/srv/irclogs.ubuntu.com/2012/11/30/#kubuntu-devel.txt

yofeluploads rejected: ffmpegthumbs, kfloppy, kiten and pairs. Someone else will have to upload those00:02
yofel(not in packageset for quantal)00:02
ScottKyofel_: Yes.  Also need meta-kde.04:51
ScottKyofel_: I'm rejecting packaging that don't have any upstream diff.  Don't be alarmed.04:52
ScottKyofel_: kdenetwork needs a reupload to not overwrite the debian/changelog entry for the SRU that's already in updates.05:01
ScottKyofel_: The rest is fine.05:03
=== valorie_ is now known as valorie
shadeslayermmm09:42
shadeslayerpeople here asking for PA3 for quantal :)09:43
shadeslayerwe should get cracking on that ASAP09:43
Riddellyeah plenty of requests09:43
* shadeslayer is at the KDE Miniconf @ FOSS.in \\o/09:44
shadeslayerwhoops, sebas can't load the RSS widget in touch mode09:57
shadeslayer!find plasmate10:15
ubottuPackage/file plasmate does not exist in quantal10:15
=== Tonio_aw is now known as Tonio_
=== Tonio__ is now known as Tonio_aw
agateauafiestas: ping10:30
=== Tonio_aw is now known as Tonio__
=== Tonio__ is now known as Tonio_aw
=== Tonio_aw is now known as Tonio__
BluesKajHowdy all13:28
Riddellhmm, kubuntu-devel hasn't been moderated in a while13:35
* ScottK didn't know it had a moderation queue.13:37
Riddellit usually doesn't since any address not subscribed of ubuntuy gets rejected13:37
Riddellbut some merge requests don't get through and laney's post there was over the size limit13:37
Riddellso remember to look at the date on those e-mails :)13:38
ScottKyofel_: How goes finishing up 4.9.3?14:19
yofel_bad, I'm at work, so not really much time for it right now14:20
yofel_won't get anything done for at least another 3h14:21
ScottKOK.  I'd like to wait for powerpc to catch up some more before accepting, so no rush.  I'd just like to get it sorted today.14:21
ScottKRiddell: Could you upload ffmpegthumbs, kfloppy, kiten and pairs for yofel (4.9.3 for quantal-proposed) so I can accept them later.14:22
yofel_sure14:22
afiestasagateau: pong14:27
RiddellScottK: oh yes I forgot to note that down last night on my todo14:28
ScottKThanks.14:28
agateauafiestas: hey, I have been thinking more about our yesterday discussion regarding packages for beta versions14:28
agateauafiestas: the more I think of it, the more I think beta tarballs are a practice of the past14:28
agateauDuring beta period, master should be as stable or more stable than beta release14:29
agateauMeans people can use packages built from master to test instead of packages build from beta tarball14:29
agateauSo something like project-neon would be good enough for testers, maybe even better since it does not mess with "stable settings".14:30
agateauit also means it is easier for a tester to test a fix after a bug report14:30
agateauno need to wait for beta+114:30
agateaufreeze schedules still makes sense, but I am not sure creating tarballs and even tagging beta is useful nowadays14:31
agateauthat is: if we get other distros to provide nightly packages as well14:31
agateauafiestas: what do you think?14:31
afiestasagateau: I think that if you rewatch the KDE tea time we did a few days back you'll see that I hold that opinion as well14:31
agateauafiestas: yes, that's why I was a bit surprised you wanted packages for beta releases14:32
afiestasagateau: for the future I want a lot of things, but for tyhe present I want different stuff14:32
agateau:)14:33
agateauafiestas: project-neon is not the future, though14:33
afiestasin the future, yes we need a more modern release cycle for multiple reasons, but until then Kubuntu should package things before anyother distro14:33
afiestas*any other14:33
afiestasI was a little bit "shocked" when jussi said neon is not good enough to have real packages14:34
afiestaswe need real packages for master, always asap14:34
agateauthey are not good enough to get into the main archive, but they are good enough for testers14:34
afiestasmaybe we have to make upstream modify his behaviour by announcing new dependencies and stuff like that14:34
afiestaspackages should not differ technically, they may legally (copyright and stuff like that)14:35
afiestasbut technically (dependencies, features) they should be the same14:35
afiestasif not #fail14:35
afiestas (because users won't be testing the real thing)14:35
agateauthey also differ by the amount of splitting14:35
agateaufor example: calligra in neon is one single package, whereas it is splitted in the archive14:35
afiestasI can see an user using neon oh look everything works ! I got everything!14:36
afiestasthen he isntalls the final packages, forgets to install a split package and "meh mtp is not working on stable while it was working in unstable"!14:36
yofel_project-neon differs quite a bit technically due to the idea to install it side-by-side with a stable KDE setup14:36
yofel_the 'put everything into one package' thing is mostly a thing of maintainability14:37
afiestasyofel_: that should be a PREFIX and maybe a few dependencies that can't be added to the archive/bacjkports/thingy14:37
afiestasright now on my system I have distro KDE, KDE 4.10, KDE 4.9 and Frameworks514:37
