known as txwikinger [13:18] fun [13:18] there is no topgit stuff at all -.- [13:18] Tonio_: is that code from above coming from debian? [13:19] apachelogger: yeah [13:19] apachelogger: that's what I said when I compained about the "million ways to patch" [13:19] well [13:20] apachelogger: so the package is broken, fine... I'll fix [13:20] dh --with kde --with-quilt --dbg-package=plasma-widget-networkmanagement-dbg $@ [13:20] just add that --with-quilt to the dh command [13:20] then you should be good [13:20] apachelogger: thanks [13:20] I have to learn a bit more about dh7 [13:21] well, that is particularly tricky because the package combines dh7 with source format 3 [13:21] and yet debian decides that using topgit is the way to go [13:21] to quote from 300: "this is madness" [13:22] apachelogger: changing the packaging standards every 6 month is bullshit... [13:22] apachelogger: leave for a couple of month and you don't know how to package... [13:22] we had debhelper 5, then 6, then cdbs, then dh7 [13:22] and the patch command, then dpatch, then patchsys, then quilt, then topgit [13:22] that is beacause cdbs was intermediate crap just because people could not wait for dh7 [13:22] that's crap [13:23] Tonio_: go tell debian [13:23] apachelogger: I love stability in the process, that helps a lot, even when the process is not the way to go :) [13:23] some people in their team constantly urge for the latest and greatest [13:23] apachelogger: linux was intermediate just because hurd was not ready :) [13:23] Tonio_: linux became useful [13:23] cdbs never was [13:24] just remember how long it took to track the target you would want to hook into if you need something very specific get done [13:24] apachelogger: yeah I know :) [13:24] also [13:24] that topgit crap there is uesless [13:24] apachelogger: I don't mind things changing, but too fast can be a problem sometimes :) [13:25] apachelogger: that's just my point [13:25] unless I misunderstand the included .mk it does not hook into the build [13:25] but adds the necessary targets so that you can do it [13:25] apachelogger: free software is not know for its goodproject management afaik :) [13:25] just like quilt.make [13:25] apachelogger: kk [13:34] I need a new laptop, but I dont know which one to buy ... any suggestions? [13:36] apachelogger: I'd say the lenovo x200 is the best whatyouget vs what you pay on the market :) [13:36] apachelogger: and it works perfect with linux [13:36] * Tonio_ owns a x301 which is just so lovelly [13:36] apachelogger: the x200 is almost the same thing, but half the price [13:36] hm [13:36] apachelogger: and can go up to 10 hours on battery... :) [13:37] goodness [13:37] 12 inches anyway... but with 1440x900 that's acceptable [13:37] apachelogger: it all depends on own much can you afford to pay... [13:38] <=2000 EUR [13:38] well the x200 is all for you then :) [13:38] x301 is way to expensive (3400 euros infrance, so I bought it in canada) [13:38] *nod* [13:39] apachelogger: maybe the price is different in your country anyway [13:39] * apachelogger gets student discount anyway and gets all whiny when something is too expensive so he gets all sorts of discounts ;) [13:39] it was only 2500 $cad in canada, which is much more acceptable :) [13:39] apachelogger: it's still expensive, but it deserves the price, really... [13:39] anyone got ideas for summer of code projects? [13:40] ideas is the least of problems [13:40] Riddell: get sflphone/kde to work [13:41] Tonio_: I really think about getting a laptop with SSD [13:41] Riddell: most of the code is there but it's not completly finished [13:41] or at least HDD + SSD [13:41] apachelogger: what's the most of the problems? [13:41] Riddell: the gnome sflphone is just a dream when it comes to sip or iax suppoer :) [13:41] apachelogger: I have 128 ssd here [13:41] Riddell: getting me drunk so that I come up with them ;) [13:41] apachelogger: it's frustrating cause I can't even see the ubuntu logo at boot, or just for 1/2 sec [13:41] apachelogger: it looks like a bug haha [13:42] ^^ [13:42] Riddell: wouldn't it be nice since we want a softphone for years ? [13:42] Riddell: I suppose we could just take some random people and do lowbudget usability analysis [13:42] Riddell: oh and samba support in kde, which I didn't have time to complete... lack of time for at least