[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

PD status v.s. DARPA?

Thanks so much for the extremely thorough account of the public domain
history of CMU Lisp.

Until 1987, Moore and I always distributed our prover ``without
copyright license, or fee''.  We even said so in our 1988 book ``A
Computational Logic,'' which is still the users manual for our prover.
ONR and NSF, the main sponsors, never objected at all.  We used a
requirement in the NSF grants manual (now changed by the way) to
justify giving it away, on the grounds that we had to do it.

But in 1987, we used some DARPA money, for the first time, just to
convert the code from Zetalisp to Common Lisp.  We made it clear to
DARPA that we were making the results available by anonymous ftp.  I
insisted on this.  About July of 1988, after the aforementioned book
was already in press, we started getting considerable pressure from
Scherlis at DARPA to put a license and copyright on the code and to
not distribute it via anonymous ftp.  (He never suggested we not
distribute it to whomever we wanted.)  I was never party to the
discussions with DARPA.  It was J Moore and Don Good who told me of
the pressures.  Eventually, Moore negotiated a compromise with
Scherlis to which I agreed.  The compromise was to put a license on
it, but a license that mainly only requires that Clinc be kept
notified of who has copies.  It remains available by anonymous ftp.
You can see the license at the top of the file /pub/nqthm/basis.lisp,
available by anonymous ftp from cli.com.

I suspect that Scherlis is still somewhat sore at me or even
Computational Logic over this.

I am not sure what the fundamental concern is.  But I think that there
is a fear within DARPA that it will get in trouble if it is known (by
Congress, I guess) that some software DARPA has paid for is being used
outside the U.S.A.  I think that some in DARPA regard DARPA's future
justification for existence is as the guiding ship for an U.S.
industrial policy, comparable to the Japanese MITI.

Another possibility is that DARPA quite reasonably would just like to
know who is using the stuff it has paid for, so it can get credit for
what it has funded.  I think that the only reason that there is a
license on KCL is that the Japanese bureaucracy wants to keep track of
who has it out of vanity, and insisted upon that to Nakajima.  DARPA
actually mentioned the KCL model to justify having a license for our

Thanks so much for making CMU Common Lisp public domain and available
by anonymous ftp.