Re: [GNAP] Consensus Call on Continuation Request

[Dick Hardt <[e-mail]>]

Dave: which points changed your mind?

On Tue, Dec 15, 2020 at 1:11 PM Dave Tonge <[e-mail]> wrote:

> Thanks for the detailed response.
>
> I think you make the points well and I'm now in favour of the PR.
> However I do think that to keep the consistency that keeps being
> discussed, that the tokens shouldn't be rotated.
>
> I also think it would be good to have a discussion about "grant
> management" and identifiers for a grant.
>
> Dave
>
> On Tue, 15 Dec 2020 at 17:30, Justin Richer <[e-mail]> wrote:
>
>> On Dec 15, 2020, at 10:50 AM, Dave Tonge <[e-mail]> wrote:
>>
>>
>> > The access token pattern has a lot of benefits, otherwise we wouldn’t
>> have an entire OAuth ecosystem based on it
>>
>> *But what are the benefits in this particular use case?* The
>> continuation API is not like any other API - it is integral to the AS.
>> There are quite a few extensions to OAuth that use client authentication
>> rather than access tokens.
>>
>>
>> Yes, and the propagation of that has lead to a mess in the OAuth world.
>> You’ve got re-definitions of client authentications at all different
>> endpoints, they’re technically allowed to vary between endpoints (though I
>> don’t know of it happening in practice, that feels like a downgrade attack
>> waiting to happen to someone). Then there’s the fact that all the newest
>> security mechanisms we have — PKCE, DPoP, and MTLS — don’t rely on client
>> authentication at all to achieve security. All of these work without the
>> client having credentials previously known to the AS, and we already know
>> that GNAP is going to need to live in this more dynamic world. We need to
>> think beyond what OAuth 2 has done in the past, and especially from our
>> perceptions and assumptions of the models that drive OAuth 2’s decisions,
>> lest we repeat its mistakes.
>>
>> Also, I want to challenge this idea of “integral to the AS” as a point.
>> In OAuth 1, the API was “integral” to the server side, but in OAuth 2 we
>> split that into the RS concept. Even though in practice, a lot of RS’s are
>> still integrated to the AS in some fashion because it’s a single service,
>> OAuth 2 is clear about what’s expected to be known by each component. For
>> the AS as currently defined in GNAP, I’m seeing three distinct functions.
>> As per the Terminology discussion, we don’t have explicit names for these
>> yet:
>>
>>  - starting a request; this is an endpoint to kick things off; it needs
>> to be able to look up the rights asked for by the client (if it knows the
>> client at all ahead of time) and make initial decisions about needed
>> interaction and follow up
>>  - continuing a request; this is an API that needs to know the context of
>> the request itself, including which key it’s bound to; this is separate
>> from any identity of the client and possibly the user, but an AS
>> implementation that has access to those elements can use them
>>  - interacting with the user; this is front-facing and in OAuth today is
>> already deployed as a separate service in some places, we should embrace
>> that at the very least (but doing so formally is a separate issue)
>>
>> Why separate these in this way? There is immense power in having a single
>> consistent way to start the process. The “continuation API” gives us an
>> HTTP-defined mechanism for managing a request over time, including the
>> simple case of returning information from the front channel, but people
>> have already raised the question of non-HTTP and self-hosted AS’s, which
>> would probably want a different kind of continuation API to communicate to
>> the AS. The same thing with separating out the interaction: there are going
>> to be a lot of different ways to handle interaction out there, and not all
>> of them will be “integral” to the AS in the way that a simple
>> implementation might do. We’re defining a protocol based strongly on HTTP
>> and JSON, but we should structure it in such a way that it can be extended
>> and translated elsewhere in a clear way.
>>
>> This all raises the question: if we can rely on a unique URL for
>> redirect-based interaction, why not here? The simple reason is that a URL
>> is the :only: mechanism we have in the front channel for passing
>> information, and we need to warn implementors against including sensitive
>> information in it, and we need to protect it with additional items. We have
>> access to more than just URLs when we’re dealing with the continuation API,
>> and we ought to make use of all of our tools in ways that are consistent
>> and make sense.
>>
>> From a previous email the advantages I see are:
>>  - more data can be encoded in a token than in a uri (this seems more of
>> an edge case)
>>  - it can be an identifier for the grant (I don't agree with this)
>>
>>
>> It allows the parts above to live separately, even if they don’t HAVE to.
>> And it simplifies what we’re asking the client to do by making it
>> consistent with other parts of the ecosystem. The AS offering an API
>> doesn’t :have: to be different, and OpenID Connect showed us, with the
>> UserInfo Endpoint, that that’s very much the case in practice.
>>
>> Having my client code do very similar things in slightly different ways
>> is not simpler.
>>
>>
>> Another question I have: *do we envisage granting access tokens to the
>> RC that will allow it to manage multiple grants*?
>>
>>
>> I don’t see that happening, personally, and it hasn’t been brought up as
>> a use case to date.