agateauyofel_: note that I am not against putting everything in one package for neon14:37
agateauyofel_: it makes sense for nightly builds to follow upstream split imo14:37
yofel_agateau: I didn't assume you did ;)14:38
simplewnow kdelibs lock screen appears different, with a kubuntu background and the dialog to enter the password also different, is this something done with a patch in kdelibs?14:38
yofel_any more splitting means tracking what files are installed where. That gets tricky if upstream changes things all the time14:38
agateauafiestas: set up is a bit more involved for things like akonadi14:38
afiestasagateau: why's that?14:39
yofel_simplew: raring? KDE 4.10 has a new screenlocker14:39
agateauafiestas: you need to define quite a few env vars if you don't want to mix akonadi versions14:39
simplewyofel_: raring kde 4.9.8014:39
agateauafiestas: akonadi does not use $KDEHOME for example14:39
afiestasagateau: XDG vars, so?14:39
yofel_simplew: that's just the new KDE screenlocker then14:39
agateauafiestas: yes14:39
afiestasI'm afraid of feedback not being 100% reliable because users have literally different things isntalled14:40
simplewyofel_: but that was a patch submited where, in kdelibs?14:40
afiestaswe need to not only test KDE but also Kubuntu setup14:40
yofel_simplew: I don't know - I think it's part of workspace14:40
afiestasagateau: btw, if master were stable, why do we need project neon?14:40
agateauafiestas: I did not say master is stable14:40
agateau(unfortunately)14:41
agateauafiestas: I say it is as stable as beta when we start freezing14:41
afiestasDuring beta period, master should be as stable or more stable than beta release14:41
afiestasoks, well there we differ then14:41
simplewyofel_: when i move the mouse to get the dialog to enter the passowrd, the dialog does NOT get focus, to be able to enter the password i need to first go with the mouse and click in the password field14:41
agateauafiestas: I am not saying I like it that way14:41
afiestasin my ideal world of the future, master should be prepared for release anytime14:41
agateauafiestas: I agree with this14:41
agateauafiestas: but right now (ie, for 4.10) I think we can consider master stable enough to be used by testers14:42
afiestasfor Akademy2013 I want to hold a round table or something to talk about this14:42
afiestaswe must act 14:42
agateauafiestas: I'd rather have testers use nightly builds of master than builds of beta114:42
afiestasI always use master, and only very very very few times I'm annoyed by it14:42
agateauagain, for 4.1014:42
agateauafiestas: this is a different topic14:43
agateau(and I agree with you we need an always-releasable master)14:43
afiestasfor 4.10 I'd stick with normal releases, we have developers that give value to that14:43
simplewyofel_: i just prefer to use kde default unloack screen/dialog, that one gets focus when you move the mouse, and i prefer the black background, the kubuntu background image isnt that pretty...14:43
afiestasand since ti is the stablished way we should respect that14:43
yofel_simplew: that's the default KDE background, we didn't patch anything there14:43
ScottKsimplew: It's an upstream change in KDE.14:44
agateauafiestas: What I am saying is: for 4.10 should we encourage testers to use project-neon to test, instead of beta1 packages (which do not exist)14:44
yofel_simplew: and the screenlocker you get now is the new KDE default14:44
yofel_simplew: if it has bugs report them on bugs.kde.org, but someone else will have to tell you against which component14:44
simplewScottK, yofel_: but i dont see anywhere a way to configure unlock background image 14:44
afiestasagateau: and about that I'm saying that I disagree, we should provide real packages for the releases upstream does14:45
ScottKsimplew: I'd ask on #kde.  It's not Kubuntu specific.14:45
afiestasfor various reasons, 1-Upstream developers expect bugs to have a release14:45
afiestas2-Project neon packages re different from normal packages14:45
afiestasthat's it, 214:45
afiestasxd14:45
simplewScottK: but seams that was submited by a kubuntu developer, since it uses kubuntu default background image...14:45
afiestasI was expecting more reasons :p14:45
agateau:)14:45
agateauafiestas: I think 2) is not an upstream concern14:46
ScottKsimplew: I don't think so.  I think it uses system default, which for Kubuntu would be the Kubuntu image.14:46
afiestasagateau: well, it should be ours14:46
agateauafiestas: we have a problem with versions indeed14:46
simplewScottK: hum14:46
afiestasit is conceptualy broken ask people to test something that won't be used in 13.0414:46
agateauafiestas: if we were to use nightlies we would need a way to publish their commit-id when bugs are reported14:46
agateauafiestas: I am not talking about Kubuntu here14:47
afiestasso let's wait for 4.12 and do this together with upstream14:47
agateauafiestas: I am talking about testing KDE14:47
afiestasI'm ok on doing nighties as long as packages as real ones, not neon's14:47
Riddellyofel_, ScottK: uploaded14:47
yofel_thanks14:47
ScottKRiddell: Thanks.14:47
agateauafiestas: I want people who are willing to test KDE and happens to be running Kubuntu, to have an easy way to test master14:47
