Re: [TLS] Will CAs decide server signing algorithms in TLS 1.2?

["Wan-Teh Chang" <wtchang at gmail.com>]

On 5/4/07, Pasi.Eronen at nokia.com <Pasi.Eronen at nokia.com> wrote:
> Hi Simon,
>
> This issue was discussed in Prague, and I think the conclusion
> was that the current text needs to be changed. In other words:
> if the client tells it supports RSA+SHA256, the server can
> use it even if the CA used RSA+SHA1 to sign the server cert

How does the client tells the server that it supports RSA+SHA256?
By overloading the "Cert Hash Types" hello extension or adding a
new extension specifically for this purpose?

There is a corresponding issue for the client's signature in the
CertificateVerify message.

Another related issue is the longer DSA variants defined in Draft
FIPS 186-3, which aren't hardcoded to use SHA-1.

Wan-Teh

_______________________________________________
TLS mailing list
TLS at lists.ietf.org
https://www1.ietf.org/mailman/listinfo/tls