=== rdieter_afk is now known as rdieter [02:00] someone give our logo something to eat [02:00] it is too skinny to look good [02:10] apachelogger: What logo? [02:10] the circle thing [02:11] Too skinny? [02:11] http://wstaw.org/m/2012/03/16/plasma-desktopuo1981.png [02:12] ... I still have the other blue one... [02:12] The one from Lucid when we changed logos... how does one get that one? [02:12] cuz you have not been installing my superior new theme -.- [02:12] LOL [02:12] wiki history I suppose [02:12] we definitely ahve it in the kds history [02:12] LiteDM? [02:12] that not be the problem though [02:13] master sheytan made a version in gimp that made sense [02:13] now he redid the thing in inkscape and it looks like the ungodly child of zoidberg and bender [02:13] LOL [02:14] http://wstaw.org/m/2012/03/16/plasma-desktopqW1981.png [02:14] apachelogger: You going to be able to make it to Oakland? [02:14] Looks better [02:14] unfortunately not [02:14] ENOTIME [02:14] bugger. [02:14] yeah, that latter one is with the gimp version [02:14] the former is the skinny thing [02:14] I have to call tomorrow and get my flight stuff figured out. [02:15] I wanna see the zoidberg/bender one [02:15] http://wstaw.org/m/2012/03/16/plasma-desktopqL1981.png [02:16] gimp apparently makes things bigger [02:16] BUWAHAHAHA [02:16] we should tell the world [02:16] "edit yoru nakkid pics with gimp" [02:16] I wonder who all will be at UDS [02:16] oi [02:16] that could become their new slogan [02:16] gimp - we make it look big [02:16] Darkwing: mark hopefully :P [02:17] HAH! Mark. Yeah... [02:17] I wonder if he will try and pander to me again about what a great job Kubuntu is doing and what a valuable part of Ubuntu it is like he did at UDS-P [02:18] how the flip did gimp make that happen [02:18] * apachelogger scales like a mad man in inkscape and doesn't get it to match the gimp version Oo [02:19] I don't use Gimp anymore. [02:19] ur fault then, only gimp makes things bigger [02:19] I use Inkscape like people use Gimp [02:19] gimp - size matters [02:20] It's about pixelated images vs vector images. [02:20] what good is a vector if it is a skinny one :P [02:20] oh, I know how gimp did it [02:21] the drop shadow exceeds the actual dimension of the image [02:21] Ungroup to base then you can alter it in Inkscape [02:21] now since we have a drop shaddow that looks like it goes in the drop shaddow in that case actually is like the stroke of the gear parts [02:21] hence if one places the original logo parts inside the gimp ones they pretty much match where the drop ends [02:22] Darkwing: which reminds me that sheytan did mask on all 3 parts, so ungroup actually breaks evyerthing :P [02:22] good thing yours truely knows how to replicate what sheytan did by reading the xml ^^ [02:22] code always wins [02:24] http://wstaw.org/m/2012/03/16/plasma-desktoprz1981.png [02:24] I should totally become an artist [02:25] however sheytan will have to redo the thing, I am not quite sure how he clipped the edge [02:26] and now I crashed inkscape [02:26] ^^ [02:26] <3 [02:42] apachelogger: Why not just take the existing SVG and recolor it yourself? [02:43] you realize that is more than recoloring? :P [02:44] Of course I do. ;) [02:44] if you tell me how [02:45] You are applying the mask to do what part? [02:45] the worse part of loosing everything (Including backups) is you forget how much stuff you had that you are missing... like SSH keys [02:45] Sooo true [02:46] ^^ [02:46] GPG keys, old passwords, everything. [02:46] Yup. [02:46] Had everything stolen and backups went plooey due to Dell recovery [02:46] <3 dell [02:46] Sooooooo, I'm starting from ground zero. [02:47] they put rubbish thermal jizz in my laptop [02:47] And cloud computing, here I come. [02:47] Darkwing: the engraved part is masked if I read the svg right [02:47] i.e. it is a stroke that is blured and then masked so that only the inner blur is visible [02:48] Thankfully I've ben using IMAP for some time now. [02:48] apachelogger: Yeah, there are built in filters in Inkscape [02:49] so which one do I use then? [03:00] ehm [03:00] silly me [03:00] s/mask/clip/g [03:00] apachelogger: You did something wrong... Try s/you/me/ or tell me "help sed" [03:00] FYU [03:00] FYI even [03:10] lol [03:10] Darkwing: it appears the paths in our original logo are not exactly precise either [03:10] No? [03:11] I'll have to install inkscape and take a look. [03:11] stroking them yields different parts Oo [03:11] I just did a few hundread images for the Southern California Linux Expo === JackyAlcine_ is now known as jalcine [03:12] or maybe inkscape is being incorrect [03:12] http://wstaw.org/m/2012/03/16/plasma-desktopZS1981.png [03:13] Darkwing: how would you combine object and stroke to get one object? [03:14] * ScottK struggles not to imagine apachelogger stroking objects. [03:14] good thing I put it that way ;) [03:16] Darkwing: not finding a proper option I duplicated the object, then converted the stroke to a path and then applied path combine on original object and the stroke path [03:17] the only part for which this resulted in a path with sane nodes is the left [03:17] apachelogger: Lemme reinstall it so I can see what needs to be done to get that. [03:17] the other two have some weird mixture of object and stroke nodes [03:18] oh oho [03:18] nvm [03:18] curious [03:18] my working copy has weird paths, the orginal combined version has alright ones [03:18] Oo [03:19] oh no, nvm again [03:19] the original is not combined ^^ [03:19] now I broke plasma by pasting a snapshot [03:19] sometimes I wonder what life would be like if I were using usable software [03:20] hehehe [03:24] Oo [03:24] simplyfing a path in inkscape actually introduces more nodes! [03:24] It [03:25] It's actually quite nice when you get used to it. [03:25] simplyfing mean complify? [03:25] I am not sure about that, why not call it complify? :P [03:27] hehehe [03:36] I wonder how many ppl will be at UDS from Kubuntu. [03:38] Darkwing: ask on the ml? [03:38] Darkwing: http://wstaw.org/m/2012/03/16/plasma-desktopGM9738.png [03:38] that is closer to the gimp version, no? [03:39] The one on the right is your new one? [03:40] yes [03:40] I actually like it better. [03:41] * apachelogger wonders whether it isn't still too skinny [03:46] and crash again [03:47] why this is fun [04:01] http://wstaw.org/m/2012/03/16/plasma-desktopZS9738.png [04:01] getting there? === jalcine_ is now known as JackyAlcine === JackyAlcine is now known as jalcine [04:05] I'm getting a DNS error. [04:07] Darkwing: with http://wstaw.org/? [04:08] dantti: http://wstaw.org/m/2012/03/16/plasma-desktopdW9738.png that's what you had in mind, no? [04:08] instead of dots [04:08] oh kwin [04:08] -.- [04:08] http://wstaw.org/m/2012/03/16/plasma-desktoplN9738.png [04:08] that one === jalcine is now known as JackyAlcine [04:11] hey all... I'm Steve Riley, a mod on KubuntuForums [04:12] Welcome [04:12] apachelogger: closer [04:12] need to report a curious bug about system sounds, but since it isn't related to one specific package, i need a suggestion on the best way to file [04:12] SteveRiley: Kubuntu sounds? [04:12] yes. see http://www.kubuntuforums.net/showthread.php?57819-Notification-Sounds-Disappeared-after-updates-to-KDE-4-8-1 [04:13] Darkwing broke it [04:14] SteveRiley: Give me a moment. [04:14] okie [04:14] apachelogger: Am I the new nixternal? [04:14] yes [04:14] apachelogger: I AM MADE IT! [04:14] BRB [04:15] SteveRiley: I'll be right back, just a quick smoke and I'll see how the best way to move forward is. [04:15] https://bugs.kde.org/show_bug.cgi?id=285028 [04:15] KDE bug 285028 in kdecore "Qt 4.8 QUrl.toLocalFile behavior change, impacts KUrl (and at least knotify)" [Normal,New: ] [04:16] Darkwing: best way to move forward .... talk to dfaure and rdieter find out if that is going to be fixed for .2 if not ask them what the educated nixternal replacement is meant to do to fix this [04:17] also write a mail to the list to make sure people are aware of the issue [04:17] oh, interesting. i hadn't even thought to check kde bugs, since i figured this might have been a packaging problem [04:17] if it were our fault it would be fixed already ;) [04:17] lol. [04:18] i can believe that! [04:18] ^^ [04:19] i was chatting with claydoh the other day about stepping up my involvement with kubuntu besides moderating the forum. what needs done that a non-developer can work on? === JackyAlcine is now known as jalcine === jalcine is now known as JackyAlcine === JackyAlcine is now known as jalcine [04:21] SteveRiley: Something you might want to train the kubuntuforums people is to check and use bugs.kde.org [04:21] apachelogger: righto [04:21] Darkwing: http://wstaw.org/m/2012/03/16/plasma-desktoprW9738.png [04:21] I figure it looks better because it is so pixly :P [04:21] apachelogger: Also, FYI, I'll be covering it at UDS but, this will be the last cycle of the kubuntu-docs package. [04:22] I like the one on the right :D [04:23] oh yes, that version seems perfect now [04:24] now the dots look weird ^^ [04:24] stupid dots [04:24] http://wstaw.org/m/2012/03/16/plasma-desktopD10533.png [04:24] Darkwing: i think we have an ancient post somewhere in the forum for how to report bugs. i'll search for and revive it. [04:25] SteveRiley: Thanks. Because most of the bugs "should" be handled upstream at bugs.kde.org [04:28] apachelogger: I'm getting that dumb DNS error [04:33] clearly your intartubes are broken [04:35] http://i.imgur.com/Zq4YL.jpg [04:36] proportions are all wrong [04:36] but I think that approach needs major artistic vision first [04:36] :D [04:37] I just looked at my clock... apachelogger, do you sleep? [04:38] no [04:38] maybe I am sleep drawing? :O [04:38] omg [04:38] I should see a doctor [04:39] http://i.imgur.com/bi5v1.jpg [04:40] what I find entertaining is that the out shadow makes the gear parts look bigger than the one with the in shadow [04:40] Getting there... the lower right seems to be off. [04:40] the part area in both cases is actually the same [04:40] Darkwing: yeah, that is what I am musing about ;) [04:40] Oh I know... [04:40] :P:P I'm here to annoy ;) [04:40] as an engineer I can tell you that it is perfect :P [04:41] it just happens to look odd to you [04:41] bugs.eyes.org [04:41] hehehe [04:43] apachelogger: Where can I find version information for Python, gcc and glibc? [04:44] apt? [04:45] DOH [04:45] * Darkwing slaps self [04:49] I'm glad shadeslayer asked if I'd do another arm build of digikam _if_ the boxen were free because I had something I wanted t build first ... [04:50] dagnabbit. [04:50] what is the glibc for? [04:51] man glibc [04:52] http://i.imgur.com/N8wpe.png [04:52] that be the version for broken framebuffer systems [04:52] Why can't man display version numbers?? [04:53] I'm gonna go file a bug somewhere for that. :P:P [04:54] Darkwing: man pages can have version numbers in them and if they are provided, they are displayed. [04:55] ScottK: ahhh, right. I should do this while I'm not half asleep :P:P [04:55] ScottK: You coming to Oakland? [04:55] Dude, it's barely even dark where you are, how can you be tired? [04:55] No. [04:57] It's 10p and my kinda were up all night last night with stomach flu [04:57] dark? what? [04:58] it's just getting light Oo [04:58] LOL [04:58] It's what... almost 5a there apachelogger? [04:58] so [04:58] can you two please add the experimental ppa on a precise install and upgrade kubuntu-default-settings and plymouth-something and test the splashes [04:59] Darkwing: s/5/6/g [04:59] So you are UTC+1? [05:00] in winter, yes [05:00] Yeah... [05:01] u know, the splash for broken systems actually has nice minimalism to it [05:02] * apachelogger ponders making that his default [05:05] [kubuntu-default-settings] Harald Sitter * apachelogger@ubuntu.com-20120316050523-1vk3dnnuqdgj02s1 * lib/plymouth/themes/kubuntu-logo/ (kubuntu_logo.svgz kubuntu_logo.png kubuntu_logo16.png) new logo for splash + srcs [05:05] Darkwing, ScottK: installed the splashes yet? [05:06] do we even have ksplash-qml on the CD? [05:06] apachelogger: Not yet... working on something else... [05:06] :( === jalcine is now known as webjadmin_ [05:06] apachelogger: No, but I did conclude my failed experiment and kick of shadeslayer's digkam build. === webjadmin_ is now known as jalcine [05:06] scary [05:06] oh shadeslayer is being inverviewed today [05:06] hrrrhrr [05:06] :D [05:07] yeah... in like 10 hours IIRC [05:07] I might have to revise my final question a bit [05:12] [kubuntu-default-settings] Harald Sitter * apachelogger@ubuntu.com-20120316051150-31goezv31fgbp38k * lib/plymouth/themes/kubuntu-logo/ (8 files) pngcrush++ [05:16] [kubuntu-default-settings] Harald Sitter * apachelogger@ubuntu.com-20120316051601-ea0v6shxckyzx0zm * usr-share-kde4/apps/ksplash/Themes/Kubuntu/main.qml reintroduce fallback wallpaper loading [05:24] [kubuntu-default-settings] Harald Sitter * apachelogger@ubuntu.com-20120316052414-y5f5cstr00uup31a * usr-share-kde4/apps/ksplash/Themes/Kubuntu/ (Preview.png Theme.rc) update preview and description [05:32] Hey there, I was chatting with Darkwing and he disconnected. I'm off to bed, and am hoping one of you can tell him thank you VERY much for all the help for me. [05:32] [kubuntu-default-settings] Harald Sitter * apachelogger@ubuntu.com-20120316053142-331xh8csh02i8kvn * lib/plymouth/themes/kubuntu-logo/kubuntu-logo.script add kate modeline as it fails to highlight plymouth language (Which is like javascript, so we simply make it highlight that) [05:33] littlegirl: have a good night [05:33] apachelogger: Thanks - you, too, and please give him the message for me. (: [05:33] certainly [05:36] Darkwing: Riddell: ScottK: I pushed the lastet version of the splash stuff to experimental this time also for oneiric https://code.launchpad.net/~kubuntu-packagers/+recipe/kubuntu-default-settings-plymouth [05:36] pretty plz be trying it [05:40] * jussi waves [06:06] jussi: yo === fenris is now known as Guest98977 [06:19] apachelogger: hello! [06:19] apachelogger: know any minions that want a packaging project? [06:19] apachelogger: cause this is cool :) http://www.openoctave.org/oomidi_2011 [06:21] its kinda packaged for kxstudio, but no idea of the quality: https://launchpad.net/~kxstudio-team/+archive/openoctave [07:36] apachelogger: dude, my GRUB looks horrible [07:36] I can't read shit [07:36] Riddell: pls, can you merge zanshin 0.2.1-1 from Debian [07:38] * shadeslayer goes back to reading the Debian Policy manual [07:46] Qt 5 Alpha release this month btw [07:48] I can't wait that long, I'm building it from git now :) [07:48] :D [08:47] Riddell: plasma-active image works more or less [09:43] bulldog98: :) [09:43] bulldog98: what's the more or less? [09:43] fabo: it needs a merge does it? [09:43] Riddell: it’s slow [09:43] the ui nearly does not react [09:45] bulldog98: on what hardware? [09:45] I found it ran decently fast [09:45] maybe it's a compositing issue? [10:05] Riddell: zanshin needs a sync. do you need a sync request? [10:06] fabo: yes a bug would help (the script needs it) [10:07] release team meeting update: what was done this week? [10:07] kubuntu active was activated === fenris is now known as Guest77154 [10:53] apachelogger: no [10:53] apachelogger: I tought coloring the carved part with a white gradient [10:53] not raising the carved part [10:53] I think it would look better === Riddell 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 | wiki.kubuntu.org/Kubuntu/UDSQSeries | Next Meeting: 16th March 3.30 PM UTC | Milestoned beasties http://goo.gl/9iyUD [11:15] http://goo.gl/9iyUD is bugs milestoned for beta 2 and tagged kubuntu [11:15] feel free to add to it [11:17] landing the kdm config changes before beta 2 freeze would be good [11:17] yofel: ^ [11:30] Riddell: could you please upload https://launchpad.net/~rohangarg/+archive/experimental/+files/digikam_2.5.0-1ubuntu2~ppa3.dsc ? [11:32] debfx: what changes are those? [11:33] shadeslayer: what's new? [11:34] Riddell: fixes arm/armhf builds and adds dh-apparmor to build depends [11:35] fixes arm/armhf builds by not pulling in opengl libs [11:36] lovelty [11:36] lovely [11:36] :) === amichair is now known as amichair_ [11:48] Riddell: restarting the x server after logout and something about input grabbing [11:52] uhh [11:52] Riddell: heh, I also added a patch kubuntu_fix_arm_build.patch that should go into the changelog [11:53] argh, it's already there, just ignore me [11:54] * shadeslayer needs coffee [11:55] Hey all [12:08] so that wasn't me tempering with my grub configs? [12:08] shadeslayer: ^ [12:09] dantti: what color? there is no color [12:10] apachelogger: I know, but my proposal would be to make the carved part lighter not raised [12:11] pretty much like painting with white color but has to be a gradient white to look nicer [12:12] does not compute [12:12] take the svg and make it happen :P [12:15] shadeslayer: Almost. [12:15] dh_install: kipi-plugins-common missing files (usr/share/icons/hicolor/*/actions/ogl.png), aborting [12:15] (re digikam) [12:16] apachelogger: huh? GRUB is utter shit right now, I can't read a thing [12:17] ScottK: ignore that, kipi-plugins-common is arch all and pbuilders build arch all packages by default on every arch [12:17] I've talked about this before, and the package will build just fine on i386 [12:17] shadeslayer: OK. Then I guess you win. Are we ready to upload then? [12:17] yep [12:17] https://launchpad.net/~rohangarg/+archive/experimental/+files/digikam_2.5.0-1ubuntu2~ppa3.dsc [12:17] that one [12:18] OK [12:18] ppa2 had a extra line in the debian rules file that was not required [12:18] Riddell didn't do it already? [12:18] I don't know [12:19] Doesn't appear so. [12:19] apachelogger: the current grub theme is all sorts of crap [12:19] apachelogger: I can barely read what's written [12:20] what is the current gurb theme? [12:20] lemme see [12:20] what you be whining about? [12:21] the colors [12:21] they hurt [12:21] make a picture [12:21] I'll take a pic and show, give me a minute to reboot and everything [12:21] yeah [12:21] less than a minute cuz I should have been leaving 10min ago [12:22] * Don't install firefox-kde-support anymore. The KDE integration patches [12:22] have