Second Round Candidates (new NIST hash function)

National Institute of Standards and Technology ^ | July 24, 2009 | NIST

[antiRepublicrat]

Official comments on the Second Round Candidate Algorithms should be submitted using the "Submit Comment" link for the appropriate algorithm. Comments from hash-forum listserv subscribers will also be forwarded to the hash-forum listserv. We will periodically post and update the comments received to the appropriate algorithm.

[antiRepublicrat]

This is for geeks only, but this will have a serious impact for all of us since hashes underlie much of modern computer security. It's about a competition for a new secure hash algorithm to replace the currently-used MD5 and SHA, which are showing to be less and less secure.

I'm happy to see that Bruce Schneier's Skein algorithm made the second cut. The security looks good, and the performance very good. I've been annoyed with the long waits of SHA when hashing large datasets and files.

[rabscuttle385]

-

[taxcontrol]

I have two patents with regards to hash functions. The first provides greater protection to the /etc/password, unix crypto function and the second is a hash collision avoidance means that can be used by any hash function.

[ShadowAce]

[rfp1234]

We are the Knights who say "NIST".

We are the keepers of the secret algorithm.

[zeugma]

Bruce also made the 2nd cut of the AES competition. It would be cool if he makes it on this.

[zeugma]

Too funny. However for a hash to be useful, it can’t be secret. Only fools rely on crypto that is so bad, the details have to be kept secret.

[antiRepublicrat]

Thanks for that info. I’d love it of Bruce won both, that is if his algorithms are the best. He doesn’t have a perfect record, remember MacGuffin, but he’s advanced a lot in the 15 years since.

[antiRepublicrat]

BTW, one of the advantages of Skein is that it is very easy to analyze, understand and implement. It only has three basic operations. IMHO, that makes it easier to find flaws and easier to program proper implementations. I think it would have been out of the competition at round one if it weren’t secure.

[AdmSmith]

This is a very important issue, as SHA-1 was compromised years ago. The question is if the deadline 2Q 2012 gives sufficient time to check the new algorithm? But, the “best hash function” is a moving target depending on breakthroughs in mathematics and hardware. It is not easy.

[antiRepublicrat]

I think three years is enough time for them to be vetted, and the finalists are going to get absolutely hammered for a while by most of the competent cryptographers in the world before the winner is picked.