15:05:14 <rdieter> #startmeeting kde-sig
15:05:14 <zodbot> Meeting started Tue Feb  4 15:05:14 2014 UTC.  The chair is rdieter. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:05:14 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:05:17 <rdieter> #meetingname kde-sig
15:05:17 <zodbot> The meeting name has been set to 'kde-sig'
15:05:21 <rdieter> #topic roll call
15:05:39 <rdieter> hi all, who's present for a friendly kde-sig meeting today?
15:05:46 <than> present
15:05:52 * jgrulich is present
15:05:54 <Kevin_Kofler> Present.
15:06:41 <rdieter> #info rdieter than jgrulich Kevin_Kofler present
15:06:47 <rdieter> #chair than  jgrulich Kevin_Kofler
15:06:47 <zodbot> Current chairs: Kevin_Kofler jgrulich rdieter than
15:07:49 <rdieter> #topic agenda
15:07:57 <rdieter> alrighty, what to discuss today?
15:08:05 * rdieter throws out kde-4.12.2 status
15:09:23 <ltinkl> hi
15:09:36 <dvratil> hi
15:09:51 <rdieter> #info ltinkl dvratil present
15:09:53 <rdieter> hi
15:10:13 <Kevin_Kofler> Any updates on KDE in Fedora.NEXT?
15:10:25 <Kevin_Kofler> I feel that the gnomies want to use it as a pretext to extinguish us. :-(
15:10:44 <ltinkl> the underlying DE has not been determined yet officially
15:10:50 <ltinkl> but the outcome is obvious
15:11:03 <ltinkl> now the main question is:
15:11:31 <ltinkl> do we want to try to push KDE as an alternative desktop beside Gnome, based on the Workstation product
15:11:32 <rdieter> not really, still waiting on whether we can fit under workstation umbrella (likely not), or if we need to propose our own product (procedure/protocol not yet determined)
15:11:43 <ltinkl> or do we want to come up and start working on our own product?
15:11:47 <rdieter> or some other better 3rd thing. :)
15:11:55 <rdieter> ltinkl: right.
15:12:11 <rdieter> I think it's premature to decide anything yet
15:12:12 <ltinkl> those are the 2 possibilities
15:12:12 <Kevin_Kofler> I think we really need to be a separate Product.
15:12:23 <ltinkl> Kevin_Kofler: not necessarily I guess
15:12:31 <Kevin_Kofler> Getting ourself forced onto the Workstation Product will only lead to conflicts.
15:12:45 <Kevin_Kofler> At least assuming the Product will be primarily a live image.
15:12:46 <ltinkl> depends on how it's done technically
15:13:01 <Kevin_Kofler> If we're doing installers instead, it's doable, but it'd be sad to lose the live image.
15:13:05 <rdieter> theres obviously pros/cons to any approach, but we dont know what they are yet
15:13:25 <rdieter> anyway, let's add this to the agenda if you want
15:13:29 <ltinkl> there's Fedora devconf going on during this weekend, I'd say let's discuss those things in person, mainly with the Gnome folks (most of them will be present there)
15:13:30 <rdieter> anything else to add?
15:13:43 <Kevin_Kofler> ltinkl: +1
15:13:46 * Kevin_Kofler will be there, too.
15:14:04 <ltinkl> there's a lot of misconceptions and strange ideas flying around the desktop and -devel MLs
15:14:28 <rdieter> sounds like not, adjusting topic to match reality
15:14:41 <Kevin_Kofler> That just shows what kind of a broken mess "Fedora.NEXT" is. :-(
15:14:55 <ltinkl> not broken... largely undefined
15:15:01 <rdieter> ltinkl: +1
15:15:21 <ltinkl> and we can take a proactive role (at least try to) in defining it
15:15:30 <Kevin_Kofler> error: "Fedora" has no member called "NEXT" ;-)
15:15:33 <ltinkl> because if we miss the train now... hmm
15:16:46 <rdieter> ltinkl: someone onlist suggested that Workstation include some extra runtime libraries by default, including Qt.  Id go so far as to suggest kdelibs/kde-runtime even
15:17:07 <ltinkl> rdieter: I'd go even further
15:17:20 <rdieter> ltinkl: like?
15:17:24 <Kevin_Kofler> Yeah, after all we are forced to ship various libgnome* crap as dependencies of some global Fedora stuff…
15:17:33 <Kevin_Kofler> So, why shouldn't they ship our libs, too?
15:17:38 <ltinkl> rdieter: the workstation is a product, the DE doesn't matter in this case imho
15:17:44 <Kevin_Kofler> That way they could actually ship e.g. liveusb-creator, too.
