[Openid-specs-ab] ServerMTI in Messages

Mike Jones Michael.Jones at microsoft.com
Wed Jan 30 20:23:14 UTC 2013


http://openid.net/specs/openid-connect-messages-1_0.html#sigenc.alg says "Servers SHOULD support none and RS256."  So it's already not MTI.

To Torsten's first question, the UserInfo endpoint is RECOMMENDED for static configurations as well.  The only reason it's not REQURIED is some static OPs may choose not to expose this information at all.  (They would only provide SSO but no claims.)

				-- Mike

-----Original Message-----
From: openid-specs-ab-bounces at lists.openid.net [mailto:openid-specs-ab-bounces at lists.openid.net] On Behalf Of John Bradley
Sent: Wednesday, January 30, 2013 4:38 AM
To: Torsten Lodderstedt
Cc: openid-specs-ab at lists.openid.net
Subject: Re: [Openid-specs-ab] ServerMTI in Messages

The problem with the user info endpoint is that it can't be supported by self issued providers so always making it MTI won't work.

Just because a provider doesn't support discovery and registration, I wouldn't expect them  to not have a user info endpoint.

The default signing for request objects should be RS256.  I would prefer not to make RS256 MTI as it may be legitimate for a server to support only ECDH for security reasons.

I would be OK with saying the server needs to support one of RS256, RS512, ES512.  So if they don't want to support the default they have to support one of the more secure ones.

The problem is what happens when someone only wants to support ECDSA with SHA3?   That is the problem with MTI.

John B.

On 2013-01-30, at 8:25 AM, Torsten Lodderstedt <torsten at lodderstedt.net> wrote:

> Hi all,
> 
> just re-read the MTI section of messages (8.1. specifically), which caused two questions:
> 1) Assuming the scope values "profile", "email", "address" and "phone" are required for all server implementations, how is a non-dynamic OpenID provider supposed to expose this data? I'm asking since the UserInfo endpoint is MTI for dynamic OpenID providers, only.
> 2) Which are the default signing algorithms for request objects? Discovery says "Servers SHOULD support none and RS256".
> 
> regards,
> Torsten.
> 
> _______________________________________________
> Openid-specs-ab mailing list
> Openid-specs-ab at lists.openid.net
> http://lists.openid.net/mailman/listinfo/openid-specs-ab

_______________________________________________
Openid-specs-ab mailing list
Openid-specs-ab at lists.openid.net
http://lists.openid.net/mailman/listinfo/openid-specs-ab


More information about the Openid-specs-ab mailing list