[Cryptography] IETF discussion on new ECC curves.

[Viktor Dukhovni]

On Fri, Aug 08, 2014 at 01:53:08PM -0400, Richard Outerbridge wrote:

> The more the merrier, if there were also an agreed upon way to tell
> bad randomly generated curves from good ones.

Given that there is and likely cannot be such a mechanism, and that
random curves have noticeably bad performance.  And may take
considerable time to generate, this approach seems unlikely.

We already a few more good new curves than we may need, plus all
the old ones whose reputation is now in doubt, but are needed for
interoperability.

-- 
	Viktor.