Re: [Txauth] Revisiting the photo sharing example (a driving use case for the creation of OAuth)

Justin Richer <> Mon, 03 August 2020 19:02 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 602833A0A6D for <>; Mon, 3 Aug 2020 12:02:21 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: 0.004
X-Spam-Status: No, score=0.004 tagged_above=-999 required=5 tests=[HTML_MESSAGE=0.001, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=ham autolearn_force=no
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id itydWNl6mWI4 for <>; Mon, 3 Aug 2020 12:02:19 -0700 (PDT)
Received: from ( []) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by (Postfix) with ESMTPS id DC4D33A0A6A for <>; Mon, 3 Aug 2020 12:02:18 -0700 (PDT)
Received: from [] ( []) (authenticated bits=0) (User authenticated as jricher@ATHENA.MIT.EDU) by (8.14.7/8.12.4) with ESMTP id 073J2BoC010490 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Mon, 3 Aug 2020 15:02:12 -0400
From: Justin Richer <>
Message-Id: <>
Content-Type: multipart/alternative; boundary="Apple-Mail=_44F70B26-F1AC-4B7D-992D-EE6920C1E716"
Mime-Version: 1.0 (Mac OS X Mail 13.4 \(3608.\))
Date: Mon, 3 Aug 2020 15:02:11 -0400
In-Reply-To: <>
Cc: Dick Hardt <>, "" <>
To: Denis <>
References: <>
X-Mailer: Apple Mail (2.3608.
Archived-At: <>
Subject: Re: [Txauth] Revisiting the photo sharing example (a driving use case for the creation of OAuth)
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Mon, 03 Aug 2020 19:02:21 -0000

Hi Denis,

I think there’s still a problem with the terminology in use here. What you describe as RS2, which might in fact be an RS unto itself, is a “Client” in OAuth parlance because it is a client of RS1. What you call a “client” has no analogue in the OAuth world, but it is not at all the same as an OAuth client. I appreciate your mapping of the entities below, but it makes it difficult to hold a discussion if we aren’t using the same terms.

The good news is that this isn’t OAuth, and as a new WG we can define our own terms. The bad news is that this is really hard to do.

In GNAP, we shouldn’t just re-use existing terms with new definitions, but we’ve got a chance to be more precise with how we define things. This could mean splitting up things into multiple roles, or defining new roles for aspects that weren’t in consideration in the protocol previously. Perhaps what you’re calling a client below could potentially be considered the “interaction component” or “consent gatherer”, which was previously a portion of the AS. 

I’ve never been a fan of the term “client” in OAuth because it’s too generic and confusing to a lot of developers, especially the web programmers that OAuth is aimed at. “User Agent” would be similarly unsuitable for the same reasons — both “client” and “user agent” are taken to mean “browser” in many circles. I’d love for us to come up with a new term for this, but apart from the “Resource Client (RC)” that I proposed in the XYZ Draft, I don’t have an answer.

