🚀 go-pugleaf

pdc@dcs.ed.ac.uk (Paul Crowley) writes:
> Recent research seems 
> to indicate that (1) the key is 56 bits because there's only 56 bits 
> worth of security in the algorithm anyway, and (2) the NSA S-boxes are 
> chosen to be more secure.  If they genuinely don't have the scoop on 
> DES, why should they promote it while suppressing RSA? 

This seems fairly simple to me.  With the ability to build fast custom 
hardware (certainly within the NSA's capabilities :)), a brute-force search 
of the keyspace of DES is *much* more computationally tractable than it is 
for RSA.  That is to say, DES does not represent a strategic barrier to an 
entity with the resources of a government (or these days, of a medium-sized 
corporation).  RSA evidently does, at least at current levels of knowledge.

DES was never billed as anything more than a cipher system that was good 
enough for most applications.  I think it still serves as such, and at this 
point I, at least, am confident that it's exactly what it's advertised to be.  
I am also confident that the NSA and many of their international counterparts 
can routinely read DES-encoded traffic.

Generally speaking, I find that this does not matter to me.  But make no 
mistake, I do feel that legal restraints against the use of strong 
cryptosystems by private citizens should be opposed, for much the same 
reasons that I support the right to keep and bear firearms and other weapons.

It's not the times when I don't need them that concern me--it's the 
occasional times when I do.  The most likely example for me personally is 
likely to be guarding against corporate espionage; in my particular field, 
assuming adversaries do not have a lot of knowledge and computer power is not 
a particularly safe assumption.



Amanda Walker
InterCon Systems Corporation