Re: Question about hashing algorithms

bigzaphod@gmail.com wrote:

> I was just pondering a distributed file transfer system (like
> bittorrent) which breaks down the file into many small chunks. Each
> chunk is verified with a hash so that you can get the chunk from any
> number of possible sources (and prove it is correct).

Are you trying to protect against an intelligent adversary, or
only against the possibility of two chunks accidentally having
the same hash? If the latter, what is the size of the universe
of chunks among which hash collisions would cause trouble?

-- 
Peter Pearson
To get my email address, substitute:
nowhere -> spamcop, invalid -> net