[01:16] Stupid wifi notification won't go away. [02:07] I can't think of anything to do. [02:58] manchicken: from earlier: 16:47] [18:01:17] anyone with vivid who wants to test 5.2.0? [02:58] sitter is offline, Riddell is out, etc. [02:59] I'm online [02:59] ^_^ [02:59] I asked earlier but they were out then. I sent an email. [03:00] anyway, that one ask is all I saw in the backlog [03:00] and I've not upgraded to vivid yet [03:02] I'm on vivid. It's got some weird issues still. [03:02] Which is - of course - expected. [03:02] No show stoppers, just major annoyances. [03:03] Notifications for wifi are completely nuts. I've had this wifi notification popup on my screen since I first connected. It won't go away. [03:03] I sometimes know how to fix/get around issues, but am no expert [03:03] which is why I'm holding back [03:03] Where's your sense of adventure? [03:03] I have no backup machine right now [03:04] heh [03:04] and actually, need to do a backup of /home on this one too [03:04] not enough hours in the day, really [03:04] I'm backing up my mobile now so I can reset it. Lollipop really harmed performance for me, I'm going to see if resetting it will help. [05:37] Clue me in: Is the "14.12.0" for the PPA that downgrades to KDE 4? [05:40] there is a ppa that downgrades? [05:40] um [05:41] how can that even be possible? [05:42] Maybe "downgrade" is the wrong word... [05:42] it's vivid, according to http://qa.kubuntu.co.uk/ninjas-status/build_status_14.12.0_vivid.html [05:42] I meant the KDE version, not Kubuntu itself [05:43] keep in mind that "14.12.0" is the applications [05:43] nothing to do with plasma 5 or frameworks 5 in the numbering [05:44] Oh. [05:44] Lol, clumsy me can't read the description. :P [05:46] there is an email on release-team and K-c-d about it [05:46] * valorie wouldn't touch it yet.... [05:46] called "Apps 14.12 release aftermath / Running KF5 apps in KDE 4" [05:47] of course we're not shipping "kde4" in vivid [05:49] So this is the PPA for Apps 14.12 ... feels a bit like Ubuntu GNOME where all the extra GNOME apps (that arguably give GNOME that, well, pure GNOME look and feel) get stuck in (at least) one PPA [05:52] Whoa! Why does whoopsie have a hard dependency on an old version of libwhoopsie0? [05:53] depends on 0.2.39 but installed version is 0.2.44 [05:53] oooo, please file a bug report [05:53] Will do immediately. [05:53] sitter will want to know about that [05:54] ninjas is very early stuff, SonikkuAmerica [05:54] we're not even to beta with Vivid [05:55] Well, with this at least we can fix one problem with in-place upgrades [05:56] I could also check the daily-build contents as well [05:56] (I would assume the correct assertion would be in there) [05:56] I want to be able to upgrade, yeah [05:57] not brave enough to do it yet [05:57] Well I test in odd and weird situations, so I guess this is one for you guys at Kubuntu dev team to munch on. [05:58] yep [05:58] I actually spotted where the mismatch originated. Apparently the version of whoopsie in vivid is still pointing to the version of libwhoopsie0 used in Utopic [06:00] these guys are doing so incredibly much at the same time [06:00] thanks so much for reporting the bug [06:06] Also I should mention that I already made the switch to systemd [06:08] working well for you? [06:08] Absolutely. (I had experience messing with it in ArchLinux with GNOME.) Learned the ins and outs of creating .service files too while I was at it [06:09] * SonikkuAmerica is picking up a 15.04 Alpha 2 image for a clean install [06:09] be sure to report your results to the qa tracker [06:10] Right. [06:13] https://bugs.launchpad.net/ubuntu/+source/whoopsie/+bug/1414899 [06:14] Launchpad bug 1414899 in whoopsie (Ubuntu) "whoopsie cannot be installed on upgrade to vivid (hard dependency on libwhoopsie0 0.2.39)" [Undecided,New] [06:15] thank you, SonikkuAmerica [07:13] good morning [07:14] hi soee [07:15] well my e-mail is back working today, I wonder how many I've missed and how many mailing lists I've unsubscribed from [07:15] hihi, Riddell do you have yesterdays history ? I posted 2 paste.ubuntu.com links with