Re: Comments on draft-ietf-secsh-x509-00

[Henrick Hellström <henrick%streamsec.se@localhost>]

Bill Sommerfeld wrote:

> I would have hoped to see:
>
> 1) expected relationship(s), if any between the certificate Subject 
> and/or subjectAltName fields and the identity of the server or user 
> which owns the certificate.  (This of course opens up the "Naming is 
> Hard" discussion.)


I mostly agree with these points. Without these there seems to be little
point in using X.509 certificates at all.

I can only see a point in using X.509 certificates with SSH, in case
secure distribution of public keys between the server and the clients is
not feasible.

In the case of naming, what would really make sense would be a
relationship between:

a) On one hand the URI and version string of the server, and on the
other hand the subject and/or subjectAltName of the server certificate
b) On one hand the username of the client, and on the other hand the
subject and/or subjectAltName of the client certificate.

The presumption of the secsh wg ought to be that the URI of the server
and the username of the client user already identifies these entities.

The draft should not go into details about the theory of names. If any,
such discussions should occur in the X.509 documents and/or Analytical
Philosophy 101.

> 2) text regarding KeyUsage and ExtendedKeyUsage.


This is application specific, so it should go into the secsh-x509 spec.


> 3) text regarding recommendations for certificate revocation checks.
>
> 4) discussion of how to handle certificate chains.


These two points are already covered by PKIX documents. I don't think
there are any secsh specific considerations here.


> 5) more text about the use of certificates for user authentication or
> a claim that it's entirely out of scope for the document..


Agreed.



--
Henrick Hellström
www.streamsec.com