/srv/irclogs.ubuntu.com/2015/10/26/#kubuntu-devel.txt

lordievaderGood morning07:26
amichairRiddell: Thanks for all your great work and leadership with Kubuntu during the past decade (?), it is much appreciated by lots of people whom you've never met :-)09:28
amichairRiddell: Good luck with your future endeavors, I'm sure you'll do many more great things!09:28
yossarianukhi - is this bug -> https://bugs.kde.org/show_bug.cgi?id=354230  The cause of a really slow login to kde on 15.10 ?09:29
ubottuKDE bug 354230 in general "Blocking calls from PlasmaNM to BlueZ for 30s" [Normal,Resolved: fixed]09:29
yossarianuktakes about 30 - 40 secs to get a usable desktop from login on 15.10 (didn;t happen on 15.04)09:30
yossarianukjust wanted to ensure it was this bug (as that looks fixed upstream)09:30
amichairyossarianuk: I have that bug, and it depends on what you mean by login....09:31
amichairyossarianuk: it happens after the actual login screen, when some of the other stuff already appaers (including partial panel)09:32
yossarianuki.e after logging in to sddm 09:32
amichairyossarianuk: but then it hangs for ~50 secs (in my case) until the panel is completely drawn and responsive09:32
yossarianukuntil its fully loaded the taskbar locations of clock, etc are wrong also 09:32
yossarianuksounds very similar.09:32
amichairso yes, it's probably that09:32
yofelthere's more than one reason for a slow plasma login, but in ~2h or so I could wrap up a test package with above fix09:32
yossarianukI assume the fix will be in the normal updates at some point?09:33
yossarianukyofel: I'll happily test the package (when im back home)09:33
yossarianukso far though the slow login is the only real issue with 15.10, apart from that it seems better than 15.04.09:34
yofelthe fix is in Plasma/5.4, so it should be part of 5.4.309:34
amichairyofel: any idea when that will be released and available in kubuntu updates?09:34
yofel.3 is due in ~3 weeks09:35
yofelwe can cherry-pick that patch before that if it really does help09:35
=== tazz_ is now known as tazz
amichairyofel: personally I don't log in/out too often, so a 3 weeks wait is fine. I'd think the ubiquity bug that causes the upgrade to 15.10 to crash is far more critical.09:42
yofelerm, what bug? (ubiquity is the live-installer, and has nothing to do with upgrading)09:44
amichairyofel: oh, then someone pointed me to report a bug in the wrong place, perhaps09:45
yossarianukyofel: login bug - its not exactly critical however would give bad impressions to someone checking out how plasma5 is getting along. maybe add to known 15.10 issues on the release notes ?09:45
yofelamichair: I guess update-manager would be more correct. I think I remember someone talking about a crash there09:46
yofelyossarianuk: hm, we could maybe add a generic message about that. There seems to be at least the bluez issue and an issue with akonadi that cause that09:47
amichairyofel: bug #150965509:47
ubottubug 1509655 in ubiquity (Ubuntu) "installer crashed on custom configuration file dialog" [Undecided,New] https://launchpad.net/bugs/150965509:47
amichairyofel: there was also bug #1509653 which was scary (saw it on two systems), but doesn't actually prevent the upgrade from succeeding, so not that critical09:48
ubottubug 1509653 in ubiquity (Ubuntu) "kdeinit5 crashes during upgrade to 15.10" [Undecided,New] https://launchpad.net/bugs/150965309:48
amichairyofel: but the former bug just crashed the upgrade and left people with a broken system. discussed it with two other guys on #kubuntu the other day who also suffered from it.09:49
yofelyeah, I know about the kdeinit5 one. But no idea what to do about that09:50
soee_yossarianuk: allee was thinking that the loading system lag problem could be caused by akonadi10:01
yofelI think he pretty much proved that his problem is akonadi10:03
* yofel wonders if the pyqt5 API changed...10:04
alleeYes. Unfortunately dvratil could not reproduce with master :-(10:04
alleeyofel: is there a description how to pick e.g. only KDEPIM from CI  (to test if the problem goes away with master pkgs)10:05
yofelno, your best bet would be to add the repository and selectively upgrade the packages you're interested in10:06
BluesKaj'Morning folks10:49
sgclarkmorning10:57
ahoneybuno/11:01
