15:12:53 #startmeeting kde-sig 15:12:53 Meeting started Tue Jan 9 15:12:53 2018 UTC. The chair is rdieter. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:12:53 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:12:53 The meeting name has been set to 'kde-sig' 15:12:56 hi all 15:12:58 hi 15:12:59 #topic roll call 15:13:02 who's present today? 15:13:09 hello 15:13:24 hello 15:13:36 #info rdieter lupinix jgrulich mbriza present 15:13:42 #chair lupinix jgrulich mbriza 15:13:42 Current chairs: jgrulich lupinix mbriza rdieter 15:16:35 .hello jbwillia 15:16:36 Southern_Gentlem: jbwillia 'Ben Williams' 15:16:49 info Southern_Gentlem present 15:16:54 #info Southern_Gentlem present 15:17:00 #topic agenda 15:17:08 ok, any agenda topics to discuss? 15:17:14 I have: "plasma-5.11 for f26" 15:17:18 wayland by default for F28 15:20:05 ok, let's go with that 15:20:09 #topic plasma-5.11 for f26 15:20:29 subject says it all, what say ye? Any comments, concerns, objections? 15:20:46 i'm +1 with that 15:21:09 +1 definitely, I don't have any problem with that 15:21:19 +1 15:21:34 historically, we've done at least one plasma-5.x+1 update per release, this would be f26's turn 15:22:37 One minor thing I may investigate is that dnf doesn't handle package renames 100% well yet, wrt plasma-workspace-drkonqi -> plasma-drkonqi rename 15:23:00 what is the exact issue here? 15:23:09 I *may* end up changing the name back to plasma-workspace-drkonqi 15:23:24 lupinix: for f26/f27, try doing: dnf install @kde-desktop 15:23:34 (assuming it's not already installed of course). 15:23:41 * lupinix will check later 15:24:10 It'll try to install plasma-workspace-drkonqi (and fail/complain about missing/broken deps), instead of using plasma-drkonqi (that Obsoletes/Provides it) 15:24:34 :( 15:24:53 I hit that with a box here that I initially installed workstation, then wanted to add kde-desktop 15:25:06 then changing the name back would be the best workaround, yes 15:25:10 * Southern_Gentlem testing 15:25:13 is this a known issue? Is it reported? 15:25:28 I haven't reported it yet, at least. I don't know about "known" status 15:25:47 I did support a fedora-comps pull request to modify f28 (and f27) though 15:25:53 s/support/submit/ 15:27:15 otherwise, sounds like we're agreed on moving forward with the f26 plan 15:27:39 rdieter, appears to be working under xfce so maybe you were hitting a bad mirror 15:27:40 #agreed plasma-5.11.x for f26 is a go 15:28:15 Southern_Gentlem: so installing kde-desktop group, ended up pulling in plasma-drkonqi ? 15:28:20 (on f27?) 15:28:33 (am i late?) 15:28:33 will check as soon as the install finishes 15:28:40 * rdieter double checks if comps pull request has been pulled in yet (I didn't think so) 15:28:49 pino|work: right on time :) 15:28:52 #info pino|work present 15:29:01 I'm around too 15:29:07 pino|work: currently discussing plan to bring plasma-5.11.x to f26 15:29:11 #info tosky present 15:30:02 pino|work, tosky: any related comments, concerns or objections? 15:30:19 options a) bring to f26 what's in f27+, and then update everything b) update f27+ first to 5.11.x, then update f26 15:30:22 ? 15:31:25 pino|work: F27 is already on 5.11.x 15:31:26 f27 already has it 15:31:45 yes, i know, i mean to whatever is the latest 5.11.x 15:31:56 pino|work: your plan a matches 15:32:24 already have plasma-5.11.5 f27 builds ready to submit to bodhi. 15:32:34 wanted ok to move forward with doing same for f26 15:33:03 (before you got here): historically, we've done at least one plasma-5.x+1 update per release, this would be f26's turn 15:33:13 f26 is currently at plasma-5.10.5 15:35:05 rdieter, anyway that will happen with the 4.14.12 kernel? (or whatever kernel that fixes the intel mess) 15:35:43 Southern_Gentlem: not sure what your question is, can you rephrase ? 15:37:00 (or clarify for me how that is relevant to us?) 15:37:32 rdieter, since i am the person the produce the respins i watch the kernels more than what the DE are doing so if your new 5.11.5 is pushed the sametime as the kernel i will make a new respin at that time 15:38:11 ah! thanks 15:38:54 moving on... 15:38:55 rpm -qa |grep |plasma |grep drkonqi returns plasma-workspace-drkonqi 15:39:06 Southern_Gentlem: on f27? 15:39:09 yep 15:39:31 I didn't think that was installable on top of plasma-5.11 15:39:35 xfce 27 doing dnf install @kde-desktop 15:39:58 Southern_Gentlem: thanks! 