Re: Help needed with a proof...

From: Bartosz Zoltak (X_at_vmpcfunction.com;)
Date: 04/29/04 

Date: Thu, 29 Apr 2004 20:24:48 +0200

David Wagner wrote:
> >If g is the inverse of f, then the combination
> >is the identity function, and not very "secure".
>
> Yes, if you look at the statement of the theorem that I gave,
> such a case is ruled out. The correct theorem requires that f
> be one-way and that g be easy to compute before one can say
something
> about g o f. If g is the inverse of f, then those two preconditions
> cannot be simultaneously achieved.

A cool observtion! Anton's comment still holds for my formulation of
the problem, but with David's one - when we assume the one-wayness of
f -the assumption about independence of f and g is neatly made
redundant! I like it :-)

Bartosz