webweaving: more precipitation-induced brain mush.

So if the basic DHM asymmetric arrangement where a sends b a message is (((g to the a)mod p)to the b) mod p , and out pops the shared key, couldn't we try something like:

a picks (some group of keys) as a key group.
b picks (some other group of keys).
All key group values are > 1 and < p. Otherwise, they're unrelated.

Meh. Had something in mind here. This is what happens when I don't write this down exactly when it goes flyin' by.

I think it had something to do with using a separate key for signing, and a third key for additional nonsense. Bleh.