GNU OpenStep

[Mail Delivery Subsystem by way of perkins@andromeda.rutgers.edu (Steven C. Perkins)]

>Path:
newark-news.rutgers.edu!rutgers!newsserver.jvnc.net!howland.reston.ans.net!g
atech!swrinde!elroy.jpl.nasa.gov!decwrl!netcomsv!netcom.com!keith
>From: ke...@netcom.com (Keith Mason)
>Newsgroups:
gnu.misc.discuss,comp.os.linux.misc,comp.sys.next.advocacy,comp.sys.next.pro
gramm,comp.sys.next.software
>Subject: Re: OpenStep
>Message-ID: <keithCH...@netcom.com>
>Date: 10 Dec 93 11:45:41 GMT
>Organization: Netcom Online Communications Services (408-241-9760 login: guest)
>Lines: 55

In article <2ctds8$j...@skates.gsfc.nasa.gov> cah...@clouds.gsfc.nasa.gov
(Robert F. Cahalan) writes:
>In June 1994 NeXT will release "OpenStep" to the public domain
>This will be a full and open version of their operating system
>interface, with the full object heirarchy, the Interface
>Builder, and the underlying messaging system.
>
>Also, how about a version of OpenStep for linux?
>Comments, linux developers?

I must let you all know that a "GNU" implementation of OpenStep is in the
works, and has been for a few months now.

It started from two directions: One, some people at the Stanford Linear
Accelerator Center needed to be able to compile their NeXTSTEP apps on
a other machines. They wrote quite a bit of their code to make a NeXT-
like API on top of Motif. More beyond this, I really shouldn't say, as
that wasn't my project. Two, I decided that I'd like to be able to create
programs in a NeXT environment, but couldn't afford the NeXT machine OR
their software for Intel machines (or the special hardware they require
for the Intel machines). So I decided to start drawing up my own API.

Since the announcement of OpenStep, our two teams have come together,
somewhat. SLAC agreed to let what they created into GNU, and my idea
was to make my software GNU all along. While their are still differences
in the implementation details between our groups, I believe we can
still create the "GnuStep" everyone here would like to see in a reasonably
short period of time...

The current implementation is still rather flakey, but IT DOES WORK, kinda.
The idea, at the moment is to fix up what we already have to make it
true OpenStep. The current phase is to isolate the Motif components of
SLAC's library, and set the whole thing up so that Objective-C "categories"
can be used to run OpenStep on a variety of windowing systems. The
first windowing system we are supporting will, of course, be X. Motif
support may be done at the same time, although Motif is being looked
down upon the the GNU gods (can't say I disagree!).

There has been a call for help placed in the GNU Objective-C mailing
list, and it's now being placed here. It would be a tremendous help for
those out there to be willing to spend some time to "fix" what we already
have (isolating Motif), and to build the remaining necessary components.
If all of you out there really mean what you say, that you'd love to see
OpenStep for your machine (even in a somewhat crippled implementation),
let me know! The plan is being layed out, and WE NEED HELP!

The current SLAC implementation is available at

ftp.slac.stanford.edu:pub/sources/objcX-0.5.tar.gz

Thank you.

-- Keith
--
Keith Mason
send all replies to: ke...@netcom.com