Hello Brian,

On Tue, Aug 11, 2020 at 5:55 PM Brian Campbell <bcampbell=
[email protected]> wrote:

> Hi Francis,
>
> My apologies for the tardy response to this - I was away for some time on
> holiday. But thank you for the review and feedback on the draft. I've tried
> to respond inline below.
>
>
> On Fri, Jul 31, 2020 at 5:01 PM Francis Pouatcha <fpo=
> [email protected]> wrote:
>
>> Bellow is the only remark I found from reviewing the draft draft:
>>
>> 2.1.  Request:
>>
>> requires the parameters "code_challenge" and "code_challenge_method" but
>>
>> https://openid.net/specs/openid-financial-api-part-2-ID2.html#confidential-client
>>  mentions
>> that RFC7636 is not required for confidential clients. I guess those two
>> parameters have to be taken off the mandatory list and pushed to the list
>> below.
>>
>
> The list of parameters in Section 2.1 is qualified with a "basic parameter
> set will typically include" and is definitely not intended to convey a set
> of required parameters. It's just a list of parameters that make up a
> hypothetical typical request.  Perhaps some text in the section or even the
> formatting needs to be adjusted so as to (hopefully) avoid any confusion
> like this that the list somehow conveys normative requirements?
>
>
>
>> - Using jwsreq, non repudiation is provided as request is signed (jws).
>> This section also mentions that the request can be sent as form url
>> encoded (x-www-form-urlencoded). In this case, there is no way
>> to provide non repudiation unless we mention that request can be signed by
>> client using signature methods declared by the AS (AS metadata).
>>
>
>  I am not aware of any signature methods or means of an AS declaring
> support for a signature method in metadata that are sufficiently
> standardized to be mentioned in the context of this draft. The "request"
> parameter https://tools.ietf.org/html/draft-ietf-oauth-par-03#section-3
> can be sent to the PAR endpoint and should provide the same notation of
> non-repudiation as does jwsreq. I think that's sufficient treatment of
> non-repudiation for the PAR draft.
>
This is the case when PAR uses "Content-type:
application/oauth.authz.req+jwt".

This is fine as the jws form param is signed. This is also equivalent to
jwsreq in matter of providing non repudiation.

Content-Type: application/x-www-form-urlencoded

....

request=eyJraWQiOiJrMmJ.....3Gkk488RQohhgt1I0onw


This is not equivalent to jwsreq. As request body is not signed. This
does not provide non repudiation.

Content-Type: application/x-www-form-urlencoded

....

response_type=code&
state=af0ifjsldkj


It is worth mentioning this in the draft
Best regards
/Francis

-- 
Francis Pouatcha
Co-Founder and Technical Lead
adorsys GmbH & Co. KG
https://adorsys-platform.de/solutions/
_______________________________________________
OAuth mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/oauth

Reply via email to