[00:00] tsimonq2: that's half way there then [00:01] tsimonq2: I have the impression you just have to do quilt push && quilt refresh for each one of the patches with fuzz [00:02] sometimes..... [00:03] sometimes not so easy.. [00:04] acheronuk: I know, in this case with my stuff here against git they refresh fine [00:16] I'll have a look what's happened with them in the morning [00:16] For now. goodnight :) [00:17] AHHHHHHHH NOO WHYYYYY Project merger_kde-baseapps build #18: FAILURE [00:21] trying to checkout a revision that no longer exists [00:21] yeah I'll fix it... [00:23] just make sure that the patchs after modification either (a) do what they were intended to or (b) are removed disabled if now defunct [00:23] acheronuk: I'm keeping that in mind ;) [00:24] acheronuk: thanks though [00:24] I know that's stating the obvious, but with forcing things and quilt refresh, it's easy to end up with empty patches [00:25] Ok. Goodnight [00:25] :) o/ [00:27] tsimonq2: I would like to know how you could reproduce it locally, apparently I can't tell here dpkg-source to fail on patches with fuzz [00:27] allegedly it does that by default [00:28] santa_: I'm using sbuild [00:41] YES I think I did it! [00:41] come on [00:41] go go go go GO! [00:56] :(\ [01:17] ^ I'm working on fixing that now [01:22] regarding this commit: https://git.launchpad.net/~kubuntu-packagers/kubuntu-packaging/+git/kde-baseapps/commit/?id=c6d67e5ff88332f34f11ef3d4ddaaa213b905166 [01:22] tsimonq2: ↑ are you sure this is what you actually want to do? [01:22] santa_: what about it? [01:23] because kubuntu_folderview_livecd_directory.diff is gone which seems refreshable here [01:23] also [01:23] oh okay, really? [01:23] after I fix the build errors I'll restore [01:24] kubuntu_folderwidget_drawing.diff and kubuntu_disable-failing-test.diff are also disabled in debian/patches/series [01:25] ...no? [01:25] and kubuntu_ignore-tests.diff is now on top, it seems there is no need to do that [01:25] tsimonq2: just see the diff of the link I pasted [01:25] ...you cloned from the Git source and are using a fresh Git clone? [01:26] I think so [01:27] and according to what I tested here to get the patches applied without fuzz it was enough to do "quilt push && quilt refresh" on each patch having fuzz [01:28] the orig.tar.xz I'm using is a tarball created with latest upstream code from Applications/16.08 from kde's git [01:28] oh k [01:28] let's see [01:28] the packaging I'm checking is the latest in the kubuntu_unstable branch [01:28] pull some changes I just made please [01:30] santa_: would you like to fix it? if you push it somewhere I can pull from your changes [01:31] tsimonq2: I see them, but the stuff with the patches I described is still there [01:32] santa_: I know, my changes were just dep fixes [01:32] tsimonq2: what if I guide you towards fixing it? so you deal with quilt properly [01:32] santa_: sure [01:32] what do I need set up? [01:33] tsimonq2: ok, revert the commit where you touched the patches in your local clone and paste me the result of git show [01:34] tsimonq2: the hash of the commit in question is c6d67e5ff88332f34f11ef3d4ddaaa213b905166 [01:35] santa_: so literally git revert HASH ? [01:35] tsimonq2: git revert c6d67e5ff88332f34f11ef3d4ddaaa213b905166 [01:35] if that's what you mean [01:36] yeah [01:36] so we can re-do it easily [01:36] $ git show | pastebinit [01:36] http://paste.ubuntu.com/22113039/ [01:37] allright. now let's do it properly [01:38] tsimonq2: now inside the git clone of the packaging "git clean -xdff" then "tar xvf ../kde-baseapps_16.04.3.orig.tar.gz --strip=1" [01:39] so we have the proper upstream source code [01:39] yup [01:39] done [01:40] tsimonq2: so now if you do a "quilt push -a" you must get something like this https://paste.kde.org/phkdrbmle [01:40] can you confirm this is correct? [01:41] wrong [01:41] http://paste.ubuntu.com/22113666/ [01:42] hmm [01:42] ohh [01:42] santa_: where did your upstream tar come from? [01:42] so we have something different here [01:42] yeah [01:42] from kde's git, branch Applications/16.08 [01:43] I don't think it pulls from that [01:43] I'm pulling