a year now... travelling too much :'( [13:43] right [13:43] though upstream has improved already [13:43] Tonio_: finishing sflphone sounds like something for the sflphone project to do [13:43] basically you need to install samba and kde-filsharing and then share setup in dolphin should work from what I saw [13:43] ehlo 4.4.1 is t3h fast, :) nice made guys [13:44] Riddell: make kubuntu boot faster, make kubuntu waste less memory, make qt-language-creator a useful tool [13:45] turn software-properties into policykitware and implement kauth support for KDE UI then make it a kcm and embedd into kpk [13:45] improve kpk UI to suite Kubuntus needs [13:46] kubuntu-software-center (in c++ pleaese) [13:46] ubuntu-one-kde (also in c++) [13:46] kpk needs a lot of love [13:46] janitor-kde [13:46] a lot meaning, a whole ui rewrite probably :( [13:46] basically port all the ubuntu stuff [13:47] seele: yeah :/ [13:48] Riddell: following up on the "make kubuntu waste less memory" -> improve performance of ubuntu/kubuntu apps by whatever means possible in python etc. [13:48] that probably also got a nice analysis touch to it, since one would first need to find out where resources get wasted to begin with ;) [13:49] create a kubuntu welcome widget thingy (since KDE seems stalled on that) [13:49] on a KDE POV: rewrite khelpcenter [13:50] I need to go [13:50] exam in 20 minutes [13:50] wish me luck! [13:51] good luck! [13:56] wow usa says opensource is against the intension of copyright protectet software [13:56] wth [13:57] Riddell: also getting tty is open when shutting down [13:59] http://www.h-online.com/open/news/item/US-copyright-lobby-claims-free-software-undermines-respect-for-intellectual-property-940921.html [13:59] retarded [14:00] Note that's a lobbying group. [14:01] Riddell: Just read the backscroll. Sounds like we need a way to disable the windeco animations on Intel. [14:01] I don't even understand what window decoration animations it includes [14:02] I'm hoping mgraesslin will weigh in and explain the hidden options thingy. [14:02] ScottK: one moment, I don't know it by heart [14:03] mgraesslin: OK. [14:03] mgraesslin: semi-related, do you know why the blue stripes disappeared? means you can't distinguish active and inactive windows any more [14:03] * ScottK goes to dig up the netbook so he can try it once it's found. [14:03] that's a change in 4.4 [14:03] we have this new outline active window thing in Oxygen [14:04] ScottK: https://bugs.kde.org/show_bug.cgi?id=226634#c7 [14:04] KDE bug 226634 in general "kwin has effects perfomance regression" [Normal,Resolved: duplicate] [14:05] the stripes are AFAIK completely removed [14:05] ScottK: sure its a lobby group but its stupid to say such things or even take it serious [14:06] and piliticans mostly tajke care of lobbyists [14:07] mgraesslin: Thanks. Looking [14:07] ghostcube: That's true of most things lobbyists say on most topics. [14:08] They get paid to sell a point of view, not be sensible. [14:08] :) [14:09] mgraesslin: I have 4.4.1. Is the promised graphical way to turn off the animations available? [14:10] * mgraesslin looks at the UI and can't find an option [14:10] Ah, read the rest of the bug. [14:10] I'm using branch from today [14:11] if it's important for you I'm sure hpereira can come up with a patch [14:15] I notice a definite difference. [14:15] Riddell: I think we need to do something about this. [14:16] Riddell: uploading to experimental :) [14:16] mgraesslin: I think we need to have a way for users to control this using a gui. Editing config files is fine for me, but not my Dad. [14:16] I agree [14:16] It's particularly a big deal for netbooks. [14:16] ScottK: well I turned it off but i don't see anything different (didn't notice any animations to start with mind) [14:17] Riddell: I don't notice any different look, but it feels faster. [14:17] it's noticable when minimizing windows with effects turned on [14:17] the animation laggs [14:18] The netbooks we're aiming at all have Intel graphics and not lots of spare CPU, so for netbook, this should generally default off. [14:18] It'd be nice to get a way to default it on for people with appropriate graphics cards. [14:19] which is difficult [14:19] yes. [14:19] we had complains for almoast all vendors [14:19] a widget at srat asking if intel or another cardd ? [14:19] wont