[00:38] BarkingFish: How's it going? [00:39] poorly right now [00:40] Need assistance or still working through it? [00:40] still working through it, will need to sign off soon as it's aiming for 2am here [00:40] i'll have one last go on scenic, then it's quits for the night [00:41] OK. [00:42] dammit built on all architectures, so it's officially done. https://launchpad.net/ubuntu/oneiric/+source/dammit/0~preview1-3ubuntu1 [00:42] woohoo! [00:43] * BarkingFish grabs a can of welch's grape soda out of his usb fridge to celebrate [00:55] ScottK, Scenic failed to build again, missing libv4l-dev, and it's not installing with all the other depends. I've installed the local package for natty, and that's not doing it. [00:56] myxzptlk [00:56] Add lib4l-dev to build-depends in debian/control [00:56] ok [00:56] one moment [00:56] Then add a debian/changelog entry [00:56] Then debuild -S (to make an updated source package) [00:56] Then try to build that in pbuilder. [00:59] * ScottK notices qtruby and goes to work on that. [01:00] can I just check, it's already at 1ubuntu1, moving it to oneiric, do I increment that to 1ubuntu2? [01:01] this will be my last try, it's gone 2am now [01:02] ScottK ^^ [01:05] I'll get this done, then I really have got to get to bed. [01:09] Yes [01:09] I'll increment it anyway, just to test the build. if it's wrong, i'll fix it later :) [01:09] Oh ok [01:09] dch -i should do that for you [01:10] it did, but I wasn't sure if it was right or not, because it marked the change for natty :) [01:10] If you're on natty, building for natty is the debchange default. [01:11] aha [01:11] The revision numbering heuristics are (generally) independent of the release. [01:11] I have to change that once I'm doing the entry though [01:12] i gotta admit, dch -i is easier than manually formatting the changelog :) [01:12] it's picked up libv4l-dev this time [01:12] i just saw the GET while it was building [01:17] qtruby uploaded [01:17] Progress. [01:19] Doing perlqt now. [01:20] another failure. it's moaning about something else nwo [01:20] *now [01:20] configure: error: Not all gstreamer plugins installed [01:20] make: *** [debian/stamp-autotools] Error 1 [01:21] Probably another missing build-depends. [01:21] it's mentioning a missing plugin called v4lsrc [01:22] !info v4lsrc [01:22] Package v4lsrc does not exist in natty [01:22] !info v4lsrc oneiric [01:22] Package v4lsrc does not exist in oneiric [01:22] right, I'm gonna sleep on this, Scott. It's 2.22am, and frankly, I need to get some rest [01:23] Doing smokekde. [01:23] Thanks for the help with dammit, I really appreciate getting my foot in the door :) [01:23] BarkingFish: OK. Have a good night. [01:23] I will, don't you worry :P [01:23] see ya [01:35] Anyone remember why we didn't package mobipocket? [02:14] Doing kdeadmin [02:19] Dear apachelogger, please upload stuff after you wake up. [02:28] Doing kdepim [02:44] Done. [02:56] ScottK: I'm sure you're aware you have a bunch of arm failures for kde stuff [02:56] Yep. [02:57] It's inevitable unless we want to wait the 12 hours it takes kde4libs to build on arm to upload the next package. [02:57] Soyuz build-dep handling FTW. [02:57] ah, it's archive skey [02:57] *skew [02:57] Depends on how you look at it. [02:58] I look at it as Soyuz being unable to handle common situations automatically and so it fails the build when it should really do better. [02:59] ScottK: ah, it should be depwait, but it's not [03:00] It should more properly be bd-uninstallable, but that's not impelemented in Soyuz. [03:00] couldn't you just depends on kdelibs5-dev >= VER instead of Breaks < VER? or it it not really a depends? [03:00] It would fall over anyway. [03:00] oh, there are multiple ones... [03:01] OEM services doesn't need it, so it's not an LP priority for Soyuz development. [03:01] is there a bug filed at least? [03:02] this is a common situation as you mentioned [03:02] I don't