>>
>>
>> I also think there will be confusion with the signing being used for
>> different things. Maybe it just needs to be called out in the spec that:
>>
>> Request 1: signature = client authentication
>> Request 2+: signature = proof of possession for access token
>>
>>
>> That’s more or less the intent of what’s in the specification right now,
>> and why all the signature methods, which are used for both client auth and
>> token possession, are all together in section 8 and not separated by use.
>>  (With the caveat: it’s only client authentication in the first request if
>> the AS knows about the client instance ahead of time, which isn’t always
>> going to be true.) That all can likely be made clearer, as is always the
>> case with spec text. But you can use the signature methods with and without
>> access tokens, and it only gets used without in an initial call where you
>> don’t :have: an access token to present.
>>
>> Separating the different kinds of “client authentication" out in OAuth is
>> the source of some real confusion. Like right now, what happens if you try
>> to combine a client assertion, signed request objects for PAR, and DPoP
>> proofs, all in a single request? All of these are optional and all of them
>> “do client authentication” in some arguable fashion. I’ve worked on several
>> systems and implemented these things, and their interplay is really
>> confusing to manage and can go sideways really fast. And that’s just the
>> work for a single endpoint, this gets repeated for introspection,
>> revocation, CIBA, device, and on.
>>
>> At the end of the day I’m in favor of giving the client developers a very
>> clear set of directions on what they need to do and how they need to access
>> things, and treating the continuation as a token-bound API is, to me, the
>> clearest pattern we can offer for this piece.
>>
>>  — Justin
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> On Tue, 15 Dec 2020 at 15:33, Justin Richer <[e-mail]> wrote:
>>
>>> I agree with Fabien that the persistent identifier is a separate issue.
>>> The current spec re-uses the access token for this artifact, but that’s
>>> potentially brittle and could be changed out for something else. There’s an
>>> issue asking for expanding on the use cases for this functionality (which
>>> hasn’t been addressed by this PR):
>>>
>>> https://github.com/ietf-wg-gnap/gnap-core-protocol/issues/87
>>>
>>> A potentially-rotating URI would be just as brittle, and so having a
>>> single codified identifier for this artifact would be useful, but it does
>>> assume some things about the nature of the AS. Every other use the client
>>> has to manage the ongoing request over time doesn’t need an explicit
>>> identifier. Just like with OAuth-protected APIs, the server can determine
>>> the context not just from the URI but from the rest of the request,
>>> including the access token itself. This is both more common and more
>>> powerful than a strict reading of REST designs. It also follows the HATEOS
>>> principles as the entire HTTP request is taken into account, including the
>>> access token and signature portions.
>>>
>>> I’ll also point out that the rotation of these credentials is also filed
>>> as a separate issue that’s not being addressed right now, so we can revisit
>>> that discussion separately:
>>>
>>> https://github.com/ietf-wg-gnap/gnap-core-protocol/issues/87
>>>
>>> As for the name, we could give it a different label. We have this same
>>> pattern of issuing a resource-specific access token alongside a URL in the
>>> Dynamic Registration specification, both in OAuth (
>>> https://tools.ietf.org/html/rfc7592#section-3) and OpenID Connect (
>>> https://openid.net/specs/openid-connect-registration-1_0.html#RegistrationResponse).
>>> Here it’s called the “registration access token” — but what’s important
>>> there, as it is here, is that it’s not a different kind of artifact that
>>> the client now has to figure out how to use, it’s an access token plain and
>>> simple. In the OAuth world this is a bearer token, since that’s what OAuth
>>> 2 uses. In the GNAP world it’ll be a bound token, as that’s what we’re
>>> looking to build on. This is also related to another future discussion
>>> about responses tying an access token to a specific API that’s told to the
>>> client, as we could potentially re-use those components and concepts here
>>> as well:
>>>
>>> https://github.com/ietf-wg-gnap/gnap-core-protocol/issues/69
>>>
>>> And finally, no speculation on the complexity is needed: I implemented
>>> this pattern several months ago during the design team discussions when we
>>> were considering this pattern, and the code is all online for people to
>>> see.
>>>
>>> https://github.com/bspk/oauth.xyz-java/
>>>
>>> On the AS side, the most interesting code to this discussion is in the
>>> TransactionEndpoint class:
>>>
>>>
>>> https://github.com/bspk/oauth.xyz-java/blob/master/as/src/main/java/io/bspk/oauth/xyz/authserver/endpoint/TransactionEndpoint.java
>>>
>>> Here, you’ll see that on initial request, the server looks up the client
>>> to see if it’s been registered, but after that, it makes sure that the
>>> token and key are appropriate for the ongoing request. From the client side
>>> it’s even simpler. The client’s got a small service function to manage the
>>> different signature methods that are implemented, and all of them can take
>>> in an optional access token:
>>>
>>>
>>> https://github.com/bspk/oauth.xyz-java/blob/master/rc/src/main/java/io/bspk/oauth/xyz/http/SigningRestTemplateService.java
>>>
>>> The heavy lift is doing the actual signing, and you need that in order
>>> to start the process anyway. Managing the access token as an artifact to
>>> use at the API is completely trivial since the client already needs to
>>> manage its own state internally to do any of this. And note that all of
>>> this is changed from how it was before: Previously, the XYZ Protocol had
>>> used a “transaction handle” returned by the AS that the client would use to
>>> continue the request. However, this simple model was limiting, and the
>>> design team adopted XAuth’s model of continuation being an API. In doing
>>> so, we made it like all of the other APIs the client is going to call: in
>>> the initial state, it doesn’t have any kind of access rights, it’s just
>>> calling. From that initial call forward, the AS just needs to know that the
>>> token and key match what it expects.