this be possible [14:19] I don't understand why we didn't see this issue during beta cycle [14:19] *start [14:20] sebas: I just uploaded on our experimental ppa, including your latest svn commits [14:24] apachelogger: just to finish about this patching stuff, what is very very strange to me is that according to you patches shouldn't apply... [14:25] apachelogger: I mean in the current knm package... [14:25] apachelogger: so technically it should ftbfs, unless the patches are applied in the diff.gz, which seems to be the case... [14:25] sebas: What's the advantage of switching to the plasmoid for knm? [14:25] apachelogger: that's why I got it wrong on how that worked [14:26] anyway, having patches applied in diff.gz is... pretty bad I must say :) [14:26] ScottK: nicer and sexier ? :) [14:27] Tonio_: I'll take working as the first priority. [14:27] ScottK: well imho none of them could be considered "working" right now... [14:28] ScottK: but yeah, if knetworkmanager works and not the plasmoid, no reason to switch [14:31] Switching from the plasmoid to the systray application was a good move that helped a lot. I think we should have a good reason to switch back. [14:32] mgraesslin: We can wait for 4.4.2 for the GUI to control the animations I think, so no distro patch needed, but we do need it in 4.4.2. [14:34] I just pinged hpereira to ask for news on the string freeze exception [14:35] ScottK: look at it, and you'll be enlightened [14:35] :) [14:35] nevertheless I would recommend to change the default settings for the netbook [14:35] sebas: Can it connect to hidden SSIDs? [14:35] mgraesslin: I will. [14:35] not yet, it's on my list though [14:35] good point, I was going to set up an accesspoint for this [14:36] That's the only thing I've missed in the systray app recently. [14:36] I was at at meeting and needed to get wired internet to install the Gnome nm-applet. [14:36] we have preliminary support for it, just needs fixing up and testing ("just" ;-)) === Quintasan_ is now known as Quintasan [14:36] \o [14:36] sebas: I can test that when you're ready. Feel free to ping me. [14:37] cool :) [15:02] Riddell and mgraesslin: I just uploaded an update to netbook default settings to default that to off. [15:02] Now to $WORK. [15:08] Riddell: plasma-widget-networkmanagement ready for testing in the ppa, and built [15:09] Tonio_: why should the build fail if the patches are not applied at all? [15:09] Tonio_: at least with the current package there is no implication of applied patches to be able to build [15:11] OMG I thought my Dell Mini 9 was just a bit rubbish performance wise, since I've never had anything but Kubuntu and Oxygen, but turns out it really isn't :D [15:11] apachelogger: cause the patch causes one file installation that is in an *.install file :) [15:11] apachelogger: no patching -> ftbfs [15:12] Tonio_: good point [15:12] oh wellz [15:12] apachelogger: and cause the rules didn't apply, it's obvious that the patch was in the dif.gz [15:12] Tonio_: it's a source 3 package, there should be no diff.gz [15:12] it autoconverts all direct changes to quilt patches [15:12] since untaring the orig and then copy the debian folder ended up ftbfs :) [15:12] Tonio_: link to the ppa version please? [15:13] apachelogger: hum, then how does it work if there is no way to apply patches ? [15:13] maybe there is hidden magic [15:13] apachelogger: current knm package in the repo is a mystery for me :) [15:13] apachelogger: I simply don't get how that works [15:15] ohhhhh [15:15] apachelogger: any web page about this source3 thing ? [15:15] silly me [15:15] apachelogger: it means no orig anymore ? [15:15] Tonio_: source 3 also implies that quilt patches get applied upon dpkg-source -x [15:15] jussi01: https://launchpad.net/~kubuntu-ppa/+archive/experimental [15:15] which still makes the topgit crap useless, but makes you worry less about that [15:16] apachelogger: so many changes for me :) [15:16] Tonio_: it means no diff.gz anymore, but rather a tar.gz that contains the debian/ dir and no direct source changes anymore [15:16] oh in experimental - do I need anything other than the package itsself? [15:16] apachelogger: so no orig anymore, no patching system, no way to debdiff ? [15:16] Tonio_: anyhow, that makes me wonder again why you claim that the patch was not applied :P [15:16] Tonio_: