[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Federated Authentication Beyond The Web: Problem Statement and Requirements



On 7/9/10 7:30 PM, Hannes Tschofenig wrote:

Hi Hannes,

sorry for the late response.

np


Interesting statement.

I agree that there are other approaches (and probably everyone would
agree with that; we could even list Kerberos).

However, the MOONSHOT BOF is (if I understood it correctly)
constraint to the mentioned constraints.

hmm, afaik it is a federated authentication BoF, not a Moonshot BoF


The usage of OpenID in SASL/GSS-API (like you pointed out) will be
done in KITTEN independently and has different design constraints.

I think there is a bit of misunderstanding, I gave those just as an example because I happen to be familiar with them. I am happy with those drafts being in KITTEN, if only to speed things up. So that is not the point.

My point is that I'd like the BoF to start open-minded. I believe federated authentication for non-Web is a large and interesting area. I would like to make sure that we don't rule out beforehand alternative approaches. I am happy with a charter as outcome that specifies that the WG is going to take the Moonshot approach, but imo that should be the *result* of the BoF, not a *precondition*.

Hope this clarifies the misunderstanding.

Klaas


Ciao Hannes

On 7/6/10 11:15 AM, Hannes Tschofenig wrote:

Hi Hannes,

at the next IETF meeting we are going to have a BOF about
"Federated
Authentication Beyond The Web". In case you have not noticed the
work relates to RADIUS and Diameter.

I wrote this very short problem statement document to explain
the
purpose of the BOF:
http://www.ietf.org/internet-drafts/draft-tschofenig-moonshot-ps-00.txt



Let me know if you find the description useful. Feedback about the BOF
topic would also be appreciated.

I find the description useful, however I would like to challenge
the MUST for RADIUS and/or Diamter. There are a number of
Federated Authentication for applications access protocols out
there, SAML, OpenID and others. RADIUS and Diamter are typically
associated with network access. And while I do see the
attractiveness of marrying the two (and thus leveraging existing
trust fabrics), I wonder why you want to restrict a priori to just
those. As an example draft-cantor-ietf-sasl-saml-ec-00.txt,
draft-lear-ietf-sasl-openid-00, and
draft-wierenga-ietf-sasl-saml-00 specify the use of federated
authentication in a SASL context. And services like eduroam are an
example of the use of just RADIUS to implement federated
authentication for non-web applications. I do understand that it is
not possible nor desirable to take on everything, but let's at
least have this scoping discussion in the BoF.

Klaas

-- to unsubscribe send a message to radiusext-request@ops.ietf.org
with the word 'unsubscribe' in a single line as the message text
body. archive:<http://psg.com/lists/radiusext/>


--
to unsubscribe send a message to radiusext-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/radiusext/>