Re: [TLS] Ala Carte Cipher suites - was: DSA should die

[Viktor Dukhovni <[e-mail]>]

On Mon, Apr 13, 2015 at 04:15:39PM -0700, Geoffrey Keating wrote:

> > Why?  There?s nothing invalid about using Curve25519 key exchange
> > with an RSA certificate and GOST symmetric ciphers, so TLS should
> > not provide this. Requiring all three things to be GOST is a policy
> > decision that can and should be part of server (and perhaps client)
> > configuration.
> 
> It seems like that would be an interoperability issue if the server
> and client disagree on the policy to apply.

If a client or server is willing to do non-GOST algorithms along
some axis of the cipher-suite parameter space, it should be willing
to accept mixed results.  It is no less GOST-compliant when using
a mixed suite than when using a suite that has no GOST features.

But your point is valid, and can be taken to imply that policies
which disallow valid combinations of parameters are to be avoided,
if a server selects a combination from the choices offered by the
client, the client SHOULD or MUST accept that combination, or else
why were the choices in question offered by the client?

So a key question is whether policies that rule out various corners
of the product space are legitimately required???

-- 
	Viktor.