From: kentborg@world.std.com (Kent Borg)
Date: Sun, 17 Jul 94 00:19:34 PDT
To: cypherpunks@toad.com
Subject: Re:  Hashed hash
Message-ID: <199407170719.AA03993@world.std.com>
MIME-Version: 1.0
Content-Type: text/plain


Ben.Goren@asu.edu foolishly says:
>I'm planning on implementing the "cryptographic protection of databases"

And wonders about the hash being too fast to compute, that a
brute-force traversal of the database would be too easy.  The idea is
then to hash a bunch of times to burn CPU cycles, but what if the hash
is a group, extra hashing could be reversed quickly.  (Did I get that
right?)


Well, as the LOUD proponent of making secret keys s-l-o-w-e-r to
decrypt, I have thought about this a bit, and have a suggestion:

Hash once, then do a zillion encryptions of the hash with a non-group
cypher like DES.  

Another idea (something I have thought less about): send every legit
user of the database a custom version with the parts encrypted with
that user's public key--and do the trick mailing list companies use,
scatter some dummy info in the list.  When a dummy (not just me) gets
a junk mailing, go beat up on the user who's copy had to have supplied
the junk.  Not perfect: combinations of dummies are needed in case the
junk mailer cracks multiple copies (multiple work) and then trys to
sift unique dummies that way.  Another problem: it is expensive to
monitor the dummies.  (1990's biz opportunity?, the monitoring of data
that no one is supposed to have.)


-kb, the Kent who doesn't want to be thought of as only a card player


--
Kent Borg                                                  +1 (617) 776-6899
kentborg@world.std.com                                
kentborg@aol.com                                      
          Proud to claim 35:00 hours of TV viewing so far in 1994!