been dropped from the firefox package. [12:22] the ever degrading firefox ^^ [12:26] shadeslayer: Uploaded. Thank you for your contribution to Kubuntu. [12:31] apachelogger: this : http://wstaw.org/m/2012/03/16/IMG_20120316_175309.jpg [12:32] ScottK: thanks! [12:34] You're welcome. [12:36] this is interesting [12:36] it is iinteresting because one can set the background but not the font [12:36] shadeslayer: will have a look [12:37] yay [12:37] we prolly should just make it blackkkkk [12:37] I agree [12:37] people who see that have failed in their quest for awesomeness anyway [12:37] -> uni [12:37] cya [12:37] don't be back for the next 7 hours [12:46] shadeslayer: ok [12:46] apachelogger: :P [12:46] shadeslayer: btw, does you like the splash? [12:46] apachelogger: yes, it's awesome [12:46] shadeslayer: also does you see the kubuntu splashens in the ksplash kcm? [12:46] now if only you could make it spin .... [12:46] because I fear we do not ship ksplashqml by default [12:46] which needs fix0ring [12:46] shadeslayer: define spin [12:46] apachelogger: I'm in neon ...will test once in stable [12:47] mind me, from what I have seen the stable is just like neon :P [12:47] plasma freezing on me, inkscape crashing on me, amarok crashing on exit in collection dtor, also phonon being broken and apparently randomly stopping amarok's playback [12:48] apachelogger: http://www.youtube.com/watch?v=qaSg2rRj4HQ < Kind of like that [12:49] see the ubuntu logo spinning/rotating at the end of the video ? Like that [12:50] this is pretty cool too http://www.youtube.com/watch?v=TrvlgxqKVcE [12:52] looks silly [12:53] looks even more silly with our gear things [12:53] if they were an actual cog wheel that would work, but it aint [12:54] hm, sheytan could come up with something :P [12:55] one can come up with many things, but that doesn't mean they all make sense [12:55] \o/ http://quickgit.kde.org/?p=digikam.git&a=commit&h=e0a2fcfb9f0bc75084e4bf01905524fd5e29edd6 [12:56] Finally fixed the esata to sata connection. There was small piece of plastic stuck in the mobo sata port that was preventing proper contact.trimmed it off with my trusty swiss army knife and Voila! Now I don'y have to suffer the USB 2 speeds when transferring files [12:56] shadeslayer: you realize that because you were doing arm stuff you will now get a lot of arm questions? [12:56] oh noes [12:56] I know only bits and pieces [12:57] more than rbelem then :P [12:57] ... [12:57] yay, our imaginary digikam ARM users can now run the latest version :P [12:58] :D [12:58] shadeslayer: it is either ARM or a lot about binary signatures [12:58] I'll toss a coin I guess [12:58] debfx: I am not so sure they are imaginary [12:58] apachelogger: ARM it is then [12:59] apachelogger: they are not even imaginary? [12:59] there is plenty of organizations that use kubuntu for their products which from what I have seen address a small margin market [13:00] so just because *we* do not know about digikam on arm users, it does not mean there are none, in fact it is very likely that there are [13:01] with digikam being the greatest of all photo apps and arm being the greatest of all arches and kubuntu being the greatest of all linux distros [13:01] it is a parade of greatness [13:04] It's definitely a parade of something. [13:10] agateau: do you mind if I put lightdm-kde into a kubuntu-packaging branch? [13:14] hola [13:14] ScottK, shadeslayer: digikam still need doing? (I got distracted) [13:15] Riddell: already uploaded [13:15] lovely [13:15] Altho, https://bugs.kde.org/show_bug.cgi?id=296141 needs a bit of attention [13:15] KDE bug 296141 in Portability "Make digikam compile on ARM" [Normal,Resolved: fixed] [13:15] Plus it looks like the calligra failure on armhf was just archive skew. Retried it. [13:16] shadeslayer: Did you send your changes upstream? [13:16] ScottK: it was but I'm just about to upload it with transitional packages [13:16] ScottK: yes, see that bug [13:16] Riddell: Oh. Oops. Already retried. [13:16] I've gtg for a bit, will be back later [13:17] shadeslayer: mind and be back in a couple of hours :) [13:22] Riddell: is something important happening? [13:22] "Next Meeting: 16th March 3.30 PM UTC" [13:22] apachelogger: just a boring meeting, you can bring your pitch forks incase it gets messy [13:24] cool [13:24] ah 3:30 utc [13:24] goody [13:24] * apachelogger remembered 3:30 cet and was worried he might have to interrupt his coding session ^^ [13:24] debfx: I wanted to enable it around sunday, or I'll forget it next week before beta freeze, I'm not yet clear on the input grabbing thing which I'll look at too this weekend === greyback is now known as greyback|lunch [13:56] Pitch forks? :O [13:57] did I mention telepathy-gabble is so awesome that they have ponies in their source code [13:58] This is part of the same telepathy stack that now when Ubuntu screws up a library transition, our stuff breaks now too? [13:58] yep [13:58] takes care of XMPP accounts and such [13:59] Yeah. [13:59] Thanks to that little misadventure, I now know more about the telepathy stack than I'd planned on. [13:59] It's a bit funky tho, if I send messages from Jabber to GTalk in rapid succession GTalk refuses my messages [14:00] otoh I can send a large number of messages from GTalk to Jabber [14:01] but get this, If I use psi or irssi-plugin-xmpp ... there's no rate limit and GTalk accepts all messages [14:01] That's probably a function of Google's implementation. [14:01] Hmmm. [14:01] I'm not sure if its Google or gabble [14:02] maybe google maintains a whitelist and implements rate limiting for everything else [14:02] maybe its gabble ... can't say [14:02] all weird and stuff, the jabber.org folks couldn't pin it down as well [14:20] hmm shadeslayer's meeting clashes with the moved release team meeting [14:20] oh well, two meetings at once, how hard can it be? [14:20] uh oh [14:21] I was here first, make them change the time! :P [14:21] Tonio_, fabo, debfx, apachelogger, NCommander, Quintasan, yofel_, ScottK [14:22] 1 hour until shadeslayer meeting for kubuntu-dev [14:22] * NCommander makes a load snore sound [14:22] er, what? :-P [14:22] meeting in an hour, you can go back to sleep :) [14:22] * shadeslayer bribes NCommander with cookies [14:22] * NCommander codes auto-NCommander which +1s everything [14:23] cookies, they work every time [14:25] Riddell: http://paste.kde.org/441242 :P [14:25] I wonder how many such bugs I can find in KDE === greyback|lunch is now known as greyback [14:27] shadeslayer: what is that? [14:27] Riddell: it's using a comparison operator instead of a assignment operator [14:27] shadeslayer: yes it does look like it's doing the wrong thing [14:27] shadeslayer: but what package is it from and is it upstream and do we need to upload a patched package? [14:27] ah [14:28] Riddell: konsole [14:28] so it's probably been there for years [14:28] probably [14:28] shadeslayer: sent it upstream? [14:28] yep [14:28] just emailed the konsole maintainer with a attached patch [14:28] shadeslayer: is robert still maintaining konsole or someone else? [14:29] I think Jekyll Wu is maintaining it now [14:29] shadeslayer: e-mailing the maintainer is fine if we know it has an active maintainer but it's probably better to make a bug and attach so it doesn't get lost of the e-mail goes to someone without time to care [14:29] ( Looking at bug 296116 gives me the impression that he's the one maintaining it ) [14:30] Launchpad bug 296116 in IPython "Input to %timeit is not preparsed" [Undecided,Confirmed] https://launchpad.net/bugs/296116 [14:30] good idea [14:30] he is listed as "bugs" in the about box so seems a good candidate [14:30] I'll do that if I don't get a reply in the next 30 minutes [14:31] bugzilla already has a bazillion bugs where everyone gets lost [14:31] the audacity to want to be equal to the likes of NCommander and Tonio_ [14:32] bug trackers can be dumping grounds too right enough [14:32] aka not doing no nothing *fist waving and all that stuff* [14:32] shadeslayer: the other way is just to get me to commit it, we'll do that if he doesn't get back to you [14:32] Riddell: or .. I could push it myself :P [14:33] apachelogger: if you think we should have a way of removing people from kubuntu-dev I hear we're got a kubuntu-dev meeting happening soon [14:33] shadeslayer: indeed :) [14:49] Riddell: no, what we should have is a way to force people to do stuff :P [14:54] apachelogger: yeah like NCommander could help me with this pandaboard that works with oneiric but not precice [14:56] Riddell: heh, doesn't boot? [14:56] Riddell: There's a long and well established tradition of Canonical abandoning support for arm hardware. If you want stuff to keep working, install Debian. [14:57] ^ [14:57] Like the raspberry pi [14:57] If only ARM was backward compatible [14:58] shadeslayer: that's a bit unfair! it does have a chipset we don't even compile things for [14:58] No, the bigger issue in Ubuntu is sub-archs being added and dropped with great frequency. [14:58] ScottK: pandaboards in general work fine but this model seems to have some peculiarity that's