>>>
>>> In summary, my views are:
>>>  - The access token pattern has a lot of benefits, otherwise we wouldn’t
>>> have an entire OAuth ecosystem based on it
>>>  - Magic URIs have a lot of drawbacks which are well understood; while
>>> they can be mitigated, they can also be avoided
>>>  - Continuation is an API, and treating it like the other kinds of API
>>> the client would call makes sense
>>>  - Calling this by a special name, like “grant access token” or
>>> “continuation access token” is fine, but it should function like any other
>>> access token
>>>  - The heavy lift for clients is on protecting the message
>>> cryptographically, which they need to do anyway
>>>
>>>  — Justin
>>>
>>>
>>> On Dec 15, 2020, at 7:50 AM, Dave Tonge <[e-mail]> wrote:
>>>
>>> The persistent identifier is not a different issue, the current access
>>> token is used to reference an existing grant
>>> <https://github.com/ietf-wg-gnap/gnap-core-protocol/blob/main/draft-ietf-gnap-core-protocol.md#referencing-an-existing-grant-request-request-existing>
>>>
>>>
>>> It may not be more difficult for a client to *use* an access token at
>>> the AS / RS. But there is definitely an overhead on the client to
>>> *manage* this separate access token.
>>>
>>>
>>>
>>> On Tue, 15 Dec 2020 at 12:10, Fabien Imbault <[e-mail]>
>>> wrote:
>>>
>>>> I think we always said the access token was different, and handled as a
>>>> bound token.
>>>>
>>>> But it doesn't mean it's more difficult for the client that already
>>>> needs to be able to handle tokens anyway (bearer or not, both cases could
>>>> occur). It's mostly consolidating the logic.
>>>>
>>>> You're anticipating a lot of issues which have no specific reason to
>>>> occur, such as "can't be used with the token management APIs?". The
>>>> management API is part of the same general flow.
>>>> Anticipating issues with rotation is useful, but there are also many
>>>> ways it can be hard to manage through a stateful approach too. And
>>>> fundamentally, having everything in a common model (both for the internals
>>>> of the AS and for the API calls) will help improve by a large margin what
>>>> is probably the weakest point in today's infrastructure. But it's early to
>>>> be definitive as to the downstream impact either way.
>>>>
>>>> As for a persistent identifier instead of a continuation API, and
>>>> generally the end of your message, it's a totally unrelated issue to this
>>>> PR, so I suggest we don't discuss that here, but in a separate issue if
>>>> needed.
>>>>
>>>> Fabien
>>>>
>>>> On Tue, Dec 15, 2020 at 11:08 AM Dave Tonge <[e-mail]>
>>>> wrote:
>>>>
>>>>> So we've established that this is a different access token, that
>>>>> requires different handling at the client. So keeping it could cause more
>>>>> confusion?
>>>>>
>>>>> As an RC, I will have to store the continue `uri` as although it could
>>>>> be static it could also be dynamic. Why do I need to store an access token
>>>>> as well. It brings me no benefit as an RC, in fact it brings more
>>>>> complexity. As I will now need to manage multiple types of tokens with
>>>>> different lifecycles:
>>>>>
>>>>> *Continuation token*
>>>>>   - can only be used at the continue endpoint (the name of which is
>>>>> confusing as I can use this endpoint to revoke a grant or get metadata on
>>>>> the grant).
>>>>>  - may be rotated each time it is used, or may not be
>>>>>  - provided in the `continue` section of the response
>>>>>  - must be sender-constrained
>>>>>  - can't be used with the token management APIs?
>>>>>  - can be used to identify the grant when making subsequent grants
>>>>>
>>>>> *Access token(s) to use at RS*
>>>>>   - can only be used at the specified RS
>>>>>   - may be sender constrained
>>>>>   - when used at the RS, will not result in rotation
>>>>>
>>>>> From my perspective, most use-cases will require the RC to have a
>>>>> persistent identifier for the grant. Why not bring this into the protocol
>>>>> and let the AS provide this persistent identifier (through the form of the
>>>>> continue uri). Using a rotating access token as a persistent identifier
>>>>> doesn't seem like the right choice.
>>>>>
>>>>> I see no security benefit to having the continuation access token. It
>>>>> doesn't matter if the continue uri leaks as it is useless without an
>>>>> accompanying signature, i.e. any security benefit of having an access token
>>>>> is already provided by having a signature.
