[noise] Noise test vector format

[Alex]

On Thu, 12 Nov 2015 11:04:32 -0800
Trevor Perrin <trevp at trevp.net> wrote:

> On Thu, Nov 12, 2015 at 10:31 AM, Alex <alex at centromere.net> wrote:
> > On Thu, 12 Nov 2015 10:25:18 -0800
> > Trevor Perrin <trevp at trevp.net> wrote:  
> >>
> >> I was imagining random bytes are taken sequentially, so as long as
> >> random_i is long enough it would be used for the static, then
> >> ephemeral, and other random fields if we add those.
> >>  
> >
> > But then it's ambiguous (from looking at the format alone) which key
> > comes first (static vs. ephemeral). I think it would be more
> > readable to have four 32-byte values (assuming Curve25519) rather
> > than raw output from a RNG.  
> 
> Maybe, sounds reasonable - if you write this tool first, people will
> probably follow your lead on minor points like this.
> 
> Treor

I've created a sample test vector for Noise_NN. Before I spend time on
the rest of the handshakes, does this look like a reasonable format?:

http://lpaste.net/8887995761745199104

In the specific case of NN the init_static, resp_static, and
resp_ephemeral keys are superfluous given that they are not passed to
Initialize().

Also, it would be awesome if someone could verify this one vector to see
if it is correct.

-- 
Alex