On Wed, 2 Aug 2000, Bo Elkjaer wrote:
>United States Patent 6,097,812
>Friedman August 1, 2000
>
>
>Cryptographic system
>Abstract
>
>The crytographic system automatically and continuously changes the cipher
>equivalents representing plaintext characters so as to prevent any
>periodicity in the relationship. The system has a series of juxtaposed,
>rotatable, connection changing mechanisms to provide a large number of
>alternative paths for the passage of an electric current corresponding to
>a message character.
Dude, this is a patent on the rotor cipher machine. The Axis used
them through WWII - There were similar designs in production in the
US and europe for commercial traffic through the fifties.
More to the point, there are now algorithms for breaking it -- pretty
easily, actually -- and the patent doesn't cover the identical
algorithm implemented as software.
So what we're looking at here is a patent on an obsolete piece of
cipher hardware implementing an insecure cipher, which does not
cover the software implementations currently in use in basic crypto
histories and courses.
They couldn't turn down the patent, because it *was* original work
and it *did* have real value -- but at the same time, they couldn't
grant the patent *at the time* because military crypto and the
fate of the "free world" rested on keeping it secret and patent
papers are public docments.
So what we're seeing is that the NSA has finally noticed they have
no reason to hold up the process any more and they've released it
-- with a 17-year delay - to the public domain.
I dunno if Mr. Freidman is still alive, or whether he has any heirs,
but if so it appears they now own a patent on an "invention" which is
now a museum peice. I doubt anyone will be rushing to license it.
Ray
