[Openid-specs-ab] WG Meeting notes 25th April 2025
Andy Barlow
0xandybarlow at gmail.com
Thu Apr 24 23:31:00 UTC 2025
Hi all, please find my notes from todays WG call, I hope I didn't miss
anything important - please correct if I have missed anything you wanted
recording!
Andy
Attendees:
Michael Jones, George Fletcher, Aaron Parecki, Marcus Almgren, Andy Barlow,
John Melati, Filip Skokan, Edmund, Chris Filips, Brian Campbell (briefly)
------------------------------
1. Introductions & Housekeeping
-
Call opened with participant introductions.
-
Catch-up on recent events.
-
Dietary requirements reminders were noted.
-
Agenda review
Announcement:
-
Mike shared that the OAuth 2.0 Protected Resource Metadata spec (
https://datatracker.ietf.org/doc/rfc9728/) has been published.
-
Congratulations to Aaron on publishing his first specification — with
much more to come!
------------------------------
2. OpenID Connect Claims Aggregation
-
The spec was highlighted in an issue as being currently dormant (last
update in 2021).
-
An objection to the dormant classification from Nat was noted, and Nat's
proposed changes were merged.
-
Action: Michael Jones will contact the authors to confirm readiness of
draft publication.
Further notes from a later second pass discussion:
-
Mike suggested running the draft through OpenID’s spec tooling:
-
Publishing Tooling Guide
<https://openid.net/wg/resources/publishing-specifications/>
-
Tooling GitHub Repo
<https://github.com/openid/publication/blob/main/README.md>
-
Aaron spoke about how the tool performs checks on PRs before merging.
-
Edmund asked if possible to add the tooling to CI.
-
Mike suggested consulting Mark H.
-
Action request: Edmunds use of the tooling is a good opportunity to
provide feedback on the tooling.
-
Edmund noted a new PR simplifies the aggregation draft (reverting some
VC language, simplifying, reusing userinfo/claims parameters, adding
verification text).
-
Mike provided some context on the previous removal of VC text.
------------------------------
3. OpenID Connect Relying Party Metadata Choices
-
Discussion focused on GitHub PR #6
<https://github.com/openid/rp-metadata-choices/pull/6>.
-
PR introduces implementation considerations.
-
Filip requested some changes in the PR, which led to discussion
regarding normative language and the location of the guidance within the
text.
-
George (via chat): Suggested keeping non-conflicting content, even if
redundant.
-
Filip: asked for more prescriptive (normative) guidance.
-
Mike: Proposed accepting Filip’s changes and continuing discussion
within the WG.
------------------------------
4. Federation Wallet Architectures
-
An editorial pass PR is still open.
-
Action: Efforts will be made to move the PR forward.
------------------------------
5. OpenID FederationIssue #202 – Issuer Identifier Normalization
-
From Michael Fraser: should an issuer of google.com and google.com/ be
treated the same?
-
Mike clarifies that yes, they should resolve to the same well-known
endpoint.
-
Mention of possible need for normalization (e.g., lowercasing).
-
Filip: Raised issues with normalization when URLs include paths (such as
google.com/a-path/ and spoke about two common implementation approaches:
1.
Strict equality check (===).
2.
Normalize then compare.
Issue #194 – Trust Mark ID Field Clarification
-
John noted difficulties modeling the trust_mark_id field as it doesn’t
always behave like a URL.
-
Mike: Acknowledged it’s not always a traditional URL (in terms of being
treated like an issuer (e.g., can represent a class etc) and acknowledges
the confusion this could create currently.
-
Action: Mike to clarify the interpretation of trust mark in the issue
thread and will attempt to get consensus but notes that spec is nearing
final.
Issue #192 – Test Update
-
Marcus: Will update the relevant test (noted that the test was an early
revision).
-
Mike: Will comment on the described edge case.
Issue #193 – Entity Declaration Behavior
-
Mike: Entities should be declared even when metadata is inherited.
-
Chris: Asked about interoperability in cases with multiple entity types
— how to resolve those dynamically?
-
Mike: Resolution is dynamic. Clients should only implement types they
understand.
-
Action: Mike to follow up with implementers in Sweden for additional
clarity and has made a note to do so.
------------------------------
6. Native SSO Spec Status
-
George: No real updates to share with WG.
-
Action: Will talk to Mark H. about migrating the spec from Bitbucket to
GitHub to support better collaboration.
-
Mike agreed that GitHub is more collaborative.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openid.net/pipermail/openid-specs-ab/attachments/20250425/674d6a4c/attachment-0001.htm>
More information about the Openid-specs-ab
mailing list