Re: [Cfrg] KCipher-2

[Joachim Strömbergson <[e-mail]>]

Aloha!

On 2012-12-07 20:00 , Jon Callas wrote:
> Yes, put the table in. It's a long-standing frustration of mine that
> crypto documents seem to think that part of the fun is to guess about
> implementation until you finally get it right. A sort of
> gameification, I suppose, that makes a spec a choose-your-own
> adventure. This gets boring, fast. The vast majority of us
> implementors want to get it done, fast and accurate. Anything that
> helps that is a Good Thing.

(Slight OT)

+100

At the absolute minimum add test vectors. Adding a clean, well 
documented reference implementation that uses std types (stdint.h) is 
makes it even better and if you also include internal states during 
processing of at least one test vector you make the implementers life 
_much_ easier.

The siphash specification includes good test vectors and a great walk 
through that made the implementation pretty comfortable. The C reference 
code is, shall we say, terse. Also it really helps if one defines 
operation symbols before using them in the spec...

https://131002.net/siphash/

-- 
Med vänlig hälsning, Yours

Joachim Strömbergson - Alltid i harmonisk svängning.
========================================================================