Re: [TLS] Signed messages should be prefixed with a NUL-terminated context string.

[Eric Rescorla <[e-mail]>]

On Tue, Nov 25, 2014 at 11:40 PM, Nikos Mavrogiannopoulos <[e-mail]>
wrote:

> ----- Original Message -----
> > This seems like a good idea. Thanks for raising it.
>
> An additional issue of switching to the new CertificateVerify
> format for client and server, is that privilege separation for
> authentication is not possible any more.
>
> On a previous mail I was asking for that feature to be extended for the
> server as well:
> http://www.ietf.org/mail-archive/web/tls/current/msg13398.html
>
> Was the removal of that protocol feature intentional?
>

Not really. I was aiming for unifying the structures.



> If not I'd suggest the following format for the signatures:
>       struct {
>            digitally-signed struct {
>                opaque client_random[32];
>                opaque server_random[32];
>                opaque handshake_messages[handshake_messages_length];
>            }
>       } CertificateVerify;
>

I've filed https://github.com/tlswg/tls13-spec/issues/97 so this doesn't get
lost.

If someone wants to contribute a PR so that we can have something
concrete to look at that would be even better.

Best
-Ekr