Conversation
|
Hi Waldo, That is interesting. I hadn't thought of testing that kind of multi-protocol multi-federation approach. A few comments:
So yeah, I do see the value in having tests about subject identifier generation for OPs, but we might not need active dual federation stacks and topologies to support that, not sure. Probably overlaps with the policies WG as well. |
|
Hi Phil, Thanks for your thoughts. I agree that it would be better scoped to being able to generate `suitable/appropriate/saml-identical' subject identifiers. I've updated the docco c9eaf49 The main use-case I was considering was that since IdPs will support both protocols, it would be reasonable to ensure that SPs should compatible with the generated identifers from both stacks, so that when the time comes migrating from SAML -> OpenID Federation is simpler. SP Implementation would vary wildly, but having the generation customizable from the OP side would be enough to satisfy this test case |
|
Having tests for identifiers seems like a good idea to me. The new test in your updated pull request looks great. |
Given that the Shibboleth IdP will be running with this plugin: https://shibboleth.atlassian.net/wiki/spaces/IDPPLUGINS/pages/4500914216/OPFederationWIP
It would be good to add a section of SAML/OpenID Federation interoperability. I have started with the a simple? ( maybe not ) case where a relying party is both registered to SAML and OpenID Federation. The end user should appear as the same user, regardless of auth mechanism.
cc @philsmart for additional thoughts