know. [03:03] I know that the Soyuz developers, back when there was such a thing, were well aware of it. [04:00] Doing kdeaccessibility [04:07] kdegames too [04:18] kdepim-runtime. [04:22] kdemultimedia [04:36] pykde [05:16] ScottK, micahg: Solving archive skew dependencies is not trivial. [05:16] wgrant: I have faith in you :) [05:17] It requires us to regularly load the indices and run a full apt resolver for every depwait build. [05:17] Having bd-uninstallable like Debian owuld help. [05:17] And, well, there is no us any more. [05:17] That's the biggest problem. [05:17] Right, same thing. [05:17] Difficult for us, because we have so many more builds and many archives. [05:17] So just do it for the primary archive. [05:17] The primary archive? [05:17] There will soon be dozens. [05:17] The Ubunt uone. [05:18] An Ubuntu special case is unlikely to be prioritised. [05:18] However, Ubuntu could do this through the API. [05:19] Yes. I hear that supporting Ubuntu is a priority for LP, except when it actually requires something to be done. [05:19] (not blaming you) [05:19] Heh, indeed. [05:20] Would it really be possible to achieve automatic retries of bd-uninstallable packages through the API? [05:21] Sure. There are API methods to identify failed builds, read logs, and retry builds. [05:21] Then you just need to plug that into edos-debcheck or whatever Debian uses now. [05:21] I wonder if that code is around somewhere. === amichair_ is now known as amichair [07:45] does kde4's replacement for kicker have a specific name? (wikipedia only shows plasma) [07:57] morning [08:19] scar[w]: replacement can be called a "Plasma panel" I think [08:19] scar[w]: the panel being a container for plasma applets [08:20] thank you, want to used the correct terms when submitting bugs [09:35] ScottK, debfx: question: where do you track what is uploaded and what is not? [09:42] apachelogger: http://felix.fobos.de/kubuntu/kubuntu-buildstatus.htm gives you an overview [09:43] didrocks: has micahg talked to you about the diginotar qt patch? [09:43] why you no use software to track actual uploads? [09:43] not landing in lp [09:43] on a related note [09:44] why you no use software to upload all the stuff at once? [09:44] debfx: yeah, I already fixed it yesterday when I saw the blog post btw :) it will be in today's upload (just after unity-2d is uploaded) [09:45] ah, great :) [09:45] apachelogger: if only all packages were properly reviewed [09:46] ah [09:46] ok [09:46] easy [09:46] introduce a review state [09:46] so one core-dev reviews, the other takes care of upload and rebuilds where necessary [09:48] * apachelogger thinks a uscan'd list of core packages would also be good [10:04] apachelogger: I don't see the advantage of a batch upload. once you've reviewed a package you can just upload it [10:06] except you need to update changelog, debcommit, build source package and upload source package [10:07] that said, of course the person doing uploads can go ahead and do reviewing should they run out of reviewed packags [10:07] major advantage is that reviewing can be done independently of whether we can upload or not [10:08] i.e. reviewing can be concurrent with packaging AND uploading [10:17] when are you not able to upload packages? [10:17] due to limited upload bandwidth? [10:24] debfx: for example, also when frozen, also when not everything is packaged yet [10:25] some packager updates kde4libs -> core dev could immediately review while some packager continues with kdepimlibs [10:25] when some packager is done and everything is tested some other core dev starts uploading kde4libs while core dev is still reviewing [10:26] chances are that core dev will be done reviewing before some other core dev is done uploading [10:26] so we'd have buffers on each side of the pipeline [10:26] and reviewing would become a non-blocking task for the better part [10:27] you can upload all of kde sc even when kdelibs isn't built