Re: [OAUTH-WG] Error Encoding: Conclusion

[Hannes Tschofenig <[e-mail]>]

Hi Mike, 

thank you for compiling the text. It looks good to me. I have not seen anyone from the working group screaming either. 

Eran, can you incorporate these changes into the next draft version?

Ciao
Hannes

On May 30, 2012, at 2:10 AM, Mike Jones wrote:

> I’ve made another set of updates to a copy of Core -26 to address the questions raised by Eran and David below (attached).
>  
> An unrelated change that you should probably pick up, Eran is adding this to the <front> section, so that the heading shows that the draft is a product of the “OAuth Working Group” rather than the “Network Working Group”:
>     <area>Security</area>
>     <workgroup>OAuth Working Group</workgroup>
>  
> One change I didn’t make, but that should be considered, is to delete the reference to OASIS.saml-core-2.0-os, since it is used by no <xref> in the document.
>  
> The new proposed text for Section 7.2 follows:
>  
> 7.2.  Error Response
>  
>    If a resource access request fails, the resource server SHOULD inform
>    the client of the error.  While the specific error responses possible
>    and methods for transmitting those errors when using any particular
>    access token type are beyond the scope of this specification, any
>    "error" code values defined for use with OAuth resource access
>    methods MUST be registered (following the procedures in
>    Section 11.4).
>  
>    Specifically, when the OAuth resource access method uses an "error"
>    result parameter to return an error code value that indicates the
>    resource access error encountered, then these error code values MUST
>    be registered.  Values for these "error" codes MUST NOT include
>    characters outside the set %x20-21 / %x23-5B / %x5D-7E. When an
>    "error" code value is registered for use by an OAuth resource access
>    method, should that same code already be registered for use by
>    another OAuth resource access method or at a different OAuth error
>    usage location, then the meaning of that error code value in in the
>    new registration MUST be consistent with the its meaning in prior
>    registrations.
>  
>    The OAuth resource access error registration requirement applies only
>    to "error" code values and not to other means of returning error
>    indications, including HTTP status codes, or other error-related
>    result parameters, such as "error_description", "error_uri", or other
>    kinds of error status return methods that may be employed by the
>    resource access method.  There is no requirement that OAuth resource
>    access methods employ an "error" parameter.
>  
> Hopefully incorporating these changes will enable us to close the remaining DISCUSS issues on both the Core and Bearer drafts.
>  
>                                                                 Thanks all,
>                                                                 -- Mike
>  
>  
> From: Eran Hammer [mailto:[e-mail]] 
> Sent: Wednesday, May 23, 2012 11:45 PM
> To: David Recordon; Mike Jones; Hannes Tschofenig
> Cc: [e-mail] WG
> Subject: RE: [OAUTH-WG] Error Encoding: Conclusion
>  
> With the exception of section 7.2, the changes look reasonable and will be applied in the next revision.
>  
> The new section 7.2 is confusion and does not explain the new registry. The section introduces a new requirement to register ‘any error codes defined for use with OAuth resource access methods’. This requirement is too vague.
>  
> I have no clue how to (for example) apply this text to the MAC draft. Adding to David’s list below:
>  
> * Should the HTTP status codes used by the MAC spec as currently written be registered (since no guidance is given to the use of an error parameter)?
> * Does this introduce a requirement to add an error parameter?
> * Does the parameter need to / should be called ‘error’?
