Re: [CFRG] compact representation and HPKE

[Stephen Farrell <[e-mail]>]

Hiya,

On 12/02/2021 23:06, Loup Vaillant-David wrote:
>> I need to check (and haven't) but IIRC for an earlier draft
>> I had a problem that the OpenSSL APIs didn't support the
>> compressed form for NIST curves,
>> [...]
>> for me, this isn't a case of encouraging cfrg curves but rather
>> one of enabling simpler and more broad implementation.
> 
> Not quite sure what you meant, but I am extremely wary of making things
> "simpler" by taking on gigantic dependencies such as OpenSSL.

Wasn't that, no. My implementation uses OpenSSL and I didn't
find a way to do what Dan suggests using the obvious higher
level OpenSSL APIs.

In the meantime, Dan tells me offlist that his OpenSSL code
does use such a way, which is great, so I'll take a peek and
get back in a few days, likely to admit he knows more about
all this than me (but I knew that was true before I started:-)

So if there's no real difficulty in implementing what Dan
suggests, I'd be fine with it. If OTOH, it makes it harder
to implement (my position before I take that peek at his
code:-), then I'll continue to be against the change.

Cheers,
S.

> 
> We need to consider the complexity of the whole system, as if we were
> rewriting everything from scratch. Because over time, bit by bit, we
> eventually will.
> 
> Loup.
> 
>