13:58:53 #startmeeting kde-sig -- https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2010-03-16 13:58:54 Meeting started Tue Mar 16 13:58:53 2010 UTC. The chair is rdieter. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:58:55 Useful Commands: #action #agreed #halp #info #idea #link #topic. 13:58:57 #meetingname kde-sig 13:58:58 The meeting name has been set to 'kde-sig' 13:59:20 #topic roll call 13:59:26 who's around today ? 13:59:39 * jreznik is here 14:00:23 I'm here 14:00:51 Present. 14:01:04 moin 14:01:29 Note: I'll be in and out for the meeting, as will mchua (hi!) beause we are filming b-roll for the FAD 14:01:38 hot 14:01:58 b-roll? 14:02:00 * Sho_ is lurking 14:02:08 jreznik: b-roll = optional footage 14:02:28 alright, let's roll 14:02:43 #topic agenda 14:02:50 any topics to add to https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2010-03-16#Agenda ? 14:02:57 yeah 14:03:34 Which ones? :-) 14:03:37 I talked to spevack and stickster_afk about a few things yesterday that we can do to get better kde integration in the dvd and liveimage and first boot 14:03:46 just want to blah about it :) 14:03:57 with a "do you want to help 14:04:02 " at the end. 14:05:02 #topic rrix blah'ing about dvd/liveimage kde-integration 14:05:08 rrix: ok, blah away. 14:05:10 oh the noes 14:05:29 * than is present 14:05:37 okay, so yesterday at lunch, the topic came up on how we can improve Fedora KDE's stance among developers and users 14:06:07 the consensus was "hey Fedora isn't really seen enough as a KDE distro as opposed to kubuntu or oopensuse, even though we arguably have a better distro" 14:06:44 so, basically, it was posited that we rely too much on the live image, and don't use the DVD. Which, i explained was for known and much complained about reasons 14:07:09 But Kubuntu also relies on the live image. 14:07:10 well, stickster_afk told me that colin walters was looking at fixing that, and looking for help. 14:07:21 i think that's a fishy argument because in my experience most people go only for the live images today, anyway 14:07:30 In fact, Kubuntu basically IS the live image, there's an alternate (also KDE-only) CD, but everyone uses the live image. 14:07:32 kubuntu basically only has a live image, and opensuse's has proven popular, too 14:07:52 Sho_ Kevin_Kofler: it wasn't really a point we made saying "this is why they're better" only a way that we could improve upon ours. 14:08:10 But yeah, the DVD needs fixing. 14:08:10 well sure, as long as the DVD doesn't go away, improving it is a good idea 14:08:21 we want people to have a good time no matter which medium they pick, of course 14:08:24 Kevin_Kofler: well, anyone wanting to help should mail walters@redhat.com :) 14:08:41 rrix: any details on it? or I guess we can just fine out... 14:08:43 what we need to improve DVD? comps? 14:08:48 comps, yeah 14:09:11 comps and splitting a few of the default packages so they don't pull in gnome-world 14:09:16 Re firstboot, the main issue there is that it requires metacity. 14:09:24 Kevin_Kofler: and I was coming to that now :) 14:09:25 This also bloats our live images with unwanted dependencies. 14:10:16 Was any solution suggested? 14:10:34 Kevin_Kofler: notting brought up a few days ago (says stickster_afk, i haven't looked) that they were looking at creating a way for firstboot (and eventually anaconda if necessary) to use another window manager besides metacity. it shouldn't be /too/ hard to do i'd imagine 14:10:45 lenmme grep the list 14:10:46 hh, we should split packages to not pollute gnome word - we can do it... ask gnome folks for same :( 14:10:56 I'd suggest making it support kwin as well and having it require a virtual Provides provided by metacity and kdebase-workspace instead of metacity directly. 14:11:14 (gotta step out for a moment, ping me when the qt 4.7 topic comes up though, got my 2ct on that) 14:11:25 anaconda team is already working on metacity replacement 14:11:34 But maybe doing away with requiring desktop WMs entirely and make it use some small minimalist WM is the best solution. 