From: George A. Gleason <gg@well.sf.ca.us>
Date: Sat, 14 Nov 92 04:11:21 PST
To: hughes@soda.berkeley.edu
Subject: Re:  Rander box
Message-ID: <199211141210.AA21995@well.sf.ca.us>
MIME-Version: 1.0
Content-Type: text/plain


Re Eric's quote of me agreeing with him that OTPs are "expensive to make
relative to other forms of security."  A point of clarification seems in
order.

I do agree that OTPs are more expensive and less convenient to use than
PKSs.  However, I also believe that the public interest would *best* be
served by having *many* different kinds of cyphers available, including
OTPs, PKSs, and various conventional cyphers, historic cyphers with
relatively little current security value (for educational purposes) and so
on.  The main advantages of OTPs are provable absolute security and the fact
that the basic technique is so straightforward that it probably could never
be banned and put out of circulation.   The time may come when we *need*
OTPs, and we ought to have them ready beforehand, and have them in use in
appropriate situations long before any crisis comes (to gain operational
experience which could lead to improvements).  

With regard to PGP, I am not sure what the copyright status is on that one;
and if there is any doubt about it, the govt could screw a lot of people to
the wall on copyright-related charges if they so chose.  I would like it
very very much if PGP was free & clear public domain.  The last thing we
need is for the first warning of tyrrany to be a wave of hardware seizures
on the grounds of having unauthorised copies of copyrighted material.  Now I
may be off base on this point, but the key here is the idea that many
different kinds of cyphers, like many different varieties of plants and
animals, make for a robust ecosystem which can't be wiped out by one plague.

-gg