Re: Some thoughts about GS

[Dennis Leeuw]

> Frederic <frederic.chauvin@noos.fr> wrote (Tue, 02 Jan 2001 00:21:03 +0100):
> > > - in the meantime other technologies became very popular *and* powerful, 
> > > such as GNOME and KDE
> > True for KDE, wrong for GNOME. IMHO GNOME is a definitive failure, from
> > a technological point of view, as well as project management and
> > end-user point of view. I spent a lot of time studying GNOME, KDE and of
> > course GNUstep technologies and there is no doubt for me that GNOME is a
> > mess. This may be a chance for GNUstep as GNOME disappoints more and
> > more people that soon or late will have to find a replacement that
> > *cannot* be KDE for *political* reasons...

I agree on the GNOME part, but KDE's political reason (for not using it) is 
gone now Qt is GPL or LGPL...

> > The GNUstep team is already reusing code where possible (libxml,
> > configure scripts, etc...) but further integration does not make sense
> > for me (eg: merging AppKit and Gtk)

I agree partially, what I think would be nice is to incorporate GTK and Qt as a 
service/bridge or somethink like that. I don't understand all that is needed, 
but somehow there should be an interaction between the three that solves the 
look and feel issues and application message problems.
Where Qt and GTK are close to oneanothers look and feel GNUstep is entirely 
different, and there a GTK AppKit can help.


> > GNUstep is worth the investment for numerous reasons on its own and more
> > and more people will understand this. Nevertheless, GNUstep probably
> > needs more advertisements to speedup this movement.

How do you want to accomplish this. I have tried several announcements on 
slashdot and none came through. What other ways are there to promote GNUstep, 
except for talking with people (mouth-to-mouth).
Just popping up: maybe we should start the holly war ;) Since the one between 
GNOME and KDE doesn't seem to take off. Maybe we could write an article for the 
largest Linux sites which compares GNUstep with GNOME/KDE and Windows and Mac 
OS. If we could get that posted, we will sure draw some attention...


> And I do not talk about world domination at all,

Why not? Put your stakes high ;)


> I just would like to see GNUstep play a more important role in the free 
> software movement - with one goal in mind: more developers working on it!
> Personally I think KDE and GNOME are good, they deliver what they promise. 
> Perhaps they are not as elegant as GNUstep, but they provide working 
> solutions! Eg. you can install GNOME over the net using a graphical 
> installer, you can use apps like Gimp, Abiword, dia, evolution, nautilus, 
> mozilla/galeon and so forth (not to mention all the KDE solutions). Actually 
> I am "forced" to use Sylpheed to write my mails because there is no GNUstep 
> Mail.app... again, don't get me wrong, I don't blame anyone for not having 
> all of that in GNUstep, I just think it would be better to 'share' instead of 
> reinventing the wheel!

That's true. There is no need to write another GIMP, but is a GTK AppKit the 
only solution?
What I see as a problem is the fact that you will then need a gtk AppKit, a Qt 
AppKitt, and probably more.


> After all, it is not about KDE vs. GNOME vs. GNUstep. vs. ... but free 
> software vs. closed software (or should I say M$ ...grin). I still believe 
> that it is better to have good software written with a not so good technology 
> than no software at all!
>
> But to be more on topic:
>
> - What would it take to write a 'gtk-backend' for GNUstep? Can anyone give 
> some predictions?
> - What would it take to include/integrate bonobo into GNUstep?
> - How are we going to solve the problem that ObjC scares developers who do 
> not know that language (believe me, it does, I have seen that a lot in the 
> OPENSTEP/Mac OS X business)?

That is one of the major issues. Maybe we should rename it to something fancy 
like perl, or phyton ;)
Serious though, I named the above two because some fine coders I work with on a 
daily basis advise everyone to start using languages like these. Languages that 
you don't need to compile. For them that is the future.
And that is what I hear a lot these days.
So maybe a lack of marketing again...


> - How do we get more attention? I mean, now that MOSX is here (at least PB), 
> we cannot have a better time to do marketing, no?
> - What would it take to integrate rpm package creating in the makefiles? What 
> would it take to build GNUstep snapshot rpm's?
> - In general, what integration would be useful, what should be avoided?

I think that ProjectCenter and Gorm are the GNUstep killer apps to be. They 
make a GNUstep desktop environment into a GNUstep development environment. They 
could make ObjC into the VisualBasic of the opensource community.

IMHO what is useful:
1. Getting GNUstep to 1.0 and the same goes for PC and Gorm.
2. Building RPMs for SuSE and Red Hat (TurboLinux, Mandrake?), or contact those 
vendors to have them do it.
3. Building tgz and deb packages for inclussion in Slackware and Debian (I 
could contact Patrick Volkerding about the Slackware part...)

This would get people started. Which is the most important thing. If it is not 
on the CD they have they tend to overlook things.

Greetings,


Dennis

dleeuw.vcf
Description: Card for Dennis Leeuw