RE: Proposed Ballots for Issue Groups 6, 7, 8, 9

["Edwards, Nigel" <Nigel_Edwards@hp.com>]

> 
> >>>>> "EN" == Edwards, Nigel <Nigel_Edwards@hp.com> writes:
> 
>     EN> I agree with the sentiments expressed
>     EN> [UC-8-05:AtomicAssertions].  I think SAML assertions should be
>     EN> atomic. I think managing valid signatures over assertion
>     EN> fragments is an unnecessary complexity.
> 
>     EN> However, I think an intermediary might also in some cases
>     EN> legitimately remove an atomic assertion, provided no signature
>     EN> was invalidated by doing so.  An example is pointed out in
>     EN> [UC-8-03:IntermediaryDelete]. In this case the intermediary
>     EN> does it to make a purchase order anonymous, once it has
>     EN> validated that purchase order. I would be happy to see a
>     EN> modified version of [UC-8-03:IntermediaryDelete] that dealt
>     EN> with atomic assertions.
> 
> So, there are a couple of options here. One is that the intermediary
> simply drops the identifying authn assertion altogether (making any
> authz decision assertions invalid, as far as I can tell). Or, it
> replaces the authn assertion with one of its own.
> 
> Both make sense to me -- but it seems hard to state them as
> requirements.
> 

Either option will work. If authz assertions are bound to principal
identities
and not authn assertions, removing authn assertions, does not necessarily
invalidate authz assertions.

I am not arguing for a new requirement, merely suggesting that the submitter
of the use case, modify it.

Nigel.