broken [14:58] I see. [14:58] Riddell: but look at debian, they provide armel builds that will just work fine on the RasPi [14:59] the chipset is all good but something is making it read the disk as ext3 for no reason and that breaks something [14:59] and they have armhf for newer boards [14:59] Yep. [14:59] shadeslayer: ubuntu isn't an embedded distro [14:59] Neither is Debian. [14:59] armv6 is an embedded chip but that's not where ubuntu is at [14:59] ScottK: hey! you stole my line [14:59] (there's a derivative called Embeddian for that) [15:00] ubuntu would be happy to have the buildds supplied and armv6 builds looked after of course :) [15:01] but until that happens we'll live with what canonical can do which is what they get funding for which is armv7 [15:01] My issue isn't the arm version, but the subarchs. [15:01] My stack of arm boxes won't run Ubuntu. [15:01] what [15:01] ScottK: so they're running debian right now? [15:01] They currently have a vendor kernel and Karmic/Maverick user spaces. [15:01] shadeslayer: It's not pure Ubuntu. [15:02] O_O [15:02] In my copious free time I intend to switch them to Debian armhf as there's Debian kernel support for the hardware/sub-arch. [15:04] ScottK: which hw do you have? [15:04] fabo: Efika mx smarttop [15:05] we have builds for efika [15:05] You mean linaro? [15:05] That would be my other option. [15:05] I started precise based images this week (see my report) [15:05] Interesting. [15:05] Maybe I'll do that instead. [15:05] hwpack is missing but I can bug the right guy :) [15:06] to make it happen [15:06] Getting the smarttop's to boot to the SD card is a pain. [15:08] hmm [15:08] apachelogger: coming to akademy? [15:13] Riddell: no reply, I'm pushing the fix [15:13] shadeslayer: that's only an hour or two you gave him! [15:14] :P [15:14] go ahead and push if you're sure, but it's a risk [15:14] ah well, I'll wait then [15:17] uh here [15:17] Hiho [15:17] ok [15:17] :D [15:18] so yeah, following up on xmpp [15:18] shadeslayer: nothing much, doing maths all the time to learn as much as possible [15:18] Quintasan: any calculus in there? [15:18] nope [15:18] 48 days left [15:18] to exam [15:18] ah, trivial stuff then [15:18] 60 days till mien [15:19] *mine [15:19] well, it seems trivial but if I do poor on that exam then no uni for me [15:19] and then I'll be free from exams \o/ [15:19] I'm not a math genious by any means [15:19] Quintasan: heh, same here [15:19] I *suck* at maths [15:19] I'm mediocre at it honestly [15:20] yofel_: ping [15:20] whee [15:20] Riddell: Jekyll replied [15:20] I'm pushing it [15:21] shadeslayer: when is Akademy? [15:21] Quintasan: 30th June onwards [15:22] shadeslayer: push hard! [15:22] oh, after my exams [15:22] shadeslayer: Can one get sponsorship for going to Akademy or I need to work part time? [15:23] Quintasan: yes you can, the first deadline is 1st April [15:23] I'll apply then, if not then I will go part-time for a month after my exam [15:24] meeting in 5 mins people [15:24] I am in the middle of taking jedis apart, that will take longa [15:24] urgh we need to slap p-n back in shape [15:24] Nooo ... not the jedis [15:24] Quintasan: whats wrong with p-n? [15:25] yofel_: set an auto-bugger to Quintasan from the end of May onwards [15:25] shadeslayer: Well it's like, we need to check through build-logs to find if we need any new deps [15:25] fix fixable build fails etc [15:26] ah [15:28] NCommander: see what's wrong in this precise boot log on a pandaboard? http://starsky.19inch.net/~jr/tmp/precice-arm-boot.txt [15:29] NCommander: this looks weird "[ 6.134063] No filesystem could mount root, tried: ext3" [15:30] Meeting time [15:30] Riddell: ScottK apachelogger NCommander ping [15:30] pong [15:30] Quintasan: yofel_ ^ [15:30] pong [15:30] pong [15:30] Tonio_, fabo, debfx, apachelogger, NCommander, Quintasan, yofel_, ScottK ping [15:30] Wiki Application here : https://wiki.kubuntu.org/shadeslayer/KubuntuDevApplication [15:31] re-pong [15:31] Wiki Page here : https://wiki.kubuntu.org/shadeslayer/ [15:31] pong in a minute === yofel_ is now known as yofel [15:31] ScottK: do you remember the quorum rules? [15:32] shadeslayer: how do I find out when beta freeze is and what does it mean? [15:32] Riddell: There's no quorum rule, there's a voting rule. The minimum vote is more than 50% of those present and at least 3, so minimum number is 3 - 5 depending on votes. [15:32] groovy [15:33] shadeslayer probably doesn't mind if apachelogger fell asleep anyway. [15:33] Right, so at the start of every release we get a release page, for precise you have thi : https://wiki.ubuntu.com/PrecisePangolin/ReleaseSchedule?action=show&redirect=PreciseReleaseSchedule [15:33] ScottK: nope, not at all :D [15:33] I just look up the dates on that page [15:34] What's OK to upload during a freeze? [15:34] As for what it means, it basically means that we now have to stablize the archives and start fixing bugs [15:34] ScottK: Critical bugs that would make ISO's unbootable [15:34] or stuff that would make your installer crap itself [15:35] I told you I was right in the middle of a game, now I am all yours [15:35] oh noes [15:35] ohshi- === Quintasan is now known as trolllogger [15:35] shadeslayer: what's your favourite debian/rules build system? [15:35] shadeslayer: What's the difference between a hard freeze and a soft freeze and when do you see them? [15:35] * trolllogger whispers cdbs to shadeslayer [15:36] Riddell: I'd have to say debhelper [15:36] Don't all packages have to use debhelper? [15:36] shadeslayer: which debhelper, there are 9 of them :) [15:36] I started out when debhelper was introduced and cdbs was being deprecated [15:36] ScottK: not all, look at hello [15:36] ah, compat 7 and above [15:36] * apachelogger has 26 questions and supposes people would not appreciate if he used all of them [15:36] shadeslayer: well 7 is very different from say 9 [15:36] Riddell: I knew that, it was a question for shadeslayer. === trolllogger is now known as Quintasan [15:37] ScottK: oh sorry [15:37] Riddell: stop helping shadeslayer :P [15:37] * apachelogger cherry picks question [15:37] shadeslayer: Go back and answer my freeze question then. [15:37] shadeslayer: What does a deb file contain? [15:37] ScottK: sorry, don't know about hard freeze and soft freeze, but I'm guessinghere [15:38] Soft freeze : You can upload stuff if the release team approves it === apachelogger changed the topic of #kubuntu-devel to: Kubuntu: Friendly Grilling of Rohan Muhahahaha | 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 | wiki.kubuntu.org/Kubuntu/UDSQSeries | Next Meeting: 16th March 3.30 PM UTC | Milestoned beasties http://goo.gl/9iyUD [15:38] hard freeze : You can't upload stuff, and you have to wait till archives open again [15:38] It's sort of the other way around. [15:38] 0.o [15:38] In a soft freeze, the archive is in a normal state and devs are trusted not to upload wrong stuff. [15:39] In a hard freeze, it's frozen and only what the release team approves gets in. [15:39] apachelogger: binaries that you can run , shared libararies, and conf files [15:39] Alphas are soft freezes. [15:39] huh, I wasn't aware of that ... [15:39] shadeslayer: yes sure, but I mean technically [15:39] what makes a deb file [15:39] what is a deb file [15:39] apachelogger: then there's stuff like post and pre installer files that are run before and after a package has been unpacked/installed [15:40] actually that is a good one, let's start with What is a deb file technically. [15:40] it's just a tarball that dpkg unpacks [15:40] shadeslayer: or what's a good way to look inside one? [15:40] shadeslayer: is it a tarball? [15:40] and then it has rules which allow dpkg to unpack it ... [15:40] Riddell: dpkg --contents [15:41] shadeslayer: How do we maintain binary compatilbily between new versions of packages? [15:41] compatibility* [15:41] shadeslayer: that still doesn't answer the question... what is a deb file and what is inside it? [15:41] apachelogger: more or less, it's not your conventional tarball, but has functions like 'unpacking' [15:41] * fabo just joining [15:41] shadeslayer: oh is that it, I always wondered, I just use lesspipe :) [15:41] lol [15:41] shadeslayer: well, it is not a tarball :P [15:41] shadeslayer: it is an ar archive :P [15:41] apachelogger: one of them is a tar [15:41] new things to learn everyday [15:41] but I never remember which [15:41] Riddell: not the deb [15:42] it's a tar inside an ar I think [15:42] shadeslayer: so what is inside this ar archive? [15:42] oh um sorry [15:42] I'll shut up :) [15:42] Riddell: you are not being grilled :P [15:42] The best meeting ever [15:42] :D [15:42] Riddell: the deb file is an ar archive ;) [15:42] apachelogger: wait, so whats the difference between a archive and a tar? [15:42] ar is a format [15:42] Riddell: there is not that much difference between debhelper v7 and v9. [15:42] tar is a format [15:42] they are not the same [15:42] shadeslayer: so if you were to decompress this ar archive... what will you find inside? [15:43] shadeslayer: you may try to use ar to look at deb you have lying around ;) [15:43] debfx: I beg to differ :) (but I can be off topic) [15:43] shadeslayer: when you change the debhelper compat level, how do you know which changes in debhelper you need to be aware of? [15:43] apachelogger: binaries, symlinks, stuff that dpkg places in / [15:43] In fact, when Ian picked ar, tar existed, but it was all new fangled and stuff. ar was something everyone would have tools to deal with. [15:43] fabo: I meant how debian/rules looks like [15:43] debfx: ok, thanks for the context [15:44] shadeslayer: that is not within the ar [15:44] FWIW, ark is one of my favorite ways to look inside .debs. [15:44] uhh [15:44] shadeslayer: how did you fix digikam? [15:44] shadeslayer: so within the ar is in fact one text file and two tarballs [15:44] debfx: I'd probably refer to the debian new maintainers guide to look at what has changed [15:45] shadeslayer: the text file defines the deb binary format of this file, one tar is control.tar and contains the actual packaging information and the other is data.tar which contains what you were talking aobut ;) [15:45] apachelogger: but dpkg --contents foo.deb says thats what inside it [15:45] you are not listening to what I am trying to teach you [15:45] that is not the content of the ar archive [15:45] shadeslayer: it lists the files which are unpacked, not what the ar contains in reality [15:45] it is the contents of the data.tar [15:45] it's just apachelogger nitpicking stuff [15:45] shadeslayer: I don't think it's documented in debian new maintainers guide [15:46] no, that is quite fundamental to how dpkg works [15:46] Riddell: Well there were 2 issues in digikam, a) no opengl support on arm and b) methods returning doubles [15:46] http://tldp.org/HOWTO/html_single/Debian-Binary-Package-Building-HOWTO/ [15:46] apachelogger: eeep ... I knew that, I just thought you were only talking about data.tar [15:46] Riddell: btw, if the binary format were to change the assumption that there are tarballs within the deb might not be true anymore :) [15:46] >Revision 4.0 2005-08-09 [15:46] cool story Harald [15:47] shadeslayer: I asked you what the contents of the ar archive is ... [15:47] I consider this question dealt with [15:47] do we have other pending ones or can I continue? ;) [15:47] shadeslayer: You might want to answer my question at some point [15:47] fabo: oh ... No idea then, I've only dealt with compat 7 so far, I've never handled a compat 8 or 9 package [15:47] shadeslayer: how did you fix each of those? [15:47] Quintasan: looking [15:47] How do we maintain binary compatilbily between new versions of packages? [15:48] shadeslayer: you can find it with man debhelper 7 [15:48] Quintasan: that question is sloppy :P [15:48] yeah [15:48] Quintasan: I don't see your question [15:48] let me try [15:48] Let me rephrase that [15:48] my screen is full of orange [15:48] because I have one along those lines [15:49] shadeslayer: What are reasons to bump a soversion? Do we have a system to detect ABI changes? Do we have a system to prevent ABI incompatibility despite upstream not changing the soversion? [15:49] Riddell: as for opengl, I just disabled opengl libs on arm and armhf, they were optional [15:49] shadeslayer: what's symbol files and which document you'll recommend to read? [15:49] Riddell: as for double I replaced double with qreal [15:49] Quintasan: that covers your question I blieve? [15:49] apachelogger: PRECISELY [15:49] pun intended [15:49] perfect :) [15:49] Quintasan: as always ;) [15:49] shadeslayer: how did you track down and fix that konsole issue? [15:50] fabo: symbol files are files that contain library symbol files and are used by dh_makeshlibs to genrerate the values for shlibs:Depends [15:50] fabo: as for what document to read, I read the dh_makeshlibs man page [15:51] ehm [15:51] apachelogger: whenever upstream breaks ABI, it bumps its so version, [15:51] shadeslayer: dh_makeshlibs reads the symbols from the binaries directly [15:51] shadeslayer: does it? [15:51] shadeslayer: "whenever upstream breaks ABI, it bumps its so version" do all upstreams always do that? [15:51] or rather [15:51] shadeslayer: can you be a bit more specific about how shlibdebs uses the symbol files? Where does it look? [15:51] do they? [15:52] right-o, stop the flood a little bit, let him answer some and let apachelogger copypasta some more [15:52] Riddell: not necessarily, at times they break ABI and that causes all sorts of problems downstream [15:52] ^^ [15:52] naughty upstream [15:52] shadeslayer: now please answer my questions, because they directly relate to what you just said :) [15:53] Riddell: re konsole:I've switchec to clang which produces beautiful errors [15:53] uh, that is creative [15:53] yes, give me a second to reread the backlog :P [15:53] sure [15:53] apachelogger: so, ABI right? [15:53] apachelogger: we track ABI using symbol files [15:53] shadeslayer: What are reasons to bump a soversion? Do we have a system to detect ABI changes? Do we have a system to prevent ABI incompatibility despite upstream not changing the soversion? [15:54] so for eg if upstream changes public symbols, we see a bunch of new symbols at the end of the build [15:54] which is cause for concern since new symbols were introduced and/or old symbols were removed [15:55] righto [15:55] we then go ahead and poke upstream as to what the hell happened [15:55] and proceed to bump the so version .... [15:55] shadeslayer: are added symbols a reason to bump the ABI? [15:55] yofel: not *every* time, I usually do it on a case to case basis [15:55] *soversion [15:55] Quintasan: mind how my questions cover everything that is getting asked regarding that topic ;) [15:56] shadeslayer: so when does the soversion *need* to be bumped? [15:56] well the concern is symbols were removed without so bump [15:56] apachelogger: I never bothered to create a copypasta database of questions :P [15:56] apachelogger: well, usually what happens is that the so version is appened to a library [15:56] Quintasan: it aint a database, it is a file called rohan.txt :P [15:57] so when a build fails and you see the same library with a new soversion at the end of it, you bump it in the package as well [15:57] shadeslayer: no, what symbol changes *require* a bump of the soversion? [15:57] oh [15:57] * shadeslayer thinks for a bit [15:57] apachelogger: s/rohan/IamLazyAndEverybodyKnowsIt/g [15:57] ;) [15:57] lulz [15:57] shadeslayer: yes, but as upstream ... when do you bump [15:57] (this relates to the earlier conclusion that upstream likes to not bump soversions at times, despite it being necessary) [15:58] and when is lethal force appropriate? [15:58] lol [15:58] ^^ [15:58] ScottK: lethal weapon? [15:58] shadeslayer: Which file from debian/ is your favourite? [15:58] * Quintasan notes there is no right answer to that [15:58] <3 [15:58] hahah [15:58] really? [15:58] * apachelogger thinks there is [15:58] Quintasan: funny question :) [15:59] which do I have to look at closely as an archive admin? [15:59] * fabo me too [15:59] okay so, whenever public functions change, that would constitute a ABI breakage [15:59] wait no [15:59] Riddell: debian/compat, obviously, no? [15:59] * shadeslayer rethinks that a bit [15:59] rethings/google [15:59] no [15:59] I believe it has something to do with dpointers [16:00] yes, but that is too detailed already [16:00] trying to think back when I last did a soversion bump [16:00] don't try to understand binary compatiblity rules, you can't :) [16:00] as a packager you care about symbols [16:01] hmm .. [16:01] I believe we utterly confused him :D [16:01] yeah, really [16:01] heh [16:01] yeah, I'm a bit lost as to what exactly is it that you want about symbols [16:01] shadeslayer: let me give you an example ... there is a class Foo it has a function foobar soversion is 0 ... then upstream adds a function funbar and releases ... then upstream removes funbar and releases [16:01] you have new symbols, and you bump the soversion [16:02] what must the soversion be at least after the latest release? [16:02] 2 ... [16:03] nope 1 [16:03] beacuse adding funbar doesn't break compatibility? [16:03] adding a function is binary compatible (with a whole set of exceptions of course, but generally speaking here) [16:03] shadeslayer: read full example again [16:03] removing a function however is definitely breaking the ABI [16:04] derp @_@ [16:04] that's a little unfair since adding a function can well break ABI as you just said [16:04] thats what I always thought ^ [16:04] assuming the simplest example it cannot [16:04] apachelogger: nitpicking, it can be 2. that's a so bump :) [16:04] as a packager we don't care about the exact rules of c++ ABI, we just look at the .symbols files [16:04] fabo: at least I asked [16:04] it can also be version 300 [16:04] ah, right, at max it can be 2 [16:05] apachelogger: true, you made your point [16:05] atleast it should be 1 [16:05] exactly [16:05] I think we can move on from all this ABI mess [16:05] oh no [16:05] the last important thing [16:05] shadeslayer: last question of my paste [16:05] I thought it should be ideally 2 [16:05] apachelogger: yay [16:05] so we established that upstream sometimes does not bump the soversion even though they should [16:06] what do we do? [16:06] (I am talking about core KDE specifically here) [16:06] Back to my lethal force question, I see. [16:06] ^^ [16:06] well, you add/remove new symbols and rebuild everything that depends on the lib [16:06] shadeslayer: does that change the soversion? [16:06] because if you don't the packages still link to the old lib [16:07] apachelogger: so version of the library? Nope [16:07] so how does that solve the issue? [16:07] upstream did not bump even though they needed to [16:08] * shadeslayer thinks for a bit [16:08] apachelogger: you rebuild everything and convince upstream not to do it again [16:09] thats what I was thinking as well, but I think apachelogger is hinting at other stuff here :P [16:09] tsdgeos: ssh it's for shadeslayer's exam! [16:09] actually we have a better thing for core KDE :) [16:09] http://packages.ubuntu.com/oneiric/libkwineffects1abi2 [16:09] Riddell: damnit :D [16:09] ever wondered about that package name? :P [16:09] hah [16:09] * tsdgeos hides [16:09] oh god this [16:09] apachelogger: that's most interesting [16:09] http://packages.ubuntu.com/oneiric/i386/libkwineffects1abi2/filelist [16:09] IT'S NOT INTERESTING [16:09] ffff [16:09] it very much is :P [16:09] notice how it matches the library soversion perfectly? :P [16:09] damn [16:10] it becomes tricky [16:10] that one was a hell when doing gles stuff [16:10] yeah [16:10] well I think we can conclude the entire ABI business with the hint that shadeslayer should look a bit into the things highlighted :) [16:10] particularly why this package has such a funny abi name [16:10] well [16:10] because it is very handy [16:10] shadeslayer: one more thing, what is the symbol file used for during the package build other than making sure upstream doesn't break the ABI? [16:10] yofel: it generates the value of shlibs:Depends [16:11] shadeslayer: how does it do that? [16:11] so that other stuff that uses that library can link to that library [16:11] or rather [16:11] any thing that links to that library pulls in said library when installing [16:11] yofel: dh_makeshlibs ? [16:11] well, it can do that from the library too, and from the auto-generated symbol file, no need to use the one you ship with the package [16:12] why is it useful for *dpkg-shlibdeps*? [16:12] well, or dh_.. [16:12] ah [16:12] I'm not sure what you're asking here [16:13] * apachelogger points out that makeshlibs does not use the symbol file but gets the symbols directly from the binaries! [16:13] shadeslayer: hint: versioned dependencies [16:13] apachelogger: right, dh_shlibsdeps uses symbol files ... argh [16:13] we should probably begin wrapping up soon [16:14] * apachelogger still has 4 questions he'd like to ask ;) [16:14] yofel: so that it can automatically depend on newest lib? [16:14] shadeslayer: read the symbols part of the dpkg-shlibdeps manpage, first sentence [16:15] ah [16:15] dpkg-shlibdeps remembers the (biggest) minimal version needed for each library [16:15] when reading the symbols file [16:15] shadeslayer: What editor do you use? [16:15] apachelogger: vim [16:15] lol [16:15] What language are debian/rules files written in? And does vim have highlighting for that language? [16:16] apachelogger: make, and yes [16:16] Upstream doesn't ship a manpage, so you write one, how do you install it via the packaging? [16:16] shadeslayer: if you're running out of patience feel free to ask for a vote, the gambol is just have you convinced everyone who needs convinced :) [16:16] apachelogger: dh_installman [16:17] someone read up on dh it seems ^^ [16:17] so, final one [16:17] Imagine, if you will... bulldog89 rewrote a lot of Ubiquity using QtQuick and refactored almost all the remaining code. It looks much hotter, has improved accessibility and from internal testing it seems that it works at least as good as the current version. A feature and UI freeze exception was already granted in Feburary. Release freeze is only a few days away and all Kubuntu devs are at Riddell's to attend a pirate themed party. There [16:17] is no time to waste and you need to decide whether you want to push this new thing in or postpone it. [16:17] What do you do and how do you defend your decision when apachelogger comes back and starts whining about how your decision was not the right one. [16:17] you put it in debian/foo.manpages [16:17] Riddell: nah, I'm good [16:17] hah, I remember that one from yofel's interview [16:17] I'm coming over from #ubuntu-bugs because I was surprised by one thing that I'd like to confirm. [16:18] no it is a different one, similar scenario though ;) [16:18] Is it true that you guys close valid bugs as invalid if you prefer to have them handled upstream? [16:18] yeah .. [16:18] Laibsch: Yes, but we're in the middle of a meeting right now. [16:18] apachelogger: can I ask follow up questions? [16:18] procedure most everywhere else seems to be to leave it open until fixed in both ubuntu and upstream [16:18] shadeslayer: sure [16:19] apachelogger: what kind of testing did it go through? [16:19] Laibsch: Find someone to maintain all the bugs in LP and we'd be glad to do that (Debian Qt-KDE team does the same, BTW). [16:19] did it cover *all* cases that the current installer handles? [16:19] without any crashes [16:19] shadeslayer: all devs tried it prior to leaving for the party and a public call for testing was done on the ml [16:19] no negative feedback [16:19] I'd say it's good to go then [16:20] shadeslayer: so, which file in debian/ is your favourite? [16:20] now the interesting part... why do you think so? [16:20] has loads of new features, specually a11y and has had no negative feedback, was tested widely [16:20] Quintasan: hmm ... I'd say control [16:21] I though so, they are, well, so precisely constructed [16:21] :P [16:21] shadeslayer: just to be sure... so you'd risk breaking the installer in the release incredibly close to release? [16:21] yep [16:21] Quintasan: and rules are not? :P [16:21] apachelogger: but you just said it was tested properly and there was no negative feedback [16:22] apachelogger: look at files using qmake and then ask me again [16:22] s/files/apps [16:22] but yes, I'd be willing to take that risk === JackyAlcine_ is now known as jalcine [16:22] shadeslayer: well, there is no right or wrong here [16:22] You might want to know how it would affect CD size too. [16:22] apachelogger: yep, but that's what I think, our current installer is ... well ... not as pretty as it could be [16:23] * apachelogger wouldn't do it because he is reasonable certain that the lack of negative feedback is rather an indication of not very good testing rather than it being perfect software :) [16:23] ScottK: hmm, if it's written in QML, I think it'll be smaller [16:23] not necessary though [16:23] It'd be good to know that though. [16:23] well, desinger files = xml = more text [16:24] apachelogger: more certainly smaller than binary no? [16:24] XML FTL. [16:24] well, I have all my questions answered unless I should start asking about udeb files... ;) [16:24] shadeslayer: What's a seed and why do we care? [16:24] * ScottK jumps in. [16:24] shadeslayer: we don't have binary in ubiquity [16:24] which is why it starts so slowly [16:25] ScottK: okay, seeds are packages that go onto the CD [16:25] seeds *must* be in main ( as of right now ) [16:25] Sort of. [16:25] Seeds have lists of packages. [16:25] right, those packages go onto the CD and they must be in main [16:25] What tool do we use to get from those lists to metapackage? [16:26] (e.g. the seed doesn't list all the dependencies) [16:26] germinate [16:26] whoa [16:26] :O [16:26] Oo [16:26] I read about that like a year ago [16:26] shadeslayer++ [16:26] apachelogger: I'm surprised I remember that myself [16:27] So, let's say that since you just fixed up digikam, you've decided it really, really needs to go on the CD, what do you do? [16:27] It can't it's in universe :P [16:27] That can be changed. [16:27] but if I *had* to, I'd have to file a MIR [16:27] No, actually you wouldn't. [16:27] get it into main, add it to the seed file, germinate [16:27] It's been in Main before, so a MIR isn't required. [16:27] * apachelogger just wanted to say that :) [16:27] but the source changed ;) [16:28] you could have new potential security stuff that needs review [16:28] Sure, there are exceptions, but as a rule re-MIR is not required. [16:28] new build depends ... [16:28] shadeslayer: by that standard every new version entering main would need a mir [16:28] apachelogger: better safe than sorry imho [16:28] OK, so now the CD is 30MB to big because you added digikam without removing anything else. [16:28] What do you do next? [16:29] shadeslayer: feel free to answer that at the end: what's your workflow for new packages, let's say you have this new awesome app - QHaraldface which adds a spinning image of Harald wearing a trollface moving around your desktop, it uses qmake and you see there is a huge demad for it, how do you do it? [16:29] hmm, drop stuff that's not needed [16:29] if you remove more artwork I'll come after you with a franklinator [16:29] hahaha [16:29] FTR: http://www.youtube.com/watch?v=vy43iZCPeLI [16:29] shadeslayer: There's a step you missed here I think. If you're going to make a major change in the seeds, you really ought to discuss it with your fellow developers first. [16:29] ScottK: there's no right answer to that, you have to make tough decisions as to what goes onto the CD [16:29] * yofel needs to run, sorry (I'll leave it at +0, so it's up to the rest) [16:30] Right, but don't make it by yourself. [16:30] ScottK: I assumed that was already done [16:30] You know what happens when you assume ... [16:30] :) [16:30] So you change the seed, run the germinate update script, then what? [16:31] (hint: what package are you changing?) [16:31] you respin the ISOs [16:31] uhh [16:31] kubuntu-desktop I believe [16:32] kubuntu-meta. [16:32] kubuntu-desktop is the binary. [16:32] (-desktop is a binary pacakage of -meta| [16:32] ah, ok [16:32] I could probably use apt-cache show to figure that bit out [16:33] shadeslayer: You should read https://wiki.ubuntu.com/SeedManagement before you touch any of this stuff. [16:33] * Quintasan bookmarks that [16:33] ofcourse, I don't touch anything without consulting stuff [16:33] Quintasan: nvm that you will not find it, google will help ;) [16:33] apachelogger: >implying I have more than 10 bookmarks [16:33] I've realised that if I rush things I make mistakes [16:34] primarily why digikam took a week to fix [16:34] Quintasan: if you are serious about bookmarking policy pages, then you have a lot more than 10 bookmarks :P [16:34] Ah. Reminds me of my other question ... [16:34] yes? [16:34] shadeslayer: What's the last thing you had uploaded and the sponsor found something significant wrong with it? What did you learn from this? [16:34] hmmm [16:35] that would be Quintasan [16:35] but he didn't find the problem [16:35] :P [16:35] it was fabo who found the problem I think [16:35] Which package are we talking about? [16:35] sec, let me check [16:35] What was it? What did you learn? [16:36] it was ktp-accounts-kcm [16:36] not me then :) I didn't touched this one [16:36] and the tar was tainted [16:36] How so? [16:37] I used the wrong upstream tar [16:37] oh wait [16:37] yeah, I remember that I built it with no problems [16:37] not accounts kcm [16:37] it was tp-qt4 [16:37] https://launchpad.net/ubuntu/+source/telepathy-qt4/0.9.0+repack-0ubuntu1 [16:37] the problem was I accidently went to the telepathy-qt4 page whereas upstream now uses telepathy-qt [16:38] it was debfx! [16:38] :D [16:38] It's not my fault then :P [16:38] right, it was my fault :) [16:38] don't they use telepathy-qt4 which is renamed to telepathy-qt? [16:38] Riddell: upstream is named telepathy-qt, source package is called telepathy-qt4 [16:38] back from mobile [16:39] + In debian/ubuntu [16:39] anyway, that's over an hour [16:39] :D [16:39] I suggest we move to a decision [16:39] * Quintasan hands yofel a spare battery [16:39] ah yes, it's council vote or what? [16:39] it is developer vote [16:39] Quintasan: kubuntu-dev [16:39] * shadeslayer grabs dinner [16:39] * Quintasan can't remember that voing stuff [16:39] voting* [16:39] you did not read the intro apparently :P [16:40] How do you feel about essentially having root permissions on a users system? [16:40] shadeslayer: ^^^ come back from dinner [16:40] Rarely do I have to vote. [16:40] Quintasan: because you don't bring in minions [16:41] yofel: godlike :P [16:41] I'm +1 for decent knowledge and good application, but plenty more to learn so be cautious on the uploads [16:41] * ScottK imagines apachelogger will -1 purely to keep the minion. [16:41] Are you implying we got no artwork from shaytan apachelogger ? [16:41] Quintasan: no, in case you did not read the backlog, I was doing the new splash logo last night :P [16:42] * ScottK +1 as well. Still more to learn (as do we all), but knows enough to be trusted). [16:42] +1 for shadeslayer, he is a good developer and gets his job done and learns from his mistakes (or my mistakes :P) [16:42] :D [16:42] Quintasan: we all learn from your mistakes :P [16:42] apachelogger: I'm not sure I want to see it ;P [16:43] * ScottK looks at fabo, debfx, apachelogger, Quintasan, and yofel for votes. [16:43] Quintasan: you'll have to, it'll be default :P [16:43] Oh god, [16:43] I'll +1 for keen contributer and good general knowledge of how things are done (with lots to improve on though) [16:43] By my count we have 8 kubuntu-dev here, so one more +1 needed. [16:43] There we go. [16:44] \o/ [16:44] It would still be good to hear from the rest of you. [16:44] hard to decide it is [16:44] shadeslayer: quick question in C is Array[] the same as Array*? [16:45] the 1 is round(1 - 0.5 for the last answer) - please don't be *too* enthusiastic ;) [16:45] Anyone knows status ARM image of Kubuntu/Kubuntu Active? I tried recent ones of Kubuntu image for omap4, but they all froze after some time, not like previous ones which are quite stable [16:45] gosh, C, I can't remember the last time I had to care about that [16:45] uhm [16:45] ezjd: meeting, give us 2 mins [16:45] apachelogger: could you give me a more specific example? [16:45] char foo[] and char *foo ? [16:46] shadeslayer: for example [16:46] * Quintasan looks for his trollface.svg and wonder how to replace default splash logo === vivek is now known as pvivek [16:46] kubotu: order party for everyone [16:46] * kubotu is going to his secret storehouse to get party for everyone - might take some time. [16:46] * kubotu is back and slides party down the bar to everyone [16:46] Quintasan: /lib/plymouth/themes/kubuntu-logo/kubuntu_logo.png [16:46] durr [16:46] sudo update-initramfs -u [16:46] apachelogger: how does one invoke party time in kubotu ? [16:46] shadeslayer: so, what's the word? [16:47] apachelogger: huh? I'm waiting for your example ::P [16:47] or did you mean char foo[] and char *foo was the example [16:47] yes [16:47] I believe they're the same [16:47] oh my [16:48] +1 [16:48] FWIW the answer was wrong :P [16:48] heh, foo is a array and *foo is a pointer to a string? [16:48] they can behave the same way but fundamentally are different data types ;) [16:48] ah [16:48] an array is always on the stack [16:48] * Quintasan was about to type tha [16:48] and *foo is on the heap? [16:48] pointer is on heap isnt it? [16:48] whereas a pointer may not eve be an array [16:49] shadeslayer: it may point to the address of an array on the stack ;) [16:49] so pointer, is a pointer [16:49] :P [16:49] kubotu: order party [16:49] * kubotu gives everyone a party hat and a hand full of conffeti. [16:49] * kubotu turns on tha most funky party music as well as the all shiny disco ball. [16:49] Quintasan: wanna dance with me? :-) [16:49] * kubotu starts shaking her tight ass [16:49] still they are differnt in that an array is always a linear memory range of a known size e.g. you could sizeof the array and get the actual size [16:50] So did shadeslayer make it into kubuntu-dev? [16:50] yes [16:50] ages ago [16:50] others apparently do not want to vote [16:50] Yes, but we're still wanting to hear from debfx and fabo. [16:50] apachelogger: Change the topic then durr [16:50] apachelogger: I was more of the opinion : char foo[232]; char *foo1; foo1 = foo; [16:50] I vote +1, basics are there. but as mentioned by some, there's things to learn [16:51] shadeslayer: that is exactly the case where your pointer points to memory on the stack [16:51] apachelogger: yeah, so, that's what I thought about when you asked the question ;) [16:51] both will behave the same in this case [16:51] ( when using with printf and %s ) [16:51] aye [16:51] thanks everyone! [16:51] :D [16:52] so yeah, gratz shadeslayer [16:52] * apachelogger only asks because last night he and his mates noticed that in the C course at uni they tell students that an array and a pointer are all and the same [16:52] hah [16:52] * debfx emerges and votes +1 [16:52] all +1 [16:52] shadeslayer is a great new member [16:52] hooora [16:52] oh wow === Riddell changed the topic of #kubuntu-devel to: Kubuntu: Friendly Grilling of Rohan Muhahahaha | 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 | wiki.kubuntu.org/Kubuntu/UDSQSeries | Next Meeting: 16th March 3.30 PM UTC | Milestoned beasties http://goo.gl/9iyUD | welcome shadeslayer! [16:52] congratz and what not [16:53] * Quintasan hands shadeslayer the Stick of Pointing Harald Around === apachelogger changed the topic of #kubuntu-devel to: Kubuntu: Friendly Rohan must read up on ABI and AR and DEB | 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 | wiki.kubuntu.org/Kubuntu/UDSQSeries | Next Meeting: 16th March 3.30 PM UTC | Milestoned beasties http://goo.gl/9iyUD | welcome shadeslayer! [16:53] shadeslayer: congrats! [16:53] s/Pointing/Ordering [16:53] shadeslayer: Where are your minions? [16:53] And I'm off again, congratz shadeslayer and laterz [16:53] I wish I could do this with my new found powers now : http://xkcd.com/797/ [16:53] No, that's why we have archive admins [16:54] darn ... [16:54] anyone took care of Laibsch