with source format 3 you have: [15:16] .dsc .debian.tar.gz .orig.tar.gz [15:16] with source format 2 you have: [15:16] apachelogger: well... get a fresh svn snapshot, copy the debian folder, and try to build the package [15:17] .dsc .diff.gz .orig.tar.gz [15:17] apachelogger: it doesn't work [15:17] though with source format 3 you can also have .orig.tar.bz2 [15:17] at some points I understand the patched get applied during the source package build right ? [15:17] Tonio_: yeah, because source format 3 does the patching upon extraction [15:17] well debuild -S didn't do it [15:17] Tonio_: so you need the --with-quilt to make it work anyway [15:18] apachelogger: so how did the current plasma-widget-networkmanagement package in the archive worked out ? :) [15:18] apachelogger: I get your point, but it should have fail [15:19] ahhhhhhhhhhhhh I get it [15:19] apachelogger: it patches upon extraction [15:20] apachelogger: then something got wrong when I used pdebuild... [15:20] apachelogger: I understand... I had to patch manually before creating the source package, so that all changes would go in a patch [15:20] and would be applied [15:21] apachelogger: it's very very different from what I know... [15:21] apachelogger: that will be a real pleasure for sysadmins who maintain debian in production environment and have to backport :) [15:21] apachelogger: retro-compatibility doesn't seem to be important for debian :) [15:22] I know quite some devs who gave up on debian because of that... I can understand them :) [15:23] apachelogger: do you have a good link for debian source 3 format and so on ? [15:24] davmor2: yep... got a bug? [15:25] http://launchpad.net/bugs/532533 [15:25] Launchpad bug 532533 in ubiquity "Kubuntu ubiquity crashed after selecting the language" [Undecided,Confirmed] [15:42] ScottK, Riddell: hum the plasmoid seems to work pretty decently here :) === Tonio__ is now known as Tonio_ [15:43] Riddell: awesome, I will apply for sponsorship though I think it will end up the same way as last year (in queue) :) [15:50] davmor2: thanks...will look into that [15:52] shtylman_: the ugliness is down to kvm we think, but I've not had time to test on hw yet [16:00] Riddell: Are you covering the release team meeting? [16:01] ScottK: yes, company always welcome of course [16:01] Riddell: I'd really like to get some action assigned on Bug #531697. [16:01] Launchpad bug 531697 in gcc-4.4 "ICE building qt4-x11 in IA64" [High,New] https://launchpad.net/bugs/531697 [16:02] Tonio_: hmm, it's connected to my network but it doesn't list any connections in the left hand pane [16:03] Riddell: kquitapp knetworkmanager && sleep 3 && qdbus org.kde.kded /kded loadModule networkmanagement [16:03] Riddell: it has do be done by hand for now... [16:03] Riddell: you have to use the this kded module instead of knetworkmanager [16:04] Riddell: http://toniox.org/temp/knm-plasmoid.png [16:04] Riddell: seems to work pretty well here, except that I can't disconnect anything... I told sebas about that issue [16:06] Any ideas why touching the power button on my netbook now just turns it off instead of bringing up the powerdevel dialogue? [16:07] Tonio_: I take it VPN works? :) [16:07] sebas: pptp, yes [16:07] sebas: vpn, not tested atm [16:07] Cool. :) [16:07] sebas: but no way to disconnect, is that a known issue ? [16:07] nope, that's why I needed it tested :) [16:08] I'll think about something smart for that [16:08] sebas: okay so here is myfeedback then :) [16:08] 1/ it connects [16:08] 2/ the should be a better way to get the connection status (bold font isn't enough), especially to know what happens, like telling me "running dhcp" on hover for example [16:09] 3/ there should be a way to get a few infos on the connection, like the IP address of something, on hover, don't you think ? [16:09] 4/ and also a way to disconnect :) [16:09] sebas: I had to go on http://www.whatismyipaddress.com to figure out I was connected :) [16:10] 2/ that's in the interfaceitem [16:10] so maybe a little more infos would be nice... [16:10] 3/ is being worked on, that will show up when you click on an interface [16:10] Tonio_: we want to keep it accessible for "normal users" as well [16:10] try to use it for a while, it shows more than you think [16:10] sebas: sure [16:11] sebas: about 3/ the vpn isn't a physical interface, that's