Re: Enigma machine strenght using a computer

On 30 Dec 2006 23:52:40 GMT, Unruh <unruh-spam@xxxxxxxxxxxxxx> wrote:

"=?iso-8859-1?q?Jean-Fran=E7ois_Michaud?=" <cometaj@xxxxxxxxxxx> writes:


Ignacio aecbi wrote:
The analog these days is point of sales authentication. You can limit
the amount of time an intruder has to steal or break the
authenticating key that secures the transaction.

Maybe, but I'm talking about the algorithm in itself. If fine tuned,

I'm certain it could be fairly efficient and secure.

Enigma is not efficient. To have to do 50 table lookups as suggested in the
original post is not efficient. And the fact that each table is fixed and
assumed known to the enemy makes the procedure much dicier. You have 50
transopision cyphers going with the transopositions stepping by a single
wheel at each step (also known to the attacker). The only thing unkown is
the order of the wheels (lookup trasposition tables) and the initial
settings of them. It sound both very slow and insecure to me.


For higher security needs solutions like IPV6 and modern methods like
closed loop quantum cryptography give better security.

closed loop quantum cryptography is what exactly?

A physical method that reduces the chance of a man in the middle
attack. Fiber implementations appear to use a polarized low threshold
photon signal in a complex tightly time shuttered carrier to prevent
unauthorized access.

But aren't we talking about an idea for an algorithm versus a whole
security scheme here? I'm talking about the specifics of an algorithm
and you broadened the scope by bringing up a communication protocol and
(as far as I know), a cutting edge cloud shoveling method?

Regards
Jean-Francois Michaud

.