On 2-Jun-06, at 9:00 AM, Sam Hartman wrote:

Hi.  As far as I can tell, the use cases draft does an excellent job
of describing all the possible things that could work in this space.

Thanks.

I think it is important to do a much better job of describing which
use cases you actually see as important in the first version.

OK. I agree that's not clear. Perhaps the DIX protocol draft
should reference which use cases it's trying to cover?

I think that many of the use cases that involve communications between
the identity agent and parties who identity claims have been given to
in the past should be ruled out of scope a they hugely complicate the
solution.

Which use cases are those? There's only B14 I think? And, that's
not about claims.

I also think that it is unlikely that Beth will have a single identity
agent.

Doesn't B22 address this?

The user should be able to have as many Identity Agents as
they want... 0 to infinity....

If this is unclear then I need to fix this.

For example if Beth works for the NSA, I doubt they will let
her use the same identity agent for her work persona as for her home
persona.  I think the implications of this for use cases need to be
explored.

Or is this a case I've missed?

I really like the draft though as it allows me to concretely
understand what we're talking about here.

Thank you! Your feedback is most welcome.

John


_______________________________________________
dix mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/dix

Reply via email to