Re: [OAUTH-WG] I-D Action: draft-ietf-oauth-jwsreq-13.txt

Nat Sakimura <> Tue, 04 April 2017 15:41 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 4B8D112944E for <>; Tue, 4 Apr 2017 08:41:30 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -1.22
X-Spam-Status: No, score=-1.22 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, FREEMAIL_FORGED_FROMDOMAIN=0.197, FREEMAIL_FROM=0.001, HEADER_FROM_DIFFERENT_DOMAINS=0.001, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H3=-0.01, RCVD_IN_MSPIKE_WL=-0.01, RCVD_IN_SORBS_SPAM=0.5, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=no autolearn_force=no
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id blzE2VH5bM4R for <>; Tue, 4 Apr 2017 08:41:22 -0700 (PDT)
Received: from ( []) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by (Postfix) with ESMTPS id 5C6BE129445 for <>; Tue, 4 Apr 2017 08:41:21 -0700 (PDT)
Received: by with SMTP id r45so141975476qte.3 for <>; Tue, 04 Apr 2017 08:41:21 -0700 (PDT)
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=ofIZZRs+f+YhxasLuQ/h1fo6h0rL5Tsu5Rg4nF2uvsI=; b=oKTuW93kC5UNm62DyITQ1DjSNCbqPqQd3+nDmt1MqiVlV+jxjJjlNyNLCFRRq6+jjU GMQozNZtpLE7umM8nzgIhBY8bIYHZQGGlpgi36o2Noi1at+2p9r2vYpVmdO9ZfH6LiaE /4omQnE/wDLF91RAYbPX2mWzvaU/4egyijhrkchON1s5NB7t8m3v6lKT//MdHksTVOY6 J9t/5VeFC0tLUVU5/Ei2tHdC+QmNfqJDS8+pPeEhj1tFIfo3P/GZXtmi2tYDnevtp21U ZSJxG/Xkd+z4Kyd13mxzoJCmaurnjs/N0CyaPNUvNQobZezT+eBQZ/So1dmChJUR0zut yioA==
X-Gm-Message-State: AFeK/H2ekxKXPANA256hQ+eZSmmvx9FhYhEQ6P5HX4d2K45Od8ZY6hk0I8DGZ0eXERQwh/ldThqZIOvw4+Wj+w==
X-Received: by with SMTP id d33mr25702280qta.198.1491320480201; Tue, 04 Apr 2017 08:41:20 -0700 (PDT)
MIME-Version: 1.0
References: <> <> <> <> <> <> <> <>
In-Reply-To: <>
From: Nat Sakimura <>
Date: Tue, 04 Apr 2017 15:41:09 +0000
Message-ID: <>
To: Brian Campbell <>, John Bradley <>
Cc: IETF oauth WG <>
Content-Type: multipart/alternative; boundary="001a1140298c782ee4054c591bd8"
Archived-At: <>
Subject: Re: [OAUTH-WG] I-D Action: draft-ietf-oauth-jwsreq-13.txt
X-Mailman-Version: 2.1.22
Precedence: list
List-Id: OAUTH WG <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Tue, 04 Apr 2017 15:41:30 -0000

Thanks Brian for spotting these. I will make the corrections in -14.



On Fri, Mar 31, 2017 at 10:40 PM Brian Campbell <>

and a typo - "If thie location is" should say "If this location is"

On Fri, Mar 31, 2017 at 8:37 AM, Brian Campbell <>

BTW, the intro still has text about 'dynamic parameters such as "state"'
that need to be cleaned up.

On Fri, Mar 31, 2017 at 8:36 AM, Brian Campbell <>

"The current text causes the AS to ignore them and not return a error. " -
except that I don't believe the current text actually specifies that
anywhere. And I think that the intent of Mike's original comment was that
-13 doesn't specify the behavior but that it needs to be revised to do so.

I'd suggest that the doc say that the client must include in the request
object (request or request_uri) all the oauth parameters that it sends. And
when request or request_uri is sent, that the AS must/should only rely on
parameter values from the request object.

I think being semi or somewhat compatible or tolerant of the Connect
variation or request/request_uri is good because it uses the same parameter
names, the same endpoint, and the same metadata names.

On Thu, Mar 30, 2017 at 11:14 PM, John Bradley <> wrote:

They are mutually exclusive.

However there are two options as to how the authorization endpoint would
treat extra query parameters like state if they are sent.

The current text causes the AS to ignore them and not return a error.  This
would be more backwards compatible with the request object in OpenID
Connect, however in reality it may cause connect clients to send parameters
as query parameters  that would be processed by a connect server that would
be ignored by a OAuth server without any obvious error.  There may however
be subtle errors downstream from missing parameters.

The other option is to have a cleaner breaking change from Connect and have
the Authorization endpoint return a error if anything other than the two
new parameters are sent to the authorization endpoint.

I am leaning towards the latter as it is easier to debug,  and wont allow
incompatible Connect requests to be accepted without a error.   We would
have done this in Connect but couldn’t drop required parameters from OAuth
in a Connect spec.

The downside for the latter is that the client would need to know if the AS
is supporting The Connect version or the OAuth version.

One of the typical conundrums around how to deal with doing the best going
forward thing vs not blowing up older implementations.

In the current proposal a client could put the required parameters both
places and the same request would work on servers supporting both the
Connect and OAuth versions.

John B.

