Re: New hash contest by NIST, similair to AES competition

Mike Amling <spamonly@xxxxxxxxxxx> writes:
the (possibly known) input data as keys for the block cipher, and
generic block ciphers are not assumed secure when used that way.
Davies-Meyer does, but the others don't.

It looks like they all do, though I should have said the block cipher
keys are derived from the input data. The input data is in general
not used directly. Is there a security proof for Miyaguchi-Preneel?
My guess is there can't be one without stronger assumptions on the
block cipher than that it's a PRP.

Anyway, these constructions seem to me to be going in the wrong
direction, since hash functions should be faster than block ciphers
rather than slower.
.

Relevant Pages

  • Re: RARC C++ code
    ... Well, when I wrote Caesarion, my goal was to find something secure. ... I chose ARC4 in preference to Block Ciphers (which are complicated ... And when you use a good CSPRNG like ARC4 with my Cipher-Packet ... inspiring. ...
    (sci.crypt)
  • Overly-Secure Encryption Mode
    ... it is not assumed that block ciphers are totally secure. ... Encipher the message using Electronic Code Book mode, ... hash function, nothing much has to be assumed about the security of the ...
    (sci.crypt)
  • Re: How fast computers have made ciphers unbreakable
    ... current block ciphers are believed to be 'secure' in the sense ... keys for security in such a case leads to a false sense of security. ...
    (sci.crypt)