Re: [OAUTH-WG] Assertions: Client authentication for non-token endpoints?

[Hannes Tschofenig <[e-mail]>]

I will leave it to Justin to come up with the right wording.

Ideally, we shouldn't have too many ways for the resource server to
authenticate to the authorization server to avoid interoperability
problems.



On 04/24/2014 02:44 PM, Brian Campbell wrote:
> Perhaps it'd be more appropriate for this introspection endpoint to say
> that it accepts the same client authentication mechanisms as the token
> endpoint rather than describing specific methods itself in the document?
> 
> 
> On Thu, Apr 24, 2014 at 3:09 AM, Hannes Tschofenig
> <[e-mail] <mailto:[e-mail]>> wrote:
> 
>     Hi Brian,
> 
>     does it sound reasonable for you to add text to the token introspection
>     endpoint regarding the use of the JWT bearer assertion for the token
>     introspection endpoint?
> 
>     Ciao
>     Hannes
> 
>     On 04/24/2014 12:58 AM, Brian Campbell wrote:
>     > Just to pile on here - the Assertions draft(s) do define client
>     > assertion authentication only for the token endpoint (and register
>     token
>     > endpoint parameters). But it certainly doesn't preclude it from being
>     > profiled for use elsewhere.
>     >
>     > FWIW we used the token endpoint in our implementation of token
>     > introspection/validation partly because all supported forms of client
>     > authentication come along for free by doing so. My esteemed colleague,
>     > Dr. Paul Madsen, posted a rough draft of what we've implemented in
>     > product a while back:
>     > http://www.ietf.org/mail-archive/web/oauth/current/msg08607.html
>     >
>     >
>     > On Wed, Apr 23, 2014 at 10:49 AM, Justin Richer <[e-mail]
>     <mailto:[e-mail]>
>     > <mailto:[e-mail] <mailto:[e-mail]>>> wrote:
>     >
>     >     For introspection, we really just wanted to say "you can
>     >     authenticate the caller (client or RP) just like you would to the
>     >     token endpoint". So if you've got the means to do that with the
>     >     assertion draft or with client secrets or TLS certs or anything
>     >     else, go for it. I would not read the text of the assertions draft
>     >     as restricting this other use case.
>     >
>     >      -- Justin
>     >
>     >
>     >     On 04/23/2014 12:42 PM, Mike Jones wrote:
>     >
>     >         The assertions draft is only trying to describe how to perform
>     >         assertion-based authentication at the Token Endpoint.  Other
>     >         drafts, such as the introspection draft, could explicitly say
>     >         that this can also be done in the same manner there, but
>     that's
>     >         an extension, and should be specified by the extension
>     draft, if
>     >         appropriate - not in the assertions draft.
>     >
>     >         Justin may have more to say about the applicability or lack of
>     >         it to the introspection draft, but I'm personally not familiar
>     >         with it.
>     >
>     >                                         -- Mike
>     >
>     >         -----Original Message-----
>     >         From: OAuth [mailto:[e-mail]
>     <mailto:[e-mail]>
>     >         <mailto:[e-mail]
>     <mailto:[e-mail]>>__] On Behalf Of Hannes Tschofenig
>     >         Sent: Wednesday, April 23, 2014 5:09 AM
>     >         To: [e-mail] <mailto:[e-mail]>
>     <mailto:[e-mail] <mailto:[e-mail]>>
>     >         Subject: [OAUTH-WG] Assertions: Client authentication for
>     >         non-token endpoints?
>     >
>     >         Hi all,
>     >
>     >         in a discussion about re-using the client authentication
>     part of
>     >         the assertion framework for other specifications currently in
>     >         progress I ran into the following question:
>     >
>     >         Section 6.1 of
>     >         http://tools.ietf.org/html/__draft-ietf-oauth-assertions-15
>     >         <http://tools.ietf.org/html/draft-ietf-oauth-assertions-15>
>     >         talks about the client using the assertion with the **token
>     >         endpoint**.
>     >
>     >         Now, it appears that one cannot use the client authentication
>     >         with other endpoints, such as the introspection endpoint
>     defined in
>     >        
>     http://tools.ietf.org/html/__draft-richer-oauth-__introspection-04#section-2
>     >        
>     <http://tools.ietf.org/html/draft-richer-oauth-introspection-04#section-2>
>     >
>     >         Am I reading too much into Section 6.1 of the assertion draft?
>     >
>     >         Ciao
>     >         Hannes
>     >
>     >         _________________________________________________
>     >         OAuth mailing list
>     >         [e-mail] <mailto:[e-mail]>
>     <mailto:[e-mail] <mailto:[e-mail]>>
>     >         https://www.ietf.org/mailman/__listinfo/oauth
>     >         <https://www.ietf.org/mailman/listinfo/oauth>
>     >
>     >
>     >     _________________________________________________
>     >     OAuth mailing list
>     >     [e-mail] <mailto:[e-mail]> <mailto:[e-mail]
>     <mailto:[e-mail]>>
>     >     https://www.ietf.org/mailman/__listinfo/oauth
>     >     <https://www.ietf.org/mailman/listinfo/oauth>
>     >
>     >
> 
>