afiestasagateau:  and you have neon for  that, what's the matter?14:48
afiestasmaybe we are talking about different things14:48
agateauagateau: I am talking about getting more people to test KDE master, whatever distro they run14:48
agateaudamn, pinging myself14:48
afiestasneon is a wonderful thing for users that want to be in the cutting-edge without risking too much, but that's about it14:48
afiestasbeta testing and master testing right now should not be messed imho14:49
agateauafiestas: why?14:49
agateauright now master == beta + some fixes14:49
afiestaswell because that's not how upstream works, that's not wht some upstream expect14:49
afiestaswe should not decide those things, (we as in Kubuntu) upstream should14:50
agateaummm14:50
afiestasand it will be a waste of time make users test neon instead of the real packages that we will ship 13.04 with14:50
agateaumaybe I should have pinged you on #kde-release then14:50
agateau(is there such a channel)14:50
afiestasimho there is nothing to do for 4.10, we are in beta already14:51
yofel_IMO people should test the beta + check with neon if a bug still exists. Everyone will agree that we need to get faster at building the packages, which is work in progress. But I'm not too much a fan of yet another master build14:51
afiestasthis is a discussion either for 4.11 or 4.1214:51
agateauagain, I am not concerned about kubuntu there, I just want to take advantage of one kubuntu community product (neon) to get more tests of the upcoming kde sc release14:51
afiestasso they will have real apckages soon (Riddell said today)14:52
afiestasadn they will be able to test all releases upstream does14:52
agateauI (as a KDE upstream developer) would rather see people test my latest code than code from 3 days ago14:53
afiestasbut you are only one, maybe others don't, so the right thing to do imho is discuss this for 4.1114:53
afiestasand not mess with 4.10 since it is already going on14:53
BluesKajwell, this doesn't follow the topic , but I'm concerned about the HW recognition of the 3.7 kernel . It's freezing at my wirelessKB and mouse , with absolutely no response . 13.04 here , but had to regress to the 3.5.0.17 kernel to make things work14:54
afiestasor you can move gwenview to extragear adn do your own releases :P that's what I do with my stuff since I dont' want to follow SC14:54
agateauafiestas: I think the 4.10 release schedule does not suggest master is supposed to break between beta1 and beta214:54
agateauafiestas: but maybe you are confused about what I want to do for 4.1014:55
ScottKBluesKaj: We can't do anything about kernel stuff here.14:55
agateauafiestas: I am not saying we should review our release schedule or strategy14:55
agateauafiestas: at least not for 4.1014:56
afiestasyou are saying that we should encourage users to test KDE 4.10 with neon14:56
agateauafiestas: I think we should suggest testers to run nightly builds, because those give us the most valuable output14:56
agateauafiestas: yes14:56
afiestaswhich I'm against because users won't know how the hell report bugs14:56
agateauand how is it different with muon than with beta packages?14:57
afiestasthey wuill say "Beta1" while in reality it won' tbe Beta114:57
afiestasbut Beta1+patches14:57
afiestasthat makes the user feedback less valuable and less trustworthy because we won't know (we as in upstream) in which commit exactly the user is14:57
BluesKajScottK,  I'm not asking for a fix , let me rephrase , I guess this is a warning and have you guys seen or heard anything about this kernel module problem , because #ubuntu+1 is strangely silent on this issue.14:57
afiestasbesides, giving test to neon packages haas a serious downside for Kubuntu, which I'm also concerned about14:58
ScottKIt sounds very hardware specific.14:58
afiestasif you read for example Martin G blog you will see many references to beta releases14:58
afiestasasking specially for testing iun "Beta2 packages" and things like that14:58
afiestasif you make Beta2 not be Beta2 anymore but instead beta2+patches it may get messy14:58
agateauafiestas: well, you can replace those references with "YYYYMMDD"14:58
agateauwhich is the info you get from neon packages14:59
afiestasYYYYMMMDD of when neon packaged, not corresponding exactly to git14:59
afiestassince neon is slow according to shadeslayer 14:59
afiestasthe packaging of neon I mean14:59
yofel_there can be quite a bit of delay14:59
agateauversion says git20121022, is it the date of the git import?14:59
afiestasand btw, why would an upstream developer have to care about something kubuntu (only) does?14:59
afiestasthis is adding overhead to upstream14:59
afiestasleave things as they are for now and fix upstream, that's what I say14:59
agateauthat is why I say:15:00
yofel_first the bzr importers run periodically, which means up to (IIRC) 4h delay, then the recipes need to build and packages need to build (add at least twice the i386 builder queue time + build time)15:00
agateau1. testers should provide correct version numbers (if not possible we are screwed)15:00
agateau2. we should encourage other distro to have similar nightly packages15:00
yofel_doesn't opensusue have nighlies?15:01
