[prev in list] [next in list] [prev in thread] [next in thread] 

List:       gimp-developer
Subject:    Re: [gimp-devel] technical issues of Kimp vs Gimp vs GnomeGimp
From:       Avus <top249 () g26 ! ethz ! ch>
Date:       1998-07-08 12:02:51
[Download RAW message or body]

Having read the replies to my post I think this is at least as much an
emotional issue than a technical. Many people on both sides seem to feel
stepped on their toes, and, as usual, everybody feels innocent ;)
It's important to examine what of these opinions and
accusations/reprovals are justified, based on false information or just
due to misunderstandings.

Manish Singh wrote:
> [...]
> I have not seen anything, either code or screenshots, that tell to what
> extent the Kimp provides in terms of KDE integration. I'd say it would
> be even *more* confusing if Kimp only goes part of the way (or did they
> port all the plug-ins that come with GIMP as well?).
>
I think the main program is a good start, as it enables DnD and offers
common L&F for the most needed functions.
> 
> Once again, having only heard heresay and not seen anything real, it is
> essentially vaporware in my eyes.
> 
Just say one word, and you shall see all that you desire;-) (i.e. let
Kimp appear)

> > Going further I think the Gimp could act as a link between KDE and Gnome
> > development on the 'below-GUI' level, e.g Corba. Thus the Gimp
> > developers could prove that a pragmatic cooperation is possible among
> > all those flame wars.
> 
> The KDE people haven't done too much in the way of "pragmatic cooperation"
> from the beginning. Not informing us about Kimp in the first place is
> an irksome breach of developer etiquette, which has aggravated things
> more than they would have been if open communication was done before the
> project was started. I still haven't seen any efforts by the Kimp people
> to rectify or even apologize for this mistake. Cooperation goes both
> ways. The Kimp people got off on the wrong foot with us, and it's up
> to them to make amends.
> 
The point is: there is no (official) Kimp. Kimp was never announced
(AFAIK), it was just shown as an example for Matthias Multi-toolkit
programming technique at Linux Kongress. He used is as an eyecatcher to
demonstrate the power of his MT programming.
That's why I called it a programming example (as opposed to a project).
He was certainly aware of the mixed feelings of the Gnome developers
about a qt port and IMO that's why he did *not* make it publicly
available. 
And nobody can critizise him for making the port for his own
purposes/fun (that would be censorship), the question is only about
distributing it, which he didn't do.
In fact, rumours of Kimp have spread through mailing-lists, and Kimp was
mentioned in  Linux-Magazin in a report about the Kongress. So the
demand for Kimp comes from the users, not from KDE developers in the
first place.
You could still argue that Matthias intentionally provoked this demand,
but that would be a) speculative and b) not a reason to be really angry,
as it just provokes a discussion, which is a good thing when done
rationally.

In short: There is *no reason for amends*, just for talking.


[Miguel wrote]

> > I agree with Matthew Wilson that it is a dire necessity to implement
> > Gnome/KDE DnD support, and so is Corba interoperability. But in the
> > meantime, it would be a sin to prevent users from enjoying these
> > advantages from within KDE. 

> This is harder than you might think as KDE is using their own DnD
> protocol, a modified OffiX protocol.  Now, as far as I remember, they
> will move towards Qt's DnD protocol (Xdnd) at some point.  

Now, then when can we expect a Legacy KDE-DnD support or Xdnd support in
Gnome?
The Gnome FAQ doesn't say anything at this point, but probably Gnome
already uses Xdnd. 
As it can coexist with other DnD protocols, why don't you implement
KDE-DnD instead of saying that some time in the future we'll all use the
same one?

> > If lyx and Netscape can do multiple toolkits, Gimp can do it too!

> Actually, the LyX developers were very angry at Matthias for doing
> KLyX, as he basically ignored all of the design that had been going on
> to make LyX toolkit-independant.  Instead, Matthias just did a

As Matthias is both the founder of the KDE project and the original
author of lyx, this is a somewhat special case. As he designed and wrote
much of lyx, it's not the dev's being angry at him, but *different
opinions* between two groups of developers.
Although he doesn't lead the lyx development any more, he has IMHO still
the moral right to have his own opinion on lyx and to act according to
it.

He said he used the tk-indepentence where possible, but didn't wait for
it, as he felt a 'strong sense of urgency'. 
If you always wait for other people to finish, you'll get nowhere.

> KDE-centric port. Currently there are two versions of LyX: the
> original LyX which uses XForms and is moving towards toolkit
> independence and KLyX which is bound to KDE/Qt.  
That's actually the point of MT programming: You don't have to wait
until there is a toolkit-independant interface, but you can start right
now, change L&F, add DnD and very quick and replace the rest of the code
later on.


Concluding I'd say there are still good reasons for a cooperation of KDE
and Gimp developers, and it would be great if the Gimp could be a model
for cooperation in the free software society (esp.KDE-Gnome), beyond
ideological wars.

Perhaps we should give the KDE developers time until the version 1 is
released (they're pretty busy at the moment), and then start talking
about how such a cooperation could look like.

Regards 
Avus

[prev in list] [next in list] [prev in thread] [next in thread]