>>>>>
>>>>> The only benefits that I can see are:
>>>>>  - If the AS wants to be fully stateless, then you can encode more
>>>>> data in a token than in a uri
>>>>>  - If the AS wants to have a static endpoint for CRUD operations on
>>>>> the grant
>>>>>  - To allow the AS to identity a previous grant
>>>>>
>>>>> If we dropped the access token for the continue endpoint and rather
>>>>> mandated a dynamic uri this would make things conceptually easier to
>>>>> understand, easier for the RC to implement, easier to debug and less chance
>>>>> of errors when rotating tokens (i.e. race conditions could be quite likely
>>>>> if the AS always rotates the token)
>>>>>
>>>>> *One-off grant with no continuation or ongoing management:*
>>>>> RC sends signature and metadata, no `continue` response provided,
>>>>> therefore no grant management possible
>>>>>
>>>>> *Grant with ongoing management*
>>>>> RC sends signature and metadata, AS responds with a continue uri that
>>>>> has these purposes:
>>>>>  - can be used by the RC to continue/update, read or revoke the grant
>>>>>  - can be used by the RC when making a new grant to identify the
>>>>> previous grant
>>>>>
>>>>> As an RC the only permanent items I need to store are:
>>>>>  - the continue uri associated with the grant
>>>>>  - any access tokens I receive for the grant
>>>>>
>>>>> Dave
>>>>>
>>>>> On Tue, 15 Dec 2020 at 10:11, Fabien Imbault <[e-mail]>
>>>>> wrote:
>>>>>
>>>>>> Hi Torsten,
>>>>>>
>>>>>> You're right on both accounts.
>>>>>> - for the first remark, it fits quite nicely the init request  /
>>>>>> continuation pattern
>>>>>> - for the second remark, it is a sort of handle for the
>>>>>> continuation request, which will eventually lead to the issuance or refresh
>>>>>> of standard access tokens
>>>>>>
>>>>>> Having a specific name is a possibility, I actually suggested that
>>>>>> too at some point.
>>>>>>
>>>>>> Fabien
>>>>>>
>>>>>> On Tue, Dec 15, 2020 at 9:50 AM Torsten Lodderstedt <
>>>>>> [e-mail]> wrote:
>>>>>>
>>>>>>> Hi Fabien,
>>>>>>>
>>>>>>> > Am 12.12.2020 um 12:06 schrieb Fabien Imbault <
>>>>>>> [e-mail]>:
>>>>>>> >
>>>>>>> > Hi,
>>>>>>> >
>>>>>>> > On the contrary your feedback is most welcome.
>>>>>>> >
>>>>>>> > It doesn't accept any token, it needs the particular token as
>>>>>>> described in 3.1 and which is not a bearer token (that's what the "key" :
>>>>>>> true parameter is supposed to convey).
>>>>>>> >
>>>>>>> > Let us know if you need more clarifications.
>>>>>>>
>>>>>>> Thanks for the clarification. I think only accepting this kind of
>>>>>>> token at the continuation is a good idea otherwise the AS would need to be
>>>>>>> able to parse and understand all sorts of access tokens.
>>>>>>>
>>>>>>> Conceptually, I like the idea to treat the continuation as another
>>>>>>> kind of resource. However, here are some observations I want to share with
>>>>>>> you:
>>>>>>> - This resource is different as it will issue other access tokens
>>>>>>> (of this kind) to be used in subsequent continuation requests. This
>>>>>>> requires different handing on the client side.
>>>>>>> - This access token (if I understand correctly) is (or at least
>>>>>>> feels like) a handle for the underlying grant. So it is kind of the super
>>>>>>> access token to obtain other access tokens.
>>>>>>>
>>>>>>> I would consider using a different term to refer to this special
>>>>>>> access token, grant token or grant handle for example, in order to prevent
>>>>>>> confusion.
>>>>>>>
>>>>>>> best regards,
>>>>>>> Torsten.
>>>>>>>
>>>>>>>
>>>>>>> >
>>>>>>> > Best
>>>>>>> > Fabien
>>>>>>> >
>>>>>>> > Le sam. 12 déc. 2020 à 11:33, Torsten Lodderstedt <
>>>>>>> [e-mail]> a écrit :
>>>>>>> > Hi all,
>>>>>>> >
>>>>>>> > I didn’t follow GNAP closely so bear with me if me question seems
>>>>>>> naive.
>>>>>>> >
>>>>>>> > After having skimmed through the current draft and the PR, I‘m not
>>>>>>> sure whether the continuation requests accepts any access token issued to
>>>>>>> the RC or the particular access token returned in the „continue“ element in
>>>>>>> section 3.1..
>>>>>>> >
>>>>>>> > Can you please shed some light on this?
>>>>>>> >
>>>>>>> > kind regards,
>>>>>>> > Torsten.
>>>>>>> >
>>>>>>> >> Am 12.12.2020 um 03:35 schrieb Fabien Imbault <
>>>>>>> [e-mail]>:
>>>>>>> >>
>>>>>>> >> 
>>>>>>> >> You're completely right. Allowing the dev to be lazy is a very
>>>>>>> good thing in general, because it's what we know will work :-)
>>>>>>> >>
>>>>>>> >> Le sam. 12 déc. 2020 à 03:15, Stephen Moore <[e-mail]>
>>>>>>> a écrit :
>>>>>>> >> Hi Fabien,
>>>>>>> >>
>>>>>>> >> For #3) Even after I typed out the hypothetical attack, that was
>>>>>>> sort of in the back of my mind, it isn't a huge risk there. So I actually
>>>>>>> agree with Dick there. Something doesn't sit right with me for the unique
>>>>>>> URL solution, so I don't like it and came up with a hypothetical that seems
>>>>>>> like it could be a down side.