agateauI don't know15:01
agateauI'd like to gather such information on community.kde.org15:01
agateauin the Get involved page15:01
BluesKajcould someone expalin the references to KDE 4.10? I think it's confusing some people including me 15:01
afiestasagateau: I can agree with what you propose, but not for 4.10, again because it is not the STABLISHED things15:01
afiestasI'm all up for modifying things, you know that better than anyone xD15:01
afiestasI want to do releases each 3 months damn it...15:02
agateauyofel_: a build delay is ok, as long as the reported dates match those of the git import15:02
afiestasbut things should be done right, not just taking things into our own15:02
yofel_BluesKaj: we're talking about the beta testing and neon, with would both currently be 4.1015:02
agateauyofel_: iirc recipes even have the ability to include the git commit-id in the version number, right?15:02
yofel_agateau: well, the package version has the bzr revision in it, you can then compare bzr log and git log to find the hash out15:02
yofel_but that's not really trivial15:03
agateauyofel_: yes, too complicated :/15:03
BluesKajyofel_,  ahh neon ... ok , didn't work for me last yr 15:03
agateauafiestas: ok, then I am going to do it in my little corner15:04
agateauafiestas: asking kubuntu users willing to test gwenview from kde sc 4.10 to use neon15:05
agateauafiestas: and we can revisit this discussion for 4.1115:05
afiestasagateau: you can do as you please, but since you are in SC because you want to, you shouldn't do that imho15:05
afiestasif you are in SC you should follow  whatever SC says, and try to change SC ways if you want to15:06
agateauafiestas: actually that could provide some valuable feedback for 4.1115:06
afiestasif not, you should come to extragear with me, where is sunny etc15:06
agateauand where it's a mess to deal with translations15:06
agateau:/15:06
afiestasI don't see how it can, but well do as you want xD15:07
agateaufeedback on benefits/drawbacks of using nightly builds15:07
Riddellsome extragear packages depend on some SC packages which change API and ABI meaning we can't do backports easily15:07
agateau*cough* libkdcraw *cough*?15:08
afiestasagateau: well, then you should at least send an email somewhere to say "I'm going to do this with this objective"15:09
afiestaswe all should go at once in all this, that's why it is a SC 15:09
agateauafiestas: define "somewhere". kde-testing@ kde-release-team@?15:10
agateau<sarcasm>we are a SC, that's why we all use the same VCS</sarcasm>15:11
afiestasagateau: release team and kde core devel? or only release team15:12
afiestasthe thing about sending it to release team is that only a few people are there15:14
afiestasbut anyway, do as you please dude15:14
afiestasall this is a bloody mess already15:14
afiestasnobody will notice a little bit more mess, really15:14
agateau:/15:15
agateauI assume people on kde-core-devel build kde by hand, so they don't really care about nightly builds15:16
afiestasthey do care15:16
afiestasbecause they are the ones that are going to fix the bugs reported with "BEta1" though they won't be beta1 code but beta1+patches15:16
afiestasnobody is going to tell you "Nitghtly are bad" they are not15:17
afiestasthey wil (or I will) tell you, say to the user "Test BEta1 but instead test a nightly" is bad15:17
agateauif someone else but me and Benjamin committed code to Gwenview I would agree15:17
agateaubut they don't15:17
afiestasthen I go to the point before, you are part of SC15:17
afiestasyou are not releasing software by your own15:18
agateauno, I am just writing it15:18
afiestaswhich btw I would prefer if you did that (I'd like to have gwenview features when ready, nto every 6 months)15:18
afiestasbut right now you are part of SC, and SC should have an unified way of doing things, including asking user feedback15:18
ScottKYes.  Please.15:18
afiestasbut as I said, this is a bloody mess already everybody deos whatever they please15:18
afiestasso you are in your rights to do whatever you please as well15:19
ScottKThe betas are still important though for distros.  Don't forget that.15:19
ScottKWe can't redo the whole KDE SC in the archive every day or two.15:19
agateauScottK: sure, but would you be able to do it, say every week, or every two weeks?15:20
ScottKTwo weeks we could probably do.15:20
ScottKI guess it depends.15:20
ScottKWe're working on automating a lot of it.  There are just too many packages now.15:20
ScottKOnce that's working better, the bigger issue will be build capacity.15:21
ScottKI'm sure if we tried weekly, we'd get yelled at.15:21
ScottKBi-weekly we could probably get away with.15:21
agateaummm, actually every two weeks is roughly the time between sc unstable releases so that would not change much I guess15:22
ScottKIf there was an easy way to know the subset of packages to update, we could do it more often.15:22
ScottKSome kind of an API to query where the KDE maintainer would flip a "worth updating" flag or something.15:23
yofel_we could re-use the diff-check for SRU's for the normal packages15:23
agateauI wouldn't trust KDE maintainers to do this reliably15:23
ScottKYes, but I'm trying to get away from us having to read the diff.15:23