from master [01:43] and the kci pulls from master to create the tarballs? [01:44] I believe so [01:47] tsimonq2: lets work on that asumption then, let me recreate my tarball to see what's up actually [01:47] alright [01:48] santa_: I'm really sorry but I have to go to bed [01:48] I'm really tired [01:49] santa_: if you leave me something for tomorrow, I'll be sure to take a look [01:51] tsimonq2: allrigh, I'm investigating the thing [02:12] ok, just for the record [02:13] this is the impression I have after a quick look (needs more proper investigation) [02:13] enable_debianabimanager.diff - must be refreshed [02:13] enable_dlrestrictions.diff - must be refreshed [02:15] kubuntu_folderview_livecd_directory.diff - this patch patches something which doesn't exist anymore in -baseapps, but may be needed to apply it in other package or other file [02:15] kubuntu_folderwidget_drawing.diff - same as the previous one [02:17] kubuntu_disable-failing-test.diff - I think this one should be dropped, because the test in question doesn't exist anyore, not even in the file lib/konq/autotests/CMakeLists.txt [02:19] kubuntu_ignore-tests.diff - I have the impression this one should be dropped too, because... [02:19] Description: ignore failing tests [02:19] ignore tests failing due to switch to plasma 5 [02:19] Author: Jonathan Riddell [02:19] Origin: me [02:19] Forwarded: nope [02:19] Last-Update: 2015-09-01 [02:20] it says in the description the tests fail due to plasma 5, but apparently the test were ported to frameworks [02:20] * the tests [02:21] disclaimer: all what I wrote above is based on the assumption that KCI grabs the source code from kde's git master [02:22] whenever you wake up, yofel ↑ [02:23] context: patches of kde-baseapps/kubuntu_unstable they don't apply well anymore, making the KCI fail [07:12] santa_ tsimonq2 yes, unstable is builds of master [07:13] think yofl said he is trying to work out how to reinstate stable branch builds on KCI [07:37] reminder: 14.04.5 needs testing http://iso.qa.ubuntu.com/qatracker/milestones/365/builds [07:45] DLing now to test install in VirtualBox [07:51] Good morning. === Blizzzek is now known as Blizzz [11:11] 'Morning folks [11:18] hiho BluesKaj [11:20] Hey soee [11:38] aaand hello :) [11:39] hi jimbo! [11:40] Morning! [11:40] afternoon :) [11:40] Feels good to be back on a US time zone! [11:40] :D [11:40] \o/ [11:41] Hi there.. [11:41] Hi Rick :) [11:41] (sleep schedule, I didn't travel anywhere :P) [11:41] Hey Rick! [11:41] How are you? [11:42] PST or EST? ^^ [11:42] CDT [11:43] So an hour before EDT [11:43] oh right, it's summer [11:46] clivejo: you said you are packaging kdiagram? [11:46] acheronuk: nods [11:46] great :) [11:46] I was attempting it [11:46] I need a chroot [11:46] https://launchpadlibrarian.net/276641695/buildlog_ubuntu-yakkety-amd64.kdiagram_2.6.0-1ubuntu1~ubuntu16.10~ppa1_BUILDING.txt.gz [11:48] I cant remember what packages contain what [11:49] I git cloned the kde repo and built that, and it just built straight off with whatever -dev packages happened by chance to be on this machine, so looks nothing too complicated [11:53] Failed call: qmake-qt5 -query "QT_INSTALL_PLUGINS" [11:53] same here [11:53] is that qtchooser? I can't recall just this sec? [11:53] it built fine locally [11:54] and I cant remember how to configure a new pbuild [11:54] and of course the instructions are a great help [11:55] I had that same error in something a couple of weeks ago, and damned if I can recall what! [11:55] maybe one of the packages that ported to KF5? [11:59] the same? https://launchpadlibrarian.net/274921440/buildlog_ubuntu-yakkety-amd64.kolourpaint_4%3A16.04.3+p16.10+git20160725.0750-0_BUILDING.txt.gz [11:59] yes [12:00] its needs qtbase5-dev? [12:00] not quite "QT_INSTALL_PLUGINS" vs "QT_INSTALL_PREFIX" [12:01] but this sorted my one https://git.launchpad.net/~kubuntu-packagers/kubuntu-packaging/+git/kolourpaint/commit/?h=kubuntu_unstable&id=e85e0616c75d20ecde7f2138100537e6af84160e [12:02] Im "trying" to create a chroot locally [12:02] its not going well [12:02] qtbase5-dev at least, as I seem to recall they did port to qt5 [12:02] not well, how? [12:02] * clivejo kicks pathetically slow internet [12:03] oh. that! [12:03] linode? [12:03] I cant work properly from Linode [12:03] shame [12:03] I dont want to upload my keys to it :/ [12:04] not that I dont trust yofel [12:04] shh or gpg? or both? [12:04] I have my machine configured how I like it [12:05] and I need a local chroot [12:05] true. It was a slight faff to sort even a local reinstall the other week, even being able to copy config files straight over [12:05] just instructions for how to set it up are so confusing [12:06] instructions are all over the place [12:06] we need a set of instruction for just Kubuntu [12:07] that is pretty simple http://packaging.ubuntu.com/html/chroots.html [12:08] but depends what you want to do [12:08] I had a great set of tools for xenial [12:08] but I cant seem to get it to work with yakkety [12:09] everything seems as it was, in place upgrading to YY here [12:10] but I bet if I tried to transfer to my laptop it would all go pear shaped [12:12] do you have eatmydata installed? [12:13] yes [12:13] I think [12:13] is it working correctly in yakkety? [12:14] my packages dont seem to be cached [12:14] seems to want to download everytime [12:15] * clivejo leaves it downloading and goes for some lunch [12:15] I mostly use pbuilder-dist and get the cached on my home partition/folder [12:16] setting APTCACHEHARDLINK=no so it can link accross [12:16] where is that set? [12:17] .pbuilderrc [12:19] whats the key combin to drop out of the pbuilder environment [12:19] CTRL D [12:19] thanks! [12:20] I've not got pbuilder set up quite the way the kubuntu docs said, but it works for me and I'm loathed to change it now [12:21] beside tweeks [12:24] oh I think that is working now [12:24] not downloading all the packages everytime [12:25] last thing you would need on a slow connection! [12:25] yup [12:26] especially as they are here locally [12:35] I use sbuild and apt-cacher-ng on my host system [12:36] Also, I can do SHM builds with it too [12:37] So if I already downloaded the package files and I have some RAM free (I have 16 gigs) builds FLY [12:37] https://wiki.ubuntu.com/SimpleSbuild has instructions [12:38] clivejo: ^ [12:40] I set up a special KCI sbuild that I use that has the PPA enabled [12:44] must try that sometime [12:46] Must try now :P [12:47] Don't keep putting it off [12:47] It is absolutely amazing [12:48] pbuilder is v swift here. limit is on cpu(s) not much else [12:49] v swift? [12:53] quick [12:54] * yofel has pbuilder with ppa on shm - works the same ^^ [12:54] though eatmydata is easier to do with sbuild [12:54] not that it has much of an effect here ^^ [12:55] yofel: do you have apt-cacher-ng or something like it? [12:55] if think any further tweaks I made here would only be marginal speedwise [12:58] tsimonq2: yes [12:58] I do occasionally use sbuild after all [12:58] oh cool [13:08] yofel: would you be able to confirm what santa_ said yesterday re: kde-baseapps ? [13:09] tsimonq2: regarding what exactly? I read the discussion and found it a bit confusing [13:12] yofel: kde-baseapps was failing because of patch fuzz. I committed this, which I thought was correct: https://git.launchpad.net/~kubuntu-packagers/kubuntu-packaging/+git/kde-baseapps/commit/?h=kubuntu_unstable&id=c6d67e5ff88332f34f11ef3d4ddaaa213b905166 [13:12] yofel: santa_ stopped me and said that it wasn't correct [13:13] yofel: he left some comments last night telling me what he believed should happen [13:13] yofel: I'm all for fixing my mistake, I just want to understand why :) [13:13] (if I even made a mistake) [13:14] yofel: I'm asking, what's your opinion on this? [13:16] hm, so the debian abi manager fuzz part is correct [13:17] great [13:17] he does have a valid point about reordering series, please don't do that. It might work this time if the patches are independent from each other, but as soon as you need a patch order that would've broken something [13:17] it's also an unnecessary diff [13:18] yeah, I must have did that when I was working with the patch locally, I'll revert that [13:19] refreshing enable_dlrestrictions.diff isn't