Plasma 5.2 upgrades problems [07:19] ah libkwinxrenderutils5 -> libkwinxrenderutils6 transition [07:19] ok I'll tidy that up [07:33] Good morning. === greyback__ is now known as greyback [09:41] :O [09:42] I always marvel at the ppa-build-status script [09:47] sitter: so what to do with all these backports? [09:47] sitter: are they ready for testing and moving to kubuntu-ppa/next-backports? [09:48] staging utopic still [09:48] shadeslayer didn't run the backport :@ [09:49] Riddell: vivid is good for testing though [09:49] minus the not released bits, minus muon (deps still stuck in proposed), minus bludevil (bluez5 migration, when is that happening anyway?) [09:52] sitter: bluez, rather like qt, seems to be stuck on unity crap [09:52] hassle didrocks about it https://launchpad.net/~ubuntu-desktop/+archive/ubuntu/transitions/+packages [09:52] kubotu: order vodka [09:52] * kubotu slides vodka down the bar to sitter [09:53] not until midday! [09:54] bluez - blocked; qt - blocked; packagekitqt - blocked; telepathy-qt - blocked; [09:55] * sitter fails to read the build-status script [09:56] which happens in a large project with lots of different dependencies, it just seems to happen a lot now, and it kindae predictable with canonical wanting to write their own desktop that they don't have the resources to look after all of it [09:56] I totally fixed it at some point, why is it broken again -.- [09:56] Riddell: nm /minor/ release landed some 1.5 years after its upstream release :P [09:57] on the plus side I have been told nm 1.0 ought to land in vivid as well [09:59] sitter: minor release of what? [09:59] networkamanger [10:00] which I reckon isn't really minor considering it is networkmanager xD [10:07] kde-runtime uploaded with qapt depends removed, maybe that'll get muon into the archive [10:07] sitter: for the trello, what is telepathy-qt blocked on again? [10:08] * Riddell just puts "waiting on unity stuff" [10:09] Quintasan: don't forget you're due to sort out ksnakeduel name [10:09] Riddell: canonical specific public API patches [10:10] I did not take a look since last time though, so maybe it can get unblocked but I distinctly remember that there was an API addition that didn't seem to map to any of the upstream changes [10:10] mm yes [10:10] ah ah ah [10:10] * sitter blind [10:11] kfilemetadata and baloo are red because their versions are different [10:11] d_ed: also said he'd look at upstreaming it but it's not like it's fun stuff to work on [10:11] such smart script [10:11] yes the script needs a fix there [10:11] I think this script needs more fixes than that [10:11] well yes [10:14] does python have a static global variable type? [10:15] goodness no [10:15] but that sounds mildly evil [10:16] what's wrong with having a variable that only is initialized once? :O [10:17] global variables are usually frowned upon [10:18] http://bazaar.launchpad.net/~kubuntu-packagers/+junk/kubuntu-automation/revision/504 [10:18] .... [10:18] Riddell: the important part is the staticness not the globalness :P [10:18] interestingly I didn't need globalness anyway because the script is very efficient spagetthi holding everything in the main scope... [10:26] kglobalaccel_5.6.0-1_amd64.changes is NEW says Debian FTP masters [10:36] Riddell: when are you done with 5.2.0 release procedure? [10:36] maxy uploaded lots of stuff in the last days [10:40] sitter: wanting to tidy up the announce and make publish shortly [10:40] sitter: why do you ask? [10:40] Riddell: I want to merge documentation fixes into releaseme [10:40] and then carry the entire thing into master [10:40] oh go ahead [10:41] but I have a lot of plasma release crap in that branch so maybe make a plasma branch for me to keep that in? [10:41] Riddell: if you were to respond to my mail about plasma bits :P [10:41] sitter: which one? [10:42] the one were I said it should be rubeh :P [10:42] well rubeh or yaml [10:43] I rather see the benefit in composing a meta release as a simple yaml config and have tarme deal with how to best get the things [10:44] anyway, before it goes to master I want to finish up manual configuration, the projects