Long crypo keys (was Re: YA cracker challenge)

Peter G Capek capek@us.ibm.com
Tue, 12 Mar 2002 09:32:28 -0500


Eugene Leitl wrote:

>  A stream cypher with a lot of internal state seems to asymptotically
> approach properties of an one-time pad.

Very important to remember: this statement would only be true if the bits
of internal state are independent,
as they would be with a properly done one-time pad.   The observation that
a long key approaches a one-time
pad may be true, but it ignores the reality of generating, distributing,
and dealing with such a long key, and it's clearly
only true to the extent that key information is used once.   If key bits
are recycled (even with some fancy feedback shift
register to introduce apparent complexity), or (much worse) in conjunction
with multiple messages, it's a big problem.
Read the history of crypto.

              Peter

Peter G. Capek
IBM Thomas J. Watson Research Center
Yorktown Heights, NY   10598-0218
(+1 914) 945-1250       IBM Tieline: 8-862-1250            Fax:  X 4426