fd.o disast^Wdiscussion

The "*nods in agreement*" comment at the bottom of this post is required reading. For all of you. Go Eike.


Care to clarify what you did and do intend to happen at all?

All I got to know through your comments is

  • that you are emotional about the issue and think it's depressing (neither one I'm actually interested in to know),
  • that you think that "everyone" at fd.o is from GNOME (which certainly isn't true, but I guess you want this to be this way and thus prefer to spread this FUD?),
  • that you state fd.o is creating standards by referring to their widely usage, contradicting the fact that the fd.o platform contains tools which are not widely used at all as of now (and with the way this issue is tackled at the moment this might not change at all),
  • that following this up you apparently don't care to clarify what the fd.o platform actually is about (What makes intltool and scrollkeeper qualify as part of the fd.o platform?),
  • that while you keep talking about the lack of contributions from KDE individuals you so far completely ignored what could be invited to fd.o right away (QtGK, KIO Fuse Gatway, the GTK-Qt Theme Engine are clear candidate for fd.o for which respective counterpart could be created to make the solutions work both ways, in a true supporting "interoperability and shared technology" fashion even, the way everyone here initially thought fd.o is all about).

By Na at Fri, 01/30/2004 - 13:19

that while you keep talking about the lack of contributions from KDE individuals you so far completely ignored what could be invited to fd.o right away

Invited by who? Do the the gnome people who are active on fd.o know about those technologies. Isn't this what Daniel is complaing about. Kde developers should be promoting those technologies on the fd.o maillinglists.

By jas at Fri, 01/30/2004 - 19:33

Daniel frequents this site and several KDE mailing lists and thus can't say he didn't know about or wasn't aware of it. But instead early starting a positive discussion about what part could be hosted on fd.o he apparently preferred to make an announcement of the fd.o platform which includes nothing which is of interest for KDE at the moment, and then complained afterward about the lack of reactions. Heck, he could have simply added a tiny question to the announcement asking if there is no single piece KDE wants to contribute to the fd.o platform, and the discussion would look much different from how it does now. And from following the xdg list I can't say that there is a lack of involvement of KDE developers, there where interoperability issues are know unifying standards are being discussed regularly. But there where no neccessity for an implementation is seen developers don't take part since they are not interested, I think that's natural. The latter might well be also the reason why eg. QtGTK wasn't mentioned at fd.o even though its author Zack regulary did contribute there, Qt is a very KDE specific technology after all, and fd.o should be a place for shared stuff, shouldn't it?

I really think the issue is known to the most by now, we should probably better import all KDE technology to fd.o so others don't miss it even though it might be deemed too much KDE specific. But the way how Daniel handled the issue so far publically is certainly not encouraging anyone to start anything like that.

By Na at Fri, 01/30/2004 - 22:51

I don't "frequent" the site. I've been reading it in the last couple of days - I posted "*sigh*", and then didn't check until about a week later, when I saw a massive flamewar erupt about it on IRC.

Also, I am subscribed to three @kde.org lists - kde-debian, kde-debian-devel, and kde-debian-private. That's it, seriously. I don't think I've had write access to -core-devel for about 18 months now, something like that.

There doesn't really need to be a huge discussion about what's suitable for fd.o. Is it vaguely portable? Is it useful? Does it help bridge the gap for desktops? If you answered positively to all of the above, email me an SSH key, preferred username, and with a couple of details about your project. It'll take about 10min to set you up, if that. Seriously.

I don't claim how I've handled this to be any sort of banner for how things should be done, but I've spent years banging my head against a brick wall, and I'm sick of it. You know, if I had've posted saying 'Hey guys, I like fd.o. Does anyone want to move anything there?', no-one would've said anything. It took a pretty strongly-pointed post (and an inflammatory, unproductive reply; at least it got discussion going) to actually get things going. How much does that suck?

By daniels at Sat, 01/31/2004 - 13:08

I'm not emotional about it, so please stop trying to troll me with that word all the time. I do think it's depressing, yah. That doesn't make me some unstable 'emotional' maniac whose thinking is clearly not to be trusted.

As for the fd.o point - who's there from KDE? As I said before, Waldo's there in the core group, and that's it. Zack and Alex have been doing some binding work (though I don't know if Alex's work is hit-and-run, or not), and Scott has commit access to gstreamer, but I don't know how frequently he commits.

So we have one active KDE dev on the fd.o core group (as in, people who work on either core fd.o projects, or work on very general aspects of the site). Can you name more and prove me wrong?

The platform contains a lot of stuff that isn't standard yet, but a lot of stuff that either is, or only needs widespread deployment (in something like a platform, perhaps), to change this. If you have comments on the platform, again, please join the list and elaborate your specific concerns. It's not like I've decided everything and am ruling with ye olde iron fist, and am going to release tomorrow. Seriously, if you don't like the way it's going - help shape it! Make it better, don't just sit on the sidelines going 'boo hiss sucks'.

As for Navi's post, I seriously didn't see it - my list subscription was messed up for a while. I have to go over the posts I've missed and reply; thanks for this! scrollkeeper and intltool were only proposed; very few things are definite (xlibs, D-BUS, HAL).

And have you listened to a word I've said? I've been inviting anyone who will listen to fd.o, and I've been pimping it on the kde-debian list. Maybe I need to make myself really clear. Again.


By daniels at Sat, 01/31/2004 - 13:02

So we have one active KDE dev on the fd.o core group (as in, people who work on either core fd.o projects, or work on very general aspects of the site). Can you name more and prove me wrong?

David Faure, mime-spec and I think icon-theme-spec
Fredrik Höglund, cursor spec

I also saw other devs in discussions about unifications of
- documentation (Cornelius Schumacher)
- xclipboard handling (IIRC Lubos Lunak)

I saw George Staikos post in one of the HAL discussion, I think Ellis Whitehead in a discussion about unified key bindings, etc

By krake at Sat, 01/31/2004 - 15:46

OK, so I'll pay dfaure and maybe Fred. But I'm talking *actively* involved, all the time - not just a hit-and-run on a list discussion.

By daniels at Sun, 02/01/2004 - 03:38

I see, so providong input ont things that do not have a specification workgroup doesn't count as a contribution.

So if one of the participants in the documentation discussion comes up with a spec or implementation, then, from your point of view, none of the KDE developers provided any input.

I guess Cornelius should consider to label out documentation system as the new universal implementation and demand the others adopt it, because they failed to provide any input.

By krake at Sun, 02/01/2004 - 17:53

I said 'more than hit-and-run contributions'. I'm also not incredibly pedantic and rigid.

But you don't have to listen to me; make up your own minds on who's involved in fd.o, seriously. You can say that anyone who has an account is involved, or only those with sudo is involved, or anyone who's subscribed to any of the lists is involved, or whatever you like. Seriously.

By daniels at Sun, 02/01/2004 - 23:01

So we have just different limits of what we consider a contribution.

You count it if it changes something already on freedesktop.org, I count it if it clarifies KDE's position and details about our implementation of a new, not-added-yet, shared standard.

To clarify that on an example:
as long as there is no document describing the shared documentation system, Cornelius input does not count as contribution, but the moment such a document is accessible via the freedesktop.org website, it is.

So in order to increase KDE's contributions, we should add random parts of KDE stuff to fd.o whether GNOME, ROX or XFCE people agree or not?

By krake at Mon, 02/02/2004 - 21:28