xml does not always scale as well as it should [10:56] Riddell: what goes into the next-backports ppa, I forgot [11:00] sitter: utopic qt5/kf5/plasma5 backports [11:00] was my plan [11:01] and applications too [11:01] was there a reason not to put it into next? [11:01] ah, qt54 possibly wouldn't be too good [11:01] right, I'd worry that would break something [11:01] k [11:02] so we need update test from next to -> stage0+stage1+stage2+utopic-qt5 [11:02] actually [11:02] applications need to get split [11:02] oh meh, faff [11:02] the majority of them have nothing todo with next [11:08] Riddell: I don't get why muon doesn't migrate [11:08] everything is marked as valid candidate now [11:09] oh oh, driver-manager maybe [11:10] if britney output wasn't so atroxious [11:10] maybe even correctly spelled [11:10] * sitter hungry [11:10] sitter: https://www.youtube.com/watch?v=t3otBjVZzT0 [11:15] ah [11:15] Riddell: I think kde-runtime is blocking it [11:16] seems you uploaded a fix [11:16] I did [11:38] Hi shadeslayer, can you please quickly review http://mitya57.me/tmp/pkg-kde-tools_0.15.15ubuntu1_0.15.16ubuntu1.debdiff which I am going to upload? [11:38] Most is trivial except maybe kf5.pm part [11:39] Also qt-kde-team/1/policy.mk delta was mentioned in changelog, but was lost in previous uploads for some reason. [11:40] sitter: tagme runs ruby at 100%cpu and doesn't seem to do much :( [11:41] sitter: https://paste.kde.org/pcu7yfuay [11:43] Riddell: for how long? [11:43] it was still doing project resolution that can take a while [11:43] actually [11:43] mh [11:43] Riddell: you did not specify a name now did you? [11:43] oh with tagme you wouldn't [11:44] Riddell: are you sure your release_data is correct? [11:44] http://starsky.19inch.net/~jr/tmp/release_data [11:45] --MARK-- [11:46] ewww [11:46] cabbage juice was the worst idea ever === kbroulik is now known as kbroulik-lunch [11:47] Riddell: I think it is just really very slow [11:48] it does walk the entire xml document for each project [11:50] sitter: yeah it's going now [11:50] sorry for the noise [11:52] well it is a bit of a problem [11:52] alas, not one that is necessarily easy to work out [11:54] the problem is that release_data holds each invidiual project name, so where tarme would make one query "give me all workspace things", tagme needs to do "give me plasma-workspace, now give me plasma-desktop, now give me ksysguard..." and each of those need to traverse the entire xml because each of those identifiers could be either a component or a module or a project or a subproject of a project [11:54] it's why I raged about the projects xml when I redid the parser :P [11:55] the probably best solution is to expand release_data, make it yaml or json and basically marshal the project class which encapsulates the information found in the xml [11:55] that way nothing but tarme needs to even look at the xml [11:56] alternatively release_data could use the xml xpath thus making each of those queries an immediate hit through xpath lookup [12:09] mitya57: -DECM_MKSPECS_INSTALL_DIR=/usr/lib/ [12:09] mitya57: that should really be in kf5flags [12:09] oh [12:09] mitya57: I read that wrong [12:09] nvm [12:10] mitya57: so what happens to ECM_MKSPECS_INSTALL_DIR ? [12:11] or not required? [12:12] mitya57: and FWIW we can keep the maintainer checking, the maintainers are combined now [12:12] so it's Debian and Kubuntu Qt/KDE team [12:13] morning [12:14] morning sgclark :) [12:22] it's a mix of "Debian and Kubuntu" and "Debian and Ubuntu" and "Debian/Ubuntu" and maybe something else and sometimes there's a X-Ubuntu-Maintainer tag [12:25] on utopic something seems to drag in qt5.3 still [12:25] brr [12:26] !newversion choqok 1.5 [12:26] Riddell: I am only a bot, please don't think I'm intelligent :) [12:26] kubotu: newversion choqok 1.5 [12:26] https://bugs.launchpad.net/bugs/1415022 [12:36] shadeslayer: it was you who dropped ECM_MKSPECS_INSTALL_DIR option: http://anonscm.debian.org/cgit/pkg-kde/pkg-kde-tools.git/commit/?id=1387e463e59a9bef [12:37] aha cool [12:37] (sorry, was afk) [12:37] I'm awesome [12:37] mitya57: ship it [12:37] shadeslayer: