Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Status
colourYellow
titleDRAFT

...

  • Comment Reference: The GitHub repository issue number of the comment and possibly a reference to a subpart of that issue. All issues relevant to the Public Comment and IPR Review period use both the label "V2.0" and the label "public comment period". The content of all comments has been stored in GitHub.
  • Issue Description: Characterization of the issue in a short phrase. May be different from the issue name in GitHub.
  • Specification Reference(s): A reference such as "Grant Sec n.n" or "FedAuthz Sec n.n", indicating actual sections that were edited. "Grant" refers to User-Managed Access (UMA) 2.0 Grant for OAuth 2.0 Authorization revision 05 and "FedAuthz" refers to Federated Authorization for User-Managed Access (UMA) 2.0 revision 05, the Draft Recommendations under review. Issues relevant to each specification were labeled "grant" and "fedauthz", respectively. Some new sections were added and some sections were rearranged in the course of editing, but only rev 05 section numbers are used so that commenter references will be normative.
  • Editorial/Technical: Whether the comment involves an editorial change (a change to interpretive wording, generally minor) or a technical one (a change to normative language that requires substantive specification change). A label of "editorial" was applied to issues that appeared, at first blush, to be editorial. Note that applying these categories itself requires interpretation, and there is some gray area between them. Implementers seeking to understand changes from the pre-Public Comment period to the final Draft Recommendation and Recommendation stages are encouraged to study all changes carefully.
  • Disposition: The Work Group's conclusion about the action to take in response to the comment. "Commit" links go to specific GitHub commits showing exact specification text changes.
  • Report Out: Whether the commenter submitted comments through the official Public Comment period channel, and requires reporting back of the disposition. Kantara staff should take action on this column.
  • Notes: Context that may be helpful for the Leadership Council.
Comment ReferenceIssue DescriptionSpecification Reference(s)Editorial/ TechnicalDispositionReport OutNotes
#326Improve and reorder definition of permission ticketGrant Sec 1.3EditorialCommit
 

Editorial improvement to a spec definition suggested; discussed by several WG participants and consensus rapidly gained.
#327Rationalize usage of "object" vs "parameter" labelingGrant Sec 3.3.6EditorialCommit
 

Simple editorial wording fix suggested; discussed by several WG participants and consensus rapidly gained.
#328Interpreting how client-contributed scopes are mapped to resources during authorization assessmentGrant Sec 3.3.4EditorialCommit
 

Interpretation issue raised; discussed thoroughly by the WG and an editorial enhancement removing ambiguity adopted by the WG.

#329

Typo: Fix cross-reference internal section targetGrant 7.4.1EditorialCommit
 

Incorrect cross-reference noted; fix applied without WG intervention required.
#330Typo: Fix cross-reference external specification targetFedAuthz Sec 9.2 EditorialNo change
 

Simple editorial correction suggested; fix overcome by events (#334).
#331Typo: Fix reference to item being registeredFedAuthz Sec 9.3EditorialCommit
 

Simple editorial correction suggested; fix applied without WG intervention required.
#332Test whether definitions of PCT are sufficientGrant (various)EditorialNo change
 

Interpretation question raised; WG decided to keep the existing wording.
#333Enhance redirect_user exampleGrant Sec 3.3.6EditorialCommit
 

Simple editorial enhancement requested; fix applied without WG intervention required.
#334Treating token introspection response claims as generic JWT claims in local token validation environmentsFedAuthz Sec 1, FedAuthz Sec 9.2EditorialCommit
 

Issue raised due to lack of implementation experience; discussed thoroughly by the WG and an editorial resolution adopted by the WG involving removal of the IANA registration request section for the JWT claims.

#335a

Distinguish UMA versions of terms from OAuth onesGrant (various), FedAuthz (various)EditorialNo changeYesEditorial improvement requested; editor recommended no change.
#335bSpell out key UMA terms more fullyGrant (various), FedAuthz (various)EditorialCommitYesEditorial improvements requested; small edits applied without WG intervention required.
#335cSeparate out sequence diagrams according to resource owner-side vs requesting party-side interactionsGrant Sec 1.3EditorialCommitYesEditorial improvement to diagram(s) requested; introductory text edits made instead.
#335dSimplify sequence diagrams, e.g. to separate claims pushing from interactive claims gatheringGrant Sec 1.3EditorialCommitYesEditorial improvement to diagram(s) requested; added only clarification to existing single diagram after WG consultation.
#336Clarify that RPT takes a token_type hint of access_tokenFedAuthz Sec 5.1EditorialCommit
 

Editorial improvement requested; discussed by several WG participants and consensus rapidly gained.
#337aRationalize/complete definitions of token introspection response claimsFedAuthz Sec 5.1.1EditorialCommit
 

Clarification requested; WG determined an editorial improvement.
#337bClarify the situation with respect to client types in the UMA grantGrant Sec 3.3.3EditorialCommit
 

