Bjoern Labitzke <hermit@labitzke.isdn.cs.tu-berlin.de> writes:
> * Perry E. Metzger (perry@piermont.com) [990701 00:44]:
> > Not really. Everyone pretty much knows how many CPU cycles cracking
> > RC5 of a particular depth requires, and frankly, all this does is help 
> > RSA DSI's marketing department. Sorry to be a party pooper, but RC5
> > cracking is damn useless. It was one thing to show DES could be
> > cracked, but showing large key RC5 requires half a year of heavy work
> > from hundreds of thousands of machines isn't exactly a help.
> 
> Well, it IS usefull for some applications. I liked the examples on the
> RC5-mailing list. Consider e.g. the recipe for Coca Cola. Even if it takes
> you a year or two to decrypt it, that would matter.

Are people working on cracking the coca cola keys? No. They're working 
on cracking random keys as "challenges" -- i.e. makework.

The RC5 people know how many keys they can check per second. There is
no challenge here -- you can figure out how long this makework takes
by multiplying. Big Deal.

OTOH, the guys at SETI@HOME have a *real* task. They have enormous
amounts of data to process, and they *really* need the CPU time. This
isn't some "the answer is sitting in a sealed envelope in a vault"
type of situation.

Perry