[21:01:12] Hello [21:01:16] hola [21:01:17] I am here [21:01:23] Hallihallo [21:01:26] --> scarabeus (~scarabeus@gentoo/developer/flyingspaghettimonster/scarabeus) hat #gentoo-meetings betreten [21:01:32] sup [21:01:36] though I may be going soon, so lets get this party started :) [21:02:50] 1. roll call [21:03:01] !herd kde [21:03:01] (kde) abcd, ago, alexxy, creffett, dastergon, dilfridge, jmbsvicetto, jmorgan, johu, kensington, mschiff, patrick, reavertm, scarabeus, thev00d00 [21:03:07] here [21:03:12] here [21:03:15] here [21:03:15] here [21:03:16] -*- johu here [21:03:18] -*- alexxy here [21:03:32] wow, 6 people :D [21:03:45] +scarabeus [21:03:53] 2. Optional semantic-desktop support [21:04:18] Some parts of KDE upstream no longer have optional support for semantic-desktop. We currently work around this by using ugly hacks but it is fragile causing bugs like bug #468104 and bug #465322. It is apparently easy to disable semantic-desktop features at runtime. [21:04:19] johu: https://bugs.gentoo.org/468104 "kde-base/plasma-workspace-4.10.2 build error with Nepomuk2"; Gentoo Linux, KDE; CONF; christopher.smith:kde [21:04:24] the herd is populated by 15 people...I wondering if we need to talk about the hour preference of the meeting [21:04:29] Options: [21:04:29] Do nothing and add more fixes/hacks as bugs crop up [21:04:29] Track upstream more closely and support optional semantic-desktop only where supported by upstream (noting that find_package() can be controlled these days unless it explicitly has a REQUIRED) [21:04:30] Drop semantic-desktop USE flag completely [21:05:00] ago: wrong topic [21:05:08] -*- dilfridge suggests "drop useflag completely and focus efforts on more worthwhile issues" [21:05:09] -*- Thev00d00 votes for: Drop semantic-desktop USE flag completely [21:05:17] TBH, I have kde on my netbook because of -semantic-desktop, with semantic-desktop enabled it was too much slowly...so I'd like to see it as -is [21:05:35] ago: just disable the file indexer, then you are fine [21:05:37] but you can disable it on runtime [21:05:43] go with dropping [21:05:47] nowdays it is not so broken [21:05:50] ok lets vote [21:05:52] and the off switch finaly works [21:05:53] I'm also not really using semantic-desktop [21:06:06] who is for drop? [21:06:11] -*- ago votes do nothing [21:06:12] -*- dilfridge is for drop [21:06:24] -*- johu +drop [21:06:48] and clearly also scarabeus and Thev00d00 as seen above [21:06:59] do nothing [21:07:23] (better would be to kill semantic-desktop, but that isn't an option) [21:07:27] i guess kensington is for it too because he bring it up on the table [21:07:36] ok, so lets drop [21:07:49] jmbsvicetto: the switch now really works, so you have it built but disabled :-) [21:07:50] ok whats the course of the action then? [21:07:59] do we need an news item? [21:08:06] remove the useflag in -4.10.49.9999 [21:08:15] no in 9999 [21:08:18] johu: no because no action on user part is required [21:08:29] scarabeus: yeah and I also should systemd when I build udev, but don't have to use it ... [21:08:37] -*- jmbsvicetto refrains from making more comments [21:08:58] -*- dilfridge does not want to hear the word systemd anymore today... [21:09:00] i would suggest to drop it with 4.11 [21:09:16] ++ [21:09:26] we could blog about how to disable it at runtime [21:09:32] time enough to get the people informed by news item etcs [21:09:36] dilfridge: good idea [21:10:18] ok i will blog about it [21:10:48] we can decide on a news item in the next meeting (june) [21:11:49] any other actions we need to do for the drop? [21:12:21] simplify code, afterwards [21:12:35] check dependencies in external packages [21:12:54] [semantic-desktop] -> [semantic-desktop(+)] [21:13:06] (we can already do that now) [21:13:25] yes [21:13:47] 3. Drop oldpim? [21:14:12] yes / no / maybe [21:14:12] mask now immediately? [21:14:12] See also this link http://dilfridge.blogspot.de/2013/05/personal-experience-and-opinion-kmail2.html [21:14:12] against [21:14:19] yes [21:14:20] i.e. no [21:14:36] sorry, -EDON'TCARE [21:14:36] we do not gain much by dropping it [21:14:47] -*- Thev00d00 -EDON'TCARE [21:14:57] (the only real cruft is the -l10- splitting) [21:15:16] its not much effort to keep it in tree, so also EDON'TCARE [21:15:19] impact on eclasses is 0 [21:15:33] real question, is there someone that uses it? [21:15:42] me :) [21:15:45] from the comments yes [21:15:57] --> reavertm (~reavertm@gentoo/developer/reavertm) hat #gentoo-meetings betreten [21:16:10] dilfridge: what is the state about the bugs you blogged before? [21:16:21] kleopatra works fine again [21:16:26] ago: I don't use kdepim and only have parts of it installed because of semantic-desktop [21:16:34] the other stuff still exists but is nonfatal [21:16:45] I meant if there is someone apart dilfridge [21:17:01] only real bug is that you can't set sent mail folder per identity anymore [21:17:09] --> mschiff (~mschiff@gentoo/developer/mschiff) hat #gentoo-meetings betreten [21:17:16] ^ ^ [21:17:51] --> Guest43157 (46c082af@gentoo/developer/creffett) hat #gentoo-meetings betreten [21:17:54] i would suggest as everytime it comes on table, lets wait until its not working with new kde sc release [21:18:18] any objections, if not i move to the next topic [21:18:21] ? [21:18:41] yep. postpone. if anything pops up, assign bug to me with kde in cc [21:18:46] <-> Guest43157 heißt jetzt creffett|mobile [21:19:04] 4. Handling of systemd patches [21:19:10] let's just agree on a consistent way of handling it (as long as noone from kde team runs systemd) [21:19:19] I run it [21:19:23] :) [21:19:25] Yay [21:19:25] -*- creffett|mobile is here, sorry. Dinner running late [21:19:40] i already run it on my netbook [21:19:43] If there is somthing to test, feel free to cc me everytime [21:19:43] BUT [21:20:20] what i miss in the whole discussion is the fact that systemd is about standardization [21:20:33] and not about boot time [21:20:36] My general opinion about systemd unit files is that we should wait / let it up to upstream [21:20:56] and if we add thousands of patches downstream the goal is totally failed [21:21:03] I personally am unwilling to work on it (kde or elsewhere on Gentoo) [21:21:19] well [21:21:33] and we never add patches to add new features in kde packages... [21:21:45] I'd suggest also to forward upstrem the patch we see in the bugzilla [21:21:50] let kde upstream handle the stuff [21:21:53] I also don't like getting tons of bugs to add systemd support to X, Y, Z. Those wanting systemd should be convicing X, Y and Z to support it [21:22:06] how about "feel free to add patches that have been accepted upstream" [21:22:50] whats the point to backport this when the release is in 8 weeks? [21:23:24] dilfridge: If the patch is going to be added by upstream, if someone else (not me) is doing that work and it doesn't break or affect those running openrc, no objection from me [21:23:56] jmbsvicetto: ++ [21:24:24] jmbsvicetto: well, that comment was intended as "reply to our systemd guys" [21:24:35] But people will still complain "I gave you the file, it's trivial to add" [21:24:42] I use systemd [21:24:56] I am happy to support the systemd patches/unit files [21:25:03] possibly with ago [21:25:16] I think we're running into the same trap as the dev ml [21:25:24] creffett|mobile: I think