Not to bicker, but the resources needed to use a database of all
possible passwords even with alphanumerics and salted is very finite
-- albeit large.
OpenBSD's blowfish passwords have 128-bits of salt. A table of all 8
character (lower-case only) alphanumeric passwords would require
2^128 *
(26+10)^8 ~= 9.6*10^50 entries. Being ``very finite'' is
irrelevant at
this order of magnitude.
The term used earlier was nearly infinite, I used very finite because
it is bounded -- which infinities are not. There are as you know
multiple infinite sets that have no common members.
Just don't want people to think that they are safe as is not an NP-
complete problem. It is an NP-hard problem however.
You are aware NP-complete problems are, by definition, reducible to
NP-hard problems, right? In other words, NP-hard problems are
``harder'' than NP-complete ones.
I should have properly stated that it is not an NP-complete problem
but an NP one. NP-complete problems are the most difficult complexity
problems.
CU
