15:01:15 #startmeeting kde-sig -- https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2012-06-19 15:01:15 Meeting started Tue Jun 19 15:01:15 2012 UTC. The chair is jreznik. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:15 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:01:23 #meetingname kde-sig 15:01:23 The meeting name has been set to 'kde-sig' 15:01:31 #meetingname kde-sig 15:01:31 The meeting name has been set to 'kde-sig' 15:01:38 #topic roll call 15:02:37 * rnovacek is here 15:03:00 * ltinkl is here 15:04:27 Present. 15:04:33 * than is here 15:04:54 oh, hi 15:06:32 #chair ltinkl rnovacek than Kevin_Kofler rdieter 15:06:32 Current chairs: Kevin_Kofler jreznik ltinkl rdieter rnovacek than 15:06:47 #info Kevin_Kofler jreznik ltinkl rdieter rnovacek than present 15:06:55 #topic agenda 15:07:08 Live image bloat 15:07:19 (MiniDebuginfo and other creeping biggerism.) 15:07:27 InitialExperience feature? 15:07:43 https://fedoraproject.org/wiki/Features/InitialExperience 15:07:53 ye :/ 15:08:15 and we should discuss our opinion on https://fedoraproject.org/wiki/Features/OfflineSystemUpdates too 15:09:08 So it's Feature FiaSCo day today? 15:09:10 ;-) 15:09:14 yep, these two features 15:09:18 3 15:09:19 also MiniDebugInfos... 15:09:40 Is FESCo out to sabotage KDE? ;-( 15:09:43 do we have any status updates? 15:11:28 #topic status updates 15:12:04 any status update? kdemm split etc. 15:12:22 only kscd, ffmpegthumbs (rpmfusino) left 15:12:59 just wanted to inform we (dantti, rhughes, me) some of the battery applet/upower bugs yesterday 15:13:20 we ... some ? ... = fixed? :) 15:13:33 created? :) 15:13:43 decapitated 15:13:44 LOL rnovacek. :-) 15:13:44 * jreznik is finishing kscd review 15:13:54 fixed :) 15:14:02 dantti created it tbh xD 15:14:14 another f word is on my mind :) 15:14:50 The one Linus said to NVidia? ;-) 15:15:10 #info kdemm split mostly done, missing kscd and ffmpegthumbs (rpmfusino) 15:15:28 oh, and a couple of new ktp-0.4 related pkgs, call-ui logger-qt 15:15:44 #action jreznik to finish kscd review 15:15:56 call-ui is done 15:16:03 #info dantti, rhughes and ltinkl fixed some of the battery applet/upower bugs yesterday 15:16:09 (review I mean) 15:16:12 rnovacek: wow, fast. thanks 15:16:26 #info a couple of new ktp packages to be reviewed 15:16:28 I'm on logger-qt too 15:17:24 Great. 15:17:46 nice to see ktp chat content not get lost anymore. :) 15:17:50 rdieter: kscd-4.8.90-2.fc17.src.rpm 404 - mind upload it to finish review? :) 15:17:56 hrm, oops 15:17:58 ok 15:18:22 is that all for updates? 15:18:26 jreznik: uploaded 15:18:35 rdieter: thx 15:18:39 rsync'd actually, result should be the same. :) 15:22:14 can't think of anything else, moveon++ 15:22:26 #topic a new features approved by FESCo that hits KDE spin 15:22:51 #info there are several new features approved by FESCo that could affect KDE spin 15:23:10 MiniDebugInfo is easily the worst. 15:23:20 Extra bloat which WILL make us fail the size target. 15:23:25 imo, easfix for minidebuginfo is to (finally) make a bigger spin 15:23:33 People keep saying we should target a larger size, but WHAT size? 15:23:52 The next larger size in the QA procedures is DVD size, which is going to be too much for the Multi DVD. 15:24:04 I propose: kde spin officially target ~1gb size (or bigger, don't care about the details right now). and unofficially maintain a cd-sized one in spin-kickstarts 15:24:09 How do we pick a size guaranteed to fit on the Multi DVD? 15:24:11 rdieter: the thing is - bigger image is excuse but actually nobody acks tham before approving this change... 15:24:11 Who enforces this? 15:24:17 In particular, which spin gets to pick what size? 15:24:25 CD size was a really nice standard target. 15:24:36 rdieter: we can't be too big - due multi desktop DVD we want to be on! 15:24:40 Having to drop it because of useless bloat just sucks. 15:24:50 DVD multidesktop folks didn't mind, when I mentioned it 15:24:55 +1 for useless bloat... 