Moving kdcop

[Adriaan de Groot]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Tuesday 03 February 2004 15:54, Jason Keirstead wrote:
> Perhaps what this discussion has brought about, is the need for a *new*
> base package, kdepower (or some hopefully better name), that can take out
> some of the more "power user" KCMS and apps from kdebase, kdenetwork, etc.
> and put them in a separate package that would be highly recommended for
> "power users", but not newbies.

[ugh, rambling thread from hell] There are two things to distinguish:

- - KDE as packaged by packagers
- - KDE as structured in CVS

These are different animals. In CVS we have traditionally not cared very much 
for exactly what gets installed by what module - developers will check out as 
much as they need, and users, well, they get what the packagers think is 
best. In my case, the packager thinks that per-CVS-module is best, so, well, 
whatever. I'm told some packagers split it down to application level. As Andy 
has already pointed out, _packagers know what their users expect_. There's 
not much sense in screwing them over.

The previous paragraph is confused, fuzzed, and discombobulated by the 
increasing number of build-from-CVS scripts, distros providing CVS packages, 
and whatnot. I hear SorcerorLinux can do the equivalent of an `emerge 
kde-cvs' and it fetches and builds CVS. Yay.

I had a long talk with Ralf about modifying the CVS modules so that they match 
"user expectations" more. I don't think that's a particularly good idea 
because the source repository _doesn't know who the users are_. The 
packagers / distro's are in a far better position to say who their target 
audience / market is. So I managed to agree with Ralf that improving 
applications in CVS - even unpopular ugly ones like KUser - and replacing 
objectively broken ones - like KPaint - is a good thing, shuffling things 
around in the name of "users don't need this part of kdebase" is a bad thing, 
because it presumes knowledge of the users that we just don't have.

Remember: KDE CVS is not KDE as packaged. Do not meddle in the affairs of 
packagers, for they are subtle, and quick to anger. And Lauri the packager, 
when angered, is not subtle and will tell you to FOAD in less acronyms and 
more words.



> Good candidates I see would be:
> - DCOP functionality of KHotKeys
> - KDE Resources KCM

This might not even be a bad idea. It is, after all, something closely related 
to the PIM stuff. That would allow it to be updated for PIM 3.3. 

> None of these are things I can see the average user even wanting to look
> at.

The average user of the distro you have in mind, no. The average user of my 
distro (graphic artists with lots of PIM thingies, working in a groupware 
commune) does. Who's to say you are more important than me? Hence, since 
we're already both doing the job of picking and chosing those parts of CVS 
that make sense for our users, let us get on with our jobs. 

> What do other people think of this idea? Should it be floated further?

Only with spears stuck in it, till it washes out to sea.

- -- 
       pub  1024D/FEA2A3FE 2002-06-18 Adriaan de Groot <groot at kde.org>
The users that I support would double-click on a landmine to see what it did.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (FreeBSD)

iD8DBQFAH/U1dqzuAf6io/4RAleuAJ4gzbYjJfF80vWpxQeYm9BN5xpiaACfVIRK
Q5Lw7deiFrFDbLWxYlIxJDI=
=0V6w
-----END PGP SIGNATURE-----