> * What about future methods in which errors are not simply expressed in the form of a fixes string?
>  
> EH
>  
>  
> From: David Recordon [mailto:[e-mail]] 
> Sent: Wednesday, May 23, 2012 11:38 PM
> To: Mike Jones; Hannes Tschofenig; Eran Hammer
> Cc: [e-mail] WG
> Subject: Re: [OAUTH-WG] Error Encoding: Conclusion
>  
> Honestly still trying to fully wrap my head around what's going on here since it seems far more complex than the threads are alluding to. In any case, does Mike's text address what Eran brought up as needed in the thread Hannes referenced or is Eran wrong?
>  
> The core spec currently provides full guidance and definition for error extensibility. Extending the registry's scope means the need for non-trivial new text that:
>  
> * explains the process of adding new errors for endpoints not defined by this specification,
> * finds a common ground for value restrictions beyond what is already listed,
> * guide authors of future HTTP authentication schemes meant for use with OAuth (e.g. MAC) for their requirements for using the error registry, and
> * address the very likely scenario of the same error code carrying different meanings in different endpoints, or an extension that adds a location to a code already defined elsewhere - something very likely to happen if you cross the two very different domains (OAuth endpoints, Protected resource endpoints). This requires changing the entire structure of the registry to create separate records for each code/location pair.
>  
> Thanks,
> --David
> 
> On Wed, May 23, 2012 at 10:22 PM, Mike Jones <[e-mail]> wrote:
> Thanks Hannes.  In the interest of hopefully completing the edits to remove the DISCUSS issues for the Bearer and Core specs in the next few days so that we can send the docs to the RFC editors, I'd like to propose specific language for the Core spec to address both of the consensus call issue resolutions.  After there's consensus on the specific text for Core, it will be easy for us to add a reference in Bearer to the language in Core for the error syntax restrictions and to use the OAuth errors registry.  I'll do that in parallel with the discussions on the proposed core language changes.
> 
>  
> 
> A summary of the changes I made in response to the consensus call conclusions are:
> 
> ·        Add syntax restrictions for “error”, “error_description”, and “error_uri” from Bearer to Core
> 
> ·        Add section 7.2 about error responses from resource access requests
> 
> ·        Add “resource access error response” to the category of OAuth errors that can be registered
> 
>  
> 
> Additional editorial changes that I made as I encountered issues in the document were:
> 
> ·        Updated out of date references, especially the draft-hardt-oauth-01 reference, which contained an invalid link
> 
> ·        Added Derek Atkins to the list of chairs
> 
> ·        Added Yaron Goland’s middle initial Y. (since he prefers to include it in publications)
> 
> ·        Replaced use of the deprecated <appendix> element, which prevented the spec from building with strict checking, with a <section> element in the <back> section (which creates an appendix)
> 
>  
> 
> To make it easy to incorporate these changes into the document and so the proposed changes are unambiguous, I produced an edited version of Core -26 containing these changes.  The xml, txt, and html versions are attached to facilitate review.  Pertinent diffs from the .txt version follow.
> 
>  
> 
>                                                             Cheers,
> 
>                                                             -- Mike
> 
>  
> 
> 683c683,684
> 
> <    notation of [RFC5234].
> 
> ---
> 
> >    notation of [RFC5234].  Additionally, the rule URI-Reference is
> 
> >    included from Uniform Resource Identifier (URI) [RFC3986].
> 
> 1441c1441,1442
> 
> <          REQUIRED.  A single error code from the following:
> 
> ---
> 
