GEANT OIDC plugin

[Cantor, Scott]

> Subject is generated by the same mechanisms as SAML2 nameid is generated with. The configuration is of course 
> separate from SAML2 but the principle is the same.

I think design-wise we're all still chewing on that question. Since ultimately this is still a "claim", and thus an Attribute, I think a case could be made to move that back into the AttributeResolver and handle it like any other attribute in the OIDC space, but just encode it from a Computed/Stored-backed AttributeDefinition, which is the same thing we'll have to do for the pairwise-id SAML Attribute.

-- Scott