BluesKajnothing in the pipe for 4 days , that's unusual for a new release, still plenty of unsolved issues 11:03
d_edcan you guys look at https://bugs.launchpad.net/ubuntu/+source/plasma-nm/+bug/1509334 I've linked to a patch, if you could backport it, that'd probably be useful11:17
ubottuLaunchpad bug 1509334 in plasma-nm (Ubuntu) "KDE/Plasma very slow to launch (Kubuntu 15.10)" [Undecided,Confirmed]11:17
yofeld_ed: on my todo list, thanks for the bug link11:23
yossarianukd_ed: cheers - thats affecting me11:24
d_edit's in 5.4.3, which is ETA mid Nov11:24
* ahoneybun has a ssd so he is not feeling that11:26
d_edactually, you have bluetooth so you're not feeling that11:26
d_edwe just sit idle for 30s doing nothing11:27
ahoneybun?11:27
BluesKajyeah , but removing NM from the panel is not a solution here, since I use vpn a lot , and I have no intentions of trying to configure openvpn and server in network interfaces11:27
d_edwell, that's why we fixed it :D11:28
BluesKajd_ed, I must be missing a repos because nothing upgrades here11:30
d_edah.. When I say "we" I mean KDE upstream11:31
d_edI'm here liasoning.11:31
clivejoyofel: seems to be on the case :)11:36
clivejoyofel: do you create a quilt patch for plasma-nm and put it in the packaging?11:44
clivejoto patch libs/handler.h and libs/handler.cpp11:45
yofelclivejo: yes, do you want do prepare the SRU? I probably won't get to it until the evening11:45
clivejoyofel: never done it before so unsure of the process11:46
clivejotrying to follow what you do to fix it11:46
yofelhm, then it'll have to wait until later. Then I can tell you what needs to be done11:46
clivejodoes the patch go into debian git?11:47
clivejoand if so what branch, now that wily is release?11:47
yofelyes, but you also need to branch of kubuntu_xenial_archive, put the patch there first, then prepare a similar upload with different version in kubuntu_wily_archive, update the bug with the SRU information and get things uploaded11:48
clivejoeak11:48
clivejothe prepare wily version is were I think you lose me11:49
clivejoversion numbers still confuse me11:50
clivejoyofel: maybe give me a shout when you are free, Id like to go through it with you if thats ok11:50
yofelwell, here it's easy. Say you have -0ubuntu1 in wily now, xenial would get -0ubuntu2, wily-updates -0ubuntu1.111:51
yofelsure, will have to wait until after work, so in ~6h11:51
clivejois there a script you run to start all the new xenial branches?11:52
clivejoor are they done manually as needed?11:53
yofelgood question.. no idea right now. I think manually is fine for now11:53
yofelRiddell: what was the process for vivid->wily back then? ^11:54
Riddellsitter had a script somewhere I think11:55
Riddellmanually is also fine11:55
clivejoRiddell: would you show me the script?11:57
Riddellwhen I say somewhere it's because I don't remember where :)11:57
clivejowhere is sitter these days, havent seen him about?11:58
ahoneybunRiddell: how is that docs.kubuntu.co.uk server?11:59
clivejowhen does development on xenial official start?11:59
yofelonce the toolchain upload is done. There'll be an announcement somewhere (ubuntu-devel ML I think?)12:00
Riddellsitter said he was away on friday, I guess he's taking a long weekend12:01
Riddellahoneybun: still doing fine :)12:01
ahoneybunRiddell: could you update it :)12:02
Riddellahoneybun: update what about it?12:02
BluesKajsaw something about the toolchain for Oct 29th12:02
ahoneybunthe content?12:02
ahoneybunhttps://github.com/ahoneybun/kubuntu-manual12:03
ahoneybunmake it like this: http://192.254.78.155/12:03
yofeloh, that came out really nice :)12:04
ahoneybunthanks yofel :)12:05
* ahoneybun flys off to work12:05
yofeland me to lunch12:06
yofelbbl12:06
Riddellahoneybun: feel free to update it, not my responsibility any more I'm afraid12:06
mhall119hi, does anyone here involved in plasma mobile want to host a session talking about it for the convergence track of UOS 15.11?13:04
yofelclivejo: actually, I just noticed that #ubuntu-devel says "Archive: open" - so xenial is already open for dev14:16
clivejoyofel: Im trying to find the scripts sitter used14:18
clivejonot having much luck14:19
yofelshadeslayer: do you know where those are ^14:19
clivejoRiddell: suggested the channel logs14:19
clivejobut cant find anything yet14:20
