XRID proposal for YADIS

Peter Davis peter.davis at neustar.biz
Mon Nov 7 17:16:06 PST 2005

On 11/6/2005 12:05 AM, "Drummond Reed" <drummond.reed at cordance.net> wrote:
>> 4) I'm a little queasy about the term "Service" -- for example, a
>> YADIS implementor might want to advertise that it understands
>> credentials of type A, B and C -- and use the capabilities
>> advertisement to do so. Credential types aren't really services ...
>> which is why we called it "capability" in the first place. This isn't
>> major, I just wanted to point it out.
> I don't think the XRI TC is wedded to "Service". It's a legacy of a similar
> element in XRI Resolution 1.0. One version we discussed this week introduced
> the word "Capability", so I can check and see what people think.

Capabilities is a property of the service. When you say 'credentials', what
do you mean?  My first response is 'things I present to an Identity Provider
to authentication to them'.  Is this what you meant?  If so, SAML calls this
an authentication context. In which case, credentials would be mechanisms
supported by the service endpoint.

I think 'service' is pretty well suited for XRID's, as that is really what
they are describing.  Services available for the requests XRI (or in this
case, URI).  

Of course, if my assumption of your meaning of capabilities is different,
than we should continue this thread.


More information about the yadis mailing list