yofel_the scripts have a diff check15:24
ScottKIt could be a git tag.15:24
yofel_automated15:24
ScottKOh.15:24
ScottKThen how come you uploaded stuff with no diff?15:24
ScottK(for 4.9.3)15:24
yofel_anything that starts wth lib is not included there becaues some packages might build-dep on >= 4.9.3 of them15:24
simplewhow can i downgrade to a previous kernel version?15:25
ScottKagateau: Maybe a transitional approach would be to keep the beta/RC schedule, but define a git tag format maintainers could use, if they choose, to signal it's worth taking a snapshot.15:25
simplewwrong channel, sorry15:26
yofel_we should re-check that for those packages that should have a stable ABI15:26
ScottKThen a commit hook could monitor for the tag and maybe mail the pacakger list.15:26
Riddellafiestas: 15:27 < seb128> Riddell, ScottK: does https://code.launchpad.net/~stijnbrouwers/ubuntu/quantal/kamoso/fix-missing-icons/+merge/134772 seems fine to you?  it's basically making an icon change "webcamreceive" -> "digikam"15:27
agateauScottK: interesting, so your point is that you'd rather avoid rebuilding everything15:28
ScottKafiestas: I told seb128 it needed to go upstream.15:29
ScottKagateau: It takes a LOT of build time to rebuild the whole SC.15:29
agateauScottK: sure, and it is the same problem for other distributions, so that's something to take into account15:30
ScottKI think it'd be cool to do more updates and keep things closer to upstream if there's a good way to find the interesting times to do it.15:31
* yofel_ perpares kde-l10n and wonders what happened to the 4.9.4 tagging15:38
ScottKI accepted kde4libs for quantal so it'll have time to build everywhere.15:39
ScottKyofel_: don't forget meta too.15:39
yofel_do we really need meta?15:40
ScottKI think there's still something we use from it.15:40
Riddellyofel_: I guess the same thing happened to 4.9.4 as 4.10 beta 2 tagging15:40
* yofel_ is worried that, as some of the meta packages use >=, our no-diff rejects might not really work15:41
ScottKyofel_: Good point.15:41
ScottKMaybe we don't need it.15:41
yofel_I'll take a look at it and check if we need it after all15:41
ScottKThanks.15:42
Riddellkubuntu peeps, last chance to decide do we want an alpha next week? we do but it'll get in the way of 4.10 beta 215:42
yofel_hm, 4.9.85 had the same tagging date, right15:42
yofel_geez, time flies...15:42
ScottKYou're point about the >= is a good one.15:42
ScottKRiddell: Yes.  We want it.15:42
afiestasScottK: that seems like a workaround for a workaround 15:43
afiestasmeaning, depending on digikam icon instad of webcamreceive15:43
Riddellnew kde sc and an alpha, a busy week :)15:43
afiestasanyway, tell him to make a reviewboard and we w'll check it asap15:43
ScottKWe need it to work out the machinery of a milestone without the Canonical flavors participating.15:43
yofel_any other flavours planning to have an a1 ?15:44
ScottKafiestas: I didn't look at the content of the change just enough to see it wasn't something I'd distro patch.15:44
ScottKyofel_: No.15:44
Riddellyofel_: I think we'll be all alone15:44
yofel_that'll be fun15:44
Riddelli386/amd64 desktop images?15:44
ScottKYes.15:44
Riddellstgraber will be driver15:44
yofel_is there a reason why we still have alternate dailies?15:45
yofel_or did someone just forget to kill them?15:45
ScottKNo.  Those should go away.15:45
yofel_k15:46
=== yofel_ is now known as yofel
micahgRiddell: for owncloud, maybe you want to add a NEWS entry as that's a major change?15:48
ScottKyofel: killed.15:48
micahgRiddell: for the SRUs I mean15:48
Riddellmicahg: owncloud has no NEWS file and I'd think all users will read the index.php15:49
ScottKAdding a NEWS file is easy.15:49
micahgRiddell: index.php is too late :)15:50
micahgRiddell: you want to warn users that you're pulling the rug out from under them15:51
micahgat least I would think you'd want to :)15:51
Riddellmicahg: how will a NEWS file do that?15:51
Riddellit's in the changelog which is what muon etc shows15:51
micahgRiddell: well, in cases of cli tools, it's displayed on install, it's e-mailed to admins on servers15:52
micahgidk how muon handles it TBH15:52
ScottKRiddell: server installs don't have muon.  NEWS is the right tool.15:52
RiddellI've never seen that15:52
Riddelland I've been running a debian server for a decade :)15:52
Riddelldoes apt-get do this or something else?15:52
ScottKApt will show it.15:53
ScottKIt's only used in cases where there's a significant incompatibility.  I've only ever done it once.15:53
Riddellinteresting15:54
Riddellso I want a file in /usr/share/doc/owncloud/NEWS detailing the change?15:54
micahgdch --news?15:55
yofelwhere the hell is that documented? I couldn't find it in the debian policy15:56
micahgnot in policy, but developers reference: http://www.debian.org/doc/manuals/developers-reference/best-pkging-practices.html#bpp-news-debian15:57
yofelthanks15:58
micahgI'm shocked it's not in policy15:59
ScottKPolicy would be use a NEWS file when ...., not here's how you do a NEWS file.16:07
micahgScottK: right, but I couldn't seem to find that either16:08
ScottKAs with many things in Debian, I think it's left to the maintainer to decide.16:09