really neccessary, unless that has fuzz too. Offsets we don't usually refresh [13:20] oh THAT is where that folderview OEM patch was [13:21] tsimonq2: yes, that patch needs moving to plasma5 [13:21] once someone finds out where [13:21] hi everyone [13:22] hey santa_! we're just discussing kde-baseapps :) [13:22] that's supposed to show you the "OEM setup" icon on the desktop after you did an OEM install [13:22] oh yeah [13:22] ok [13:22] which I think is missing currently? [13:22] yofel: as a possible stop in gap solution we could do this with the patches: [13:23] refresh the abi manager and dl restrictions, as it's obvious that ones must be refreshed [13:23] drop the ones disabling the tests as it seems they are not needed anymore [13:23] that ignore-tests patch I'm not sure what to do with, does that even do anything? [13:24] that removes a test which doesn't exist anymore [13:24] kill it [13:24] as I mentioned yesterday [13:24] and regarding the foldevview patches we could comment them out temporarily [13:25] no, that can go, that's useless in that package and needs to be added elsewhere [13:25] not sure if in plasma-framework or plasma-workspace though [13:25] so delete them? [13:26] yes, esp. as the other one seems like a Qt4 workaround o.O? [13:26] oh, one cuestion more, the KCI grabs the upstream source code from git master right? [13:27] unstable == master, stable == branch (but those jobs are MIA) [13:27] allright [13:27] so at this point I could either: [13:28] a) prepare a merge request with the agreed changes to the -baseapps changes [13:28] b) mentor tsimonq2 to do it [13:28] whichever you prefer [13:29] I'd kinda like #2 :) [13:29] santa_: but there are a bunch more fuzz issues in KCI [13:30] I can help you with quilt if you are in trouble too [13:34] yofel: so remove folderview, what about folderwidget? [13:35] I assume I can remove that? [13:35] yes, that plasma widget isn't even in use anymore as that's kde4 [13:35] alright [13:36] the only 2 ones which must survive are the abi manager and the dlrestrictions one iirc [13:37] kubuntu_disable-failing-test.diff [13:37] that seems outdated ^ [13:37] it can't find the input file [13:38] yes, that test doesn't exist anymore, that patch should be removed [13:38] alright [13:38] kubuntu_install_flash.diff is commented out, I can remove it, right? [13:40] we also have upstream_Use-user-places.xbel-instead-of-bookmarks.xml-in-pla.patch but it's not in series [13:40] can I remove those two? [13:40] I would leave the flash one like that for know, unless you want to investigate properly why it should be finally removed [13:41] the upstream one sounds safe to remove [13:42] tsimonq2: indeed the upstream one was already applied upstream according to debian/changelog, so I think it's safe to remove it [13:43] great :) [13:43] probably someone forgot to "git rm" it after it was deleted from series [13:46] santa_, yofel: how does this diff look? http://paste.ubuntu.com/22178816/ [13:48] tsimonq2: if I'm not overlooking anything seems good - i.e. does what we discussed here [13:49] ahoneybun: https://twitter.com/feralgames/status/761127268227747840 [13:51] They are on fire [13:51] I've got 3 games that they ported [13:53] yofel: are you enabling stable again in KCI? [13:56] clivejo: once I have a place to store the configuration, and the tooling can read that... yes [13:56] For now I'll probably make that all a pre-source static configuration in kci.yaml, which should be fairly easy to adapt to [13:58] pushed! \o/ https://git.launchpad.net/~kubuntu-packagers/kubuntu-packaging/+git/kde-baseapps/?h=kubuntu_unstable [13:58] although I still think the build fails [13:59] LOL ''So you tried to chicken out from symbol validation, not going to happen'' [14:01] ARGH [14:03] la la la la [14:03] lol [14:04] o/ clivejo [14:04] * acheronuk hums a tune [14:04] doh ray mee [14:04] o/ Jimarvan [14:05] doh [14:05] yo :) [14:05] double doh [14:05] lol I'm listening to some Eminem loudly, I can't hear you singing lol [14:06] how many retries? [14:06] hm? [14:06] grrrr this IP camera is really annoying me now [14:07] piece of *beep* [14:14] grrrr