Re[3]: Gnustep + mac + windows?)

[Manuel Guesdon]

On Wed, 25 Sep 2002 12:14:22 +0200 Andreas Heppel <aheppel@web.de> wrote:

 >| On 2002-09-24 13:44:56 +0200 Philippe C.D. Robert <probert@sgi.com> wrote:
 >| 
 >| > On Tue, 24 Sep 2002, Pascal Bourguignon wrote:
 >| >> Anyway,  I can  think of  dozens of  programs to  write  before adding
 >| >> themes...  For example, a  MS-Windows backend would probably have more
 >| >> impact on potential users of GNUstep.
 >| > 
 >| > Why this?
 >| Because there lots of Windows users and developers who might be interested 
 >in a framework that is easy to use _and_
 >| portable.
 >| Coming from the Windows world, I know the existing frameworks there and 
 >know now how much easier it is to do a
 >| graphical application using the GNUstep framework than using C++ and the 
 >MFC (Microsoft Foundation Classes) or
 >| Borland's frameworks. Not to mention the fact that they cost lots of money.
 >| Tell those Windows developers how easy it is to learn Objective-C and how 
 >easy GNUstep development is, and even if
 >| only a very small fraction starts to develop using GNUstep this will be a 
 >lot, because there are millions and
 >| millions Windows developers out there.

May be I'm a marginal case here and sometimes I fell a bit suprised by all 
these discussions but I think there's a lot of
people who doesn't care about GUI, backends,... One of the thing which first 
attracted me was base classes and 
their coherence. I used to work with borland classes and MFC (long time ago) 
and never found something like
OpenStep/GNUstep base objects. For example MFC didn't provide real basic object 
manipulation: no copy, no real 
memory management,... I've even implemented concepts from OpenStep into MFC 
based objects to make life 
easier (and next switched to GNustep :-).

Todays, thre's a lot of people who's work is mainly oriented on servers, web 
applications and back office. I think they
don't care GUI, all they want is web application framework, database framework, 
network related objects and business
objects. 


Another thought is that GNUstep like lots of 'free' projects is based on people 
interests. If someone want a feature,
it code it. If he's the only one who want it and don't want to make it, the 
feature never come. I think GNustep is like
this: people have different interests and desire and make things to follow 
that. Windows people and Mac OS X
ones may choose to be interested and participate to GNUstep development but why 
one should do it for them ?

I've followed many Mac OS people discussion and near the only thing I've seen 
is, to summary: "Is GNustep ready 
so we can use it without participation to it's coding ? No, it don't have this 
or that. Okay, I'll check next year if it
willMany of t be ready.". They seems to don't care about all Apple trashed 
'standards' and other forgotten projects. New
marketing orientation is always better :-(
Windows people keeps working with MS tools even if they are buggy and expensive 
(MSDN cost is really expensive for
independent or small sized companies), even if MS don't care about   ascending 
compatibility and so on.  Why should we 
try to make them switch to *nix, GNUstep,... ?

Just few evening thoughts...

Manuel

BTW, GNUstep will now be used to make postscripts files to print SNCF (French 
national railway company) official
trains/hours indicators for Paris and it's region. No GUI, just base classes 
and GNustepWeb :-)
--
______________________________________________________________________
Manuel Guesdon - ORANGE CONCEPT <mguesdon@orange-concept.com>
14 rue Jean-Baptiste Clement  -  93200 Saint-Denis  -  France
Tel: +33 1 4940 0997  -  Fax: +33 1 4940 0998