14:12:25 eh, I can't find it 14:12:33 Both Metacity and KWin can be used standalone, but aren't really designed for that use. Plus, Metacity is deprecated in GNOME now that gnome-shell is coming. 14:12:38 ok, sounds good. Can we have someone from our team willing to be comps-kde-czar and lead these coordination efforts? 14:13:12 what's actually our target for comps? 14:13:21 I can get with walters, after we discuss what we want 14:13:43 does it mean different comps? some selection in Anaconda to choose DE? 14:13:48 * rdieter dubs rrix, sir comps-kde-czar , be brave and strong 14:13:50 I don't know what walters imagines. My idea is to have a desktop selection screen like in OpenSUSE prepended to the package selection and to have conditionals based on DE in comps. 14:14:07 * mchua here, lurking 14:14:13 But walters may be thinking of something different. 14:14:17 we'll have to see how thigns develop as the details roll in 14:14:31 mchua: hi! 14:14:32 I'll get with him in the next day hopefully 14:14:41 will be busy Doing Awesome Things today, but 14:14:46 eh, brb, sec 14:14:58 Sho_: ping, if you're around, we can do qt-4.7.0 next 14:15:14 ok 14:15:24 #topic qt-4.7.0-tp ready for devel/ import 14:15:37 so I've been running Qt 4.7tp1 on F12 for a few days now, via rex' packages 14:15:45 I've got a first crack at packaging qt-4.7.0-tp done. 14:16:05 IMHO we should get this into Rawhide ASAP. 14:16:07 awesome! 14:16:15 the only problem I've had is a build problem with Amarok git master, which was their fault since they were doing something stupid, namely initializing QStrings with ints (i.e. QString foo(0) basically) which 4.7 no longer likes due to changed constructors 14:16:16 If it's not horribly broken ship it to rawhide! 14:16:18 The qt-assistant-adp compat package is now imported (but not built yet). 14:16:51 I fixed that in their git, but it was probably after the 2.3.0 tag, so if they do 2.3.1 from a branch rather than master and don't backport it, amarok 2.3.1 might have trouble building with 4.7 or so 14:16:51 my only/primary concern is that this will inevitably leave PyQt4/PyKDE4 broken 14:16:58 other than that, the experience has been regression-free for me 14:17:10 Re qt-assistant-adp, should we keep the < versioned Conflicts (which are more correct) or use >= versioned Requires instead? 14:17:10 i don't see any problem to ship 4.7 in rawhide 14:17:13 it's also been fix-free (e.g. the disappearing mouse cursor bug that entered the scene in 4.6.0 is still there) 14:17:50 Kevin_Kofler: I'd prefer the versioned requires myself 14:17:56 The Conflicts are really correct because it really does conflict with the old versions of Qt. 14:18:03 But Requires will have almost the same effect. 14:18:44 ok, looks like we have no dissenters, let's do it then 14:19:03 I noticed qt4-x11 is not provided with ISA, let's please fix that for 4.7 so I can Require qt4-x11%{?_isa} >= 4.7. 14:19:07 I'll commit and get builds going after the meeting. 14:19:24 doing just qt4 dep should be sufficient 14:19:49 -x11 will get pulled in implicitly, and it has versioned requires to make sure things are sane 14:20:18 Yeah, OK. 14:20:56 #action rdieter to commit/build qt-4.7.0-tp1 in devel/ 14:21:06 #topic kde-4.4.1 status 14:21:29 well, our week passed, and -testing has been good I think. 14:21:40 Yeah, let's please push this to stable now. 14:21:47 We've been sitting on it for way too long. 14:21:59 And future bugfix releases should be pushed out much faster. 14:22:07 No issues here so I'm good for a push to stable 14:22:14 patience. 14:22:29 but here, indeed, +1 stable 14:22:32 4.4.1 fixes some regressions in 4.4.0. Hopefully that SIGBUS we're getting abrt-spammed with should be fixed too in it. 14:22:46 * rrix nods 14:22:47 than? 14:22:54 +1 to stable 14:22:59 +1 from me 14:23:14 +1 too 14:23:16 I'm using 4.4.1 on all of my systems, looks ok, much more stable than 4.4.0 (from abrt perspective) 14:23:18 +1 14:23:25 #agreed kde-4.4.1 is ready for stable updates 14:23:48 #topic triaging trips/tricks for commonly reported abrt bugs? 14:23:50 I'm pushing it to stable right now. 14:24:12 speaking of abrt bugs, we're gettings lots of dups, Kevin_Kofler's been very good at catching/triaging those 14:24:31 rdieter: abrt people are working on dups checker but it's not easy to check kde bt for duplicates 14:24:43 but I'm constantly spamming abrt people with our duplicated 14:24:44 2 things come to mind for me: 1. how can we do a better job of triaging these 14:24:44 Well, basically I've caught all the dupes of that KPixmapCache bug because they were very easy to recognize. 