>>>>>>> >>
>>>>>>> >> I still think the access token model with the signed request is
>>>>>>> the way I'd like to go, because again, it's a mechanism I'd be implementing
>>>>>>> anyway to talk to any 'normal' resource. The fact is there is _something_
>>>>>>> representing context that has to pass back and forth here, whether that is
>>>>>>> an access token (which I feel like is more flexible for extensions etc), a
>>>>>>> unique url, or even a cookie sent in the cookie header. So just to
>>>>>>> re-iterate, I'm a +1 on this pull request, speaking as a lazy developer ;)
>>>>>>> >> -steve
>>>>>>> >>
>>>>>>> >> On Fri, Dec 11, 2020 at 8:51 PM Fabien Imbault <
>>>>>>> [e-mail]> wrote:
>>>>>>> >> Again speaking in my own name here.
>>>>>>> >>
>>>>>>> >> Dick, we know you'd prefer to have a different design, but this
>>>>>>> PR shouldn't be about that.
>>>>>>> >>
>>>>>>> >> Back on your 3 items :
>>>>>>> >>
>>>>>>> >> 1) yes we could make pre-register mandatory, but we already
>>>>>>> decided that wouldn't be how that would work. We have a client instance
>>>>>>> that allows a more generic and flexible pattern (which BTW also allows what
>>>>>>> you want)
>>>>>>> >>
>>>>>>> >> 2) instead of blame arguments of who's less
>>>>>>> restful/HATEOAS/whatever that have the tenancy to flame conversations, I
>>>>>>> suggest we speak in less abstract terms and ask ourselves what that means
>>>>>>> in practice for devs. Stephen and several others (myself included) have
>>>>>>> expressed that it wouldn't be harder to implement, it would even simplify
>>>>>>> things quite a lot. If you disagree please send us a code sample to really
>>>>>>> show that point by example, because that's really not obvious.
>>>>>>> >>
>>>>>>> >> 3) "If someone has the client credentials, they can impersonate
>>>>>>> the client, and all bets are off." Are you seriously making this argument?
>>>>>>> Because if you have a better proposal than using cryptographic keys, I'm
>>>>>>> all hears. You make it look like there's a problem, while in reality we're
>>>>>>> only relying on the basic assumption of all modern digital communications.
>>>>>>> >>
>>>>>>> >> And more importantly you never responded to the issues of how to
>>>>>>> avoid the security pitfalls of what you proposed.
>>>>>>> >>
>>>>>>> >> Fabien
>>>>>>> >>
>>>>>>> >>
>>>>>>> >> Le sam. 12 déc. 2020 à 00:35, Dick Hardt <[e-mail]>
>>>>>>> a écrit :
>>>>>>> >>
>>>>>>> >>
>>>>>>> >> On Fri, Dec 11, 2020 at 2:53 PM Stephen Moore <[e-mail]>
>>>>>>> wrote:
>>>>>>> >> But from the spec:
>>>>>>> >> "
>>>>>>> >> When sending a non-continuation request to the AS, the RC MUST
>>>>>>> identify itself by including the client field of the request...
>>>>>>> >> ...
>>>>>>> >> key (object / string) : The public key of the RC to be used in
>>>>>>> this request as described in {{request-key}}. This field is REQUIRED.
>>>>>>> >> ...
>>>>>>> >> "
>>>>>>> >> So on the initial request, the key will be there.
>>>>>>> >>
>>>>>>> >> The client field can be an object or a string. If the client is
>>>>>>> pre-registered, then a string could be provided instead of an object.
>>>>>>> >>
>>>>>>> >>
>>>>>>> >> If you don't have the access token, then how do you differentiate
>>>>>>> between two requests from the same web application by two different users?
>>>>>>> Is the web application supposed to have different credentials for every
>>>>>>> request?
>>>>>>> >>
>>>>>>> >> The AS returns a URI for manipulating the request. I would change
>>>>>>> the spec so that each request would have a unique URI. This is the usually
>>>>>>> RESTful pattern that the resource (the grant request) has an URI.
>>>>>>> >>
>>>>>>> >>
>>>>>>> >> So in this case, the easy way out is to pass the access token to
>>>>>>> the client, who then, as i stated before, treats the continue request as a
>>>>>>> RS call (albeit a specialized version of the RS where the RS is the AS) OR
>>>>>>> to use the unique URL,
>>>>>>> >> but that seems open to a brute force attack by a malicious RC.
>>>>>>> (What would be the point of that attack, I don't know, I guess if someone
>>>>>>> had the client credentials but not any subjects/resources they could try to
>>>>>>> intercept the grant via continue... I just don't feel right locking things
>>>>>>> down to unique URLs that way.)
