Cryptology and knot theory?

[Krister Walfridsson]

On Sat, 15 Jun 1996, Perry E. Metzger wrote:
> I would have my doubts about how interesting the direction could get,
> since knot theory is a dead area. The classification problem was fully
> solved, and after that things got boring...

Well...  I think the theory has become much more interesting after 
the classification, because we know that our problem _can_ be solved,
and our only problem is to do it faster and to get a better understanding
for the subject (there are lots of conjectures which seems simple, but 
whose state is unknown.)
 
I agree that it doesn't look good (since most of our invariants are NP-hard)
but the vassiliev invariants might be used to approximate the other
invariants... (I do not know what has been done in such approximation
theory the last couple of years.)

   /Krister