Riddellhe may have just put it in a pastebin14:20
yofelnot sure I would find anything there either. I believe the CI repo is on Alioth, but I don't remember how it's called14:20
yofelOTOH, scripting creation of a branch in all repos is something you can do with a 5-line bash script or so..14:20
clivejoI been looking at logs after the 23rd April 201514:21
clivejowould hardly be done before it?14:21
yofelvery unlikely, right14:22
clivejohttp://anonscm.debian.org/cgit/pkg-kde/applications/amor.git/log/?h=kubuntu_wily_archive14:24
clivejodoes it log when the wily archive was started?14:24
clivejothat period of time looks like sgclark was working on it14:25
Riddellach it's just one command, just do it by hand14:30
yofelconsidering our workflow, santa_'s new scripts should probably have a failsafe check that makes sure the branch exists before it tries to check it out14:37
yofelso just leave it for now and just create the branch you need by hand14:38
pursuivantmuon (master) v5.4.2-137-gcdc24d3 * Aleix Pol: discover/qml/PageHeader.qml15:11
pursuivantFix margins in the page header15:11
pursuivantOnly leave them on the side15:11
pursuivanthttp://commits.kde.org/muon/cdc24d3516b565bbdbe906147ee6cdd8480ccd0b15:11
pursuivantmuon (master) v5.4.2-138-g1e8f40a * Aleix Pol: discover/qml/SourcesPage.qml15:12
pursuivantFix sources page display15:12
pursuivanthttp://commits.kde.org/muon/1e8f40a82e577dfb9428579e817132936d75426b15:12
sgclarkclivejo: all by hand, I did not have a magic script.16:25
kustodianI noticed that Telepathy auth for Google Hangouts has been changed, so now you have to login with your google account and than it justs uses the auth token16:32
kustodianbut the problem is that it doesn't work16:33
kustodianafter I login into google without any errors, when I try to go online it doesn't work16:33
kustodiantries for a few seconds and than it stops without any errors16:33
kustodianI'm on 15.1016:33
kustodianupgraded today16:33
marco-parillokustodian: Can you go to you GMail, and see if you have an e-mail about some kind of insecure, deprecated access attempt?16:49
clivejosgclark: wow, you must be very patient!16:49
yofelyou may imagine a world without the automation scripts :P16:51
sgclarkheh yes16:51
sgclarkalso had to do all the debian merges at the same time16:51
yofelhm, we have to do those too, right..16:52
sgclarkyeah16:52
* yofel imagines a 300-or-so long todo list16:52
yofelsomeone motivate me...16:52
sgclarkI am working on trusty and vivd backports right now, can after though.16:52
santa_yofel, clivejo: what you are trying to do? creating new branches for xenial?16:52
yofelsanta_: that was the original topic, yes16:53
clivejoyes, automatically16:53
santa_for frameworks/plasma/apps for example?16:53
clivejoall of them16:53
yofelFWIW, as scarlett said, we have to do that anyway when we merge, so the point is moot in some sense16:54
santa_clivejo, yofel: give me a couple of minutes and I will send a terminator to the rescue16:55
santa_actually an early version of its firmware16:55
yofelXD16:55
kustodianmarco-parillo: I didn't receive an email like that16:56
kustodianI checked in the list of allow applications and "KDE Online Accounts" is there and it has access to a lot of stuff16:57
santa_yofel: https://gitlab.com/kubuntu-clones/kubuntu-automation17:09
santa_in the automation-ng branch I have the first version of git-clone-all and do-all17:09
santa_so you can do this17:10
santa_$ mkdir ~/kde-all/17:10
santa_$ cd ~/kde-all/17:10
santa_$ git-clone-all17:11
santa_$ do-all "git checkout kubuntu_wily_archive"17:11
santa_$ do-all "git branch kubuntu_xenial_archive"17:11
santa_$ do-all "git push origin kubuntu_xenial_archive"17:12
santa_and I think that should do the thing if that's what you want to do17:12
yofelthat page says 504, gitlab, what's wrong with you..17:12
santa_sigh17:13
santa_let me create a clone in github then17:13
santa_yofel: https://github.com/jmsantamaria/kubuntu-automation-work17:18
yofelthanks, I'll try it out when I'm home17:19
yofel~2h17:19
clivejoyofel: will you please have a wee look at https://launchpad.net/~clivejo/+archive/ubuntu/wily/+build/8193598 when you get time18:06
clivejoor even santa_ ^^18:06
clivejothink the version number is wrong :(18:13
clivejoneeds to be 4:5.4.2-0ubuntu1.1 for wily?18:14
