Rotor in Jython?

[brueckd at tbye.com]

On 3 Mar 2002, Paul Rubin wrote:

> > > > Put another way: sorting by key length, AES beats rotors hands down.
> > > > Sorting by computational cost, rotors always win because you can just make
> > > > the keys longer (without having to reimplement your crypto).
> > >
> > > Longer keys are only relevant if breaking the cipher requires searching
> > > for the key.
> >
> > Heheh... that's essentially what cryptanalysis is - finding the key since
> > in most cases the algorithm itself is known. If you're referring to
> > non-brute force key searches, then key length is relevent there as well
> > (unless the algorithm itself is flawed).
>
> Yes, that's the point.  AES appears to be very attack resistant.

No, it's better than DES but it's widely recognized that AES is actually
susceptible to some of the same types of differential cryptanalysis.
Personally I think it's plenty strong, but it's not immune.

> Compared to AES, rotors have a pretty sorry history of falling to
> cryptanalysis.  There's no reason to think brute force is the best way
> to solve a rotor cipher.

But the sorry history of rotors has nothing to do with rotor strength and
everything to do with short/obvious keys, the obvious starting text of the
messages (e.g. "No news to report"), and the limitations of mechanical
implementations. You can *always* make a rotor stronger by increasing the
length of the key - and at no additional computational cost.

-Dave