>>>>>>> >>
>>>>>>> >> If someone has the client credentials, they can impersonate the
>>>>>>> client, and all bets are off.
>>>>>>> >>
>>>>>>> >> LOTS of RS servers return a resource specific URL -- my proposal
>>>>>>> is no different.
>>>>>>> >>
>>>>>>> >>
>>>>>>> >>
>>>>>>> >> -steve
>>>>>>> >>
>>>>>>> >> On Fri, Dec 11, 2020 at 5:33 PM Dick Hardt <[e-mail]>
>>>>>>> wrote:
>>>>>>> >> Hi Stephen
>>>>>>> >>
>>>>>>> >> The client is signing the first request. The key *might* be in
>>>>>>> the body. The client is signing all the subsequent requests as well. The
>>>>>>> "access token" is not needed by the client to prove it is authorized as the
>>>>>>> client is proving it is the same client again.
>>>>>>> >>
>>>>>>> >> In other words, I don't see the need for an access token, so it
>>>>>>> does not need to be put in a URL or an auth header.
>>>>>>> >>
>>>>>>> >> If a developer really, really wants to hand context back to the
>>>>>>> client for subsequent calls, they can put it in the URL or some other
>>>>>>> method. Putting it in the HTTP Authorization header is confusing because it
>>>>>>> is NOT an access token -- it is the context of the request.
>>>>>>> >>
>>>>>>> >> ᐧ
>>>>>>> >>
>>>>>>> >> On Fri, Dec 11, 2020 at 2:01 PM Stephen Moore <[e-mail]>
>>>>>>> wrote:
>>>>>>> >> Even though I've only been lightly following things, I feel the
>>>>>>> need to voice my preference as a developer since I will probably someday
>>>>>>> have to either write a RC or RS...
>>>>>>> >>
>>>>>>> >> The way I see it is the RC makes the initial request to the AS as
>>>>>>> part of this request, it provides it's key in the body... (So no use of the
>>>>>>> Authorization header)
>>>>>>> >> At this point that request, represented by the continue URL +
>>>>>>> "Access Token", from my lazy developer standpoint, is a Resource Endpoint
>>>>>>> and Access Token, and the AS is acting as a specialized RS in this case.
>>>>>>> >> So my client posts to whatever URL with the 'access token' in the
>>>>>>> authorization header, just like acting on any other resource I have a token
>>>>>>> for. YES, I get a new token value to use every call, and there is a
>>>>>>> decision point of "Do I have another continue, or do I have a real token
>>>>>>> for the resource..." But the mechanism is the same to me in the client.
>>>>>>> >> Personally I like that, because if I have an access_token, I
>>>>>>> already think "Put it in the auth header."
>>>>>>> >>
>>>>>>> >> So my vote would be +1 for the pull request at this time.
>>>>>>> >> -steve
>>>>>>> >>
>>>>>>> >> On Fri, Dec 11, 2020 at 3:04 PM Dick Hardt <[e-mail]>
>>>>>>> wrote:
>>>>>>> >> inline ...
>>>>>>> >>
>>>>>>> >> On Fri, Dec 11, 2020 at 9:58 AM Justin Richer <[e-mail]>
>>>>>>> wrote:
>>>>>>> >> Others had already responded to this previous thread, but I
>>>>>>> wanted to add a couple points to clarify some things.
>>>>>>> >>
>>>>>>> >>> 3) What the client has to do with the "access token" is not the
>>>>>>> same as access tokens for an RS. The client gets a new "access token" for
>>>>>>> each grant request, and for each API call to the AS, and the client learns
>>>>>>> it can not make any more API calls for that specific request when it does
>>>>>>> not get an "access token" back. This is a completely different design
>>>>>>> pattern than calling an RS API with an access token, and is a new design
>>>>>>> pattern for calling APIs. This adds complexity to the client that it would
>>>>>>> not normally have, and I don't think GNAP is the right place to start a new
>>>>>>> design pattern.
>>>>>>> >>>
>>>>>>> >>
>>>>>>> >> I’m not sure what you mean by these being different — the whole
>>>>>>> point of the design is that the client would be doing the same thing with
>>>>>>> the access token at the AS that it does with the RS by re-using the access
>>>>>>> token structure. Can you please describe what the differences are, apart
>>>>>>> from the rotation? Presentation of the token and signing of the message are
>>>>>>> identical.
>>>>>>> >>
>>>>>>> >> The client is getting the "access token" from its API. It is not
>>>>>>> using an "access_token" in other API calls to the AS.
>>>>>>> >>
>>>>>>> >>
>>>>>>> >> Rotation of the access token and artifacts for ongoing
>>>>>>> continuation responses is a separate issue to be discussed:
>>>>>>> https://github.com/ietf-wg-gnap/gnap-core-protocol/issues/87
>>>>>>> >>
>>>>>>> >> And for what it’s worth, GNAP is absolutely the right place to
>>>>>>> have new designs — not that this is one.
>>>>>>> >>
>>>>>>> >> You are proposing a new way for an API to provide context for
>>>>>>> subsequent API calls. Looks out of scope to me.
