[OpenID] Interop (was: RE: Conformance and Interop...)

[Guido Sohne]

On 6/8/07, Fen Labalme <fen at 2idi.com> wrote:
> Don't give up on XRI - it's (IMO) a superior identity standard for user
> centric services.

And if it is really superior, then it should be what we use defacto,
not as an add on.

> However, the XRI spec does not empower any particular 'global' namespace, and
> therefore conceivably communities could create their own "global" namespaces
> that resolve locally rather than sending all names out to e.g. the
> http://xri.net proxy for resolution.  While these names may not be globally
> unique and inter-community standards for namespace management (aka peering)
> have not yet been developed, there's a straight-forward and time-tested path
> to managing namespace collisions.  The results provide all the power and
> flexibility of XRI/i-names while creating a somewhat subversive framework for
> user centric digital identity.  Further, peering arrangements between closed
> community "super-nodes" will provide an ideal environment for the development
> of the most important service when people truly control their identity:
> reputation.

If I understand you correctly, we could have an openid:// namespace
which would be a non-commercial, (or more 'open') alternative to the
xri.net proxy? That's not so bad, but it does raise the bar for
implementations significantly, if they are to provide these facilities
natively (i.e not by proxy) ...

> Bottom line: what's important to me is not that I have global name ("=fen")
> that people I'll never know in some_remote_location can use to address me.
> Rather, what I want is the people in the communities I am part of to be able
> to refer to "fen" and know they're talking about the same person.  Names just
> make more sense in a local context.

So what about OpenID 1.x referring to URLs as identifiers and OpenID
2.x referring to XRIs as identifiers? Is the problem that a 2.x would
never take off because a 1.x would kill off its ecosystem?

I'm not for or against any one of these approaches in particular, it
is just that I think they don't belong together ... but what do I
know, really??

-- G.