Riddellownclouds reuploaded with NEWS16:09
Quintasan_\o16:24
=== Quintasan_ is now known as Quintasan
yofelevening queuebot16:24
yofel...16:24
yofelgood evening to you too Quintasan16:24
QuintasanT_T16:25
* Quintasan throws bricks at queuebot16:25
Riddelloy, no violence in this channel!16:26
* Quintasan thorws more bricks at queuebot16:28
QuintasanIt's just a bot16:28
agateaus/bricks/lego bricks/16:31
agateaulet's nerf it16:31
* Riddell stands infront of queuebot in an act of nonviolent resistance16:32
* yofel sends a creeper in queuebot's direction16:33
RiddellI don't like the sound of that!16:34
QuintasanRiddell: Got some time to review maliit?16:43
RiddellQuintasan: throw it at me and see how far I get :)16:44
Quintasanmikhas: Could you get someone to write more extensive descriptions for packages in maliit? I only want descriptions since the current ones are not really good and I don't feel like figuring out which module does what16:44
Quintasanduh16:45
QuintasanRiddell: Wait, I accidentaly overwrote the damn changelog16:45
shadeslayersebas is going to test Kubuntu active, he says to expect a long email about broken things if he finds any issues :)16:46
Quintasanshadeslayer: how did the newpackage for our bot work?16:47
Quintasanubottu: help16:48
ubottuPlease don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) See also !patience16:48
Quintasanduh16:48
shadeslayernewpackage foo ver 16:49
shadeslayerI think 16:49
shadeslayer!newpackage 16:49
ubottuThe packaging guide is at http://developer.ubuntu.com/packaging/html/  - See https://wiki.ubuntu.com/UbuntuDevelopment/NewPackages for information on getting a package integrated into Ubuntu - Other developer resources are at https://wiki.ubuntu.com/UbuntuDevelopment - See also !backports and !sponsoring16:49
shadeslayer~newpackage 16:49
QuintasanXDD16:50
Quintasanyofel: help us!16:50
shadeslayerno kubotu 16:50
shadeslayerso ...16:50
yofeltalk to agateau16:50
yofeler, apachelogger_16:50
yofeltoday's my tab-complete failure day...16:50
shadeslayertab fail day for yofel 16:50
agateau\o/16:50
shadeslayer:P16:50
Quintasanyofel: but what was the commad format again16:51
yofelkubotu: newpackage foo ver [add. descr]16:52
yofelthe bot only runs 'newpackage' from kubuntu-dev-tools, so just use that16:52
Quintasanapachelogger_: Where be me bots16:52
ScottKshadeslayer: Can we get PA3 done first and he tries that?16:53
ScottKWe know there's breakage in the earlier stuff.16:53
mikhasQuintasan, you need a package description for each individual package?16:53
shadeslayerhuh ? I got most requests for PA3 today 16:54
shadeslayerso we should prioritize that16:54
Quintasanmikhas: Where applicable, let me send you what I have so there is something to expand16:54
shadeslayerthere's a packaging tutorial tomorrow, will try and recruit people for doing kde games 16:54
Quintasan[tools]% newpackage                                                                                                     (quintasan@demonbane:~/Sauce/tools)16:55
QuintasanTraceback (most recent call last):16:55
Quintasan  File "/usr/bin/newpackage", line 26, in <module>16:55
Quintasan    from KubuntuDevTools.launchpad import KDTLaunchpad16:55
Quintasanwat16:55
yofelScottK: kde-l10n coming, you said kdenetwork needs to be redone?16:55
yofelset pythonpath to the pylib folder or just install the package from the PPA16:55
yofelQuintasan: ^16:55
Quintasanwhat PPA now16:56
QuintasanT_T16:56
ScottKyofel: yes.  there was already a kdenetwork SRU in quantal and you dropped the changelog entry for it.16:56
yofelQuintasan: https://code.launchpad.net/~bulldog98/+archive/kubuntu-dev-tools16:56
yofelbulldog98: could you please enable raring for the recipe? thanks16:56
yofelScottK: we should probably add a check for that to the scripts16:57
ScottKYes.16:57
yofelI just took the 4.9.3 packages as we did them16:57
ScottKThat's why I get to read ALL the diffs.16:58
tsimpsonyou have kubotu now16:58
yofeltsimpson++16:58
yofelQuintasan: there be bot ;)16:58
Quintasantsimpson: Thanks!16:59
Darkwingyofel: Why did I read that is Scottys voice from ST:4 "There be whales!"16:59
yofellol17:00
Quintasanmikhas: http://paste.kde.org/61762417:01
Quintasanmikhas: for example libmaliit-glib117:02
shadeslayerhahaha 17:02
QuintasanI could leave the description like this but I can imagine people whining about the descriptions17:02
Quintasanshadeslayer: ?17:02
shadeslayerQuintasan: there be whales :P17:03
Quintasanoh17:03
QuintasanRiddell: dget -xu http://people.ubuntu.com/~quintasan/uploads/maliit-framework_0.93.0-0ubuntu1.dsc17:04
QuintasanRiddell: I think everything apart from descriptions and tests not building should be alright17:04
DarkwingFor those of you who didn't catch the there be whale ref... http://www.youtube.com/watch?v=4CM8tTG9Yig17:04
=== Blizzzek is now known as Blizzz
QuintasanMFW 600MB OF DEPS FOR MALIIT17:06
Quintasanwhat17:06
Quintasangraphvis17:07
Quintasanmikhas: Why do we need graphviz for maliit?17:07
Quintasanor it's doxygen pulling so much texmagic17:07
mikhasQuintasan, OK17:08
Quintasanmikhas: It doesn't have to be very extensive, just to let admins know that "Hey, this package does X or has files required by X"17:09
