> CRCs aren't uniformly distributed though, so this might not mean much :)

Don't the common CRC's in use have a two cycles, one of (2^n - 1)
length and the other pathological cycle with one state?  

So while they aren't strictly uniform (they are missing one of the
possible states), they should be "close enough".

-wolfgang
-- 
We are from the U.N. and we are here to help you.

spider food: http://www.wsrcc.com/baddream/usenet/
(NOTE: The email address above is valid.  Edit it at your own peril.)