can I add back maintainer checking everywhere? (1,2,3) [12:37] mitya57: only in 3 [12:38] ok [12:38] I don't think our older packages have co maintainer ship with the Debian team [12:38] so that'll fail there [12:39] yep === kbroulik-lunch is now known as kbroulik [12:46] Riddell: utopic seems good for copy unless you want more testering? [12:48] sitter: how about kate? [12:49] that's applications [12:49] sitter: what are you proposing to move? [12:49] plasma-frameworks-qt [12:49] why not applications? [12:50] because I have not tested them [12:50] also they need splitting IMO, no point in putting the kde4 apps into next [12:51] I think apps should have all go into next-backports and kde4 ones into backports [12:51] if you're ready to go with plasma-frameworks-qt then ok [12:52] and we can fight over who gets to finish off applications [12:52] Unpacking kate5-data (4:14.12.0-0ubuntu1~ubuntu14.10~ppa1) ... [12:52] dpkg: error processing archive /var/cache/apt/archives/kate5-data_4%3a14.12.0-0ubuntu1~ubuntu14.10~ppa1_all.deb (--unpack): [12:52] trying to overwrite '/usr/share/icons/hicolor/64x64/apps/kate.png', which is also in package kate-data 4:4.14.1-0ubuntu1 [12:52] dpkg-deb: error: subprocess paste was killed by signal (Broken pipe) [12:52] I am not sure how that backport is supposed to work anyway [12:52] installing that will break dolphin and kdevelop and whatnot [12:52] the icon should be removed from kata-data is all [12:52] that doesn't solve anything :P [12:53] well anyway that needs more testing and tidying as I say [12:53] plasma 5.2 on vivid has the issue soee reported earlier with overlapping kwin libs http://paste.ubuntu.com/9885905/ [12:54] if you install kate (kf5) it replaces kate (kdelibs4) and all things that use kpart(4) depend on kate so they'll be satisfied by the kf5 version but that has an incompatibru with kdelibs4 apps, so they have no katepart [12:54] same for konsole [12:54] Riddell: I blame that on you [12:58] fixed in git [12:59] Uploading kwin_5.2.0-0ubuntu1~ubuntu15.04~ppa2_source.changes: done. [12:59] Successfully uploaded packages. [12:59] lovely [12:59] time to log out and see what breaks [13:01] yay it works! [13:02] sitter: let's coordinate on a plan? plasma 5.2 needs uploaded to vivid? plasma-frameworks-qt backports copied over? apps backports need copied over? who's doing what? [13:03] Hey folks [13:10] Riddell: doesn't the app backport need testing first? (looking at what sitter posted above it doesn't seem to work?) [13:10] Riddell: apps needs fixing [13:10] plasma-framework-qt backport I am on [13:10] * sitter writes script :O [13:11] also, please only copy full PPA's so no packages get missed (like with the 4.14 precise backport) [13:11] if you copy subsets copy that somewhere else first and re-test that [13:12] (or at the very least, re-test *after* the copy has been done) [13:14] hence why I made 3 stages :P [13:14] sitter++ [13:14] Riddell: qtquick1 backport is still missing it appears :P [13:16] pft [13:16] shadeslayer: oauth gem is not threadsafe xd [13:16] yay [13:16] such bastard [13:16] mutex all the things \o/ [13:16] My mozilla CI scripts are becoming awesomer by the minute [13:17] is that "awesome" or "monkey awesome" ? [13:17] monkey patch all the things \o/ [13:17] not monkey patching [13:17] I found this thing called filterdiff [13:17] it is the most magical thing ever [13:18] filterdiff used to be big in the days when people still used debdiff [13:18] people still use debdiff [13:19] yeah, there's no need for filtering tho [13:20] xD [13:20] sitter: http://paste.ubuntu.com/9897989/ [13:20] so far [13:21] not that it'll make any sense :P [13:22] * sitter has threaded ppa copy \o/ [13:22] all hail the mutex [13:23] pft [13:26] sitter: ok shall I upload plasma 5.2 to vivid? [13:26] yes :) [13:32] Riddell: all goody from my side [13:32] Riddell: while you are at it, maybe make muon migrate :P [13:32] sitter: rewrite dci/mozilla.rb (91%) [13:32] xD [13:33] it has no classes [13:33] I find this highly suspicous [13:33] sitter: I'm trying! (on muon) [13:33] don't need no classes [13:33] sitter: though it feels like