15:24:57 ie, they'll likely kick on DE out 15:25:14 rdieter: sugar will be killed probably 15:25:16 s/on/one/ but that's not specifically our problem 15:25:16 rdieter: Well, they need to define WHAT size we need to target so that we are not the spin they'll kick out. 15:25:30 Kevin_Kofler: yep 15:25:40 FWIW, I wonder if we could get away with just rm-ing the MiniDebugInfo in the spin kickstart's %post. :-/ 15:25:45 1gb was the target size when I mentioned it 15:25:45 cwickert: are you around as multi desktop dvd guy? 15:25:55 Though deleting files owned by RPMs strikes me as a nasty hack. 15:26:25 seems there's at least some support in the gnome/desktop camp to do the same (increase size) 15:26:52 Last time they wanted to do it, they tried it in Rawhide and then reverted it at the last moment. :-/ 15:26:56 And it hasn't been mentioned ever since. 15:26:59 Kevin_Kofler: imo, deleting files is probably not practical 15:27:05 ok, let's try to kick off the size discussion once this feature was approved... 15:27:11 There were a lot of complaints about the spin not fitting on a CD! 15:27:25 jreznik: It was approved yesterday. :-( 15:27:37 Kevin_Kofler: we'll be different, we will still provide a cd-sized image 15:27:55 Kevin_Kofler: it was approved but nobody approved a new image sizes 15:27:56 rdieter: With the bloat coming in, I wonder if it'll even be possible. 15:28:00 though it'll likely be relegated to alt content or torrent-only or something 15:28:10 jreznik: huh? 15:28:13 jreznik: That's exactly the problem. 15:28:35 cwickert: how much space do we have on multi desktop dvd available now? 15:28:36 We don't even have a size target such as "1 GiB" to pick from, QA only lists CD size and DVD size as possible size targets. 15:29:12 Kevin_Kofler: yep, that's the problem right now... we probably can't force FESCo to -1 this :( with retrace server it's nonsense but... 15:29:17 jreznik: http://serverbeach1.fedoraproject.org/pub/alt/releases/17/Multi/ 15:29:52 cwickert: we're considering increasing spin size (partly in response to the minidebuginfo feature). 15:30:02 and it's impact on the multidesktop DVD 15:30:19 as CD option is out... 15:30:21 jreznik: desktop is at 6.1 and we have 8.5 on a DL 15:30:58 cwickert: ok, so there's still some space available 15:31:32 Only 1.2 GiB per arch for all spins compounded. 15:31:36 That's not much. 15:31:42 and probably xfc, lxde, sugar will never use CD+ size 15:31:52 (my guess) 15:32:13 But they'll come closer to the 700 MiB, so they'll also contribute to endangering the Multi Desktop Live's size. 15:32:18 jreznik: xfce already is at 700, but we are trying to get smaller by getting rid if gnome stuff 15:32:35 how about you target 800? 15:32:38 lightdm should help 15:32:46 nirik: exactly 15:33:10 800 MB can still be burned on a CD, at least there are 800 MB media available 15:33:12 * jreznik hopes for soon lightdm for kde spin too 15:33:39 oh? moving from kdm? 15:33:39 cwickert: The problem is, last I checked, QA and rel-eng didn't want to put the "CD" label on anything exceeding 700 MiB. 15:33:43 800 MB means minidebug infos + some more packages (but few) 15:34:59 Kevin_Kofler: we do not want to call it CD anymore (just for people who wants it on CD, it can help) 15:36:02 well, let's (re)start serious discussion on -devel - there's now argument why to do it... 15:36:57 * rdieter would be happy to do baby steps and target 800mb 15:37:05 rdieter: +1 15:37:17 jreznik: +1, this feature shouldn't be approved without this consecutive decisions 15:37:28 as we already have found that 1 GB is nonsense - with 1.5 one 15:37:40 rnovacek: +1 for your comment... 15:37:43 rdieter, why not target 1gb and be done dvds are the same prices as cds and most people use usbkeys anyways 15:37:58 rnovacek: They already did approve it. :-( 15:38:30 Southern_Gentlem: we have to make sure to fit everything on multi desktop DVD 15:38:34 They ignored our size concerns entirely, saying they're not relevant and implying between the lines that we're dinosaurs for wanting to target CDs. :-/ 15:38:58 Kevin_Kofler: I know, but I think they shouldn't have (not sure about correct tense :) ) 15:39:03 Southern_Gentlem: 1gb was my initial proposal. :-/ 15:39:04 Kevin_Kofler: so let's be raptors and make 800 MiB possible! 