15:17:45 <rdieter> ok, fair enough
15:18:00 <ltinkl> and, if a certain DE meets the technical criteria, it should be possible to switch between them
15:18:06 <rdieter> ltinkl: though a good starting point is default runtime library support
15:18:21 <rdieter> at a minimum.
15:18:31 <Kevin_Kofler> rdieter: We still need a spin with our workspaces!
15:18:45 <Kevin_Kofler> So to be honest, I don't care that much about what runtime libraries ship on the GNOME image.
15:18:46 <rdieter> Kevin_Kofler: thats a separate topic, but yes, of course
15:18:51 <ltinkl> the various technical criteria will be imho: systemd, logind, upower, udisks, bluez5, etc
15:18:55 <rdieter> Kevin_Kofler: ok, then *you* dont care. :)
15:18:56 <ltinkl> all of which KDE does meet
15:19:02 <Kevin_Kofler> After all, libraries are normally dragged in as dependencies.
15:19:18 <Kevin_Kofler> It's not like "yum install amarok" won't work even if those are not installed by default.
15:19:33 <rdieter> I just want to take "but this kde app foo pulls in all these extra dependencies" excuse off the table as valid
15:19:47 <ltinkl> technically, it's just a matter of a checkbox in the installer (Gnome x KDE)
15:19:49 <Kevin_Kofler> (unless they screw up things with product-specific packages (which people have suggested will be allowed) in a way that breaks things)
15:19:53 <ltinkl> and then adjusting the comps
15:20:18 <Kevin_Kofler> ltinkl: That assumes the product is delivered as an installer rather than a live image.
15:20:44 <rdieter> so vendors, isv's, 3rd-parties can have a reasonable expectation that Qt/kde-runtime is available by default in fedora workstation
15:20:45 <ltinkl> ye, it's been also said that the Workstation won't likely target CD-sized live images
15:20:56 <ltinkl> and rather go with a 2GB USB disk images
15:21:04 <Kevin_Kofler> We haven't been targeting CD-sized live images for ages.
15:21:11 <ltinkl> should be fine then
15:21:13 <rdieter> Kevin_Kofler: +1, true
15:21:32 <Kevin_Kofler> It stopped being realistic when "Mini"DebugInfo was introduced.
15:21:45 <rdieter> even before that, imo, but we digress. :)
15:21:48 <Kevin_Kofler> (10% bloat for no practical benefit whatsoever, sigh…)
15:22:03 <Kevin_Kofler> The release before "Mini"DebugInfo, we delivered a CD just fine.
15:22:29 <Kevin_Kofler> Anyway, other bloat has been introduced since, too.
15:22:52 <Kevin_Kofler> So even if "Mini"DebugInfo were dropped, we probably wouldn't fit on a CD with any reasonable package set.
15:23:01 <ltinkl> is it realistic to think that the current gnome+kde live images would fit in 2 gigs combined?
15:23:12 <ltinkl> after all, lots of the RPMs are share
15:23:23 <Kevin_Kofler> (unless maybe if the ban on removing selinux-* gets lifted :-p )
15:23:24 <ltinkl> d
15:23:32 <rdieter> ltinkl: I think that is doable
15:23:58 <Kevin_Kofler> Well, the thing is, how do you define "combined"?
15:24:14 <Kevin_Kofler> So far, the approach has been to take the separate images and make a multiboot image out of them.
15:24:21 * rdieter assumed a single live image, with both DEs
15:24:31 <Kevin_Kofler> If you merge the images into one, then you get the problem that you have (at least) 2 apps for everything.
15:24:40 <rdieter> doesn't matter a *lot* except for DM choice
15:24:59 <Kevin_Kofler> And then the GNOME folks are going to yell loudly at the "menu bloat" and impose OnlyShowIn= abuse that I really DON'T want to see.
15:25:19 <rdieter> Kevin_Kofler: it's an issue that needs to be addressed sooner or later
15:25:21 <Kevin_Kofler> (I want people to be able to use KDE apps in GNOME and the other way round. But a default menu with 2 entries for everything isn't that great a solution either.)
15:25:34 <rdieter> but that's something to worry about later, imho
15:25:41 <rdieter> there are bigger issues
15:25:44 <Kevin_Kofler> There's a reason the Multi-Desktop ISO is done the way it is.
15:26:21 <rdieter> sites (like mine) deply fedora with multiple DE's all the time, it's time that use-case is addressed better
15:26:23 <Kevin_Kofler> rdieter: I disagree. As soon as you suggest merged live images, this becomes a very practical issue.
15:26:42 <rdieter> yes, I'm not saying it's not an issue, I'm saying lets worry about it later
15:26:43 <Kevin_Kofler> (I disagree about the "worry about later" thing, that is.)
15:27:12 <rdieter> we dont even know how kde can/will get shipped ... yet.  and you're worring about app overlap?
