When is UNIX UNIX

[Armando P. Stettner]

This is a reposting as news got scrambled recently on decvax.
		aps.


>From its beginning, UNIX was in a constant state of change.  I do not
feel that any single version or release typifies UNIX.  I feel that
UNIX represents a philosophy of implementation and architecture more so
than its common services.  However, I feel that there is a set of
characteristics which is required in order consider a thing a member of
the UNIX lineage: open, read, write, seek, close, fork, exec, getuid,
setuid, a real hierarchical file system, a shell, etc.

What other capabilities a "UNIX System" is given is not important to
the question of whether or not the system is UNIX.  What maybe more
important is how a capability is implemented and how it is presented to
a user [process?].  Adding VMS type RMS capabilities (file access
methods) to UNIX is not necessarily a bad idea.  Doing so will not
change whether or not it is UNIX.  However, if the implementation means
that all files have an RMS file type and files can not be reproduced
*exactly* by simply copying their bytes because the system knows more
about their structure to begin with or if the implementation means that
the output of one program can not be piped into the input of another,
it probably isn't UNIX anymore.

While I'm on it, 4.2BSD isn't UNIX; neither is 4.1BSD, System V or
System III or Version 7 (although I think it is the "real standard").
Even the /usr/group "standard" is not UNIX.  These are all UNIX.  They
are all part of the evolution of UNIX.  As I have said before, I
believe that UNIX is evolution.  To nail down the idea UNIX to some
specific implementation is not good.  UNIX is a direction or set of
directions.  For me, it is a way of being.  (How's that for
existentialism.)  If I can get real-time capabilities or the ability to
share resources across several machines running UNIX, more power to me,
a user.

			Armando P. Stettner
			UNIX User.