15:39:27 or even 1 GiB but carefully 15:39:31 jreznik, then why is a design spin on the multi live in the first place 15:39:51 it should be all the DE 15:39:58 800MiB doesn't make sense to me, to big for fitting to regular CD, to small to contain cool stuff 15:40:13 s/to big/too big/ 15:40:29 lol, and s/to small/too small/ 15:40:32 rnovacek: but same was for 1 GB - too small for real cool stuff 15:40:47 Even 1 GiB is too smal for even just fitting kde-l10n-*. 15:40:51 *small 15:40:54 Argh! 15:41:04 jreznik: but it has at least some target media - 1GiB flash drives 15:41:21 Kevin_Kofler: indeed 15:41:22 rnovacek: 800 MiB does too, there are 800 MiB CDs. 15:41:38 And also 1 GiB USB - some space for a persistent overlay. 15:41:53 ok, let's try to ask people around what they think... we have now minidebuginfos argument to make it possible 15:42:04 Kevin_Kofler: yeah, but I personally haven't seen any 800MiB CD 15:42:11 I think 800 MB is a good compromise 15:42:13 rnovacek: They sell them in Vienna. 15:42:36 I have 3 left from my pack of 10. 15:42:49 800-850 is a reasonable initial target 15:43:01 Kevin_Kofler: the persistant overlay is a good point 15:43:15 I suppose 800, to keep open the possibility of those 800 cd's 15:43:26 who agrees? 15:43:29 * jreznik is +1 15:43:31 +1 15:43:33 +1 15:43:53 * rdieter is agreeable to any reasonable increase in spin size 15:44:16 +1 15:44:18 +1 to 800, discuss going further up if the bloat makes us exceed 800 MiB too. 15:44:45 btw: do we have precise data how much bigger are the images with the MiniDebugInfo? 15:45:17 no 15:45:20 We'll probably need to get QA on the board for any target size between CD and DVD though. Right now they only check either or. 15:45:30 #agreed to try to make 800 MiB goal possible 15:45:45 jwb: That should have been one of the requirements before approving the feature! 15:45:53 our general f18/kde feature page should mention this explicitly, so there's no surprises 15:47:40 we need a feature page asap, I'll take a care 15:47:44 One big issue with the new target is that people are going to be much less willing to fix or revert their bloat to make us fit 800 MiB than for 700 MiB, because 700 MiB is a common hardware limit, 800 MiB not so much (only CD-R90). 15:48:05 But it looks like they stopped caring even about standard CD-Rs, so… :-( 15:48:09 jreznik: gold star and cookie for jreznik 15:48:42 Kevin_Kofler: I am prepared to maintain a 700mb spin too, mind you 15:48:55 for 1 release at least 15:49:29 Coming soon: The Fedora KDE Live BluRay. Additional 100 exabytes of software are available from the repository. 15:49:42 #action jreznik to create feature page asap and mention our size intentions there 15:49:50 ok, let's move on 15:50:17 #action rdieter to help maintain cd-sized (700mb) kde spin 15:50:51 rdieter: +1 15:51:22 * Kevin_Kofler remembers the times when 1.44 MiB were considered a lot of memory… 15:51:39 (or were those decimal MB? I don't remember.) 15:52:18 :) 15:52:39 * Kevin_Kofler still has a RHL 5.2 boot floppy on his desk, it recently resurfaced from who knows where. ^^ 15:53:02 we're running out of time - offline updates are really out of scope now and we can opt-out (and dantti is going to work on support, so we can get into - not sure) 15:53:21 but initial experience and unclear future of firstboot could hit us 15:53:26 Yes, no offline updates until Apper supports them, of course. 15:53:36 (I think that's a given.) 15:53:50 again - the feature was approved without consensus :( with excuse it will not hit 100% :((( 15:53:55 Kevin_Kofler: yep 15:54:19 my impression was that initial experience was gnome-only, and that firstboot would continue to exist and be used as before 15:54:28 #info defer offline updates until Apper supports them (dantti will implement it) 15:54:35 Yes, it's even called gnome-initial-experience by upstream. 