15:27:13 <Kevin_Kofler> And how do you propose resolving it?
15:27:31 <ltinkl> technical stuff that will have to be resolved later
15:27:37 <rdieter> ltinkl: +1, exactly
15:27:40 <Kevin_Kofler> With a hack like kmenu-gnome that puts all GNOME apps in a submenu in KDE (and something similar the other way round)? That's only marginally better than NotShowIn. :-/
15:27:59 <Kevin_Kofler> ltinkl: Having a solution to this problem is a premise for even thinking of doing merged ISOs.
15:28:03 <ltinkl> Kevin_Kofler: you're going into much detail now (eventho valid)
15:28:08 <Kevin_Kofler> I think it cannot be solved.
15:28:17 <Kevin_Kofler> And I'm explaining why I think it's unsolvable.
15:28:24 <Kevin_Kofler> And thus IMHO merged ISOs are a no-go.
15:28:33 <Kevin_Kofler> (merged live ISOs, that is)
15:28:45 * rdieter doesn't consider that a blocker, imo, but again, that's something to be decided later
15:28:48 <ltinkl> you opinion, fine
15:28:50 <Kevin_Kofler> Of course, something with a package-based installer can offer both options, but then we lose the live image.
15:28:51 <ltinkl> your *
15:29:19 <rdieter> any new topics wrt Fedora.Next now, or can we move on?
15:29:23 <ltinkl> so... before we move on, I'd like to ask others
15:29:25 <Kevin_Kofler> When the OnlyShowIn= pollution will start, I'll be able to tell you "told you so". :-/
15:29:42 <rdieter> ltinkl: ok
15:29:53 <ltinkl> dvratil, jgrulich, than, Kevin_Kofler, rdieter: which path would you prefer?
15:30:17 <ltinkl> a) trying to pursue the "one" Fedora Workstation way
15:30:26 <ltinkl> b) come up with a separate product
15:30:29 <Kevin_Kofler> Or the alternative, a "KDE" desktop with not only Firefox and LibreOffice, but also Evince, Gedit etc. as apps, no thanks!
15:31:41 <jgrulich> probably (b), but I'm not much familiar with Fedora.Next
15:31:51 <dvratil> b) sounds like more work for us, on the other hand it gives us possibility to hand-craft the system to our needs
15:32:26 * rdieter prefers a  (slightly, has better opportunities for sharing resources)
15:33:10 <ltinkl_> jreznik_: too :) your opinion is very sought after
15:33:53 * than prefers (b)
15:34:08 * jreznik_ is at meeting :(
15:34:44 <ltinkl_> and the rest? (sry if I missed it, I got dropped)
15:34:56 <jgrulich> [16:31] <jgrulich> probably (b), but I'm not much familiar with Fedora.Next
15:35:54 <ltinkl_> then there's a question of (mostly web) exposure, do we want to get buried deep and hidden somewhere as an nth option?
15:36:06 <jgrulich> ltinkl: dvratil is also for (b) and rdieter is for (a)
15:36:08 <Kevin_Kofler> I'm also for (b).
15:36:18 <Kevin_Kofler> I don't think (a) is practicable.
15:36:28 * rdieter is worried that b will end up being more work than we'll be able to accomodate
15:36:39 <ltinkl> rdieter: yup, that too
15:36:50 <rdieter> we're stretched thin as-is
15:37:10 <Kevin_Kofler> I think the technical constraints of (a) are going to be incompatible with the design goals of the GNOME people, and to be honest probably ours too.
15:37:15 <ltinkl> see, we can at least _try_ to go with a) and fallback to b) if that fails or turns out unfeasible
15:37:26 <rdieter> ltinkl: thats fair
15:37:56 <Kevin_Kofler> As I said, (a) means either no live image, or a bad live image with messy menus, or a bad live image with only GNOME apps under "KDE", or abuse of OnlyShowIn/NotShowIn that breaks the experience for people who want to use an app from the other desktop.
15:37:58 <rdieter> given that many here expect a to fail for one reason or another
15:38:25 <Kevin_Kofler> So in practice, IMHO, (a) means no live image.
15:38:31 <Kevin_Kofler> The other stuff all sucks.
15:38:44 <ltinkl> define the failure :) I think a failure would be if Fedora as a whole fails to get more relevant and more contributors/users
15:38:52 <Kevin_Kofler> And no live image means we lose a very popular deliverable that we have so far focused on.
15:39:15 <ltinkl> my fear is that going the "Gnome OS" way is exactly that for Fedora
15:39:18 <Kevin_Kofler> ltinkl: KDE with GNOME apps would be a bad failure.
