From: nobody@REPLAY.COM (Anonymous)
Date: Tue, 29 Jul 1997 10:07:19 +0800
To: cypherpunks@toad.com
Subject: Re: NSA leak (fwd)
Message-ID: <199707290150.DAA04489@basement.replay.com>
MIME-Version: 1.0
Content-Type: text/plain



Ian Goldberg wrote:
> >To: fight-censorship@vorlon.mit.edu
> >Subject: NSA leak
> >
> >WASHINGTON (AP) - In a rare moment of openness bordering on glibness, 
> > >senior official at the super-secret National Security Agency was overheard
> > >at a White House press conference concerning current bans on the export of
> > >enryption technology saying, "It would not take any twelve times the age of
> > >the universe to decrypt a 128-bit message.  Thirty-three minutes is more
> > >like it."

> Uh-huh.  Unless the Administration has granted a secret Executive Order
> repealing the Laws of Physics for the NSA, the above statement, if true,
> would imply one of the following things:
>
> 1. The NSA has a reversible computing machine with at least 2^128*128 bits
>    = 5.44*10^39 bytes = 4.95*10^27 TB of memory.
> 2. Their cracker changes the state of 2^128 bits in 33 minutes.
> 3. They have a quantum computer, or some alien technology, or something
>    else we know pretty much nothing about.

  This above is true when analyzed within the boundaries of current
cryptographical theory, but has about as much meaning as would the
discounting of Einstein's work by an old-school physicist.
  There was a rather bizarre post to the list a few months ago which
layed out a conceptual schematic of new research being done in the
area of encryption. Although the post was bizarre, the information
it provided was not. As a matter of fact, it seemed to be wider in
scope and more advanced in its application than the level currently
achieved by NSA and related covert government organizations.
  {And although the 33 minutes quote may lean toward bragging, it
is not off by much.}

  Research currently being done in the encryption field by those who
do *not* tell contains a smattering of new mathematical theory and
a significant quantity of analysis which is a close cousin to traffic
analysis. The result is that a person using _all_ of the security
and encryption capabilities of a strong crypto product is likely to
have secure communications, but if they neglect to perform even one
of the _minor_ security steps possible, then there is a crack in
their communication which can be expanded to a size large enough
to drive a tank through.
  LEA's are not battling to squelch the use of crypto that they
cannot readily break--they are battling to slow the growth of
strong encryption in order to maintain their current ability to
break encryption which is commonly thought to be secure.

D r . R o b e r t s
> ~~~~~~~~~~~~~~~~~~~