Clarification requested; WG determined an editorial improvement.
#337c,dCreate and document a concrete method to require and enable clients to pre-register claims redirection URIsGrant Sec 2, Grant Sec 3.3.2, Grant new Sec 7.3TechnicalCommit
 

Request for new mechanism for dynamic client registration mechanism and clarity; WG agreed. Mechanism requires a registration request to IANA.
#337eTypo: Example that should show a response message is a request messageGrant Sec 3.3.3EditorialCommit
 

Simple editorial correction requested; fix applied without WG intervention required.
#337fClarify how permission requests with multiple permissions in them contribute to set mathGrant Sec 3.3.4EditorialCommit
 

Clarification requested; WG determined an editorial improvement.
 
#337gEnsure authorization servers apply the maximum security checks on permission ticketsGrant (various)EditorialCommit
 

Additional security considerations requested; WG agreed to add a form of security considerations that gives more discretion to the authorization server than was requested.
#338Typo: Example shows the wrong endpoint path componentFedAuthz Sec 3.2.1EditorialCommit
 

Simple typo correction requested; typo fixed without WG intervention required.
#339Clarify whether an array can be used for a request for a single permissionFedAuthz Sec 4.1EditorialCommit
 

Clarification requested; WG confirmed the correct interpretation and clarification text.
#340A unique error should be available for a definitive policy-failed errorGrant Sec 3.3.6, Sec 7.4.1TechnicalCommit, commit, commit, commit
 

Change requested; WG ultimately reintroduced (and renamed) an UMA1 error code that was previously removed: was not_authorized, now called request_denied.
#341The request_submitted error should be allowed to be terminal (no permission ticket)Grant Sec 3.3.6, Sec 5.6TechnicalCommit
 

Change requested; WG made a different change, adding an optional new interval polling hint feature.
#342Security considerations need to be made clearer and more to the point, especially Grant Sec 5.2Grant Sec 5EditorialCommit
 

 Clarification requested; WG determined an editorial improvement.
#343Refer to RFC 6759 token endpoint error codes specificallyGrant Sec 3.3.6EditorialCommit
 

Clarification requested; WG confirmed the correct interpretation and clarification text.
#344Explain what to do if the client presents an invalid or expired claim tokenGrant Sec 3.3.6EditorialCommit
 

Clarification requested; WG confirmed the correct interpretation and clarification text.
#345Explain what to do if the client presents a claim token in a format the authorization server can't handleGrant Sec 3.3.6EditorialCommit
 

Clarification requested; WG confirmed the correct interpretation and clarification text.
#346ClientedRegistered scopes should not be a first-class set math citizenGrant Sec 3.3.4EditorialNo change
 

Clarification requested; commenter decided to close own issue without action
#347The authorization server should be given discretion to determine if it's an error when the client requests a scope it did not pre-register forGrant Sec 3.3.6EditorialCommit
 

Change requested; WG gave the authorization server discretion to report the requested error.
#348On the refresh flow, the authorization server should be given discretion to perform authorization assessmentGrant Sec 3.3.1, Sec 3.6, new Sec 6.1, FedAuthz Sec 1.4.1, Sec 8EditorialCommit
 

Clarification requested; WG confirmed the correct interpretation and clarification text.
#349Explain what to do if the client presents an invalid or expired RPT for upgrading(see above)EditorialCommit
 

Clarification requested; WG confirmed the correct interpretation and clarification text. (See #348 for details.)
#350Explain what error code to return when CandidateGrantedScopes < RequestedScopesGrant Sec 3.3.4EditorialCommit, commitYesClarification requested; WG confirmed the correct interpretation and clarification text.
#351Variety of editorial issues in FedAuthzFedAuthz (various)EditorialCommit
 

Variety of editorial comments, typo corrections, and the like made; implemented without WG intervention required. Note that the original form of the text in Sec 3.2, since corrected, could have led implementers astray, implying that a field was required when it was clear in a different context (Sec 3.2.4) that the field would not appear.
#352The PAT is not needed for the permission and token introspection endpoints, so use client credentials insteadFedAuthz Sec 1.4.1, Sec 1.5EditorialCommit
 

Change requested; WG made some clarifications instead.
#354-1What should the AS do if the resource registration endpoint gets a request message with a bad/broken request body?FedAuthz Sec 3.2TechnicalCommit
Change requested; editor, in discussion with some WG participants, proposed and implemented an HTTP status code and optional UMA-level error code that tracks the design of OAuth and OAuth bearer token error codes.
#354-2The PAT should definitively be a bearer tokenFedAuthz Sec 1.3EditorialCommit
Change requested; editor, in discussion with some WG participants, agreed that this corrected an interoperability oversight in having previously removed a requirement to support bearer PATs in UMA1.