15:39:19 <ltinkl> (imho)
15:39:20 <rdieter> ltinkl: in my context "fail" as in fail to satisfy our requirements/constraints
15:40:17 <Kevin_Kofler> Imagine, you select the Plasma desktop, you open a Browser, you get Firefox if you're lucky, Epiphany otherwise. Then you open a file manager, it's Nautilus! And you open a text file, it's Gedit. PDF? Evince. Aaaaargh!!!
15:40:54 * rdieter does that (to an extent) everyday at my site, it's not that bad.  users largely dont care
15:41:06 <ltinkl> Kevin_Kofler: again, technicalities, if you selected the KDE desktop in the installer, you'd get KDE apps as defaults
15:41:09 <rdieter> they just want stuff to work
15:41:42 <rdieter> ltinkl: +1
15:42:15 <ltinkl> look at Mandriva or SUSE
15:42:20 <ltinkl> they do exactly the same
15:42:52 <ltinkl> and with Ubuntu and their Unity based on Qt/QML, fedora would be the only major distro to default to Gnome shell
15:44:26 <Kevin_Kofler> ltinkl: Again, you're assuming the (main) deliverable will be an installer.
15:44:45 <Kevin_Kofler> As I said, your proposal basically disallows live images.
15:44:57 <ltinkl> Kevin_Kofler: right, what else could it be that wouldn't fit us?
15:46:11 <Kevin_Kofler> It could be a live image.
15:46:18 <rdieter> though, we havent discussed or agreed-on what our requirements/constraints are yet, so postulating what solutions would satisfy those are... again... premature.
15:46:24 <Kevin_Kofler> And see what I wrote about merged live images.
15:46:49 * Kevin_Kofler wonders if he is talking to a wall. :-(
15:47:12 <ltinkl> we replied to that very argument
15:47:12 <rdieter> you're creating an imaginary wall, then talking at it, yes.
15:47:44 <rdieter> unless I've missed something
15:48:26 <ltinkl> [16:27] <ltinkl> technical stuff that will have to be resolved later
15:48:44 <rdieter> can we move on now?  (I suspect we're going in circle now)
15:48:59 <Kevin_Kofler> ltinkl: We cannot plan something that is not possible technically.
15:49:22 <ltinkl> [16:28] <ltinkl> Kevin_Kofler: you're going into much detail now (eventho valid)
15:49:43 <rdieter> Kevin_Kofler: if there are irreconsilable problems, we'll deal with it when the time comes (like resorting to creating our own product)
15:49:53 <ltinkl> yup
15:50:17 * Kevin_Kofler gives up… :-/ Can we move on?
15:50:21 <rdieter> ok
15:50:40 <rdieter> kde-4.12.2 status:  builds are almost done in f20-kde target
15:50:43 <Kevin_Kofler> #topic kde-4.12.2
15:50:58 <rdieter> (crap, I used wrong command, thanks)
15:51:13 <rdieter> most of it I stuffed into kde-unstable already
15:51:33 <rdieter> once the last ones, calligra/digikam are done, will move over to kde-testing for some smoke-testing, then off to bodhi
15:52:04 <Kevin_Kofler> Great!
15:52:08 <rdieter> there's a kde-workspace-4.11.6 build too, but we can do that separately
15:52:21 <Kevin_Kofler> What's the plan for F19?
15:52:44 <rdieter> I only did f20 so far, fwiw.
15:53:15 <rdieter> and we already did a 4.10 -> 4.11 update for f19
15:53:38 <than> i would say onyl 4.11.6 for f19
15:54:13 <rdieter> I did unofficial kde-unstable 4.12.1 f19 builds, I could do that for 4.12.2 too
15:54:39 <Kevin_Kofler> One thing to be aware of is that the next KDevelop needs libkomparediff2 from 4.12, but we can do a grouped libkomparediff2-kompare-kdevplatform-kdevelop update, there aren't really any changes other than the external libkomparediff2 in the new Kompare and it should build fine against older libs.
15:54:58 <Kevin_Kofler> And AFAIK it's not released yet anyway.
15:55:08 <rdieter> ok, that's doable
15:57:00 <rdieter> alrightly
15:57:04 <rdieter> #topic open discussion
15:57:05 <Kevin_Kofler> Of course if it were just me, I'd just push 4.12 to F19 too…
15:57:17 <rdieter> anything else for the last few minutes?
15:57:25 <Kevin_Kofler> But we've had that discussion often enough.
15:57:26 <Kevin_Kofler> ;-)
15:57:49 * rdieter is open to the idea, but not pushing for it either
15:58:12 <Kevin_Kofler> So who will be at DevConf? I suppose the RH Brno folks will all be there. :-)
15:58:25 * Kevin_Kofler is going to DevConf too.
15:58:37 <dvratil> see you there :)
16:00:50 <rdieter> our time is up, thanks everyone
16:00:57 <rdieter> #endmeeting