yet [10:27] they will just fail to build and we need a script to retry those anyway [10:30] debfx: not building! [10:30] debfx: packaging! [10:30] lemme make you a graphic :P [10:33] * apachelogger constantly gets distracted [10:46] good lord kdeartwork is fat [10:49] debfx: https://docs.google.com/drawings/d/17X0nfJFv-bNCHI3GFYpta2Om3pCt3x15gMKxdwuI2SQ/edit?hl=en_US [10:59] yeah, I just think that not being able to directly upload after reviewing doesn't happen very often [11:04] debfx: so what is the workflow you have in mind? === rickspencer3__ is now known as rickspencer3 [11:12] apachelogger: I think our current workflow is fine except that the wiki is a bad place to track who is working on what [11:13] because clicking the edit button, scrolling to the package you want and saving the whole thing is very inconvenient [11:28] debfx: but our current workflow is totally non-concurrent :S [11:30] debfx: https://docs.google.com/drawings/d/17X0nfJFv-bNCHI3GFYpta2Om3pCt3x15gMKxdwuI2SQ/edit?hl=en_US [11:30] even the graphic shows why that is bad :P [11:58] hallo everyone [11:58] hi [11:58] i won't be available for the next 10 days, should i like send a email to kubuntu devel? [11:59] lp spams masivly I got ~200 mails of lp because of my packages [11:59] would make sense [11:59] bulldog98: heh, say hi to rosseta [11:59] * bulldog98 wants lp to stop spaming each and everyone [12:08] shadeslayer: who is community council? [12:10] * bulldog98 closes the poll [12:11] bulldog98: You mean, who's in the community council? [12:11] nigelb: yeah, but I already found the lp page [12:12] Riddell: do you have time for a meeting at 10 pm 13th of september? [12:19] bulldog98: yes should be fine [12:25] * bulldog98 finalises it than [12:26] bulldog98: pad.lv/~kubuntu-council i think [12:26] shadeslayer: ? [12:27] bulldog98: invite everyone who is in that team [12:27] ScottK: pokety poke [12:27] shadeslayer: you mean lp.net/~kubuntu-council i think :P [12:28] bulldog98: copy paste that link into yer browser [12:28] and see the magic [12:28] pad.lv is the launchpad url shortner service thingy like bit.ly and goo.gl [12:29] ScottK: ever setup debian using the buisness card ISO ? [12:29] * shadeslayer is lost in the setup [12:43] shadeslayer: Nope. Sorry. [12:43] apachelogger: We've just been coordinating on IRC. [12:44] ScottK: ah alright [12:45] debfx: Did you make any progress with Quintasan_ on -runtime? [12:46] apachelogger: Please do some uploading. Everything to the left of pimlibs (except korundum and perlkde) is already uploaded. [12:46] ScottK: yes, but it still needs fixing [12:46] OK. [12:47] the install files need to be updated now that active is actually separated [12:47] Ah. Right. [12:49] how did that even end up in the archive? [12:49] ScottK: I did artwork and then ended up being busy with phonon stuff again [12:49] it is madness really [12:50] Come on, we need to get this done. [12:54] * ScottK does marble [12:59] btw here's the qtwebkit log : http://cl.ly/1p1t411J2g0a2l1b0t1w [13:02] Now libkexiv2 [13:02] shadeslayer: We want to update, I'm reasonably sure. Please make an FFe. [13:03] ScottK: right, but what all info am i supposed to provide, Riddell and apachelogger tested the package and they said the upgrade went fine, i'll also attach the qtwebkit log, anything else? [13:03] Build logs. [13:03] Any notable changes in the packaging. [13:04] Also noting that it fixes a beta 1 bug is good (Riddell's form thing) [13:04] ok, i'll file the FFe then [13:04] yep [13:05] apachelogger: I have to thank you for pointing out debcommit -r -R. For once someone made a tool that actually does exactly what I want. [13:34] shadeslayer: Can we have some bug numbers for bugs it fixes? [13:34] Not sure if they actually got filed or not. [13:35] [klettres] Michal Zajac * 10 * debian/changelog releasing version 4:4.7.1-0ubuntu1 [13:35] Doing kalgebra [13:35] ScottK: i think they're mentioned in the