the deal [16:11] \4 is only a problem for VPN, no? [16:11] sebas: well yes, cause I can disconnect from an interface, but not a network [16:11] Probably we should make VPN a "virtual interface", in the UI sense [16:11] sebas: yes, this way it would be nice [16:11] right, ok, so that's on my radar now as well :) [16:12] sebas: the point is that the disconnection thing is upon an interface, not a connection [16:12] I know :> [16:12] sebas: do you ? ^-^ [16:12] hehe, I think I do :P [16:13] well technically there is no reason to check for a virtual interface, there is a physical one [16:13] gotta clean up a bit here, K is coming home soon and we're having friends over tonight [16:13] ppp0, tap0, tun0... [16:13] networkmanager hacking on the couch == cables + usb interfaces and stuff in the living room [16:13] there should be a way to list them through network-manager right ? [16:13] I'll have a look in how far Solid shows them [16:13] kk [16:14] sebas: ping me if tests needed, I'll package in the hour and test it for you, and will test openvpn soon too [16:15] anyone ever tried using krunner in gnome? === plopplop is now known as Tonio_ [16:15] ll [16:15] oops [16:16] sebas: hum, it looks like clicking on an established connection disconnects too... at least works for the wireless, so there are 2 ways to disconnect right ? [16:16] from the interface and the connection [16:30] arm images have been building for the last 40 minutes, so that's promising [16:31] Riddell: No plasma-netbook autostart after the last updates. Any idea what that might be? [16:31] ScottK: update from what to what? [16:32] Updated my Lucid system an hour ago for the first time in a couple of days. [16:32] new kdebase-runtime? [16:32] Yes [16:32] New metapackage (kubuntu-netboo) too. [16:32] we need to think more about that, currently it's an autostart file set to false which ubiquity sets to true on install [16:33] OK. [16:33] but that won't work because the upgrade will just overwrite it as you found out [16:33] I think it needs a script to decide if netbook or desktop should run [16:33] but not run both [16:33] I had hoped that upstream would fix it sensibly but they haven't [16:34] maybe it should be an xsession [16:35] Riddell: are you working on soprano 2.4.1? [16:36] Lex79: no, I think it's for the taking [16:37] ok, I will do [17:10] apachelogger: ping [17:12] apachelogger: nvm :P [17:13] * Trouble pokes apachelogger [17:13] Oh actually don't worry :D [17:16] sebas: I was just thinking... is it fine to consider disconnection should be done upon an interface or upon a connection ? [17:17] what I want to kill is a connection, not an interface, right ? [17:17] sebas: I can get it, but since you want it usable by normal users... it sounds logic to me to disconnect from within a connection, not from an interface... [17:17] sebas: my 0.02$ [17:18] sebas: seele's opinion would certainly mean a lot more :) I'm not a usability expert at all [17:20] hiya Tonio_ [17:21] hey maco2 :) long time no see [17:29] apachelogger: best advice today was - use boost.random since rand() sucks [17:33] Riddell: soprano https://launchpad.net/~alessandro-ghersi/+archive/staging/+packages [17:36] Someone packaging http://michalm.wordpress.com/2010/03/05/k3b-2-0-release-candidate/ === yofel_ is now known as yofel [18:16] Tonio_: source format 3 is explained in the debian wiki IIRC [18:23] http://wiki.debian.org/Projects/DebSrc3.0 [18:26] hrmhrm, k3b release candidate [18:26] awsum [18:26] lol: http://michalm.wordpress.com/2010/03/05/k3b-2-0-release-candidate/#comments [18:27] he's being a bit hard on Michal, I think. [18:27] hmm just a simple question, on a dice, you have wall or sides (where numbers are)? [18:28] JontheEchidna: I just uploaded to bzr [18:28] *k3b I mean [18:28] Lex79: need a sponsor? [18:28] wait, ain't that guy from Poland? [18:28] yes [18:28] Quintasan: Polish are the only people I've seen to use that funny "l" :P [18:28] ł [18:28] :P [18:29] polish for the win! [18:29] Good grief, I didn't know that main k3b dev is from Poland :O [18:29] * Quintasan wrote a dice roll simulator using boost::random [18:29] :P [18:30] Quintasan: probably because he's the new main dev [18:30] JontheEchidna: if you have the power there is also soprano in my ppa [18:30] no soprano powars :( [18:30] :( [18:34] is the brightness