clivejoBluesKaj: ping18:15
Riddellyes use 1.1 for wily-updates18:19
clivejoRiddell: cant you check that over for me please :)18:19
clivejocan you18:19
clivejoI used the xenial packaging, to test build and upload to my PPA, do I need to push in wily archive too, with a different version number?18:23
Riddellsorry I'm about to go out, but yes push it to wily_archive with the 1.1 version18:25
clivejowith the 1.1 version number18:25
Riddellyes18:25
clivejothanks :)18:25
BluesKajclivejo, pong18:31
clivejofancy testing plasma-nm for me?18:31
soeeclivejo: this build contains teh fix for the 30 sec lag 18:40
soee?18:40
soeeor it was related to different package ?18:40
clivejoI believe so18:40
clivejoif I have done it right18:40
soee;D18:41
clivejobut the version number is wrong18:41
clivejoIm trying to fix that now18:41
clivejoeak, 1.1 version is pending for 2 hours18:51
yofelclivejo: thanks for preparing plasma-nm, a few comments though19:22
yofel- please always commit with UNRELEASED unless the version has been uploaded to the primary archive19:23
yofel- the patch is either kubuntu_* -> origin: vendor, OR upstream_* -> origin: upstream. In this case, you'll want latter19:23
yofel- The changelog message is actually rather good, but slightly wrong:19:25
yofel "will be applied upstream" is "from upstream" - it's already been committed. 19:25
yofelThe parentheses around the patch name don't really make sense. 19:25
yofelyou indicate an upload fixing a bug with: LP: #0000000 - which is usually at the end of your changelog message. (That will then auto-close the bug once uploaded)19:26
soeeyofel: is it save to test it https://launchpad.net/~clivejo/+archive/ubuntu/wily/+build/8193870 ?19:47
yofelsoee: yes19:48
soeerebooting19:50
soeeback19:51
soeeyofel: interesting, system loads now in ~8 sec19:51
soeeso no lag as before19:51
yofelok, so you did have the bluez issue. At least we know that it works fine :)19:52
yofelthanks for testing19:52
soeeyofel: one more reboot, just to be sure19:52
soeeyup, confirmed19:53
soeesystem loads in ~ 8 sec now19:53
soeeclivejo: gret work with the package19:54
soeeallee: so in my case the lag was caused by bluez19:54
soeeyofel: what bug numer/link was it ?19:55
yofellp 150933419:55
ubottuLaunchpad bug 1509334 in plasma-nm (Ubuntu) "KDE/Plasma very slow to launch (Kubuntu 15.10)" [Undecided,Triaged] https://launchpad.net/bugs/150933419:55
soeeyofel: just poted small info about it on G+, hope it wasn't to early :)20:03
soeenow there are 2 annying bugs left imo.20:04
soeefirst: missing plasma-pa icon in systray 20:05
soeesecond: baloo indexer crash after system boot20:05
yofelI don't have the second, but I've seen the first one - but it's a bit random..20:05
soeeyofel: http://paste.ubuntu.com/12973455/20:10
soeeoh and: https://plus.google.com/+JakobHarz/posts/h39zc6U85hc20:17
clivejoyofel: thanks, nice to know all those20:26
clivejoso kubuntu_ patch is only for something affecting us?20:26
yofelis only for stuff either created by us or stuff specific to us20:27
clivejoI see, I thought it was for a patch added by the kubuntu packagers20:27
clivejoI commited xenial as UNRELEASED and wily with WILY, is that wrong?20:28
clivejoregarding the patch name, should it be upstream_redhat_* to show where the patch came from?20:30
yofelno, just upstream_20:30
yofelunless you got that from some RHEL repository or so20:30
clivejoKDE commit20:30
clivejoI used quilt import *.diff20:31
yofelright, so just upstream_ (i.e. it came from the repository that the source itself came from)20:31
clivejoso at the end of my changelog I should put "* Fixes LP: #1509334" and LP will mark it as fixed?20:32
ubottuLaunchpad bug 1509334 in plasma-nm (Ubuntu) "KDE/Plasma very slow to launch (Kubuntu 15.10)" [Undecided,Triaged] https://launchpad.net/bugs/150933420:32
yofeldrop the "* Fixes"20:33
yofela) it belongs the the message above, b) the syntax already says that it fixes it20:33
soeehttps://plus.google.com/110954078302330754910/posts/EYpzxwHXPG820:35
clivejoshould I make those changes in git?20:35
clivejoyofel: also, have I broken KCI ?20:35
yofelplease do the changes20:36
yofelas for CI, it's been rather red lately... haven't looked at it20:36
clivejodo I put LP: #000000 on a new line?20:40
yofelthat's up to you.. I usually only do that if the line gets too long20:41