in a query or something? [16:54] my evil plan was thwarted [16:54] ScottK did but he departed [16:54] To the lands of #ubuntu-bugs I presume [16:55] well yes, I mean, he did not ask a real question or anything, so I was wondering ^^ [16:55] apachelogger: I didn't do anything in a query. Just what you saw in channel. [16:55] curious visit then [16:55] *shrug* [16:55] <--- hungry [16:55] * shadeslayer is waiting for someone to add him to https://launchpad.net/~kubuntu-dev [16:55] oh ezjd left [16:55] Riddell: go do something [16:56] also :* [16:56] apachelogger: like what? this oneiric->precise upgrade says it'll take 6 hours on arm [16:56] Oo [16:56] sweet baby skywalker [16:56] Riddell: add shadeslayer to the overlord launchpad team [16:56] apachelogger: what game are you playing? [16:57] shadeslayer: FYI, in addition to dpkg --contents, you should also know about debc (in devscripts) if you don't. [16:57] swtor [16:57] obviously [16:57] :O [16:57] apachelogger: where? How? [16:57] the intartubes [16:57] playable via a browser? [16:57] shadeslayer: ScottK: dpkg-deb is also an option [16:57] no? [16:57] like 30gig game :P [16:57] shadeslayer: Don't go off playing with him, he should be working [16:57] :O [16:57] swtor.com [16:58] I am on a german server anyway [16:58] no idea how that ever happend [16:58] True. [16:58] apachelogger: ark is still one of my favorites. [16:58] seems utterly silly given that I am up at night and sleep at day local time [16:58] aw man [16:58] ScottK: aye [16:59] only windows [16:59] well [16:59] ever seen a good game work well on linux? [16:59] >mfw apachelogger has windoorz installed [16:59] apachelogger: Aion Online [16:59] Durr [16:59] SMB [16:59] or wait, Heroes of Newerth [17:00] you fail to see the words good and well apparently [17:00] Riddell: lesspipe is cool! [17:01] * apachelogger needs to move out and find food [17:01] shadeslayer: no more fiddly commands to remember, it just does the right thing :) [17:01] :D [17:01] apachelogger: It seems like you fail to comprehend the word good [17:01] :P [17:01] + aion and HoN work good [17:01] obviously not on an 6 years old machine [17:03] ohohohh [17:03] someone please get the ksplash qml stuff on the CD [17:03] needed for the new kubuntu ksplash [17:03] shadeslayer: [17:03] ^^ [17:03] * apachelogger now really gooes looky looky for food [17:04] I'll do it tomorrow [17:04] shadeslayer: oh yes, you do that, if you fail you get demoted again :P [17:04] a) No battery left, b) have to code a bit [17:04] like commit/push stuff before I run out of battery powa [17:09] bah, swotr is not even available for the xbox [17:09] Riddell: ezjd is here [17:09] So meeting finished? [17:09] Yup. [17:11] my question is about ARM. the daily image freeze frequently on omap4, [17:11] and I am looking forward to a kubuntu active arm image [17:14] ezjd: iirc Riddell said that we'll focus on ARM after getting the i386 plasma active image right [17:16] shadeslayer: Since you want to be MOTU also, please look at this and propose a solution, including packaging changes if you believe they are appropriate. I'll sponsor changes, if any. [17:16] uh, what should I look at? :) [17:17] https://lists.ubuntu.com/archives/ubuntu-devel-discuss/2012-March/013427.html [17:17] sorry [17:18] * shadeslayer looks [17:19] shadeslayer: Also, we should discuss on #ubuntu-motu, not here. [17:19] right [17:20] shadeslayer: Congrats! [17:21] guys from time to time in kde 4.8.1 (oneiric backports), ark freezes when i click on "extract" and then error message appears: "kdeinit could not launch /usr/bin/ark" [17:21] where to report? upstream or lp? [17:22] I vote upstrem [17:22] though kdeinit bugs are always a mystery to me [17:22] apachelogger: ^^ [17:23] shadeslayer: OK. Thanks. Then I will try it out by myself. I am wondering if most of update is shared in irc or email. I am unable to do irc a lot and kubuntu-devel mail list seems very quite. === vivek is now known as pvivek [17:26] shadeslayer: congrats on becoming kubuntu-dev :) [17:26] micahg: thanks! [17:28] Quintasan: hm? [17:28] shadeslayer: u here? [17:28] yes [17:28] shadeslayer: what is a udeb file? [17:28] the power came back [17:29] and what does it contain? [17:29] and how is it different from a regular deb file? [17:29] I've never worked with them [17:29] any knows? [17:29] easy [17:30] shoot [17:30] "Some core Debian packages are available as udebs (“micro debs”), and are typically used only for bootstrapping a Debian installation" [17:30] udeb is a deb file but it contains only bare minimum [17:30] ex. no docs [17:30] aye [17:30] and is used for bootstraping debian [17:30] but but but [17:30] Quintasan: so pbuilder uses udeb? [17:30] udeb packages are not installable on a standard Debian system. [17:30] shadeslayer: that means? [17:30] only bootstrapping I guess [17:31] apachelogger: AFAIR no, I remeber it downloading regular debs [17:31] apachelogger: It can use them if you ask it do so IIRC [17:31] but doesn't pbuilder bootstrap? :P [17:31] * shadeslayer wants to play swtor [17:31] * apachelogger wants pizza [17:31] apachelogger: I think that when you're creating the initial bootstrap env it downloads the udeb [17:31] They're used in the installer too. [17:32] ^ or in net installs [17:32] mhh, yes, primary use case actually [17:32] shadeslayer: no, that is crap [17:32] as you said, they are not inteded for actual debian systems [17:32] apachelogger: okay, like I said, I've never worked with them [17:32] the installer itself needs to load an os and whatnot [17:32] so he uses udeb [17:32] hmm [17:32] also installer plugins would be udeb [17:33] generally everything necessary for installation itself has a udeb package available [17:33] curious how little we know about our systems ^^ [17:33] well ... how does one specify buildd to make udeb packages? [17:33] shadeslayer: there is some option to enable that in pbuilder [17:33] persia told me at one point [17:34] btw. what the hell is with persia, noone has seen him since uds, right-o? [17:34] persia also told shadeslayer to read the make manual, which surely he didn't do yet [17:34] like 2 years later [17:34] nope [17:34] I understand alot about debian/rules tho [17:34] what does .phony do? :P [17:35] oh god, this [17:35] I don't remember that one [17:36] Quintasan: read up : http://linuxdevcenter.com/pub/a/linux/2002/01/31/make_intro.html?page=2 [17:36] tells make that which targets are not files [17:36] so if you have files with the same name, make doesn't confuse them with the targets [17:36] and vice versa [17:36] u know, back in the days of the motu one had to have at least one package with a plain rules file (no cdbs magic etc.) to get endorsed for motu status [17:37] @_@ [17:37] dh++ [17:37] at least to get one from persia or me, other motus who shall remain anonymous were nicer :P [17:38] * Quintasan wonders what happened to persia [17:38] thanks to apachelogger I have a minor headache right now ;) [17:38] oh yeah [17:39] we haven't seen him since the earthquake [17:39] He wasn't employed by Canonical, was he? [17:39] nah [17:40] he was a contractor for them iirc [17:40] he has a tendency to vanish at times [17:40] kubotu: seen persia [17:40] persia was last seen 4 months, 1 day, 17 hours, 5 minutes and 5 seconds ago, quitting IRC (Quit: Coyote finally caught me) [17:40] uh oh [17:40] Who/what is Coyote? [17:41] this http://en.wikipedia.org/wiki/Coyote ? :P [17:42] that has been his quitmessage for ages [17:42] I know what kind of animal coyote is but I can't imagine persia running away from a coyote in middle of Tokyo [17:42] Ah, I see [17:42] :S [17:42] heh [17:44] hm [17:44] new horror movie [17:45] coyotes in tokyo [17:45] Probably enlarged by the radiation from the nuclear melt down. [17:45] oh my [17:46] godzilla vs. coyote [17:46] Released by the earthquake from their primordial prison ... [17:47] the script is basically writing itself [17:55] lol [17:59] apachelogger: oh btw, so lets say upstream updated the dpointer, that would not constitute ABI breakage because dpointers are meant to prevent exactly that correct? [17:59] shadeslayer: yes with a dpointer you can add new things [17:59] otoh if they add new API calls, that would constitute ABI breakage? [17:59] but stuff like virtual methods not [17:59] KDE has a wiki page [18:00] http://techbase.kde.org/Policies/Binary_Compatibility_Issues_With_C++ [19:10] shadeslayer: ping [19:10] pong [19:10] do you have regexp skillz? [19:11] *very* basic regexp [19:11] shit [19:11] like I know what * dos [19:11] *does [19:11] I can use awk a bit [19:12] Quintasan: yofel or apachelogger might be able to help [19:16] uhm [19:17] does Qt not support opengles on ARM platforms? [19:17] we have libqt4-opengl-dev but no libqt4-opengles-dev or libqt4-opengles [19:38] Quintasan: pong [19:38] dah nvm [19:38] got it [20:57] * ScottK wonders if shadeslayer broken anything yet. [20:58] heh, nope, I'm coding [20:58] ScottK, heh [20:58] Riddell, sorry i went away by time you replied yesterday :) [21:02] hi koolhead17|away [23:15] anyone else having issues with plasma freezing or being slow in beta 1?