Sent from Mail <> for Windows

*From: *Torsten Lodderstedt <>
*Sent: *March 30, 2017 11:01 PM
*To: *John Bradley <>
*Cc: *Nat Sakimura <>; Nat Sakimura <>; IETF
oauth WG <>
*Subject: *Re: [OAUTH-WG] I-D Action: draft-ietf-oauth-jwsreq-13.txt

I had assumed using the request object is mutual exclusive to use of URI
query parameters. Did I misinterpret the draft?

Am 30.03.2017 um 22:40 schrieb John Bradley <>:

It is a trade off between compatibility with Connect and possible
configuration errors.

In reality it may not be compatible with Connect if the client is sending
some parameters outside the object without including them in the object as
a Connect client might.    You would potentially wind up dropping state or
nonce without an error.

I asked Mike and he was leaning to making it a error to send them as query
parameters as that would be a clean change.

I think the choice is a bit of a grey area.

Sent from Mail <> for Windows

*From: *
*Sent: *March 30, 2017 9:57 PM
*To: *John Bradley <>; Nat Sakimura <>
*Cc: *IETF oauth WG <>
*Subject: *Re: [OAUTH-WG] FW: I-D Action: draft-ietf-oauth-jwsreq-13.txt


Sent from my Huawei Mobile

-------- Original Message --------
Subject: Re: [OAUTH-WG] FW: I-D Action: draft-ietf-oauth-jwsreq-13.txt
From: John Bradley
To: Nat Sakimura
CC: IETF oauth WG

So I think we need to make the must ignore clearer for the additional
paramaters on the authorization endpoint.

On Mar 30, 2017 17:33, "Nat Sakimura" <> wrote:

Not right now.

As of this writing, a client can still send duplicate parameters in the
query but they get ignored by the servers honoring OAuth JAR. So, it is
backwards compatible with OpenID Connect in that sense (OpenID Connect
sends duplicate manatory RFC6749 parameters as the query parameters as well
just to be compliant to RFC6749). Conversely, servers that do not support
OAuth JAR will ignore request_uri etc.

On Mar 30, 2017, at 4:47 PM, Mike Jones <> wrote:

Is there a clear statement somewhere along the lines of “parameters (other
than “request” or “request_uri”) are only allowed to be in the signed
object if a signed object is used”?  That’s the kind of thing I was looking
for and didn’t find.

                                                       -- Mike

*From:* John Bradley []
*Sent:* Thursday, March 30, 2017 4:44 PM
*To:* Mike Jones <>
*Cc:* Nat Sakimura <>; IETF oauth WG <>
*Subject:* RE: [OAUTH-WG] FW: I-D Action: draft-ietf-oauth-jwsreq-13.txt

The intent of the change is to only allow the paramaters to be in the
signed object if a signed object is used.

This requires State, nonce etc to be in the JWT.  Only one place to check
will hopefully reduce implimentation errors.

This also allows us to remove the caching text as we now have one JWT per
request, so caching won't happen.

John B.

On Mar 30, 2017 4:36 PM, "Mike Jones" <> wrote:

I **believe** the intent is that **all** parameters must be in the request
object, but the spec doesn’t actually say that, as far as I can tell.  Or
maybe the intent is that parameters must not be duplicated between the
query parameters and the request object.

One or the other of these statements should be explicitly included in the
specification.  Of course, I could have missed the statement I’m asking for
in my review, in which case please let me know what I missed.


                                                      -- Mike

*From:* OAuth [] *On Behalf Of *John Bradley
*Sent:* Thursday, March 30, 2017 3:00 PM
*Subject:* [OAUTH-WG] FW: I-D Action: draft-ietf-oauth-jwsreq-13.txt

Based on feeback from the IESG we have removed some of the optionality in
the draft.

It is a shorter read than draft 12.

John B.

Sent from Mail <> for Windows

*From: *
*Sent: *March 30, 2017 1:38 PM
*To: *
*Cc: *
*Subject: *[OAUTH-WG] I-D Action: draft-ietf-oauth-jwsreq-13.txt

A New Internet-Draft is available from the on-line Internet-Drafts

This draft is a work item of the Web Authorization Protocol of the IETF.

        Title           : The OAuth 2.0 Authorization Framework: JWT
Secured Authorization Request (JAR)

        Authors         : Nat Sakimura

                          John Bradley

           Filename        : draft-ietf-oauth-jwsreq-13.txt

           Pages           : 27

           Date            : 2017-03-30


   The authorization request in OAuth 2.0 described in RFC 6749 utilizes

   query parameter serialization, which means that Authorization Request

   parameters are encoded in the URI of the request and sent through

  user agents such as web browsers.  While it is easy to implement, it

   means that (a) the communication through the user agents are not

   integrity protected and thus the parameters can be tainted, and (b)

   the source of the communication is not authenticated.  Because of

   these weaknesses, several attacks to the protocol have now been put


   This document introduces the ability to send request parameters in a

   JSON Web Token (JWT) instead, which allows the request to be signed

   with JSON Web Signature (JWS) and/or encrypted with JSON Web

   Encryption (JWE) so that the integrity, source authentication and

   confidentiality property of the Authorization Request is attained.

   The request can be sent by value or by reference.

The IETF datatracker status page for this draft is:

There are also htmlized versions available at:

A diff from the previous version is available at:

Please note that it may take a couple of minutes from the time of

until the htmlized version and diff are available at

Internet-Drafts are also available by anonymous FTP at:


OAuth mailing list

OAuth mailing list

OAuth mailing list

OAuth mailing list


Nat Sakimura

Chairman of the Board, OpenID Foundation