that we need to check if thservice start [21:25:27] We should not worry too much about small support files [21:25:43] apart kdm do we have a lot of packages that installs an init script? [21:25:46] but just add them if they are helpful and do not hurt otherwise. [21:25:49] we dont need to discuss the small unit files.... [21:26:16] the only thing worth discussing are significant patches for code [21:26:18] its already added in kde-base/kdm [21:26:45] and i add the gentoo unit file in upstream bug [21:26:52] last kdm works for me [21:28:15] if we start to add not reviewed source code patches i will leave the herd for sure [21:28:38] its not about systemd, its about the correct way [21:29:04] johu: we suggest to add the code to kde codereview, if it is accepted then we will include [21:29:23] ++ [21:29:26] ++ [21:29:28] ago: thats what i have done in the bug [21:29:53] ok let me try to summarize: [21:30:19] So this means we as a herd from now on will only apply patches that have been accepted upstream? [21:30:24] johu: and that's the correct way, so let's continue [21:30:41] but if our packages does not install a init script, we don't need to discuss anymore [21:31:32] "Small support files as e.g. unit files that affect only systemd but do not interfere with other software can be added locally but should eventually be upstreamed. Patches for adding functionality should be accepted upstream first." [21:31:41] sounds ok? [21:32:01] dilfridge: so long as this is not only for systemd [21:32:01] sound like music [21:32:01] yes [21:32:10] good here [21:32:20] Thev00d00: basically it's already the policy... [21:32:21] ++ [21:32:47] refreshing a policy is a good thing [21:33:01] 5. Bugs [21:33:16] bug 438790 [21:33:18] johu: https://bugs.gentoo.org/438790 "kde-misc/polkit-kde-kcmodules: Store polkit configuration changes to /etc instead of /usr"; Gentoo Linux, KDE; IN_P; nikoli:kde [21:33:27] sigh [21:33:58] Nothing from upstream yet on that [21:34:31] didn't other distros patch that? maybe we could attach someone's code to that bugreport... [21:34:39] By the time they have a fix frameworks 5 will be out and this will be irrelevant [21:34:40] who was working on that bug, it was in my off time [21:34:47] Me [21:34:57] i think this is fixed in fedora [21:35:52] there was a thread about it on the release-team ml I think (but I dont have my archive here) [21:36:21] I believe so, but am not sure, and it worries me that most dostros said they don't ship that module [21:36:46] http://pkgs.fedoraproject.org/cgit/kdelibs.git/tree/kdelibs-4.6.90-kstandarddirs.patch [21:37:47] err [21:37:53] that was the link posted in the ml thread [21:37:54] ... [21:37:54] how do I get that page to show code? [21:38:33] maybe they have dropped it in the meantime [21:38:43] http://pkgs.fedoraproject.org/cgit/kdelibs.git/tree/kdelibs-4.10.0-kstandarddirs.patch [21:38:46] more useful [21:38:51] I will try that patch when I return from vacation and we can include it in the next bump [21:39:22] is there an chance to get it upstreamed? [21:39:34] we're just falling into our own trap :D [21:39:39] creffett|mobile: ^ [21:41:37] Well yeah [21:41:42] when the bug is not in hurry we could try to get it in before 4.11 [21:42:17] actually that looks small enough so it's realistic [21:42:17] But this is a fairly major issue [21:42:47] And I repeat that most distros do not ship this module [21:43:01] yes, which lets me wonder why [21:43:10] (does fedora?) [21:43:13] yeah we could drop the package... [21:43:27] -*- Thev00d00 votes drop it [21:44:24] I think it isn't officially released [21:45:21] no opinion here [21:45:28] I vote make it no longer a kdelibs dep [21:46:07] thats not a complete solution [21:46:21] either we fix it or we drop it [21:46:22] what does the package actually do? [21:47:33] ^ ^ [21:47:33] i guess you can manage the polkit rules [21:47:37] Not sure tbh [21:48:03] so we're going through a lot of pain because of something buggy that noone uses or even knows what it's good for. awesome. [21:48:39] lastrite? [21:48:51] vote! [21:49:11] options? [21:49:31] 1. get fedora patch upstream [21:49:45] (and patch polkit-kde-kcmodules as required for that) [21:50:17] 2. get rid of it [21:50:34] 3.? [21:50:53] 3. drop it as kdelibs dependency and drop it to ~arch [21:51:03] -*- ago votes 1 [21:51:04] 3 [21:51:15] -*- dilfridge does not care [21:51:15] -*- johu votes 1 [21:51:30] --> creffett (~creffett@gentoo/developer/creffett) hat #gentoo-meetings betreten [21:51:33] there we go [21:51:36] much better [21:51:59] <-- creffett|mobile (46c082af@gentoo/developer/creffett) hat das Netzwerk verlassen (Disconnected by services) [21:52:12] scarabeus, jmbsvicetto ? [21:52:24] alexxy? [21:53:00] I reluctantly vote punt it, with the caveat that we should make sure that punting it doesn't completely hose KDE before we send the lastrite [21:53:31] not for me to decide [21:53:38] you guys just voted to have policy upstream first [21:53:39] :D [21:53:45] so you cant just rape it right away [21:53:47] :D [21:54:00] it was never realy released :P [21:54:34] ^^ [21:54:41] perhaps retire it to kde overlay only? [21:54:50] thats option [21:54:56] wipe it from stable ebuilds would work [21:54:58] bugs persists in overlay too [21:55:05] people shouldn't have this in their world files to start with, the only thing pulling it in is kdelibs [21:55:22] yes, but for most people it won't matter anymore [21:55:44] creffett: you started to work on the bug you can decide on you own as lead ;) [21:56:02] we will cover your decision with full force [21:56:04] -*- creffett puts his lead hat on [21:56:16] -*- creffett thinks that we need a nice pope hat or something to mail around between leads [21:56:18] anyway [21:56:27] bug 444952 [21:56:29] johu: https://bugs.gentoo.org/444952 "Add KDE Plasma Active support"; Gentoo Linux, KDE; CONF; creffett:kde [21:57:02] unless someone seriously objects, I say move polkit-kde-kcmmodules to overlay, remove kdelibs dep, and re-evaluate if and when they get their act together upstream [21:57:15] creffett: the lead gets a beer stein on the head at fosdem... [21:57:25] whats the state? why we need to discuss this? [21:57:29] plasma-active was added to the overlay under kde-active category [21:57:43] why not kde-base/ [21:57:48] if anything needs to be discussed, it's basically "when and how to move it to tree" [21:58:05] scarabeus: no real reason besides it not strictly being part of the KDE SC or whatever they call it these days [21:58:12] scarabeus: because it's not strictly part of KDE SC? [21:58:19] nice timing. [21:58:41] kde-frameworks wont be part of kde sc too.... [21:58:44] it's not really KDE is the thing, it's a plasma product [21:59:04] (which is a sub-product of KDE, I know) [21:59:21] the kde thing is going away anyway... we should already prepare renaming ourselves... [21:59:21] well how many packages it is [21:59:28] no new category unless it is more than 10 [21:59:42] ...9. [22:00:03] next question, what's a more appropriate category? do we agree that it should go in kde-base? [22:00:15] kde-active is fine. [22:00:28] creffett: you could package the telepathy active client... [22:00:38] then you would have 10 [22:00:40] hehe, then it's 10 [22:00:53] that's the other thing, it's not entirely ready because there are a lot of related products that aren't