Quintasanmikhas: Generic copy and paste doesn't work (I tried) :)17:09
mikhasQuintasan, doxygen17:10
mikhasthat one pulls in graphviz for nicer dependency graphs17:10
mikhas600MB, nice17:10
RiddellQuintasan: mm, I'm being called away I think I won't be able to look at this tonight17:12
QuintasanRiddell: Sure thing, I'm still halfway there if we want tests17:13
Quintasanif we do not need them then it is ready17:13
Quintasanmikhas: make check <-- that's what's used for tests?17:17
mikhasyep17:17
Quintasanhmmmmmmmmmmmmmm17:17
mikhasno workey?17:17
Quintasanit works when I invoke it when building manually17:17
Quintasanor not17:17
Quintasanstill fails some tests17:18
mikhassend me the faillog by mail?17:18
mikhasor probably got an OBS link?17:18
Quintasanmikhas: QMAKE_OPTIONS = M_IM_PREFIX=/usr CONFIG+=disable-gtk-cache-update CONFIG+=notests17:19
Quintasanduh17:19
mikhasM_IM_PREFIX? that sounds wrong17:19
mikhascan you try just "PREFIX"17:19
mikhasand yes, enable tests if you want to run them =p17:19
* Quintasan facepalms17:20
QuintasanI'm sooooooo dumb17:20
mikhasI guess someone from us got fed up with failing tests on OBS or so17:20
mikhaseh17:20
Quintasanprobably17:20
mikhascould be me really doing similar stupid things17:20
QuintasanI just copypasted the whole QMAKE_OPTIONS from what you gave me17:20
mikhasin fact, I definitely have, in the past ;-)17:20
QuintasanM_IM_PREFIX as well17:20
mikhasright17:20
mikhasthat one is old and probalby only worked because it was using /usr anyway17:21
mikhas"worked"17:21
QuintasanQMAKE_OPTIONS = PREFIX=/usr CONFIG+=disable-gtk-cache-update17:21
QuintasanSo it should be like that17:21
mikhasyep17:21
mikhasQuintasan, what is the max length for a deb package description? 80 cols?17:22
Quintasanmikhas: Yeah17:22
* Quintasan tries testbuilding now17:22
mikhasand each new line starts with a space, right?17:22
QuintasanYeah17:22
ScottKActually 7917:25
ScottKAny volunteers to be the point person for kdevelop SRUs so I can ask to get it included in the micro-release exception?17:27
Riddellshadeslayer's been doing a good job of packaging it17:27
Riddellbut failing him I'll do it17:27
ScottKOK17:27
shadeslayeryeah, I can take care of kdevelop 17:28
Riddellyay17:29
shadeslayer:)17:29
Quintasancrap17:30
QuintasanScottK: Do you rembember what was the package that used xvfb to run tests?17:36
QuintasanFor some reason just installing it doesnt help17:36
ScottKI think openjdk does.17:36
QuintasanFunny thing17:38
QuintasanScottK: http://paste.ubuntu.com/1400149 <-- here is a small part of buildlog17:38
Quintasansee that previous tests work17:38
Quintasanbut17:38
Quintasanft_exampleplugin:17:39
Quintasanfails to connect for some reason17:39
ScottKDunno17:39
ScottKAsk in #ubuntu-x?17:39
mikhasQuintasan, http://paste.kde.org/617636/17:41
mikhasgotta go now17:41
Quintasanmikhas: Awesome, thanks!17:41
Quintasanshadeslayer: you got nvidia?17:53
shadeslayernope17:53
shadeslayerATi which i have disabled17:53
Quintasanyofel: You?17:53
Quintasan:/17:54
QuintasanDamn it17:54
QuintasanI wonder why compositing is SLOOOw when I have video playback17:54
* shadeslayer notes that all discrete cards are crap and one should just stick with intel cards17:55
rbelemRiddell:-)18:08
rbelemRiddell: is the icecc package ok?18:09
TygartI was trying to submit a bug, (plasma-desktop (0.4)) it says not enough information, 18:13
Tygarthttp://paste.kde.org/617666/18:13
TygartI was using "Crash Report Assistant"18:15
ScottKTygart: Use it to install debug packages and then have it regenerate the backtrace.18:22
yofelQuintasan: yes18:30
Quintasanyofel: You use mplayer or some other magic?18:31
yofelmplayer usually, yes18:31
Quintasanyofel: vdpau I guess, does you compositing slow down when you play a video?18:32
bulldog98yofel: ok18:32
yofelcan't say I notice anything with 304.64, output is xv in smplayer18:32
yofellemme try vdpau18:32
yofelQuintasan: I would need something to measure it, but it is a tad slower with vdpau18:34
yofelQuintasan: yeah, it is a bit slower, but I hardly notice it here18:36
Quintasanwhen I use xv nothing weird happens18:36
Quintasanwhen I try vpdau it's really slow18:36
QuintasanI could record a video but generally it gets really uhh18:37
Quintasanrough?18:37
yofelI don't have anything 1080p lying around to try it with. That might slow it down18:37
QuintasanDoesn't matter18:37
Quintasanyofel: It slows down even with 640x480 video18:38
QuintasanI'll try experimental18:39
yofelmy quadro nvs 3100M hardly slows down on 720p here18:39
Quintasanfeels bad mna18:39
Quintasanman18:39
QuintasanI got a GeForce GTX 56018:40
Quintasanand it's slowing down18:40
Quintasanthat's dumb18:40
Quintasanlol18:46
Quintasanyofel: Experimental solved it18:46
yofel310?18:46
Quintasanyofel: Yeah18:52
sneleQuintasan: I think this is a known problem with vdpau+kde http://www.nvnews.net/vbulletin/showthread.php?t=17351919:02
Quintasansnele: Well, good thing since 310 fixed the damn thing19:04
QuintasanBut my connection is getting slower and slower19:04
QuintasanThis ISP is retarded19:04
=== ice|away is now known as iceslide
Quintasanthe connection slows to a crawl everyday between 18:00 to 22:0019:06