osd working for anyone in lucid? [18:34] debfx: it ain't working [18:34] See Todo [18:35] agateu will know something about it :P [18:43] apachelogger: http://github.com/Quintasan/coding/blob/01071142841fe35dfa4927effa6107cbd9c752c6/gen.cpp <-- -the best app so far :P [18:45] maybe not best but most useful :P [18:46] it should go to lucid archive, but we are in feauture freeze [18:46] :D [18:46] :) [18:50] * apachelogger falls off chair [18:57] Lex79: K3b uploaded, thanks [19:05] k [19:10] Quintasan: I think I know why it doesn't work anymore [19:17] the new hal sets brightness_in_hardware to true for most (?) laptops [19:19] previously it was only set to true for laptops that emit a brightness change key event on every brightness change [19:22] I can modify the patch but then on those laptops the osd is shown even when the brightness is changed by an application [19:41] Quintasan: what does it do? [19:42] I can't read the ... whatever language that is :) [19:42] I wanna say it does a dice roll? ... [19:42] but who knows... [19:44] ... oh... I see... there is an english description... ok... so now my question is. .. why?! [19:45] shtylman_: why what? :O [19:45] why make that? [19:46] a) boredome b) I play DnD with my friend and there is a spell that requires you to roll 346 times (and many other with 100 rolls :P) [19:46] ... *sigh* ... [19:47] blame apachelogger for it, he said I should get some hands-on experience [19:47] :P [19:48] heh [19:49] though that app ported to windows fails [19:49] I wonder if it's becuase of time(NULL) [19:58] so if kdm crashes should i leave a crash report someplace or will ubuntu-bug take care of that? [19:58] * ghostcube does a mount /Quintasan -t dev /dev/Null && rm -rf /boot/brain/content.list [19:58] :D [19:59] rm: file content.list not found [19:59] :O [19:59] rofl [20:00] seriously, what were you expecting, I have been learning c++ for a month only [20:00] nah i can hardly read it :) i cant code c++ [20:01] was just joking :) [20:01] I know, I can take jokes, really :P [20:01] :) [20:03] jjesse: /var/crash ? [20:03] hrmm will have to look === gatlin is now known as gatlinAFK [20:32] so by default when I run debuild it puts the generated files in one directory up... can I cahnge that? [20:37] hey shtylman_ [20:40] txwikinger: howdy [20:41] shtylman_: Are you allowed to say that in NY? :D [20:41] txwikinger: hahaha :) [20:42] JontheEchidna: have you read? http://labs.trolltech.com/blogs/2010/03/03/qtwebkit-releases/ this means we'll can split Phonon from Qt in easy mode :) [20:45] * shtylman_ shtylman can't wait for kde to move to git... its soo close... I can almost taste it.. [20:49] Has it ever ben possible at all to run another instance of kde from inside a konsole? An user in #k is claiming 3.x had this but I never heard or saw of it [20:50] genii: I do something similar with xephyr [20:50] its opens another xsession or whatnot [20:50] and from there you can login with kdm again into another kde instance [20:51] shtylman_: Thanks [21:13] Time to get rid og that stupid Twitter and Buzz [21:13] of* [21:14] Lex79: we can split phonon out but not until after lucid, so we still need an upgraded patch to the new phonon [21:15] Is 4.4.1 64bit for Lucid OK yet? [21:19] shtylman_: I dont think you can change the directory where debuild stores its stuff [21:20] shtylman_: kde moving to git depends on there being somewhere to host the git repository [21:20] * apachelogger can host it on his netbook ^^ [21:21] I'm not sure you get 60GB SD cards yet [21:22] * apachelogger got a HDD with 130 GB [21:30] okay, Buzz deactivated [21:30] now time for forums [21:32] Riddell: ok [21:32] * apachelogger thinks about deriving the dbginstaller from kmessagebox rather than kdialog [21:36] apachelogger: why? [21:36] because kmessagebox really does everything you need anyway ^^ [21:37] well except for progress indication [21:37] maybe derive from qwidget and call KMessageBox::foo from there [21:38] i always use kdialog as it is what i am used to...was hoping you might shed some light on why kmessagebox would/might be better than kdialog [21:39] it's not, it just already implements the necessary stuff in convenience functions [21:41] as I see it the installer only needs to lookup the dbg packages, thus indicating progress meanwhile ... if not all packages could be