fully packaged yet [22:00:54] okey you suckers are covered here [22:00:55] :P [22:01:04] lets keep it in testing in cvs [22:01:08] more eyes more bugs [22:01:18] not ready to be moved yet [22:01:36] since I haven't had the time to go through all of the stuff on the website listed as "supporting plasma-active" [22:01:43] oh, and I honestly haven't actually had the chance to test it :P [22:01:58] my proposal, keep it in overlay [22:02:08] it completely screws up the display of my laptop, and I don't have a mobile device I can sacrifice to the cause [22:02:15] make an wiki article/howto and all is fine [22:02:29] okay [22:02:39] and maybe even put a little traffic on -desktop for once? :P [22:02:46] email out a request for testing [22:02:50] yeah or blog about it [22:03:16] bug 456750 [22:03:18] johu: https://bugs.gentoo.org/456750 "kde-base/plasma-desktop[-rss] unintuitive dependency on kde-base/libplasmaclock[-holidays]"; Gentoo Linux, KDE; UNCO; alpine.art.de:kde [22:03:30] I need to leave. See you guys later [22:03:42] jmbsvicetto: thanks for your presence [22:03:48] cu [22:04:19] I g2g in a minute also, anything else left on the agenda? [22:04:27] yes bugs [22:04:40] (and open floor) [22:04:48] the use should be renamed [22:04:53] i never understood why we call it rss :D [22:05:10] scarabeus: propose a name [22:05:22] -*- creffett has no idea what these use flags actually control [22:05:23] $(cmake-utils_use_with rss KdepimLibs) [22:05:29] aha! [22:05:33] it enables fucking semantic desktop [22:05:41] just bind it in [22:05:41] lol [22:05:44] AHAHA old topic [22:05:51] then yeah, calendar/semantic-desktop might be more appropriate [22:05:53] call it semantic-desktop and hard-enable it in 4.11 :) [22:06:11] just to more spam [22:06:13] rss? ( [22:06:13] we get rid of it when we do -semantic-desktop [22:06:14] $(add_kdebase_dep kdepimlibs 'semantic-desktop?') [22:06:16] $(add_kdebase_dep libplasmaclock 'holidays') [22:06:17] ) [22:06:19] !rss? ( $(add_kdebase_dep libplasmaclock '-holidays') ) [22:06:20] it was even automagic there :D [22:06:21] why does holidays actually need a use flag? [22:06:25] thats reasonf or the - [22:06:46] creffett: it crashed when you had kdepimlibs without semantic and libplasmaclock and holidays [22:06:53] ah. [22:06:55] okay [22:07:06] ok we will solve it with 4.11 ;) [22:07:14] i will take of it [22:07:16] care [22:07:39] bug 461684 [22:07:40] johu: https://bugs.gentoo.org/461684 "kde-base/kwin: backport patch to remove tearing on Intel systems"; Gentoo Linux, KDE; UNCO; mike:kde [22:08:32] yeah do it [22:08:36] we have this up for discussion because... [22:08:52] ... because it's large and intrusive [22:08:57] ah [22:09:07] that is an absurdly long discussion upstream [22:09:09] also see https://bugs.kde.org/show_bug.cgi?id=307965 [22:09:12] "I am not sure if this is a good idea since there is upstream discussion about specifically avoiding KDE/4.10 branch due to the invasive nature of the patch and potential for regressions." kensingtion [22:09:31] i would vote against it! [22:09:46] we should just add it into the next 4.10.X release, and if there are problems we could always revbump without the patch again [22:10:03] so it gets some time in ~arch for testing [22:10:11] -*- creffett has no opinion here [22:10:12] Its going to be in 4.11 so any issues will get to use soon enough [22:10:27] no strong opinion here either [22:11:21] stay with upstream [22:11:22] users can always use user patches with kde packages... [22:11:39] ok let's leave it out then [22:11:50] mm, true, and we do support userpatch [22:12:32] bug 468002 [22:12:34] johu: https://bugs.gentoo.org/468002 "kde-base/kdelibs-4.10.2 