I'm writing C code now xD [13:33] build_* suggests otherwise [13:33] sitter: build_* ? [13:33] the methods [13:34] ah [13:34] sitter: what would you recommend [13:34] I see a buildthing prototype and specifics of the kind firefox and thunderbird, so there should at least be 2 classes and a module or 2 classes and one base [13:34] *base class [13:35] mmmm [13:35] sitter: I'll spend some more time architechting it better I guess [13:35] when I have time [13:35] for now this will do [13:35] listen to robucop :P [13:35] I did [13:35] if you have more than 10 lines of code in a method you are most likely cheating [13:36] I feel that's stupid tbh [13:36] the 10 line limit [13:36] sometimes it is [13:36] most of the times it actually isn't [13:39] sitter: so something like : MozBase, which does pull-lp-source, install build deps and what not, and then specific hacks go into firefox and tb classes? [13:40] yup [13:41] Riddell: Sure. [13:42] Riddell: plasma-frameworks-qt copied, pending publication [13:43] * sitter is tempted to say that we should continue using 3 stages moving forward [13:43] makes for much easier isolated testing, unfortunately dependency management on PPAs is a bit of a drag [13:44] sitter: you didn't package plasma 5.2 bluedevil? [13:44] Riddell: bluez [13:45] sitter: right but it can be backported and the vivid packages uploaded to transition ppa [13:46] Riddell: I feel a bit uneasy about that, should bluez5 not get landed we are in a world of trouble [13:46] shrug, bluedevil for kdelibs4 doesn't do much on plasma 5 [13:47] true enough, but if we backport we need to backport bluez5 which means if a next user then upgrades to vivid they still have bluez5 with potential screwery [13:49] Riddell: maybe assert bluez5 as a hard requirement for us? [13:50] sitter: how do you mean? [13:51] Riddell: get foundations to commit to bluez5 landing in vivid, or did they do that anyway? [13:51] I understand only mobile is holding up the landing [13:51] 14.10 libbluedevil uploading to staging [13:52] sitter: my linter doesn't seem to pick up variables that aren't defined [13:52] any idea what I can use? [13:53] there no such thing I think [13:54] :( [13:54] variables can come from outofscope due to the non-finalized state of all objects, so until actual runtime there is no way to accurately tell whether a variable is defined [13:54] ruby -w might tell you at runtime though [13:55] sitter: but then I have stupid shite like http://anonscm.debian.org/cgit/pkg-kde/ci-tooling.git/commit/?id=8c69f9b885ec98ef389dc5c79acdb96ba6f5a3c5 [13:55] which plays back into the fact that you should unittest everything :P [13:55] shadeslayer: write tests :P [13:55] uf [13:55] can't, tests would need root etc [13:56] not a good test wouldn't [13:56] code needs root [13:56] in certain parts [13:56] like apt-get update and stuff [13:56] I doubt that is ever going to change btw, new_changelog in that case could be a variable defined in the super scope, a module that you mixin, a class deriving from this context etc. etc. all of which can be different files [13:57] shadeslayer: make it testable then? [13:58] not sure how to do that, but I'll try [13:58] overload Kernel::system in the test and make it noop for example [13:58] would be naughty but that's one option for example [14:00] I think we'll eventually need a class/module wrapping around Apt anyway. for testing purposes you could then open up the class in the test case and make system noop, that way the noopness is limited to apt and doesn't interfer with anything else [14:10] * sitter scratches head [14:10] ha! [14:11] status page script list for utopic is incomplete [14:14] Riddell: what do we do with kcm-touchpad and usermanager? [14:15] should I fetch new snapshots? [14:20] sitter: http://paste.ubuntu.com/9898643/ [14:20] sitter: makes me not want to use overlayfs [14:22] sitter: leave them at beta for now [14:22] kcm-touchpad will need an update but it needs testing and ktouchpadenabler removing [14:23] there are no beta backports [14:23] I broke a thing [14:23] oh noes [14:28] So I've got good news and bad news [14:28] good news is good [14:30] The good news is, I have Plasma 5. The bad news is, I have a mixed system. I filed a bug as to the reason why already. But the other bad news is, the daily build threw me "invalid magic number" errors. [14:31] So bug 1414899 , by the way [14:31] bug 1414899 in whoopsie (Ubuntu) "whoopsie cannot be installed on upgrade to vivid (hard dependency on libwhoopsie0 0.2.39)" [Undecided,New] https://launchpad.net/bugs/1414899 [14:36] 5.2.0 is uploaded to vivid [14:36] including kwin 5.2.0.1 [14:37] which we can ignore for utopic because utopic already has qt 5.4 which fixes the issue too [14:40] mitya57: re Qt 5.4 fixes for 5.3.2, I don't think it's worth the effort [14:40] mitya57: because we also want things like better high dpi support and what not [14:41] which you can probably not backport [14:43] shadeslayer: releaseme's upcoming logable module in case you want something reusable btw : http://paste.ubuntu.com/9898985/ [14:43] can be mixed into other modules and classes alike [14:47] just put it in ci-tooling/lib ? [14:48] shadeslayer: yeah [14:48] http://paste.ubuntu.com/9899074/ [14:48] this version has slightlyb etter output when used from class methods [14:49] http://paste.ubuntu.com/9899102/ [14:49] this one supports prepending properly xD [14:49] now I am done [14:51] Riddell: any reason not to pull a snapshot of user-manager/kcm-toucpad? [14:56] sitter: all that seems magic to me at the moment [14:57] sitter: any reason to do so? [14:58] sitter: ooh ooh [ubuntu/vivid] libqapt 3.0.0-0ubuntu2 (Accepted) [14:58] shadeslayer: ok [14:59] shadeslayer: in any case the patches needed for tray icons are already in vivid [14:59] mitya57: roger, thanks for that! :) [14:59] Um... is the MD5 checksum listed on cdimage.u.c correct for vivid Alpha 2? [14:59] Riddell: btw I'll need your notes from Spanish class when I get back :P [15:00] SonikkuAmerica: should be [15:01] shadeslayer: que pasa, is all you need to know :) [15:01] Riddell: Then I should be very afraid - the image I just downloaded gave me a completely different MD5 sum [15:01] xD [15:01] SonikkuAmerica: which one? [15:01] Vivid Alpha 2 [15:01] * SonikkuAmerica is DL'ing a 2nd image into another dir to double-check [15:02] >sha256sum vivid-desktop-amd64.iso [15:02] c309ef229f2ddfa59a7db5f3a346b269e51054192be3246bf0407c4d2b5ff35d vivid-desktop-amd64.iso [15:02] same as http://cdimage.ubuntu.com/kubuntu/releases/15.04/alpha-2/SHA256SUMS [15:02] shadeslayer: mh, the prepend and included functions are builtins called when one uses the include or prepend keywords on a thing, extend and include respectively then pull ClassMethods into the thing that includes or presend, and extend basically adds all the rubbish into class scope and include into instance scope [15:03] severe meta programming that is :P [15:03] I also got something different with the SHA256 sum. Probably a fluke then. The 2nd image should tell all now. > Riddell [15:04] sitter: don't try to explain it, I've only had 4 hours of sleep [15:04] Cards against humanity and the drinking went on for too long [15:04] long story short: http://paste.ubuntu.com/9899270/ [15:04] you include the logable feature into any other thing and that other thing can then use the ClassMethods defined in the Logable [15:05] alas, the name there is right now misleading because the methods get imported into class and instance scope, as seen in the class B example there you can call log_warn from an instance method and a class method [15:05] Will plasma 5.2 come to 14.10 too? [15:10] shadeslayer, Riddell: whatever is this https://launchpad.net/ubuntu/+source/muon/4:5.2.0-0ubuntu1/+build/6754738/+files/buildlog_ubuntu-vivid-i386.muon_4%3A5.2.0-0ubuntu1_FAILEDTOBUILD.txt.gz [15:10] sitter: aha [15:10] oho [15:11] sitter: mitya uploaded new pkg-kde-tools with policy check [15:11] "debian_qt_kde.mk usage denied by policy.. Stop." gosh, harsh [15:11] maybe you want to make muon shared between Debian and Kubuntu [15:11] waa, so everything is going to fail to compile? [15:11] Riddell: no, unless you changed the maintainers [15:11] it is isn't it? [15:11] from Debian Qt/KDE team to whatever [15:12]