> >          REQUIRED.  A single ASCII [USASCII] error code from the
> 
> >          following:
> 
> 1474a1475,1476
> 
> >          Values for the "error" parameter MUST NOT include characters
> 
> >          outside the set %x20-21 / %x23-5B / %x5D-7E.
> 
> 1476c1478
> 
> <          OPTIONAL.  A human-readable UTF-8 encoded text providing
> 
> ---
> 
> >          OPTIONAL.  A human-readable ASCII [USASCII] text providing
> 
> 1478a1481,1482
> 
> >          Values for the "error_description" parameter MUST NOT include
> 
> >          characters outside the set %x20-21 / %x23-5B / %x5D-7E.
> 
> 1482a1487,1489
> 
> >          Values for the "error_uri" parameter MUST conform to the URI-
> 
> >          Reference syntax, and thus MUST NOT include characters outside
> 
> >          the set %x21 / %x23-5B / %x5D-7E.
> 
> 1840c1840,1841
> 
> <          REQUIRED.  A single error code from the following:
> 
> ---
> 
> >          REQUIRED.  A single ASCII [USASCII] error code from the
> 
> >          following:
> 
> 1873a1874,1875
> 
> >          Values for the "error" parameter MUST NOT include characters
> 
> >          outside the set %x20-21 / %x23-5B / %x5D-7E.
> 
> 1875c1877
> 
> <          OPTIONAL.  A human-readable UTF-8 encoded text providing
> 
> ---
> 
> >          OPTIONAL.  A human-readable ASCII [USASCII] text providing
> 
> 1877a1880,1881
> 
> >          Values for the "error_description" parameter MUST NOT include
> 
> >          characters outside the set %x20-21 / %x23-5B / %x5D-7E.
> 
> 1881a1886,1888
> 
> >          Values for the "error_uri" parameter MUST conform to the URI-
> 
> >          Reference syntax, and thus MUST NOT include characters outside
> 
> >          the set %x21 / %x23-5B / %x5D-7E.
> 
> <          REQUIRED.  A single error code from the following:
> 
> ---
> 
> >          REQUIRED.  A single ASCII [USASCII] error code from the
> 
> >          following:
> 
> 2325a2326,2327
> 
> >          Values for the "error" parameter MUST NOT include characters
> 
> >          outside the set %x20-21 / %x23-5B / %x5D-7E.
> 
> 2327c2329
> 
> <          OPTIONAL.  A human-readable UTF-8 encoded text providing
> 
> ---
> 
> >          OPTIONAL.  A human-readable ASCII [USASCII] text providing
> 
> 2329a2332,2333
> 
> >          Values for the "error_description" parameter MUST NOT include
> 
> >          characters outside the set %x20-21 / %x23-5B / %x5D-7E.
> 
> 2333a2338,2340
> 
> >          Values for the "error_uri" parameter MUST conform to the URI-
> 
> >          Reference syntax, and thus MUST NOT include characters outside
> 
> >          the set %x21 / %x23-5B / %x5D-7E.
> 
> 2450c2460,2468
> 
> <    The method in which the client utilized the access token to
> 
> ---
> 
> >    The method in which the client utilizes the access token to
> 
> 2479c2489
> 
> <      Authorization: Bearer 7Fjfp0ZBr1KtDRbnfVdmIw
> 
> ---
> 
> >      Authorization: Bearer mF_9.B5f-4.1JqM
> 
> 2503a2514,2533
> 
> >
> 
> > 7.2.  Error Response
> 
> >
> 
> >    If a resource access request fails, the resource server SHOULD inform
> 
> >    the client of the error.  While the specific error responses possible
> 
> >    and methods for transmitting those errors when using any particular
> 
> >    access token type are beyond the scope of this specification, any
> 
> >    error codes defined for use with OAuth resource access methods MUST
> 
> >    be registered (following the procedures in Section 11.4).
> 
> >
> 
> >
> 
> 2602,2603c2624,2626
> 
> <    (Section 4.2.2.1), or the token error response (Section 5.2), such
> 
> <    error codes MAY be defined.
> 
> ---
> 
> >    (Section 4.2.2.1), the token error response (Section 5.2), or the
> 
> >    resource access error response (Section 7.2), such error codes MAY be
> 
> >    defined.
> 
> 3444c3484,3485
> 
> <       (Section 4.2.2.1), or token error response (Section 5.2).
> 
> ---
> 
