Re: [TLS] WG review of draft-ietf-tls-rfc4492bis

[Sean Turner <[e-mail]>]

> On May 4, 2017, at 08:41, Hubert Kario <[e-mail]> wrote:
> 
> On Tuesday, 11 April 2017 15:09:04 CEST Sean Turner wrote:
>> All,
>> 
>> draft-ietf-tls-rfc4492bis has been revised since it left the WG and we agree
>> with Yoav’s statement at the mic in Chicago that the WG should review the
>> changes before we ask Kathleen (our newly appointed AD) to continue
>> progressing the draft.  Please review the differences from the -12 version
>> that went through WGLC and the latest version [0] and let us know by
>> 20170426 whether there is anything that would stop progression of the
>> draft.
> 
> I know I am late with the review, but I'd like to ask two questions:
> 
> 1. In table 2, the "key authorised for use in digital signatures" was 
>    removed.
>    Does that mean that key usage extension in X.509 certificates should be 
>    ignored?

No it does not. There were changes to the table, but s2.2 is still there and that’s where it says “… MUST contain an RSA public key authorized for signing ..."

> 2. Given that RFC7919 is already accepted, standards track document, 
>    shouldn't "NamedCurve" references be renamed to "NamedGroup" (e.g. in 
>    Section 5.5.1.)

It’s either a replace or note the name change (i.e., in my mind this is a style thing).  The 1st paragraph of s5.1.1 notes them;

   RFC 4492 defined 25 different curves in the NamedCurve registry
  (now renamed the "Supported Groups" registry, although the enumeration
  below is still named NamedCurve) for use in TLS.

spt