Other Directory Sites: SeekWonder | Directory Owners Forum
The following archive was created by hippie-mail 7.98617-22 on Fri Aug 21 1998 - 17:20:26 ADT
bram (bram@gawth.com)
Thu, 16 Jul 1998 01:09:11 -0700 (PDT)
On Wed, 15 Jul 1998, John Kelsey wrote:
> Right. I will raise one point with this scheme, though: You
> actually lose a small amount of entropy here, since you can't use the
> output from this to go back to the input. I don't see this as being
> important with any real buffer size, but it's at least a theoretical
> complaint.
This raises an issue I've been wondering about for a while: how much
entropy is lost by recursive hashing? I gather that the number of possible
resulting values is about half after the first hashing, and less after the
second, but tends to hit a plateau after a couple of rounds as the chances
of a collision among the remaining possible values diminishes. Is this
correct?
-Bram
The following archive was created by hippie-mail 7.98617-22 on Fri Aug 21 1998 - 17:20:26 ADT