git log [13:35] shadeslayer: Yes, but no one will go back and close them based on that. [13:35] ah, you mean launchpad bugs [13:35] i'll have a look [13:35] most of them seem to be related to arora [13:41] apachelogger: Suitably magical. I'll try that too. [13:55] we need to get the rekonq beta into oneiric [13:56] I can do the stuff tomorrow if needed, but I won’t have time today [13:56] shadeslayer: ^^^ opinion? [13:56] bulldog98: What's their schedule for a final release? [13:56] bulldog98: please make sure that it works with the webkit from my ppa [13:56] ScottK: ask in #rekonq [13:57] shadeslayer: it does, since I use that :) [13:57] bulldog98: If you want it in, that's information that'll be needed. [13:57] especially stuff like flash, rendering of sites like youtube, ubuntu.com, kubuntu.org, kde.org etc [13:58] shadeslayer: I don’t use flash, but you can do a git clone kde:rekonq and build and test [13:58] i already use trunk rekonq in neon ;) [13:58] and it wfm [13:59] bulldog98: To get an upgrade this late in the cycle, someone will have to be found to do that kind of testing on it. [14:00] ScottK: the problem is my internet speed, but I can do that stuff [14:00] bulldog98: Excellent. [14:01] ScottK: I’ll do that tomorrow, since I have to go in about 30min [14:01] OK. [14:02] ScottK: do you by chance have a machine running debian sid? [14:02] shadeslayer: Not at the moment. I generally do stuff in a chroot on an Ubuntu system. [14:03] What do you need? [14:03] ScottK: a sources.list for upgrading to debian sid [14:03] You can get that by loging into a pbuilder chroot. [14:04] s/loging/logging/ [14:04] ScottK meant: "You can get that by logging into a pbuilder chroot." [14:04] looks like this : http://paste.kde.org/119503/ [14:04] ah ok [14:05] That looks right. [14:17] shadeslayer: What's up with the indi-apogee package in Ninjas? There's nothing private about it, so please either delete it or copy it to a public PPA. [14:18] looking [14:23] ScottK: that package needs sponsoring/uploading [14:23] Doing korundum. [14:23] should i put that that on revu? [14:23] shadeslayer: OK. How about copy it to staging. [14:23] (or your PPA) [14:23] i'll copy it to my ppa [14:24] Then delete it from Ninjas please. [14:27] ScottK: done [14:27] Thanks. [14:32] shadeslayer: I’ll try that tomorrow [14:32] bulldog98: also try commenting on a bug in staging.launchpad.net [14:42] doing perlkde === scar[w]_ is now known as scar[w] [14:47] * ScottK notices no one else is uploading and pokes at apachelogger, Quintasan, and debfx (although if they are still sorting -runtime, that's no doubt more important) [14:48] Looking at webdev [15:16] wth === Mamarok_ is now known as Mamarok [15:22] debfx: didrocks already had the patch :) === dpm is now known as dpm_ === shadeslayer is now known as shadeslayer_ === markey_ is now known as markey === seaLne_ is now known as seaLne [15:36] sdk now. === shadeslayer_ is now known as shadeslayer === apachelogger_ is now known as apachelogger [16:19] Doing gwenview [17:17] Now konsole. [17:17] * ScottK sighs at the lack of participation. [17:18] apachelogger: dchr is nice too. [17:24] alright guys, cya in 8 days :) [17:38] @_@ - core crash [18:03] ScottK: sorry, the unity-2d releases is segfaulting, and no more dx guys, I'll have to delay the Qt upload for tomorrow [18:04] OK. [18:04] sorry again, but EINMIDDLEOFTWOTEAMS :-) [18:04] Sure. [18:05] We're still uploading KDE 4.7.1 (since no one is helping me), so it's fine. [18:05] I'd rather finish that first anyway. [18:08] ScottK: urgh, good luck with that! [18:09] Only 33 packages to go. [18:14] ScottK: I'll upload wallpapers and artwork if you didn't yet [18:15] apachelogger managed to upload artwork. [18:15] I think wallpapers still needs doing. [18:15] ah, he didn't update bzr though [18:16] right, artwork is up [18:17] yofel: Almost all the ovals to the right of libkipi need uploading: https://wiki.kubuntu.org/Kubuntu/Ninjas/DependencyGraph [18:17] I've been deleting stuff from ninjas as it's uploaded, so anything still there is fair game. [18:18] ScottK: I hacked a branch overview together a short while ago that I was looking at just now http://yofel.dyndns.org/kubuntu/bzrstat/ [18:18] Nice. [18:19] [lp:~kubuntu-packagers/kubuntu-packaging/kdeartwork] Philip Muškovac * 120 * debian/changelog releasing 4:4.7.1-0ubuntu1 [18:19] refreshes every ~10m [19:05] [lp:~kubuntu-packagers/kubuntu-packaging/kde-wallpapers] Philip Muškovac * 13 * debian/changelog releasing 4:4.7.1-0ubuntu1 [19:20] oh yeah [19:21] ScottK: Fixing runtime [19:21] debfx: ping [19:21] Excellent. [19:21] That's critical. [19:22] debfx: http://paste.kde.org/119599 <- any other files that are missing? [19:23] rbelem: ^ you forgot to remove active files from the "classical" kde-runtime.install [19:23] Oh wait, now that parallel building works I can build the crap out of it [19:24] Quintasan: yes, I'll fix it [19:24] debfx: I can just push it now [19:25] debfx: Or well, I won't get in your way, if there is anything else wrong, let me know [19:26] * Quintasan cries from joy [19:26] IT'S ALIVE [19:28] [klettres] Michal Zajac * 10 * debian/changelog releasing version 4:4.7.1-0ubuntu1 [19:29] * Quintasan forgot to push [19:29] * Quintasan looks at kstars [19:29] Actually not [19:30] ScottK: Care to do kstars? [19:30] Quintasan: Not now. Doing $WORK and fixing a security bug in quassel [19:30] BTW, anyone using a quasselcore: If it's crashing, disable CTCP until you get the fix. [19:31] ScottK: Hmm, I'll do it then, it's just uploading that takes TIME [19:31] yofel: ^^ [19:31] Yes, but you can prepare the next one while that's uploading. [19:32] Quintasan: kstars: can do if you want, quassel: I already patched my core [19:33] ScottK, what is the procedure for rebuilding and fixing a package which has a missing build depend, when it's already published and in use by people? Can the package still be updated to ensure the depend gets installed at the same time? [19:34] Yes. [19:34] yofel: Cool, please do kstars [19:34] Usually what causes this is that some other build-dep used to depend on the missing one and so it worked before, but only because of the indirect connection. [19:35] So making a new revision that adds it explicitly and updating the package is the general solution. [19:35] BarkingFish: If it's in oneiric then you just update it and ask someone to upload it, if it is in natty or older then you will have to do what we call SRU [19:35] great, I've got one I need to rebuild, since it's a pain in the backside. Sweethome3d needs the sun-java6-jre, but it doesn't install it when you install the program, but it doesn't install it if it doesn't find it [19:35] Quintasan, It's in natty [19:35] I think it's SRU then [19:36] BarkingFish: it doesn't work with openjdk? [19:36] yofel, no [19:36] :/ [19:36] debfx: Actually there is another missing file ->http://paste.kde.org/119611 [19:36] It specifically asks for the sun java6 jre [19:37] BarkingFish: Wait, it's not a missing build-depend, it's a missing Depend :) [19:37] * yofel wonders what'll happen when we stop having sun-java in the archive... [19:37] well, partner [19:37] Build-Depends are installed when (obviously) building the source code and Depends are installed when the user wants to install the software we package [19:38] * Quintasan would actually name the field Runtime-Depends [19:38] Quintasan: yep, I've uploaded runtime [19:38] Quintasan, sorry, my mistake. Either way, what do I do with it? [19:38] BarkingFish: Well, you need to update the packaging, submit a SRU bug to Launchpad with a debdiff and subscribe ubuntu-sru I believe [19:39] BarkingFish: https://wiki.ubuntu.com/StableReleaseUpdates [19:39] ok [19:39] BarkingFish: also, it has to be fixed in oneiric before you do the SRU [19:39] (usually) [19:39] debfx: Thanks! [19:40] [parley] Michal Zajac * 18 * debian/changelog releasing version 4:4.7.1-0ubuntu1 [19:41] ok yofel, so in which order do I go about this? I need to get the source for it, but do I fix it and package it for oneiric first, then fix natty, or what? [19:41] Also, what exactly do I fix, since it's not a build-depend [19:41] first devel release, then stable [19:41] BarkingFish: you add it to the Depends: of the binary package [19:42] although I'm not too sure about java stuff [19:42] where am I likely to find that in the source tree? [19:42] debian/control [19:42] oh ok then [19:42] np [19:42] BarkingFish: also, join #ubuntu-motu and #ubuntu-devel [19:42] #ubuntu-packaging is also good place to hang out [19:44] BarkingFish: What package are we talking about? [19:44] ScottK: sweethome3d [19:44] Where the hell is bulldog98 [19:44] He was supposed to delete his branches which we merged [19:45] Parley uploaded [19:46] ScottK, sweethome3d [19:46] last release was sweethome3d_3.1+dfsg-1ubuntu1 [19:47] If it really needs sun java, it needs to get moved to multiverse. [19:47] Quintasan: https://code.launchpad.net/~bulldog98/+activereviews doesn't show any 4.7.1 stuff from him anymore [19:48] ScottK, ok, well I'm rebuilding it for oneiric now, will debdiff it and put it up for review, then rebuild it for natty once I know it's ok [19:49] [lp:~kubuntu-packagers/kubuntu-packaging/kstars] Philip Muškovac * 6 * debian/changelog relasing 4:4.7.1-0ubuntu1 [19:49] bah, typo [19:54] Krap [19:54] * Quintasan goes to bed [19:54] hmm [19:55] On a second thought. I'll upload one more thing [19:55] * Quintasan takes kturtle [19:56] kstars up [20:02] [kturtle] Michal Zajac * 12 * debian/changelog releasing version 4:4.7.1-0ubuntu1 [20:03] kturtle up [20:03] * Quintasan goes off to bed [20:03] Good night [20:06] * debfx takes kdenetwork [20:38] Quintasan: interesting, kde-runtime ftbfs on the archive buildds but builds fine locally [20:38] Patch kubuntu-mobile-04-Ported-the-new-plugin-system-to-active-development-4.patch does not remove cleanly (refresh it or enforce with -f) [20:49] Hey Riddell === Linkmaster_ is now known as Linkmaster [21:33] hola [22:35] I've got to find some way of patching pull-lp-source so it doesn't keep printing those +++ signs while it's downloading stuff [22:35] it's got to the stage with some packages where I have to minimise konsole while I'm doing it [22:39] BarkingFish: I think /usr/share/pyshared/ubuntutools/archive.py line 312 is to blame for the dots (and the while loop it's in) [22:40] what I want to do is basically add a command line option, like --hide-progress or something, just to switch off those crosses and their annoying nauseating repetitiveness [22:41] well, pull-*-source just use pull() from the SourcePackage class in ubuntutools [22:41] so you'll have to add it there [22:41] surely a progress counter like [50%] or something like that would be easier [22:41] and less dangerous [22:42] currently it looks like they download the files in 10KiB chunks and print a dot for each chunk o.O [22:42] which is crazy if you download something like oxygen-icons... [22:43] yeah, or the entire sun-java6-jre [22:43] which filled 12 screens with nothing but ... [22:44] well, start by filing a bug against ubuntu-dev-tools [22:45] ok [22:48] * yofel is off to bed, good night [23:03] ScottK, hi if you're about. Just a quick note on one of the rebuilds against oneiric, plee-the-bear is physically beyond help. I just tried to build it, it got to 16% and stopped dead - so many errors in the code I can't even begin to number them all. At first glance, I'd declare this "life extinct". [23:03] I've got I'd say, probably in excess of 40 compile errors just in the first 16% of the build [23:08] yofel - instead of filing it as a bug, I've started off by asking it as a question on the launchpad, as it says these can be turned into bugs later. [23:08] https://answers.launchpad.net/ubuntu/+source/ubuntu-dev-tools/+question/170609