From: peb@PROCASE.COM
Date: Fri, 21 May 93 14:40:43 PDT
To: anton@hydra.unm.edu
Subject: Re: cypto + compression
Message-ID: <9305212138.AA12666@banff.procase.com>
MIME-Version: 1.0
Content-Type: text/plain



>Why not compress this AGAIN

If your compression algorithm is any good, it should *not* be able to 
compress the output a second time. 

Compressing before encryption is vital--it makes brute force and plaintext
attacks much more difficult.

On compression forever:  I read a science fiction short story once where
(not sure of title or author, but it is a classic) a bunch of geniuses 
are ostensibly sent to another planet to "explore", but the people sending
them had a different motive: get them away from Earth and give them time
to dream up cool stuff.  Okay, so they dream up way cool stuff, but have
this problem with transmission bandwidth back to Earth.

Then they figure out that any message can be encoded in prime numbers
like:  2^a * 3^b * 5^c * 7^d...   where a b c d are the character values
(ascii or letter A==0, B==1, etc.).  After a message is encoded, the
result is a *big number*.  This number is not more compact than
the original message, but the clever geniuses flying to Tau Ceti (or 
wherever) figured out how to factor the number down to things like
M^N + P^Q, where the number of bits needed to write down the factorization
was very small, say, 100 bits or so.  THEN, they ship this factorization
back to Earth and save bandwidth and it encodes the whole Encyclopedia
Gallactica. 

This scheme doesn't work because factoring is much harder than using 
other compression techniques.  


Paul E. Baclace
peb@procase.com