clivejoand my patch file should be .patch?20:42
yofelthat again doesn't matter, just keep what you have now20:42
yofelregarding UNRELEASED, always use that unless you uploaded to the primary archive (or the target one)20:43
yofeldoesn't matter if it's dapper or xenial20:43
clivejocan I ommit an extension?20:43
clivejojust call it - upstream_fix_making_bluez_asynchronous20:43
yofelI believe yes, but that's rather unusal..20:43
clivejowhat is usual standard20:43
clivejohave to change the name anyways, might as well get it right :)20:44
yofelI think .diff is what I've seen most, but I also saw .patch or nothing20:44
clivejoIll use .diff20:45
yofelokay, what do you actually want from us...20:51
yofeloh right, the patch will cause a build failure20:52
soee;]20:52
clivejooh?20:52
yofelwell, you took a patch from upstream git, now you're building upstream git and applying an already applied patch -> BOOM20:53
clivejoah, that makes sense20:56
clivejohow does one fix that?20:56
clivejoremove the patch in git?20:56
yofelyes, in the _unstable branch once we're done20:57
clivejook I pushed those changes20:57
yofelthanks20:57
clivejohas the package been tested?20:57
clivejothe one in my PPA?20:58
yofelsoee said that it fixed his problem20:58
soeeclivejo: yup, rebooted twice to confirm it. system loads now in ~8seconds (before it was ~ 30)20:58
clivejoshould I redo a ppa2 with the new changelog?20:58
yofelwould be useful yes, then we can copy that to the updates ppa once it's done building20:59
yofelthen we can point people there20:59
yofelok, now let me upload that and fill out the paperwork21:02
clivejoyofel: can you explain?21:03
yofel!sru21:03
ubottuStable Release Update information is at http://wiki.ubuntu.com/StableReleaseUpdates21:03
yofelsee "Procedure"21:03
clivejoyou have to write an essay on why the update should go into the archive?21:04
yofelyep21:05
yofelhence "paperwork" :P21:05
yofelthe procedure comes from the very early ubuntu days where some X11 "quick fix" broke most of the user systems as an update21:06
clivejois it public?  can I have a read, just for curiosity?21:09
yofelwhat?21:09
clivejothe paperwork you submit21:09
yofelsure, it goes into the bug description21:09
yofelbut I'll do that after the upload21:09
yofelwhich comes after I testbuild - which I do once my xenial chroot creation is done ^^21:10
yofelall the stuff you have to do to fix stuff - that's why you don't break it in the first place :P21:10
clivejoonly way I learn is to do it :)21:14
clivejoand that usually means breaking it and then trying to fix it21:14
clivejoso you are currently building for xenial?21:15
yofelyep21:16
clivejoI guess I should create a xenial pbuilder-dist image21:18
clivejoand a xenial PPA 21:18
yofelnot sure why you need a seperate PPA, but the chroot you should do21:18
yofelor well, create it when you need it21:18
soeehttp://tanglu.org/blog/2015/10/tanglu-40-dasyatis-kuhlii-alpha-released/21:21
yofelxenial uploaded21:27
clivejoyipppeee21:28
yofelwily uploaded21:29
soeeany idea if there are some decisions for Martin's proposal @ Plasma bugfix releases ?21:31
yofeldon't we already have those?21:42
yofelor what do you mean?21:42
yofelclivejo: so, done updating the bug21:43
yofelnow we have to wait21:43
clivejowait on what?21:43
clivejofor it to land in main archive?21:43
yofelfor someone from ~ubuntu-sru to approve the update and an archive admin to accept the upload21:44
yofelthen the package has to be tested and tagged verification-done21:44
yofelonce that's done and at least 7 days have passed the update can go into -updates21:45
soeeyofel: Martin proposed on #plasma releases like: 1 week, 1 week, 2 weeks , 4 weeks etc.21:51
clivejoeakk21:51
soeeto get faster bug releases for users21:51
clivejosuch a load of fafff21:51
clivejohave you put it in a kubuntu_testers PPA?21:52
soee[11:12:40 CET] <mgraesslin> Riddell: I was thinking about how we can get our bug fixes faster to the user and had the idea of doing fibonaci releases21:52
soee[11:13:01 CET] <mgraesslin> that is: 1 week, 1 week, 2 weeks, 3 weeks and then depending whether we have bug fixes even more21:52
yofelhm... that does make sense actually..21:53
yofelso I wouldn't mind21:53

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