On the other hand, RFC 2523 is more relevant, Photuris being
the genus that Fireflies belong in...
At 10:32 PM 4/28/99 -0400, Vin McLellan wrote:
> The reference to "Firefly" crypto in 1217 is informative, and --
>given that the NSA's internal development of the FIREFLY protocols goes way
>back -- may have actually been intended as a comment on the Agency's design
>effort. It was, however, something less than the treasonous publication of
>classified type-1 crypto by that noted revolutionary author.
My only involvement in AT&T's STU-III design was helping keep
the typesetter alive while my co-workers did the proposal,
so I can tell you things about Imagen wet-process printers and
won't have to kill you :-) (I may have fixed some troff macros for them as
well.)
The box could do the required 2400 and 4800 baud modes, and ours also
had a 9600-baud mode for better sound and data. The guesses were that
the FIREFLY keying involved random number generator hardware and
Diffie-Hellman, but if I remember right the chip to do it came from the NSA.
The AT&T box had several optional crypto versions, including
a Wimpy Exportable crypto for commercial users, DES, an NSA hardware-based
algorithm,
and maybe something else.
Thanks!
Bill
Bill Stewart, [EMAIL PROTECTED]
PGP Fingerprint D454 E202 CBC8 40BF 3C85 B884 0ABE 4639
