[Cfrg] A draft merging rpgecc and thecurve25519function.

[Adam Langley <[e-mail]>]

Since everyone is agreed to within an isogeny of curve25519 at ~128
bits, and since there's no performance reason to pick one isogeny over
another (see my mail from yesterday), I've created an outline for what
currently appears (to me) to be the only way that we might reach
agreement.

It takes the generation procedure from draft-black-rpgecc-01,
generates the resulting curve with p=2^255-19 and then ends up with
curve25519 by pointing out that there's no difference in security
between isogenies and motivates it by compatibility with existing
practice.

Then it includes draft-turner-thecurve25519function in order to nail
down the wire-format and describe how to perform ECDH.

The resulting agglomeration is at
https://cdn.rawgit.com/agl/cfrgcurve/master/cfrgcurve.xml (requires
XSLT support in the browser) and https://github.com/agl/cfrgcurve.
(Although, until there's clarity on whether the outline is viable, the
details in the draft are unimportant.)

I have not listed the authors of the two source documents as authors
yet because that might suggest that they support the result. Instead
I've made the sources clear in section 1.

It does not suggest a signature scheme, despite several people
suggesting that it would be required in recent days, because I don't
think that we're at that point yet.


Cheers

AGL

-- 
Adam Langley [e-mail] https://www.imperialviolet.org