https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=829545 [14:14] Debian bug 829545 in lintian "lintian: [PATCH] Do not complain about new Testsuite-Triggers in dsc." [Normal,Open] [14:14] let me cherry pick that [14:21] ahoneybun: ping [14:21] pong? [14:22] ahoneybun: how did you build the code for the Kubuntu manual? (build) [14:23] sphinx [14:23] it is a python tool for making docs [14:23] sphinx-doc.org [14:23] it's an awesome idea [14:24] is it? [14:24] works pretty well so far [14:24] *worked [14:25] ahoneybun: would I be able to fork it and customize it for Lubuntu? [14:25] there's no license in it [14:25] of course [14:25] great! :) [14:25] I should add that I guess [14:26] GPL v3 would be best no? [14:26] idk as long as I can fork :P [14:26] tsimonq2: it's currently on a server like this: docs.kubuntu.org [14:26] xubuntu has a similar server name but not sure what they used [14:28] I think they used something else maybe latex [14:28] but it is html [14:28] tsimonq2: the github page for the project has some links [14:28] I *think* kde usually goes with the GFDL for documentation https://www.gnu.org/licenses/fdl-1.3.en.html [14:28] to the sphinx main page and rst [14:29] which is what the docs are written in [14:29] then a few scripts make html and such [14:29] ahoneybun: I might not fork, but I'll see if I can at least use the same style, with your permission of course :P [14:29] it can make pdf and epub as well [14:30] the site is using the bootstrap theme [14:31] LN 107 and down: https://github.com/ahoneybun/kubuntu-manual/blob/master/source/conf.py [14:33] tsimonq2: ^ [14:34] yofel: http://l10n.kde.org/docs/doc-primer/licenses.html [14:34] yep [14:38] there added a license [14:38] it allows modifing [14:55] yofel: is it the -c0 already there in override_dh_makeshlibs that is making this fail? http://kci.pangea.pub/view/yakkety%20FIX/job/yakkety_unstable_akonadi/lastFailedBuild/console [14:55] acheronuk: yes [14:55] come to think of it, that should be a bit more descriptive.... [14:56] ok. was looking at the ruby and couldn't quite see if that should fail it, or just give that message and carry on [14:57] acheronuk: yes it fails on that. We did use -c0 by default in the past, but today it's better to not have it when working on the release packages [14:58] I didn't think that anything would actually hit this..... [14:59] understood. I just couldn't work out from the code if failure was the intention there [15:02] yofel: ok. I was too dumb to notice the 'raise' part of that! [15:12] ahoneybun: https://github.com/tsimonq2/lubuntu-manual [15:30] Cool cool [15:30] That @ symbol is wrong though [15:30] hm? [15:33] The @ messes up the link [15:54] ahoneybun: Telegram did that [15:54] Mm yea most likelu [16:43] think this needs a new library package adding? libKF5MimeTreeParser https://launchpadlibrarian.net/276563376/buildlog_ubuntu-xenial-amd64.kf5-messagelib_4%3A16.04.3+p16.04+git20160803.1459-0_BUILDING.txt.gz [18:48] see ya later peeps :D [18:48] another hard working day ;) [18:48] lol 19:48 jeez... [18:49] o/ [18:49] need a swim :P [18:49] CIAO [19:49] kf5-messagelib seems to require 2 new library and associated -dev packages for libkf5webengineviewer5 & libkf5mimetreeparser5 [19:49] Q1. they don't seem to have moved any functions from the existing libraries into those, so I presume they should start afresh with their own new symbol files? [19:49] Q2. if so, then considering that these new libraries are in both master and 16.08 branches, using pkgkde-gensymbols would we still version them initially as 16.04.3+git$date if we are doing this for KCI? [20:33] and back :D [20:33] o/ [20:34] hi simon :) [20:38] hi jim [20:39] hey rik [20:39] hey rik :D [20:39] acheronuk: if they are completely new it's ok to start new symbols files. the initial version doesn't matter, as long as it <= than the initial version [20:40] ugh [20:40] let me reword it [20:40] acheronuk: if they are completely new it's ok to start new symbols files. the initial version of the symbols in the symbols file doesn't matter, as long as it <= than the initial version of the package in the archive [20:40]