14:24:47 duplicates 14:24:54 SIGBUS is something which never happens otherwise. 14:25:01 Personally I hate ABRT bugs. I've been requesting they all go upstream 14:25:04 (I did check the backtraces, it was always KPixmapCache.) 14:25:12 2. for items with lots of reports/dupes, we should make some official efforts to upstream those 14:25:28 SMParrish: That's a big issue indeed, ABRT reporting those bugs to us downstream packagers really sucks. 14:25:36 Kevin_Kofler: do we have some info from upstream? is it fixed? 14:25:46 Also most reporters never respond so almost 100% are ending up getting closed 14:25:54 jreznik: I don't know. 14:26:03 Our request to foward the bug upstream got ignored. 14:26:08 Kevin_Kofler: so rdieter is right - we should report it upstream 14:26:09 None of the reporters bothered. 14:26:16 It may be already fixed in 4.4.1. 14:26:22 We should only report it if it's still there. 14:26:31 Otherwise we should just close it. 14:26:36 SMParrish: it happens yeah. that's part of the virtue of having a *little* barrier of bug reporting. If it's too easy, no feedback. 14:27:04 ok, if we see further dups of these common issues with 4.4.1, then we'll upstream it. 14:27:05 rdieter: upstream is solving this barrier thing too 14:27:16 rdieter: good plan 14:27:36 anything else here? (else I'll move on) 14:27:41 and I try to abuse abrt people with more our dups to maka it working for us ;-) 14:28:01 I'm still for removing ABRT from the KDE spin and for autoclosing all ABRT-filed bugs with some form letter (like CLOSED UPSTREAM – please report this issue to the upstream developers, we will not evaluate autogenerated bug reports). 14:29:21 I suppose in a perfect world, ABRT clientside would be accompanied by additional ABRT server/bugzilla side tools to help deal with them all. 14:29:42 ABRT needs some work but it is here and we have to deal with it 14:29:53 #topic https://fedoraproject.org/wiki/SIGs/KDE/KDE_Target_Audience 14:30:00 ok, now for the fun stuff. 14:30:02 I get tons of ABRT reports for Gnash, almost nobody bothers forwarding them upstream when asked. 14:30:15 Some won't even retest after a Gnash update. 14:30:23 Kevin_Kofler: if duplicates - send me #bz and I'll forward it to kklic 14:30:33 And of course many crashes are not fixed because they aren't getting reported upstream. 14:30:41 jreznik: I have no idea whether they are duplicates. 14:30:48 I can't compare all the backtraces, they are too many. 14:30:56 over the past few days, I brainstormed a bit what I consider to be our target audience 14:31:01 (And some are entirely useless.) 14:31:08 and took input from a few others in #fedora-kde who happened to be around. 14:31:45 Is this close to what other's current perceptions are when it comes to the target audience we're aiming for? 14:31:57 rdieter: What is "this"? 14:32:03 https://fedoraproject.org/wiki/SIGs/KDE/KDE_Target_Audience 14:32:17 oops, sorry, reading comprehension FAIL 14:32:41 rdieter: I don't like it... looking for words how to describe my feelings ;-) 14:32:52 "more control of their desktop" -- more than what? 14:32:56 once we're close to agreeable, then I'd like to RFC from our kde list. 14:33:21 I like the Update SOP, but not sure about the target audience part 14:33:23 I'm not sure our users should be somehow special to fedora users 14:33:29 I didn't particularly like that line as worded either, but I pasted it verbatum from suggestions 14:33:37 looks like a good start to me 14:34:01 the update SOP looks very good 14:34:07 jreznik: I think it's important to differentiate from other Target audiences being bandied about, because it''s clear (to me) that we don't match 14:34:42 rdieter: now the question is - are we going to have one broad fedora audience or custom spin audience? 