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

List:       kde
Subject:    Re: GNOME Desktop Project
From:       widi () wohnheim ! uni-ulm ! de (Rainer Bawidamann)
Date:       1997-08-19 12:17:27
[Download RAW message or body]

In article <199708181512.QAA05211@diamondage.cymru.net>,
	alan@diamondage.cymru.net (Alan Cox) writes:

>> need yet another project, essentially duplicating what the KDE project has
>> already done or plans on doing? In my view, focus of forces would be called
>> for rather than dilution through yet another project. 
> 
> I would much rather see a partial Qt clone. That is to take the bottom layers
> of the KDE widgets and make those talk Gtk. Most of KDE comes from the KDE
> widgets not the Qt stuff. The only killer is the Qt signal/connect stuff.

signal/connect stuff is in gtk--0.3 - a C++ wrapper for gtk.

But I'm not really happy that gnome uses gtk. It looked at it and I
cannot say I like it: It's similar to Xt in some way but does everything
it's own way (why does everyone start from scratch? there are many
toolkits to start with and to copy from). It's slow on my machine
(486DX100), maybe because it uses many widgets in unneeded places, e.g.
a scrollbar contains 3 additional widgets for the arrows and the slider.
This could be done more efficient. The only thing is that it's GPL.

I would like to have a fast and well structured object oriented toolkit
(this might be hard with C; I would prefer C++, maybe with some
underlying C code). I wouldn't hesitate using code from other GPL'd
toolkits/programs wether it's KDE or knews or gtk. Just combine the
best to make it even bether - that's the spirit of GPL.

>> behaviour -- especially given the fact that you clearly mean to compete with
>> the KDE project. Needless to say, that we are currently investigating our
>> options  as to how we can prevent this looting from taking place. On a

I think KDE can benefit from any reuse of its code: You can get their
improvements back to KDE. Isn't that a chance?

> As it stands KDE is a nice set of tools built on the wrong toolkit.
> Duplicating them is silly, porting them makes a lot of sense, creating
> a Qt like interface and using the base KDE classes is far saner. If we
> follow that path then it should end up under the "KDE" name however.

I think we should try to make KDE more independent from the underlying
toolkit and concentrate on the main points of KDE (of any desktop
system): a good integration and the same look-and-feel of the
applications (like the common menubars, toolbars etc in KDE).

The special widgets of KDE might vanish anyways: Qt is changing
constantly. I think they will imitate some of the KDE widgets (KDE shows
them what the users of Qt want, so they will try to do it this way).

Here is one big problem with KDE using Qt: work is on hold in several
parts of KDE because of the "will be in Qt 1.3" thingy.

Rainer
-- 
                Rainer.Bawidamann@wohnheim.uni-ulm.de

-       printk("Linux IP multicast router 0.04-might-work 8)\n");
+       printk("Linux IP multicast router 0.05-maybe-works 8)\n");
[Found in a linux kernel patch]

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