[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: GNUstep Window Manager (was RE: Idea)

From: Dan Pascu
Subject: Re: GNUstep Window Manager (was RE: Idea)
Date: Sun, 7 Jan 2001 19:06:22 +0200 (EET)

On  7 Jan, Pascal J. Bourguignon wrote:
> Actually, on  NeXTSTEP and OPENSTEP,  the equivalent functions  of the
> X11 window managers are implemented, AFAIK, as PS code downloaded into
> the DPS server. That is: the "window manager" on NeXTSTEP and OPENSTEP
> is not written in Objective-C, but in Postscript.
> Actually, it seems to be that all the features we would like to remove
> from WindowMaker  to make it  a dedicated GNUstep window  manager, are
> features  needed  actually  to  support plain  X11  applications.  

Exactly. And this also brings the following question:
How do you expect gnustep to be usable with _any_ window manager out
there (at least partially)? Or it would only work with exactly one
highly adapted specifically written one?
The idea that we will only run a pure gnustep environment with only
gnustep apps is far from the true. There are plenty non-gnustep apps
that users won't throw away, and also developing their equivalent
gnsutep variants can take years. Even if there are equivalent gnustep
apps still the user may choose to use the others base on personal
reasons, habits and tastes.
And don't forget other apps also evolve while gnustep evolves. gnustep
have a lot of work to do to reach them from behind (to be able to
provide features that they already provide) and that means more time and
developers that it has, because other apps also don't just stand and
Just look at wine: they're always steps behind of the real thing. Same
with gimp versus photoshop. And they have enough developers.
So I don't think that a pure gnustep, providing all imaginable apps is a
reasonable idea and people will always use other apps to.

> I'm wondering if it would not  be better to move these features into a
> generic wrapper GNUstep application that would interface and integrate
> the X11  applications and the  GNUstep environment, independly  of the
> window manager.

I'm not sure if that would be the best way. It will work, but in my
opinion the problem gnustep has now is that is uses too many layers of
abstraction up to the real hardware. This makes it to be too slow.
Just compare how the old NeXT system run on a @25MHz machine, and how
does gnustep run on modern address@hidden machines.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]