I reviewed the section 5 text that talks about the naming issue, so here's
my opinion based on that reading.

TL;DR, don't waste time overloading something already there, just define
what you need.


While I think there might be some tactical ways to do this on the SP side
along the lines of what I said in the Jabber room last Thursday, I think
ultimately you have a parallel problem on both ends that probably just
needs to be solved consistently.

In both cases, you have an entityID you're trying to evaluate in the
context of a realm, so I think you just need to define a metadata
extension to do that.

As we've discussed, yes this looks like Scope, and in many cases it will
be the same value, but that doesn't mean it's the same thing. That also
doesn't address the SP end.

There are two obvious choices here:

- an explicit extension element
- an entity attribute

I'm completely ambivalent about which you use. Specifying an entity
attribute is less work, but the XML is slightly more verbose in the end.
An advantage that is unlikely to be all that important but YMMV is that
defining a SAML Attribute for this allows the concept of the realm to be
expressed in other contexts in SAML, and using the EntityAttributes
extension means you can (but I doubt you would) actually embed a third
party attestation to the realm information in the form of an actual SAML
Assertion inside the metadata.

On the subject of whether you should/need to do this now, I think there
are other use cases for getting what amounts to a AAA "realm" mapped in
metadata. We should just do it.

-- Scott


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

Reply via email to