From: "Perry E. Metzger" <perry@piermont.com>
Date: Thu, 20 Jun 1996 07:42:54 +0800
To: Black Unicorn <unicorn@schloss.li>
Subject: Re: SafE Mail Corporation
In-Reply-To: <Pine.SUN.3.93.960618154407.20055A-100000@polaris>
Message-ID: <199606191542.LAA25691@jekyll.piermont.com>
MIME-Version: 1.0
Content-Type: text/plain



Black Unicorn writes:
> Calling a given cypher "uncrackable" is simple fiction or ignorance.

I will point out for the benefit of all that technically there is one
cipher that is unbreakable if it is properly implemented -- the one
time pad. However, Black Unicorn's point is generally correct --
information theory dictates that any system other than a one time pad
can in theory be broken by brute force once you have ciphertext longer
than the unicity distance, which is typically quite short. Unicity
distance depends only on the redundancy of the language and the key
length in bits (well, technically, the base two log of the number
possible keys, but they are usually the same for conventional
cryptosystems -- it would make a difference for stuff like RSA but
since no one ever really cares about the unicity distance in practice
since that sort of brute force search is uninteresting...)

Perry