Wine, WIDE & Unix (was: Support for pkgconfig)

[Dimitrie O. Paun]

On April 27, 2003 05:32 pm, Francois Gouget wrote:
> Bah, I don't really care. Maybe because I use neither Gnome nor KDE.
> I see your point but I can't help but think that it's a bit of a stretch
> to say Trolltech could become another monopoly when there is a GPL
> implementation available, which is why I said the above.

I really don't understand you. It seems to me obviously clear that a
GPLed platform is as useful as a GPLed Wine. This is not something
minor, and it's no stretch -- the Trolls would have the monopoly over
all commercial apps, which is where the money is. You may not use
either, but that's no solution. People will have to choose one or
the other, they can't go back to Xaw or Motif.

> Now if you think WIDE is portable, tell us how.

Well, we do have Winelib working on non-i386 platforms, don't we?
But bottom line, people worried about portability should program 
to the Gnome/Gtk API, I haven't suggested for a moment we drop
Gtk and code everything to the Win32 API. For people like the emule
guys that like to code to the Win32 API, they can continue to
do so if they so wish.

> So the situation will remain the same: we will haev KDE applications and
> WIDE applications and they will not be interoperable. I.e. it won't be
> possible to insert an Excel spreadsheet in a KWord document.

No, it will be better: you will be able to have deep integration between
Win32 apps and Gnome apps, no small feat.

> Does that include interoperability with OLE/COM/DCOM or just the look
> and feel integration?

Well, the OLE/COM/DCOM bit is the interesting one, the look and feel
is simple and has other solutions, as we discussed.

> I believe where we differ is on the method:
>  * I think we should work on Wine to make that possible (though
>    there may be some work to do in Gnome and KDE).

For sure. But this will work for things like L&F. The difficult
things aren't even considered by anyone at this point. So no,
we are quite in violent agreement here.

>  * You think that it is mostly Gnome that must be modified to use the
>    Wine APIs and you are ready to drop/ignore KDE. I believe you also
>    proposed to replace Nautilus/Konqueror with Explorer, and to develop
>    a new window manager.

No, I'm not dropping/ignoring KDE. It's just that right now we're talking
about some deep integration issues that no one even considered before.
Right now, integration is used to describe L&F. It's fairly clear we need
to be able to integrate with things like BlueCurve so Wine integrates as
much as possible in both Gnome and KDE. 

Wide doesn't change that. In fact, being based on Wine, it will implicitly
benefit from all the integration work we do on the Wine side. But at that
point, people will ask: what is the difference between Wide, and just Wine
running in Gnome? And it's a good question. Part of the answer is that Wide
may choose to use Wine's OLE implementation to integrate 
Bonobo/Win32/OpenOffice apps. But that's really wishfull thinking, I'm not
even sure that's at all possible. 

The real answer is that Wide is just like Gnome, but with a different set 
of mentalities. Instead of picking the default apps only based on religious 
things like 'is it using Gtk?', I propose we choose the default apps from 
the set of all Gnome and free Win32 apps (and any other platform that
integrates nicely). I am sorry I can't include KDE here as well, but it's 
because their platform is not free enough, not because I don't like the Qt 
framework. If tomorrow  we  have a LGPL Qt implementation, they would be 
added to the pool of possible apps.

This should answer your Nautilus/Konqueror/Explorer. I have to religious
affiliation to Nautilus or Konqueror. If the ReactOS people build a better
(and free) browser, it goes in as the default browser. Apps should compete
on features and user experience, not on irrelevant implementation details.

-- 
Dimi.