>>>>>>> >>
>>>>>>> >>
>>>>>>> >>> 4) Clients that only want claims from the AS and no access
>>>>>>> tokens will be required to support an API calling mechanism they would not
>>>>>>> have to support otherwise.
>>>>>>> >>
>>>>>>> >> Correct, but the delta between the calls a client would make with
>>>>>>> and without an access token is vanishingly small. The client has to sign
>>>>>>> the initial request in some fashion, and it will sign the continuation
>>>>>>> request in the same exact fashion, but now include an access token in that
>>>>>>> request.
>>>>>>> >>
>>>>>>> >> Per my other point, there is no value to me in my implementations
>>>>>>> of passing context back and forth between the client and AS -- so it is
>>>>>>> extra work providing no value.
>>>>>>> >>
>>>>>>> >> Also, any client authentication mechanism that wants to use the
>>>>>>> HTTP Authentication header is precluded from using it.
>>>>>>> >>
>>>>>>> >>
>>>>>>> >>
>>>>>>> >> Clients making a request to an AS and not getting an access token
>>>>>>> is a new design pattern. I think it has value and should be included, but
>>>>>>> OAuth today shows us the immense value of getting access tokens for calling
>>>>>>> APIs, and so we shouldn’t optimize away from that pattern.
>>>>>>> >>
>>>>>>> >>>
>>>>>>> >>> 5) If the AS does not provide an "access token", there is no
>>>>>>> mechanism for a client to delete the request, as the client is not allowed
>>>>>>> to make a call without an "access token".
>>>>>>> >>
>>>>>>> >> More properly, if the AS does not provide a “continue” field then
>>>>>>> the client can’t delete the request — and yes, that’s intentional. The AS
>>>>>>> is telling this client instance that it can’t do anything else with this
>>>>>>> ongoing request. If the AS wants to allow the client to manage it, it will
>>>>>>> include the mechanisms to do so in the “continue” field.
>>>>>>> >>
>>>>>>> >> There is nuance in that intention. A related concern is that
>>>>>>> deleting a request does not seem like it is a "continue" operation.
>>>>>>> >>
>>>>>>> >>
>>>>>>> >>>
>>>>>>> >>> 6) There is no standard identifier for the request. Debugging
>>>>>>> and auditing are hampered by the client and AS having no standard way to
>>>>>>> identifying a request. While one AS may provide a unique URL for each grant
>>>>>>> request, another AS may use a persistent "access token" to identify the
>>>>>>> grant request, and other ASs may issue a new "access token" on each API
>>>>>>> call, providing no persistent identifier for the request.
>>>>>>> >>
>>>>>>> >> Debugging and auditing this kind of thing are functions of the
>>>>>>> AS. How is interoperability harmed by different ASs having different
>>>>>>> methods to identify their internal data elements? The client doesn’t need
>>>>>>> any knowledge of the AS’s identifiers, it just needs to know the next steps
>>>>>>> for continuing the negotiation.
>>>>>>> >>
>>>>>>> >> Debugging between the client and the AS was what I was referring
>>>>>>> to. How does a client developer identify the request when communicating to
>>>>>>> the AS developer. Seems complicated.
>>>>>>> >>
>>>>>>> >> ᐧ
>>>>>>> >> --
>>>>>>> >> TXAuth mailing list
>>>>>>> >> [e-mail]
>>>>>>> >> https://www.ietf.org/mailman/listinfo/txauth
>>>>>>> >> ᐧ
>>>>>>> >> --
>>>>>>> >> TXAuth mailing list
>>>>>>> >> [e-mail]
>>>>>>> >> https://www.ietf.org/mailman/listinfo/txauth
>>>>>>> >> --
>>>>>>> >> TXAuth mailing list
>>>>>>> >> [e-mail]
>>>>>>> >>
>>>>>>> https://www.google.com/url?q=https://www.ietf.org/mailman/listinfo/txauth&source=gmail-imap&ust=1608345320000000&usg=AOvVaw0r39lH4qVOu0IQPJJYtSpI
>>>>>>>
>>>>>>> --
>>>>>> TXAuth mailing list
>>>>>> [e-mail]
>>>>>> https://www.ietf.org/mailman/listinfo/txauth
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> Dave Tonge
>>>>> CTO
>>>>> [image: Moneyhub Enterprise]
>>>>> <http://www.google.com/url?q=http%3A%2F%2Fmoneyhubenterprise.com%2F&sa=D&sntz=1&usg=AFQjCNGUnR5opJv5S1uZOVg8aISwPKAv3A>
>>>>> Moneyhub Financial Technology, 5th Floor, 10 Temple Back, Bristol,
>>>>> BS1 6FL
>>>>> <https://www.google.com/maps/search/10+Temple+Back,+Bristol,+BS1+6FL?entry=gmail&source=g>
>>>>> t: [phone]
>>>>>
>>>>> Moneyhub Enterprise is a trading style of Moneyhub Financial
>>>>> Technology Limited which is authorised and regulated by the Financial
>>>>> Conduct Authority ("FCA"). Moneyhub Financial Technology is entered on the
>>>>> Financial Services Register (FRN 809360) at *https://register.fca.org.uk/
>>>>> <https://register.fca.org.uk/>*. Moneyhub Financial Technology is
>>>>> registered in England & Wales, company registration number  06909772 .