ELOG output includes Your homedir is set to ${HOME}/.kde4"; Gentoo Linux, KDE; CONF; rich0:kde [22:12:38] fast vote [22:12:40] punt it [22:12:41] can be removed [22:12:42] -*- johu ++ [22:12:44] punt [22:12:48] kill it [22:12:52] ok thanks [22:12:56] -*- Thev00d00 loves the word punt [22:13:02] on second thought... [22:13:09] nah, kill it. :) [22:13:12] bug 468894 [22:13:14] johu: https://bugs.gentoo.org/468894 "Please restore kde-base/plasma-workspace-4.10.1 because of a bug"; Gentoo Linux, KDE; UNCO; gentoo:kde [22:13:16] fast vote [22:13:18] there is absolutely no reason for people to see that message anymore :P [22:13:27] no [22:13:31] no [22:13:31] -*- creffett votes no [22:13:39] NI! [22:13:41] also looking forward to the KDE long-term support release :P [22:13:52] wrong topic :D [22:14:33] eh, it's related, since it's talking about keeping older versions [22:15:09] the bug he mentioned was already in 4.9 [22:15:23] true [22:15:25] -*- creffett shrugs [22:15:39] 6. Open floor [22:15:57] cmake-utils now has inlined patching support in overlay [22:16:27] going to eapi-bump the last few EAPI < 2 packages myself soon since it's been plenty of time for the maintainers to do it [22:16:50] which reminds me to remind you to make a tracker out of the monolithic bug mess [22:17:08] and that reminds me to say "meh, I'm bumping it all myself" [22:17:33] fine but in the future make TRACKER bugs.. [22:17:38] will do [22:18:04] i have at least 2 more topics [22:18:42] i dont like that kde-stable is not using our mail alias [22:19:33] all the things that tampakrap mentioned when the sub-projects are true in my opinion [22:20:10] hmm? anything I missed? [22:21:03] you can read the log from last october again, its only my opinion no need for action at the moment [22:22:13] and additionally i had the feeling that the stabilization of 4.10.1 was to early for the 4.10 [22:22:58] so enough bad vibrations: [22:23:27] as you may noticed i am back with full force as you may noticed in the last weeks [22:23:37] who are you, again? ;) [22:24:50] kensington's offtopic: There is a prototype reviewboard instance setup for use with the KDE overlay[1] [22:25:01] http://astralcloak.net/~kensington/rb/ [22:26:12] i would rather see running gerrit gentoo instance for all projects [22:26:19] *like to [22:27:02] gerrit++ [22:27:33] we could ask kensington if he wants to host a test gerrit instance [22:27:40] to compare [22:27:43] gerrit is overcomplicated [22:27:47] imho [22:28:13] patches of patches of patches [22:28:21] dilfridge: not really [22:28:24] worked with both [22:28:29] and gerrit is actually easier [22:28:38] for usage on both sides [22:28:41] reviewer and submitter [22:28:48] ok [22:28:53] no opinion here [22:29:09] in lo i just do "gerrit push" and "gerrit review id" if i feel like not using gui [22:29:22] and on web interface i can inline comments and approve commits directly [22:29:28] so i can do that even from tablet [22:30:21] ok [22:30:29] so who is setting up a test instance? :D [22:30:35] lets ask kensington to host a gerrit instance to test it and compare with reviewboard and then decide which we want to push into gentoo infra [22:30:57] btw did anything come from that gitlab discussion? [22:31:05] theo didnt mind if someone is willing to maintain it [22:31:58] hm he should care about his hom(i)e herd [22:32:24] any other topics? [22:32:28] nothing here [22:33:04] FYI git kdenetwork migration is in progress [22:33:39] thank you all [22:33:42] meeting closed [22:33:52] g'nite [22:34:51] nn [22:35:02] 'night [22:35:02] <-- creffett (~creffett@gentoo/developer/creffett) hat #gentoo-meetings verlassen [22:35:39] nite