Internet Privacy Guaranteed ad (POTP Jr.)

[lmccarth at cs.umass.edu]

Clay writes:
>         Given that N is the length of the message in bits. The number of
> possible combinations of bits is 2^N.  For any message length N > 1,
> 2^N < N^256.  

Uh, nope. 2^N grows asymptotically faster than N^256. Actually, for any 
constants A and B, A^N grows asymptotically faster than N^B. For A=2, B=256,
the crossover happens somewhere before N=4096. 
2^4096 = 2^(16*256) > 2^(12*256) = (2^12)^256 = (4096)^256

If the IPG people are using N=5600 (weird choice) then certainly 
2^5600 > 5600^256, for what little that's worth.

(Ah, my computer science B.S. pays off ;)

-Lewis	"You're always disappointed, nothing seems to keep you high -- drive 
	your bargains, push your papers, win your medals, fuck your strangers;
	don't it leave you on the empty side ?"  (Joni Mitchell, 1972)