14:34:45 I want it clear that our target audiences *do* include developers, contributors 14:34:51 board should decide first... 14:35:04 mefoster: More than in other offerings, like proprietary OSes, GNOME etc. 14:35:08 jreznik: the board asked what our opinions are wrt to target audience 14:35:10 this is it 14:35:26 they asks all spin owners, as a matter of fact. 14:35:57 So as-is, that section is probably worded too strongly. 14:36:02 it does not include "joe users", I'm not sure we don't want them 14:36:20 target audience isn't about excluding anyone 14:36:40 it's all about what we are focussing on. 14:36:40 They're welcome to use our stuff if they like current flexible software. 14:36:58 But we aren't going to dumb down our software or let it rot unupdated just to make them happy. 14:37:11 If that's what they want, they should find another solution. 14:37:25 well, I'd phrase in a positive way instead. 14:37:35 rdieter: positivity is good :) 14:37:37 which is what the current document is trying to do 14:37:53 So KDE spin is particularly targeted to people who want cutting-edge updates 14:37:54 I'll take out the line about "more control", its unclear and vague 14:38:00 Being current and flexible doesn't necessarily exclude anyone, but sure, if you don't want your menu item to ever move by 1 pixel, then you probably want some other distro. ;-) 14:38:02 I think jreznik's point is that we don't need to make an effort to differenciate our target audience from the audience of the default spin at all, we can define our target audience completely independently of theirs. If there's overlap, that doesn't mean we can't still be doing a better job. 14:38:13 and is of course open to be used by other people but they need to understand what they're getting 14:38:19 If you define yourself in contrast to something else, you automatically create a dependency on that something else 14:38:19 Sho_: agreed 14:38:23 and we don't need that dependency 14:38:28 rdieter: what about more configurability instead? 14:38:36 I don't like "more" in general 14:38:45 always invites comparisons 14:39:00 Sho_: exactly! 14:39:01 users who value configurability (or is that still too much the same thing?) 14:39:01 We don't really need to justify ourselves … it doesn't need to be part of that document's goals to contrast with anything else 14:39:05 What about "full control of their desktop"? 14:39:20 That's not a comparison and it gets that point across. 14:39:25 it shouldn't mention desktop at all, imo 14:39:31 it should describe the user 14:39:47 "full control of their machine"? ;-) 14:40:18 So, anyone have concrete suggestions on improving the message? 14:40:22 I like you can set KDE to be more "joe user" than gnome could ever be but still let you flexibility to do what you want... this is most exciting! 14:40:32 Or to describe the user him/herself: "users who like to be empowered, as opposed to having their computer decide for them". 14:40:48 Kevin_Kofler: but you can do this with KDE too 14:40:51 too negative still for my taste, and drawing comparisions 14:41:06 KDE can be configured both ways 14:41:34 we just need to find some median for our default config 14:41:42 "Users who want their desktops to reflect their preferences to the utmost degree." or something like that 14:41:52 * rdieter likes that 14:42:34 It gets the point across that KDE is highly configurable to a user's personal content, and doesn't make a comparison 14:42:43 Sho_: I like the first part 14:42:43 * rdieter added that one 14:43:04 Shouldn't "nitch" be "niche" instead? 14:43:13 probably 14:43:15 rdieter: I agree with jreznik that the "to the utmost degree" is a bit bulky 14:43:19 so "Users who want their desktops to reflect their preferences. PERIOD" 14:43:20 ok 14:43:22 'tweakers paradise' 14:43:26 yes, "niche" is right 14:43:32 rdieter: "Users who want their desktop to reflext their personal preferences." is probably enough already 14:43:35 *reflect 14:43:38 updated 14:44:05 mclasen: :) 14:44:21 * mclasen is in the wrong meeting... 14:44:23 Yeah, we shouldn't exaggerate, we aren't that infamous Beryl-config either. 