.       Question: should we explicitly link these to the 'user' and
'machine' TLV definitions proposed by TEAP?

.       Yes we should tie this to the Identity Type code registry from the
TEAP document

 

 

.       Figure out a way to name SAML authorities (e.g., attribute
authorities) to support synchronous requests (e.g., for assertions).

.       I don't believe that this is a current requirement.  We currently
know how to  ask the IdP in a SAML query for attributes.  I cannot think of
any reason why an RP would be able to know what SAML authority to ask a
specific authority for a SAML response in a protocol.  I believe this item
can be not done.

 

.       The document currently only discusses TLS/TCP; also should mention
TLS/UDP

.       NO!!!!!!!! Doctor it hurts when I do this.  DON'T DO THIS!!!!   With
the additional overhead from DTLS, the fragmentation is going to be even
worse that otherwise.

 

.       Include a prescription that "SAML responders SHOULD return a RADIUS
state attribute" to facilitate subsequent use of the user/machine Subject
Confirmation methods

.       YES - and this is trivial

 

.       Clarify text describing use of the SAML AuthNRequest's 'AllowCreate'
attribute

.       YES - Per mailing list conversation  - Specifically what needs to be
said is this only applies if the IdP is going to return an explicit
identifier for the client and that identifier needs to be created.

 

 

Jim

 

 

_______________________________________________
abfab mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/abfab

Reply via email to