[noise] Noise "conventions" vs the Noise core

[Jason A. Donenfeld]

On Wed, Jul 29, 2015 at 4:54 AM, Trevor Perrin <trevp at trevp.net> wrote:
>
> Presumably a good Noise library would give users access to the
> length/type convention, and the padding convention, but also allow
> users to disable them for raw access to the Noise core.
>
> What do people think?

That sounds good to me, so long as it's clear that by "convention" you
mean "not necessary to be compliant noise".

Where do ciphersuites fall in terms of core vs convention? I've
swapped out sha256(data) with blake2b(data, 32) and hmac-sha256(key,
data) with blake2b(key, data, 32), for example. (The speed
improvements are considerable.)