[plug] [semi-OT] Public/private keys?

[Mike Holland]

On Sun, 24 Nov 2002, Andrew Furey wrote:

> Hmm. Yes, I understand that side of it (and to answer
> Craig's question, I was thinking of digital signing
> when I said "vice versa").

Yes, I'm curious about that too. I had the impression that digital 
signatures work by using the keys the opposite way around.
  When you generate the key-pair, does it matter which is made public, and 
which kept private? Can you derive one way more easily?

  To answer Andrews original question, yes you can crack one key from the 
other in theory, but if the numbers are big enough, its not practically 
possible. As far as we know that is. You cant actually prove it. Maybe the 
NSA knows an easier way to factorise large numbers, and havnt told us yet.
Its a nice basis for conspiracy theories and movies.

See "Sneakers" http://us.imdb.com/Title?0105435 

> I also assumed that there was no way to reverse it, or
> else PK crypto would be basically useless.

Not no way, it just needs to be too slow.
The only code you cant crack is a one-time pad, where the key is as big as 
the text.

> I was just
> curious as to precisely _how_ it works... although

I hope Marks explanation helped. He lost me with "relatively prime" and
"muliplicative inverse".  Now where did I leave that book on number theory?


-- 
        The origin of Truth is Doubt - Socrates