QuintasanWhat the hell19:06
shadeslayerbtw I found a binary called telepathy-indicator today 19:06
shadeslayerpossibly that already does some m-i integration19:06
Quintasanhue19:06
shadeslayerneed to investigate next week 19:06
shadeslayer( though it most likely only works with empathy and ktp would need to be modified to call tp-indicator )19:07
Tygartcould someone tell me what this is? 19:38
Tygart/user/bin/virtuoso -t +foreground +configfile /temp/virtuoso_kn1964.ini +wait19:38
vHandathat is virtuoso19:38
vHandathe database used by nepomuk19:38
Tygartevery time I re-enable nepomuk it comes up and everything heats up 19:39
Tygartmy fans start running fast19:39
vHandaWhich version of KDE are you on?19:39
Tygart13.0419:39
vHandaAlso do you have akonadi enabled?19:39
vHanda13? :O19:39
vHandaYou can check for akonadi via 'akonadictl status'19:40
Tygartwhat ever version of KDE thats in 13.0419:40
TygartvHanda: it said it was not enabled19:40
Tygartenabling now19:41
vHandano no19:41
vHandadon't19:41
Tygartoop19:41
Tygartoops*19:41
vHandaCould you open an application such as Dolphin, goto help -> About KDE19:41
vHandaand check the KDE version over there19:41
vHandawell, virtuoso is knows to have problems with KDE Pim (Akonadi)19:41
genii-aroundTygart: 4.9.80 is what seems to be on my 13.0419:41
Tygart4.9.8019:42
vHandahmm, so that would be kde 4.10 beta1?19:42
vHandainteresting19:43
vHandaanyway, here is what I want you to do -19:43
Tygartok19:43
vHanda1. Run the script given over here - http://techbase.kde.org/Projects/Nepomuk/VirtuosoInternal#Connecting_directly_to_virtuoso19:44
TygartI don't know if it makes a difference but *user/bin/virtuoso -t +foreground +configfile /temp/virtuoso_kn1964.ini +wait* shows up three times in "Htop"19:44
vHandayou will get a prompt of the form SQL>19:44
vHandatype status() over there19:44
vHanda'status();''19:44
vHandahmm19:44
vHandaYou probably just have threads enabled in htop, so it is showing it a number of times.19:45
TygartvHanda: Sorry I don't understand the last part what to type into SQL19:47
vHandaSQL> status('ckrh');19:48
vHandaDo you see a prompt of the form SQL> ?19:48
TygartYes19:48
vHandacool.19:48
vHandaType - status('ckrh');19:48
* vHanda should write a script to gather all of this information19:49
TygartYou want me to paste it right?19:49
Tygartthe output19:49
vHandayup19:49
Tygarthttp://paste.kde.org/617702/19:50
vHandaTygart: $ qdbus org.kde.nepomuk.services.nepomukqueryservice19:52
vHandacould you please run that19:54
vHandaand provide me with the output19:54
TygartOk 19:54
Tygarthttp://paste.kde.org/617714/19:55
vHandaurgh, my bad19:55
vHanda$ qdbus org.kde.nepomuk.services.nepomukqueryservice /nepomukqueryservice19:55
Tygarthttp://paste.kde.org/617720/19:56
vHandaTygart: is it still consuming a lot of cpu?19:56
vHandaTygart: last one - $ qdbus org.kde.nepomuk.services.nepomukfileindexer /nepomukfileindexer userStatusString19:57
TygartNo, 19:57
vHanda:|19:57
vHandacause no queries or anything seem to be running19:57
Tygartrobert@Laptop:~$ qdbus org.kde.nepomuk.services.nepomukfileindexer /nepomukfileindexer userStatusString19:58
TygartFile indexer is idle.19:58
vHandaOkay, so I'll write a guide as to what all information you can provide when virtuoso does act up19:59
TygartThe CPU useage is a lot better and the "/user/bin/virtuoso -t +foreground +configfile /temp/virtuoso_kn1964.ini +wait" is nologer there  19:59
vHandaand I'll post it somewhere public19:59
TygartOk 19:59
vHandathat way whenever virtuoso does act up, you can provide me with that info19:59
TygartSounds good19:59
TygartvHanda: What we did,  did that correct the issue? or was that just for gathering information? 20:05
vHandainformation gathering20:05
vHandaI think since you're on beta1, it was just indexing files20:05
vHandabut it finished by the time we gathered the info20:05
Tygartok I see20:06
TygartThanks20:06
ScottKyofel: I don't see kdenetwork re-uploaded?20:08
yofelScottK: up now20:15
ScottKyofel: Thanks.20:15
* ScottK just accepted l10n.20:15
ScottKI'll start doing the others once kde4libs is published on all archs.20:15
ScottKdantti: Can you join #debian-qt-kde on OFTC?20:54
ScottKyofel: we need the newer akonadi too, don't we?21:08
yofelyou're right, give me a minute21:08
ScottKOK.21:09
yofelup21:10
* yofel wonders if there was anything else21:10
ScottKAnything you see in Ninjas you didn't upload?21:37
yofelhm, no, it's easier to see once you've accepted everything though21:39
ScottKYeah.  Powerpc buildds are way behind, so I'm trying to ease stuff in.21:39
danttiScottK: what's the irc address?22:01
yofeldantti: irc.oftc.net22:05
danttiyofel: thanks, I found in kvirc list :)22:08
kubotu::workspace-bugs:: [771661] Allow .xsession-errors to be a symlink @ https://bugs.launchpad.net/bugs/771661 (by Martin Pitt)22:19
ScottKyofel: How about smoke*22:51
yofel0-diff IIRC22:52
yofeldo we need to rebuild the bindings?22:52
=== jjesse-home_ is now known as jjesse-home
ScottKNo, if they are zero diff, leave them23:04
ScottKyofel: Those 4 packages are in the quantal packageset now, so you can upload them next time.23:04
yofelyay23:05

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!