Fwd: the new SHA-3 candidate Keccak uses Sage during its design

[Dan Drake]

----- Forwarded message from Minh Nguyen <mvngu...@gmail.com> -----
> The winner is Keccak, whose authors used Sage in the design of the
> algorithm.
>
> The news release is at
>
> http://www.nist.gov/itl/csd/sha-100212.cfm
>
> The specification and design of Keccak is at
>
> http://keccak.noekeon.org/Keccak-reference-3.0.pdf
----- End forwarded message -----

I think we should market the hell out of that fact. :) NIST-approve
hash functions are incredibly important to anyone doing anything with
computers today, and Sage had a part in designing the latest and
greatest hash algorithm. Not Maple, not Mathematica, not Matlab, not
Magma.

Perhaps someone could interview the Keccak guys about their use of Sage
for use as promotional material.

Also, maybe we should get an implementation of Keccak into Sage. (At
least until it gets into Python's hashlib.) See
http://keccak.noekeon.org/python_implementation.html.

Dan

--
--- Dan Drake
----- http://math.pugetsound.edu/~ddrake
-------

[Niles Johnson]

Indeed -- they were using some nontrivial algebra in Sage. Note that
they write SAGE, and their bibliography reference is

William Stein et al. ... 2009.

Hopefully their version of Sage isn't really that old! In the article
they mention installing "a SAGE server version 1.4". Can it be?!
Also of note, they mention on page 18,19 (at the pagebreak) a
calculation that was not supported "at the time of writing" -- finding
an inverse to a certain element in a quotient of a polynomial ring, I
think. They do some algebra by hand to simplify the question to one
which their version of Sage can solve, but I wonder if someone who is
familiar with these things can check whether the current version of
Sage really doesn't support their calculation.

-Niles

[Jason Grout]

On 10/3/12 12:07 PM, Dan Drake wrote:
> ----- Forwarded message from Minh Nguyen <mvngu...@gmail.com> -----
>> The winner is Keccak, whose authors used Sage in the design of the
>> algorithm.
>>
>> The news release is at
>>
>> http://www.nist.gov/itl/csd/sha-100212.cfm
>>
>> The specification and design of Keccak is at
>>
>> http://keccak.noekeon.org/Keccak-reference-3.0.pdf
> ----- End forwarded message -----
>
> I think we should market the hell out of that fact. :) NIST-approve
> hash functions are incredibly important to anyone doing anything with
> computers today, and Sage had a part in designing the latest and
> greatest hash algorithm. Not Maple, not Mathematica, not Matlab, not
> Magma.

+1!

>
> Perhaps someone could interview the Keccak guys about their use of Sage
> for use as promotional material.
>

+1!

> Also, maybe we should get an implementation of Keccak into Sage. (At
> least until it gets into Python's hashlib.) See
> http://keccak.noekeon.org/python_implementation.html.

+1!

So summing: +3 (too bad it's not +3! :)

Jason

[Jason Grout]

On 10/3/12 1:43 PM, Niles Johnson wrote:
> Indeed -- they were using some nontrivial algebra in Sage. Note that
> they write SAGE, and their bibliography reference is
>
> William Stein et al. ... 2009.
>
> Hopefully their version of Sage isn't really that old!

The algorithms were submitted in 2008
(http://en.wikipedia.org/wiki/NIST_hash_function_competition#Process).
Sage 1.4 was released 2 years before that---it's possible that that was
when they really started working on the algorithm.

Thanks,

Jason