UNIX System III and 'cpio'

[utzoo!decvax!ucbvax!menlo70!sri-unix!trujillo at LBL-UNIX]

From: trujillo at LBL-UNIX (Gary S. Trujillo)

     Maybe the reason that System III tapes come from Western in cpio format
is (as per a rumor I heard from a reputable source) that System III is really
a derivative of PWB.

     I would like to stimulate some discussion about the possible implications
for the future evolution of UNIX brought about by the introduction of a
version thereof which may, in a number of ways, be non-upwardly-compatible
with Version 7.  One area of non-compatibility, as I understand it, is that
various bits returned by ioctl() have significantly different meanings in
the two systems, and that the tty drivers bear little resemblence to one
another.  Further, there is something about cbreak mode not working quite
right under Sys III.

     At least as important as the above, though somewhat related, I imagine,
is the fact that the good people at Berkeley (Bill Joy, et. al.) have no
intention of backtracking to get compatible with System III, so that we
can expect future releases of their stuff to be based on the v7 and 32v
systems they started with.  The VAX version of System III does not do paging,
though I understand they're working on it.  It seems possible that the Berkeley
folks have a sufficiently great head start, though, that the Western Electric 
offering may be inferior for quite some time.

     Though there appears to be a screen editor in the works which will
presumably be released with some future incarnation of System III vintage,
it does not promise to have the functionality of the Berkeley editor.
(I know that the question of editors is a religious question, and I am
only stating my particular preference.  However, the generalized scheme
of terminal-independent screen handling provided by /etc/termcap cannot,
I feel, be so easily dismissed as religion.)

     There is much more that can be said on the subject.  I do realize that
some of what is being provided in System III is good stuff, particularly
SCCS (further evidence for the SysIII/PWB connection).  I only feel that
people should not get stuck with a "pig-in-a-poke" if it can possibly be
helped.

	Gary Trujillo