15:54:52 It's just the usual Fedora craze of selling GNOME features as Fedora features with no mention of what they really are. 15:55:06 jreznik: +1 15:55:10 rdieter: but it's chicken egg - if gnome would not use firstboot, firstboot developer is not willing to maintain it anymore (my impression after talking to grci and it makes sense) 15:55:10 This really sucks. I think it's also not in the interest of upstream GNOME. 15:55:31 jreznik: that's not the story told to fesco. :( 15:55:32 Kevin_Kofler: it's in the interested of gnome os 15:55:48 rdieter: I know, I'm in contact with all people right now... 15:56:12 trying to get more info on anaconda/firstboot rewrite 15:56:23 Another big WTF is that firstboot wants to stick with GTK+ 2 forever to support RHEL customers' custom firstboot modules forever. :-/ 15:56:29 the future of firstboot is now really unclear 15:57:29 * jreznik is not happy that fesco +1 it before actual discussion... yes, it's only for gnome spins, it's ok but other spins needs a clear firstboot situation 15:58:11 btw. I've already contacted Plasma upstream - they want similar feature too, I promised to write an email to plasma-devel 15:58:20 Yeah, FESCo has turned to just rubberstamping everything thrown at it. :-( 15:58:21 #action jreznik to contact KDE upstream about IE 15:58:55 And InitialExperience is just yet another example of how GNOME is more and more all-encompassing of systemwide stuff. So we really need KDE replacements for everything ASAP. 15:59:17 (Anyone also noticed that the acronyms is the same as for Internet Exploder? ;-) ) 15:59:19 i don't believe it was rubber stamped. it was defered twice 15:59:22 *acronym 16:00:14 jwb: But the issues raised weren't addressed in any way. 16:00:38 they were. we were told firstboot would still be a viable option for F18 for non-Gnome spins. the feature is for f18. 16:00:42 jwb: still before actual discussion... 16:01:02 jwb: That's very short-term thinking. 16:01:10 jwb: I talked to firstboot maintainer personally and it's not 100% after listening to him 16:01:36 then perhaps instead of everyone talking personally with people, it should be done on an open list 16:01:42 less wires get crossed that way 16:01:49 the only thing we need is create user... it can be done in anaconda with nonvisual firstboot and then - do anything you want 16:02:17 jwb: yep, mitr send a message to public lists after the talk... but it was approved before reply :) 16:02:24 I also remember the PolicyKit 1 story, where we and FESCo were told all the time that there'd be a compatibility PolicyKit 0.9 for at least one release, and then suddenly they changed their mind and we had to rush out a port of the client side and ship the GNOME version of the auth agent for a while. :-/ 16:03:01 Though at least that transition was discussed and brought to FESCo in the first place, unlike NM 0.9. :-/ 16:03:14 Kevin_Kofler: at least - if firstboot dies, other desktop should take care about the maintainence... 16:03:44 jreznik: what happens when nobody is willing/can't maintain it? 16:03:55 ltinkl: no spins? :) 16:04:03 jreznik: According to the discussion on the devel list, the Anaconda portions will not run in liveinst, the live images will still need firstboot! 16:04:18 ok, we are over time... let's closely watch it and comment 16:04:28 Kevin_Kofler: that's my understanding... :( 16:04:51 jreznik: I wasn't happy about how mitr handled all those features. 16:04:55 we will end with several implementation of the same problem for several spins :( 16:05:07 Also for MiniDebugInfo, he said ABRT will have no use for it, yet voted for it! 16:05:26 Kevin_Kofler: at least, he's the one who really takes at least some care about features approved... 16:05:43 it's the max we can get from fesco... 16:05:43 jreznik: But he shouldn't have voted for features he doesn't think are useful. :-/ 16:06:35 Kevin_Kofler: maybe the only way is to try to implement some feature that would break gnome spin and someone would notice it :))) 16:06:52 ok, we are over... 16:06:54 thanks guys 16:06:54 FWIW, for offline updates, I'd argue for disabling them by default once Apper supports them. 16:07:07 #endmeeting