Re: State machine hypothesis...

[Nick Pelling <incoming@xxxxxxxxxxxxxxxxx>]

Hi Gabriel,

> I don't know what is item 28 (I am off line at the moment) but if you
> want the text of the letters, send me a message and I will email it
> to you. They used to be online somewhere, but they aren't
> anymore.

These were online as part of Glen Claston's site on theglobe.com, which 
recently closed.

> Have you also read the mail archives? You can get them
> from Jim R's site. Well worth reading.

I'm working through them (specifically to find unanswered challenges), but 
it's taking time. :-/

> Are you thinking on the lines of a Porta wheel kind of encoding?

My theory du jour would look something like:-
(1) polyalphabetic, but with a non-obvious state switching mechanism.
(2) (space) is a character, not a space
(3) ends of lines padded with steganographic nulls (perhaps EVA m is involved)
(4) symbols are sorted by ranking within states for a low entropy output stream

It's the non-obvious state switching mechanism that makes it non-crackable 
using normal polyalpha means. Physical manipulations on Steve Ekwall's 
folding key (like "rotate it 180 degrees", etc) translate into the kind of 
state change I'm talking about.

As far as testing it goes: I'm currently thinking about code to map out 
symbol-to-symbol transitions within each state, as a way of 
visualising/identifying state transitions.

Cheers, .....Nick Pelling.....