Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!cmcl2!yale!leichter
From: leichter@yale.UUCP (Jerry Leichter)
Newsgroups: sci.crypt
Subject: Re: This is *stupid*
Message-ID: <4725@yale-celray.yale.UUCP>
Date: Tue, 2-Dec-86 09:04:46 EST
Article-I.D.: yale-cel.4725
Posted: Tue Dec  2 09:04:46 1986
Date-Received: Tue, 2-Dec-86 19:55:54 EST
References: <12246@watnot.UUCP>
Reply-To: leichter@yale-celray.UUCP (Jerry Leichter)
Distribution: net
Organization: Yale University, New Haven, CT
Lines: 49


The referenced article - which I'm certain will draw a large number of replies
from people happy to see their worst suspicions confirmed - is a load of non
sense.

The NSA does not protect DES.  It never has.  It certified DES as a crytographic
standard, and that certification is what is about to run out.

The NSA does not use DES to encrypt ANYTHING.  In fact, it is explictly ILLEGAL
to transmit government-classified information using only DES for encryption.
DES has NEVER been acceptable for such use.

"Unauthorized breaking of DES", whatever that might mean, is not treason.  In
fact, I rather doubt it, in and of itself, is a violation of any law whatsoever.
(The work involved MIGHT be considered classified - crytography and certain
aspects of nuclear physics having to do with bombs are in the special position
that information about them can be considered classified even if developed inde-
pendently - but even that's unlikely - the US government does not use DES for
sensitive information.)  There is plenty of open literature on methods for
attacking DES (most of which seem to indicate that it is, in fact, a pretty
strong cipher - while the practicality of brute-force attacks is known, and
there are some little hints here and there of POSSIBLE weaknesses, all attempts
at analytical attacks published so far indicate that a lot of correct decisions
went into the design of DES; variations have fallen to analytical attack, but
not DES itself.)

I find it extremely doubtful that if, indeed, the VideoCipher system has been
broken, that it's been broken by breaking DES.  A lot of very good people have
failed to break DES; I doubt a bunch of hardware hackers could do better.  That
does NOT mean that M/A-COM's USE of DES is correct; it may be that the particu-
lar way they use DES was not well thought out, and has been broken.

M/A-COM's warning about it being "possibly treasoness" to ship their boxes out
of the US is nonsense - scare tactics.  It MIGHT be illegal under various
Commerce Department regulations previously discussed in this group.  (Since it
is a decoder only, rather than a full ciphering/deciphering system only, even
that isn't clear.)

It would NOT be illegal, under those regulations, to ship decoder boxes INTO
the US.  Nor would it be illegal to sell them WITHIN the US.  These regulations
apply only to shipping FROM the US TO foreign countries.

The illegalities involved in producing and selling such boxes have nothing to
do with cryptography; they involve the same laws that make it illegal to sell
cable descramblers for simple "invert the sync pulse" encoding schemes.  That
is, what's ultimately involved is some sort of "theft of service", and the
producer/seller is prosecuted for being involved in some sort of conspiracy to
steal services.
							-- Jerry