14:44:24 for example I have really simplistic KDE - we can take a look how to tweak it to our users needs 14:44:42 not sure I like "users who appreciate keeping up with the latest versions ..." either 14:45:01 (Newer compizconfig tried to make a bit less of a messy effect, kinda what KDE 4 did to some unwieldy KDE 3 preferences dialogs.) 14:45:21 some preferences in KDE are still big mess... 14:45:28 rdieter: "Users who want access to the latest and greatest KDE has to offer"? 14:45:53 Sho_: that sounds good 14:46:16 mclasen: feel free to join our paradise ;-) beer for free ;-) 14:46:18 the reason I don't like that is because our target audience is defining our policies then (or at least too explicitly for my own taste) 14:46:34 rdieter: that's why I added "and greatest", actually 14:46:49 rdieter: if the latest turns out to suck because we (upstream) went crazy, you can still ship "the greatest" ;) 14:47:00 meh, that's better than what it says now, so... 14:47:31 added, removed commentary on rawhide too 14:47:37 Do we specifically want to mention users who are prepared to be active in testing/filing bugs ...? 14:47:58 But not only -- its more that they want to keep up with upstream developments 14:48:00 maybe, though I had that in mind by targeting contributors/developers already 14:48:05 and we hope upstream doesn't go crazy :) 14:48:37 I would not mention looking for testers and bug filers. Might give them the impression we are buggy 14:48:56 SMParrish: good point :) 14:48:58 we are cooking the dog's & cat's cake (one czech tale) 14:49:12 Just stick to "eager to keep up with latest and greatest" or whatever 14:50:06 alright, can everyone refresh the page... so are we close to something we can agree on now? 14:50:07 all buzz now is around "stable" - whatever it means 14:50:35 If so, I'll take it to the list for further comment 14:50:38 Possible can of worms ... do we want to add anything to Update SOP about Fn and Fn-1 and so on 14:50:49 mefoster: not yet, imo 14:50:52 s/kde contributor/KDE contributors/ 14:51:03 rdieter: fair enough, that can come later on 14:51:03 rdieter: "Users wishing to contribute to KDE development" @ nicer version of "kde contributor" 14:51:16 and I still miss - users who just want workstation to do their jobs 14:51:23 and generally s/kde/KDE/g and s/fedora/Fedora/g :-) 14:51:58 ok 14:52:19 maybe should've used gobby/kobby. :) 14:52:22 rdieter: "Users looking for a modern and robust desktop to get their work done" @ jreznik's wish 14:52:38 s/a/the most/ ? ^^ 14:52:57 Sho_: you are great marketing guy!!! 14:53:02 Kevin_Kofler: we discussed that above ;) 14:53:11 likie, added 14:53:35 we can probably bikeshed about ordering these later too. 14:53:54 put the ones we value most toward the top 14:53:55 yeah, I'd put that last one up top probably 14:54:02 Sho_: +1 14:54:06 me too (that's why I thought of that) 14:54:31 moved to top 14:54:34 robust and modern, personal prefs, latest and greatest would be my top 3 in that order 14:54:57 ok 14:54:57 of course it's not nice to work on something you don't like so contributors should be on top too 14:55:04 Sho_: +1 14:55:24 Put the advantages for everyone first, then all the items about contributors. 14:55:47 putting Fedora contributors on top of the software devs and KDE contributors probably makes sense in the context of Fedora's meta-goal to empower users to become Fedora contributors 14:56:05 yes 14:56:12 "Software developers" might be redundant with "get their work done" 14:56:28 past the top 3, it's pretty even to me. 14:56:30 contributors should have chance to influence it 14:56:43 it's open source world... 14:58:05 good work everybody, I really like the message now. 14:58:21 so I like "Fedora contributors who want to influence future of Fedora" 14:58:22 looks like we're about out of time, final thoughts? 14:58:44 my one insecurity at this point is that the "get their work done" skews things away from private/personal desktops somehow, not necessarily a bad thing, but KDE is not just a business desktop of course 14:58:52 to say the message - we want people with vision 14:59:09 jreznik: nice, I like empowerment 14:59:36 and we want people to influence what they get 14:59:39 Sho_: you are right 14:59:44 time's up 14:59:47 #endmeeting