From: Daniel Pezely <pezely@lazlow>
Subject: Re: programming formats
Date: Fri, 08 Jun 90 14:06:57 -0400
Organization: U of Delaware, Comp Sci Lab


I don't want to start a discussion in the newsgoup of the future of 
UNIX systems, but here's some very important industry news which has 
not made it to the monthly journals yet.

but first, this is where I stand:
Since everything is experimental at this stage and some of the code
generated will have to be rewritten at some point, we might as well use
what we have available, whether it be Unix, DOS, etc.

In article <10218@hydra.gatech.EDU> :
>
>Two perhaps controversial opinions:
>1) someone stated earlier that a standard VR development space should 
>  be MS-DOS compatable.  I disagree.  The primary "bright-eyed larvae"
>  are going to be college students.  The main op-sys on campuses 
>  today is rapidly becoming/is *nix.  I think this is a good place to
>  start.

Well, there will be a few problems with UNIX very soon.

I don't know all the details, but there is a very heated battle between
Unix International (UI) and Open Systems Foundation (OSF).  Since 
their talks broke off two or three months ago, it looks like the industry 
is going with OSF and UI's Unix is going to be killed off.  (AT&T can
keep throwing money at UI so it will survive, but SVRx will only be
used by them.)

Industry has such a little investment in UNIX, that it can afford to
scrap 99% and move to a new open operating system.  (I heard this from
someone inside one of the OSF Funding Memeber companies.)

The industry doesn't really care about the universities for the most
part right now, because they just want a standard that INDUSTRY can live
with.  They're going to ignore the universities for a while and once
industry knows what it's doing, then, and only then, they'll bring the
universities back up to speed.  This could be in a year or two.

What does all this mean?  Well, it means that working on UNIX
systems such as SunOS, Mt Xinu's BSD and MSD releases, Xenix, and SVR4
will be almost a complete waste.  OSF is completely rewriting their
system so that there will be no possibility for a law suit from any of
the UI members.  All the libraries will be redesigned.  Too much will
change too quickly for folks outside of OSF to plan ahead.

{ IBM, DEC, HP, (APOLLO), Apple, and other large vendors have announced or 
will announce that they will be dropping their own U*NX operating systems in
favor of OSF.  It's more cost effective for them to pool their money by
donating it to OSF rather than to develop their own operating systems. 
Their current systems will be supported, but expect no new versions
in a year or so. }

As was you mentioned and in another posting to sci.v-w, we must use/create
standards.  If everyone uses the POSIX.1 standard, then the software
systems will be portable to the new operating systems which will be
released in a year or so (inluding UI's UNIX).

Yes, we should get *something* running and not wait for the future
operating systems, but we have to be careful.

{ I don't think that anyone has to worry about their investment in
hardware of UI member companies (such as Sun), because someone will port
OSF/1 to it.  Especially Sun4 systems... with the Sparc processor being
licensed to non-Sun hardware developers. }

MAYBE, we should look to DOS systems too to develop a few prototype systems.
After all, some of the machines are powerful enough, and with
distributed processing capabilities added, power is not a problem.
Industry does have too large of an investment in DOS systems to ignore
them.  We will probabily see DOS shells supported by OSF implementations,
and that will make porting/upgrading easier.


>2) it seems to me that the most plausible way to go about building
>  a GOOD VR is going to be in an extremely modular fashion.  The best
>  development tools we have at present for truly modular programming
>  are object-oriented ones.  I believe that if we could get a large 
>  base of programmers to agree on a fundamental idea of object-oriented
>  VR then we'd have a much easier time of it..  Better yet, lets all
>  agree to use an industry standard that's also realitively cheap, like
>  C++.

Yes, use C++!
The software development tools on DOS systems are much cheaper than on
multi-user systems.  For example, TurboC++ (a decent compiler) costs only 
$100-$150, and some C++ compilers for SunOS start at $500.  GNU's g++ is
not bug-free enough to be considered.  Maybe next year.
AutoCAD will make a great CAD VR-prototyping front-end to anything developed, 
and it's on lots of platforms including DOS, Macs, SunOS, Ultrix, etc.

No matter what is decided here, students will use whatever they can get
their hands on.  That will include the DOS systems in some places...

-Daniel
--
Pezely@cis.udel.edu;      728 Bent Lane, Newark, DE 19711; 302/368-5931
Comp Sci Lab, Smith Hall, U of Delaware, Newark, DE 19716; 302/451-6339