found, ask the user what to do and list the files for which there were no debug packages found ... and finally ask the user whether the found packages shoud be installed, again listing the packages [21:41] so we have custom qwidget for progress indication (possibly add cancel button) + KMessageBox::warningYesNoList + KMessageBox::questionYesNoList [21:44] nixternal: also as I came to realize yesterday, reusing the same dialog for all 3 things is a) dirty b) rather unpleasent because each default button click emits the finished signal, thus ending the dialog [21:44] so for proper design you would have to create 3 classes derived from kdialog [21:44] yeah, that is a bit dirty [21:45] you couldn't just do one class with 3 different functions? [21:45] you could [21:45] but them you also need to reimplement the slots of each used button to prevent them from emitting finished upon click [21:46] otherwise your dialog destructs midway through the process because the user agreed to install the packages :) [21:46] hrmm, i see what you are getting at now [21:47] oh why [21:47] http://api.kde.org/4.x-api/kdelibs-apidocs/kdeui/html/classKProgressDialog.html [21:47] <3 KDE [21:48] http://api.kde.org/4.x-api/kdelibs-apidocs/kdeui/html/classKDialog.html - heh, you didn't look at all of the inheritance with kdialog? [21:48] not all of them [21:48] that is one of my bookmarks right there :) [21:49] also I did want to opt for the 3-functions but one dialog approach you suggested ;) [21:49] and actually some months ago in another project I also wanted to do that (I suppose apturl it was) and I came to the same realization [21:49] one-dialog-per-use [21:50] though in this particular case one could also opt for a KPageDialog I suppose [21:50] yeah, i have done the one-dialog-per-use as well [21:52] mhhh [21:52] much better with kprogressdialog [21:52] * apachelogger drops stupid ui file [21:52] did I already mention that I very much dislike ui files? [21:53] maybe it's just me, but it feels like they always get in my way [21:55] Riddell: yea... im following the mailing list on it... I got the rsync of the svn repo to play around with it ... hopefully some conclusion will be reached soon :/ [21:58] soo question... lets say I have a script which appends to the path... when it is run [21:58] if I run it again... it will append the same thing to the path.. [21:58] is there a way to prevent that? === rickspencer3_ is now known as rickspencer3 [22:00] * txwikinger wonders why the git repo cannot be hosted where the svn repo is hosted [22:00] txwikinger: it can... they would just need to install a copy of the gitorious website [22:01] and then keep that up to date and whatnot [22:01] which is just more maintenance work [22:01] well.. not really [22:01] you can have a cron job running the sync [22:01] or drop svn [22:02] txwikinger: they can't drop svn cause it will still be used for translations for the time being [22:02] ah [22:04] * apachelogger cant help thinking that the dbginstaller code is quite the mess [22:05] maybe I should not have done a straight port from bash -.- [22:54] * Quintasan goes to bed [22:54] Good night [23:23] yay ARM built http://cdimage.ubuntu.com/kubuntu-netbook/ports/daily-live/20100305/ [23:23] nice [23:43] Riddell: have any of them actually been tested on arm? or in qemu? [23:48] shtylman_: I saw it running on an arm board last month [23:48] so it does to something [23:49] cool [23:50] so does anybody know what they did to get around the kdebindings compile issue [23:52] tuxdavis: I disabled smoke from compiling [23:52] tuxdavis: I also removed the kdeedu package from the image [23:52] getting a proper fix for those segfaulting compiles would obviously be much better [23:54] well as a temporary solution to allow testing of the image this works great [23:54] yes [23:54] tuxdavis: are you able to test the images? [23:54] Besides from what I gathered during the debug process of kdebindings, the smoke part is used for developing applications using the kde api [23:55] I'm downloading it right now [23:55] great [23:55] it's used for making the ruby and c# bindings [23:55] I had to disable those from compiling too [23:55] another thing I'm trying as a possible FIX not workaround is trying the latest svn version of gcc, since there were a lot of fixed bugs on the arm processor [23:56] ah