[uf-discuss] OpenID

[Dr. Ernie Prabhakar]

Hi Christopher,

On Feb 20, 2007, at 7:00 PM, Christopher St John wrote:
> Check out the 2.0 spec. There are some changes:
>
>  http://openid.net/specs/openid-authentication-2_0-11.html
>
> It introduces XRI into the mix (think brand new, non-DNS
> naming system) Wouldn't expect it to work so well in an <a>

While I do share your concerns to some extent, I'm a bit more of an  
optimist.  While I personally find XRI's wacky, if they're  
interpretable as the "path" behind some XRI Proxy Resolvers, they're  
not much different than any "normal" URL:

http://openid.net/specs/openid-authentication-2_0-11.html

>> If the identifier is an XRI, [XRI_Resolution_2.0] (Wachob, G.,  
>> Reed, D., Chasen, L., Tan, W., and S. Churchill, “Extensible  
>> Resource Identifier (XRI) Resolution V2.0 - Working Draft 10,” .)  
>> will yield an XRDS document that contains the necessary  
>> information. It should also be noted that Relying Parties can take  
>> advantage of XRI Proxy Resolvers, such as the one provided by  
>> XDI.org at http://www.xri.net. This will remove the need for the  
>> RPs to perform XRI Resolution locally.

> I take an "extraordinary claims require extraordinary proof"
> approach to new internet-scale naming systems, and remain
> something of a skeptic. Certainly 2.0 is no longer paving the
> cow paths. YMMV, etc.

Yeah, it is an extra step, and obviously they won't work as HTML  
anchors, so I wouldn't want to use them.  But, the XRI folks -- and  
the SAML folks -- have been working at this for a long time, so  
there's always a chance they know something we don't, so I'm glad  
OpenID 2.0 is opening up to let them "into the fold."  As long as  
they don't add too much complexity into the core, open extensibility  
seems the way to go.

-- Ernie P.