'dbm' hash table software

[Guy Harris]

> AT&T (USG) never got a copy (of "dbm") from Bell, since System III -> V
> -> Vr2 -> \(if doesn't come from V7, but from V6. (I think we've been
> over this before; apologies to those who've heard it before :=) )

If UNIX/TS 1.0 -> PWB/UNIX 2.0 -> UNIX 3.0 -> UNIX 3.0.1 (S3) -> ... came
from V6, how come it has:

	1) The V7 file system

	2) The V7 "stat" structure

	3) The V7 "lseek" system call

	4) Environment variables

	5) The Bourne shell

etc., when V6 had *none* of those?  The USG/USDL UNIX series came from a
UNIX somewhere between what was sent out the door as V6 and what was sent
out the door as V7; it was closer to V7 than to V6, though.  (The fact that
it lacked "dbm" and a few other things is hardly as important as the fact
that it had the V7 features listed above.)

I'm not sure what you mean by "AT&T" and "Bell" here.  It's not relevant,
though; the Research group that did V6, V7, etc. and the USG/USDL that did
S3, S5, etc. were both part of Bell Laboratories, so they *could* have
gotten a copy of "dbm" if they'd wanted to.  The pipeline between Research
and USG/USDL was less than smooth (consider that a bug in "fgrep", fixed in
an addendum to V7, is *still* not fixed in S5!), but it could conceivably
have been made to flow better.
-- 
	Guy Harris
	{ihnp4, decvax, seismo, decwrl, ...}!sun!guy
	guy at sun.com (or guy at sun.arpa)