15:40:20 #topic wayland by default for F28 15:40:29 jgrulich: go ahead 15:40:59 yesterday I started a discussion on our mailing list about having wayland as default for F28 15:41:28 I think majority of people responded there agreed with this if there is a safe option go back to X11 15:42:09 I know it's an ambitious plan and there is lots of work ahead, but I believe that this step will help a lot to make wayland more usable 15:42:31 i'm planning to switch to wayland for qt apps under gnome, too - there will be even more testers from the other side, too 15:42:42 totally fine with that; what would it be the lastest time to revert the default, if things are not working as expected for 28? 15:42:48 in the least (I think), even if the default session is not changed, including wayland session support by default is a good baby step forwards 15:42:58 which is more or less the same as asking: how much time there is for testing? 15:43:23 I'm now running Plasma 5.11.90 (almost beta) with Qt 5.10 and it works, not perfectly and there are minor annoying issues, but I think all of them can be fixed until F28 gets released 15:43:30 * rdieter checks https://fedoraproject.org/wiki/Releases/28/Schedule 15:44:19 I'd say approximately f28 beta, 2018-03-06 is where we'll need to decide go/no-go on this feature 15:44:42 sorry, that's f28-beta freeze 15:44:47 Plasma 5.12 is going to be an LTS release, now even with LTS support for wayland 15:45:34 * lupinix uses 5.11 with wayland (on intel graphics) 15:45:45 lupinix: do you use multiple screens? 15:45:52 works fine so far (using for some weeks now) 15:45:57 no, on notebook 15:46:32 there are some issues with multiscreen, I believe that once those are properly identified then they can be fixed in time 15:47:09 and to have them properly identified we need people to test it → wayland by default 15:47:15 * rdieter looks forward to testing 15:47:47 jgrulich: curious, are you aware of any other distros condidering similar? 15:47:52 considering even 15:48:21 rdieter: nope, I just know that Neon uses it by default, but I think on the unstable version 15:48:50 ok. (was hoping for some help/collaboration) 15:48:53 * lupinix thinks we should enable as default, if things are really problematic we switch back to X for f28 final 15:49:21 i'd say we should be conservative and if there is a bunch of issues left before final, revert 15:49:23 and still have wayland preinstalled for people who want to test 15:49:35 that's why I asked when is the last day for a revert :) 15:49:37 I believe that upstream really wants wayland to move forward and they will happily help to identify and fix issues 15:49:44 sounds like we're in general agreement on moving forward with general plan 15:50:00 +1 from me, yes 15:50:03 and for sure provide the Plasma/Wayland session for session managers in any case, as rdieter proposed 15:50:14 jgrulich: once you have a change proposal in place, we can formally vote and send it on 15:50:43 rdieter: ok, I'll start working on it and we can vote on the next meeting 15:51:08 I'll prep another fedora-comps pull request to get -wayland included on spin 15:51:18 (which we'll want regardless) 15:51:29 yes 15:51:38 ideally, this change will be just changing the default in sddm 15:51:50 yes 15:51:55 rdieter: I'll also update Plasma to 5.12 as soon as possible so people get the most "stable" experience 15:52:11 beta should be released next week 15:52:54 I'll help coordinate plasma-5.12 copr availability too (as time allows) 15:53:22 for f27, though that won't be a 100% good test due to Qt 5.10 vs Qt 5.9, but it should be close 15:54:32 anything else on f28/wayland ? 15:55:13 another todo: create bugzilla tracking bug for plasma-wayland related issues 15:55:24 good idea 15:55:25 are there known issues with graphics drivers and wayland? nouveau for example? 15:55:48 I already have one in mind (pinentry fails when $DISPLAY not defined) 15:56:14 (but that's not specific to plasma, probably not worth considering a blocker either) 15:56:21 NVidia → https://blog.martin-graesslin.com/blog/2017/10/plasmawayland-and-nvidia-2017-edition/ 15:57:29 should not be a blocker as it is not shipped by us @default and X is always available as fallback 15:58:34 I think we should have checks in place like gdm does and don't allow wayland at all with problematic drivers 15:59:04 that would be nice, but no idea what that involves or how much work it would be 15:59:19 martinbriza[m]: volunteering for it? :) 15:59:57 I'll see what does it take for the next meeting 16:00:02 well, the logic can be stolen from GDM I guess, just needs to be put into a proper place in SDDM 16:01:01 Yes, I think so... I guess it's just a bunch of opengl calls 16:02:39 ok, so let's move on 16:03:01 I'll write a proposal and we will vote on it next meeting 16:03:25 #topic open discussion 16:03:28 ok, anything else for today? 16:04:45 let's go with no, we're kinda over our hour anyway. Thanks all! 16:04:50 #endmeeting