> >       (Section 4.2.2.1), token error response (Section 5.2), or resource
> 
> >       access error response (Section 7.2).
> 
> 3596a3554,3557
> 
> >    [USASCII]  American National Standards Institute, "Coded Character
> 
> >               Set -- 7-bit American Standard Code for Information
> 
> >               Interchange", ANSI X3.4, 1986.
> 
> >
> 
> 3611,3612c3572,3573
> 
> <               OAuth 2.0", draft-ietf-oauth-saml2-bearer-08 (work in
> 
> <               progress), August 2011.
> 
> ---
> 
> >               OAuth 2.0", draft-ietf-oauth-saml2-bearer-12 (work in
> 
> >               progress), May 2012.
> 
> 3616,3617c3577,3579
> 
> <               Protocol: Bearer Tokens", draft-ietf-oauth-v2-bearer-08
> 
> <               (work in progress), July 2011.
> 
> ---
> 
> >               Authorization Protocol: Bearer Tokens",
> 
> >               draft-ietf-oauth-v2-bearer-19 (work in progress),
> 
> >               April 2012.
> 
> 3620,3623c3589,3591
> 
> <               Hammer-Lahav, E., Barth, A., and B. Adida, "HTTP
> 
> <               Authentication: MAC Access Authentication",
> 
> <               draft-ietf-oauth-v2-http-mac-00 (work in progress),
> 
> <               May 2011.
> 
> ---
> 
> >               Hammer-Lahav, E., "HTTP Authentication: MAC Access
> 
> >               Authentication", draft-ietf-oauth-v2-http-mac-01 (work in
> 
> >               progress), February 2012.
> 
> 3626c3594
> 
> <               Lodderstedt, T., McGloin, M., and P. Hunt, "OAuth 2.0
> 
> ---
> 
> >               McGloin, M., Hunt, P., and T. Lodderstedt, "OAuth 2.0
> 
> 3628,3629c3596,3597
> 
> <               draft-ietf-oauth-v2-threatmodel-00 (work in progress),
> 
> <               July 2011.
> 
> ---
> 
> >               draft-ietf-oauth-v2-threatmodel-02 (work in progress),
> 
> >               February 2012.
> 
> 3468,3546d3503
> 
> <    Brian Eaton, Yaron Goland, Dick Hardt, and Allen Tom.
> 
> 3639c3609,3639
> 
> >    Brian Eaton, Yaron Y. Goland, Dick Hardt, and Allen Tom.
> 
> 3468,3546d3503
> 
> <    Yaron Goland, Brent Goldman, Kristoffer Gronowski, Justin Hart,
> 
> 3644,3645c3644,3656
> 
> >    Yaron Y. Goland, Brent Goldman, Kristoffer Gronowski, Justin Hart,
> 
> 3468,3546d3503
> 
> <    This document was produced under the chairmanship of Blaine Cook,
> 
> <    Peter Saint-Andre, Hannes Tschofenig, and Barry Leiba.  The area
> 
> <    directors included Lisa Dusseault, Peter Saint-Andre, and Stephen
> 
> <    Farrell.
> 
> 3646a3658,3661
> 
> >    This document was produced under the chairmanship of Blaine Cook,
> 
> >    Peter Saint-Andre, Hannes Tschofenig, Barry Leiba, and Derek Atkins.
> 
> >    The area directors included Lisa Dusseault, Peter Saint-Andre, and
> 
> >    Stephen Farrell.
> 
>  
> 
> -----Original Message-----
> From: [e-mail] [mailto:[e-mail]] On Behalf Of Hannes Tschofenig
> Sent: Wednesday, May 23, 2012 11:27 AM
> To: [e-mail] WG
> Subject: [OAUTH-WG] Error Encoding: Conclusion
> 
>  
> 
> Hi all,
> 
>  
> 
> on May 10th we called for consensus on an open issue regarding the error encoding. Here is the link to the call:
> 
> http://www.ietf.org/mail-archive/web/oauth/current/msg08994.html
> 
>  
> 
> Thank you all for the feedback. The conclusion of the consensus call was to harmonize the encoding between the two specifications by incorporating the restrictions from the bearer specification into the base specification. The error encoding will go into the core specification and the bearer specification will reference it.
> 
>  
> 
> Ciao
> 
> Hannes & Derek
> 
>  
> 
> _______________________________________________
> 
> OAuth mailing list
> 
> [e-mail]
> 
> https://www.ietf.org/mailman/listinfo/oauth
> 
>  
> 
> 
> _______________________________________________
> OAuth mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/oauth
> 
>  
> <draft-ietf-oauth-v2-26+mbj-2.xml><draft-ietf-oauth-v2-26+mbj-2.txt><draft-ietf-oauth-v2-26+mbj-2.html>