Splinter Unix?

Thu May 19 00:55:29 AEST 1988

In article <556 at n8emr.UUCP>, lwv at n8emr.UUCP (Larry W. Virden) writes:
> 
> Sigh!  For those of you who are as behind as I typically am, let me clue you
> in on a couple interesting articles which were passed on to me by a friend
> here at work (Hi Tom!).
> 
> Digital News, May 16, 1988 Front page:
> DEC and IBM Ponder Alternate to AT&T Unix.  Olsen says counter to Unix is 
> possible.
> 
> Infoworld, I dont have the date but would guess this or last week - front page:
> Alliance to Push AIX-Based Unix.
> 	IBM, DEC, Apollo, and Hewlett-Packard are set to announce the formation of a 
> 	consortium called the Open Software Foundation to develop an alternate
> 	Unix standard build around IBM's AIX operating system, according to 
> 	sources close to the companies involved.
> 	The consortium will be formally announced on Tuesday...
> 
> The articles go on to describe this secret alliance coming out of the Hamilton
> Groups complaints against Sun and AT&T.       
> 
> What do you folks think of this "grand" idea of the creation of a third standard 
> (previously System V and BSD could be considered standards in my estimation)?
> What impact will this have on the Posix effort?  On development of portable
> code?
> 
> Comments, flames, etc?

As an employee of the Data Systems Group of AT&T I must admit a bit of
bias on this topic.  The following comments are my own personal opinion
as an interested observer of the computer business for many years.
This opinion may bear no resemblance to the AT&T Corporate position.

I think it is clear that DEC and IBM are not interested in establishing
a new standard.  Rather, what they would like to do is destroy UNIX as
a viable competitor to their own proprietary products.  Before UNIX
began to emerge as a real threat to their fiefdoms, DEC and IBM pretty
well had their customers locked in to them.  If you had DEC gear, you
ran VMS and depended on DEC for your next generation of hardware.
Similarly, if you were an IBM customer you ran MVS and depended on IBM
(or Amdahl) for your next CPU upgrade.  Granted, I am grossly
oversimplifying here, but I think the flavor of what I am saying is
about right.

Then UNIX made its appearance in a couple of dialects.  Few machines
supported it, and little commercial software was available for it.  It
was essentially an academic playtoy.  AT&T promoted it, and a number of
software houses developed products for it, so it gradually became a
viable product in the commercial marketplace.  Also, many of the
academics who enjoyed playing with UNIX in school became employed
programmers who wanted to buy UNIX systems to do real work with.
Finally, AT&T and Sun got together to merge the two main dialects of
UNIX into a single product.  Now there is a threat of a third real
alternative to the DEC and IBM proprietary systems -- a commercially
supported UNIX which can be run on many different vendors hardware.
This is what DEC and IBM would like to destroy.

As far as POSIX goes, I hope that something useful can come out of that
effort.  However, having observed such bodies in action in the past,
what usually comes out, long after it is needed, is a standard that is
so nebulous and watered down that it is next to useless.  The standards
that have been established as defacto standards tend to be commercial
products (MSDOS, CP/M, etc.) rather than the output of international
standards bodies.  For this reason I have hope that the AT&T/Sun
efforts will really generate an industry standard that software houses
can develop their products for without worrying about the NEXT dialect
that might spring up (like AIX).

As I said earlier, as an AT&T employee I realize that I am biased.  The
personal opinions above, however, are based on over 20 years of being
in the computer industry and having worked with equipment from many
vendors (ranging from a Royal-McBee LGP-30 through Amdahl's biggest
mainframes).

				Rich Strebendt
				...!ihnp4![iwsl6|ihlpe|ihaxa]!res