Tom,

> It depends on the trust relationship the SP has with the various AAs
> in question, but in general, this is a hard problem. How does the SP
> prove to the AA that the user is present and actively involved in the
> transaction? The AA would have to have a fairly liberal attribute
> release policy to hand out user attributes to the SP without some form
> of user consent.

Right, this is an important point. I don't think consent is a showstopper in 
the sense that one can reasonably choose to not care about consent in many 
scenarios. However, I agree that it would be very desirable to have a strong 
story for consent.

ABFAB *may* provide a better approach for consent than that which can be 
achieved through the conventional HTTP-based bindings, because the architecture 
posits an agent (the EAP peer) on the subject's host which could communicate 
directly with any Issuer known by the subject (e.g. the AA uses some kind of 
RPC to request user interaction through this agent). We have previously 
considered the agent as primarily a means of managing user interaction for 
identity selection, but it doesn't seem unreasonable to extend this to consent 
management.

I think we have previously concluded that it would be reasonable to apply this 
kind of approach, but there isn't a concrete proposal on the table at the 
moment. Suggestions welcome!

Josh.

JANET(UK) is a trading name of The JNT Association, a company limited
by guarantee which is registered in England under No. 2881024 
and whose Registered Office is at Lumen House, Library Avenue,
Harwell Oxford, Didcot, Oxfordshire. OX11 0SG

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

Reply via email to