 — Justin

> On Aug 3, 2020, at 3:35 AM, Denis <> wrote:
> Hello Dick,
> This is a follow-up of the thread: "Reviewing draft-hardt-xauth-protocol-11".
> Hereafter are three exchanges between you and me which triggered this new thread:
> [Dick]    The photo sharing example was a driving use case for the creation of OAuth.
> [Denis]  We would need to revisit the original scenario and consider if it can be addressed in a different way than the original way.
> [Dick]   The use case is the same. Is there a different solution you are proposing ?
> My response is : Yes indeed, I have a different solution to address the same use case.
> RFC 6749 and draft-ietf-oauth-v2-1-00 both state:
> For example, an end-user (resource owner) can grant a printing service (client) access to her protected photos stored at 
> a photo-sharing service (resource server), without sharing her username and password with the printing service. Instead, 
> she authenticates directly with a server trusted by the photo-sharing service (authorization server), which issues the printing 
> service delegation-specific credentials (access token).
> There are no further explanations. 
> Which information will be disclosed by the resource owner to the authorization server to get "the printing service delegation-specific credentials" 
> is not described. It is a private agreement between the AS and the RS. It is more than likely that the authorization server will learn information 
> about which operation the resource owner is wishing to perform and where. Since in OAuth, the access token is supposed to be opaque to the Client, 
> the user will be unable to make sure that her instructions have been carefully followed.  
> RFC 6749 and draft-ietf-oauth-v2-1-00 both share a common point: they do not contain a "Privacy Considerations" section. 
> Thus, the leakage of information to the AS is not discussed.
> It is possible to revisit the original scenario by applying "Privacy by design" principles. 
> The scenario can be solved by using an old data transfer method that has been first described 32 years ago under the name 
> "Referenced Data Transfer (RDT)" in ECMA-131 (1988) and a few years later in ISO/IEC 10740-2:1993.
> RDT allows two servers to directly exchange large amounts of data under the supervision of a client by communicating, through the client, 
> a reference generated by one server to the other server. RDT does not use any Authorization Server (AS). This means that no AS is able 
> to act as Big Brother and this solves a major privacy concern.
> The three entities involved
> The Client supporting a User (was the Resource Owner).
> The Photo-sharing service (was the Resource Server) : RS1. 
> The Printing service (was the client): RS2.
> Flow of operations with the Photo-sharing service (RS1)
> The Client first connects to the photo-sharing service (RS1) and the User authenticates to the photo-sharing service (RS1).
> RS1 to User: "Please select the operation to be performed".
> Operation selected: "Print pictures using a third party printing service".
> RS1 to User: "Please select a set of pictures to be printed".
> The User selects the pictures.
> RS 1 User consent: "Do you agree RS1 to communicate your selection of pictures to a third party printing service" ?
> If the User consents, RS1 to Client: "Here is the reference to be used by your printing service to get the selected pictures".
> Flow of operations with the Printing service (RS2)
> The Client connects to the printing service (RS2) and the User authenticates to the printing service (RS2).
>             Note: This allows to make sure that the user has an account on RS2 so that further operations can be charged to this account 
>                       and that the prints can be sent to a known address.
> RS2 to User: "Please select the operation to be performed".
> Operation : "Print pictures to be received from a photo-sharing service ".
> RS2 to User: "Please indicate your photo-sharing service".
> The User responds: RS1.
> "Please enter the reference to be used by RS2 to receive the set of pictures from RS1".
> The Client (or the User) enters the reference generated by RS1.
> RS2 contacts RS1 using that reference in order to get the characteristics and the thumbnails of the pictures to be printed (i.e. not yet the whole pictures). 
> RS2 to client: What are your printing preferences for each picture (e.g. number of prints, size, quality of the paper, resolution, margins, colours, etc... ) ?
> The User responds to all these questions.
> RS 2 User consent: This operation will be charged XX € ? Do you agree ?
> If the payment needs to be done on-line, then a payment phase is inserted.
> Continuation of the flow of operations
> Final message from RS1 to the Client: "Your selection of pictures will be soon transmitted to RS2".
> Final message from RS2 to the Client: "Your prints will be soon on their way".
> RS2 to RS1 (asynchronous): transmit the set of pictures corresponding to this reference.
> Some of the advantages of RDT
> An end-user can grant a printing service (RS2) access to her protected photos stored at a photo-sharing service (RS1),
> without sharing her username and password with the printing service.
> Neither RS1, nor RS2 need to use or to trust any AS. This solves the Big Brother privacy issue.
> Any authentication method supported by RS1 or RS2 can be used by the User.
> The User can use any photo-sharing service (RS1) with any printing service (RS2), as long as they both support RDT.
> The User consent is first performed with the photo-sharing service (RS1) and then after with the printing service (RS2).
> The reference generated by RS1 will only be accepted by RS1 during a time period.
> The reference generated by RS1 allows RS2 to query first the thumbnails and then after the pictures selected by the User at RS1.
> The data transfer of the pictures selected at RS1 by the User is performed asynchronously from RS1 to RS2 as a back-office operation.
> Questions: What would be the full scenario of this use case using OAuth ? What about Privacy Considerations ?
> Denis
> -- 
> Txauth mailing list