>>>>> Moneyhub Financial Technology Limited 2019 ©
>>>>>
>>>>> DISCLAIMER: This email (including any attachments) is subject to
>>>>> copyright, and the information in it is confidential. Use of this email or
>>>>> of any information in it other than by the addressee is unauthorised and
>>>>> unlawful. Whilst reasonable efforts are made to ensure that any attachments
>>>>> are virus-free, it is the recipient's sole responsibility to scan all
>>>>> attachments for viruses. All calls and emails to and from this company may
>>>>> be monitored and recorded for legitimate purposes relating to this
>>>>> company's business. Any opinions expressed in this email (or in any
>>>>> attachments) are those of the author and do not necessarily represent the
>>>>> opinions of Moneyhub Financial Technology Limited or of any other group
>>>>> company.
>>>>>
>>>>> Moneyhub Enterprise is a trading style of Moneyhub Financial
>>>>> Technology Limited which is authorised and regulated by the Financial
>>>>> Conduct Authority ("FCA"). Moneyhub Financial Technology is entered on the
>>>>> Financial Services Register (FRN 809360) at
>>>>> https://register.fca.org.uk/. Moneyhub Financial Technology is
>>>>> registered in England & Wales, company registration number 06909772.
>>>>> Moneyhub Financial Technology Limited 2020 © Moneyhub Enterprise, Regus
>>>>> Building, Temple Quay, 1 Friary, Bristol, BS1 6EA
>>>>> <https://www.google.com/maps/search/1+Friary,+Bristol,+BS1+6EA?entry=gmail&source=g>
>>>>> .
>>>>>
>>>>> DISCLAIMER: This email (including any attachments) is subject to
>>>>> copyright, and the information in it is confidential. Use of this email or
>>>>> of any information in it other than by the addressee is unauthorised and
>>>>> unlawful. Whilst reasonable efforts are made to ensure that any attachments
>>>>> are virus-free, it is the recipient's sole responsibility to scan all
>>>>> attachments for viruses. All calls and emails to and from this company may
>>>>> be monitored and recorded for legitimate purposes relating to this
>>>>> company's business. Any opinions expressed in this email (or in any
>>>>> attachments) are those of the author and do not necessarily represent the
>>>>> opinions of Moneyhub Financial Technology Limited or of any other group
>>>>> company.
>>>>>
>>>>>
>>>
>>> --
>>> Dave Tonge
>>> CTO
>>> [image: Moneyhub Enterprise]
>>> <http://www.google.com/url?q=http%3A%2F%2Fmoneyhubenterprise.com%2F&sa=D&sntz=1&usg=AFQjCNGUnR5opJv5S1uZOVg8aISwPKAv3A>
>>> Moneyhub Financial Technology, 5th Floor, 10 Temple Back, Bristol, BS1
>>> 6FL
>>> <https://www.google.com/maps/search/10+Temple+Back,+Bristol,+BS1+6FL?entry=gmail&source=g>
>>> t: [phone]
>>>
>>> Moneyhub Enterprise is a trading style of Moneyhub Financial Technology
>>> Limited which is authorised and regulated by the Financial Conduct
>>> Authority ("FCA"). Moneyhub Financial Technology is entered on the
>>> Financial Services Register (FRN 809360) at *https://register.fca.org.uk/
>>> <https://register.fca.org.uk/>*. Moneyhub Financial Technology is
>>> registered in England & Wales, company registration number  06909772 .
>>> Moneyhub Financial Technology Limited 2019 ©
>>>
>>> DISCLAIMER: This email (including any attachments) is subject to
>>> copyright, and the information in it is confidential. Use of this email or
>>> of any information in it other than by the addressee is unauthorised and
>>> unlawful. Whilst reasonable efforts are made to ensure that any attachments
>>> are virus-free, it is the recipient's sole responsibility to scan all
>>> attachments for viruses. All calls and emails to and from this company may
>>> be monitored and recorded for legitimate purposes relating to this
>>> company's business. Any opinions expressed in this email (or in any
>>> attachments) are those of the author and do not necessarily represent the
>>> opinions of Moneyhub Financial Technology Limited or of any other group
>>> company.
>>>
>>> Moneyhub Enterprise is a trading style of Moneyhub Financial Technology
>>> Limited which is authorised and regulated by the Financial Conduct
>>> Authority ("FCA"). Moneyhub Financial Technology is entered on the
>>> Financial Services Register (FRN 809360) at https://register.fca.org.uk/.
>>> Moneyhub Financial Technology is registered in England & Wales, company
>>> registration number 06909772. Moneyhub Financial Technology Limited 2020 ©
>>> Moneyhub Enterprise, Regus Building, Temple Quay, 1 Friary, Bristol,
>>> BS1 6EA
>>> <https://www.google.com/maps/search/1+Friary,+Bristol,+BS1+6EA?entry=gmail&source=g>
>>> .
>>>
>>> DISCLAIMER: This email (including any attachments) is subject to
>>> copyright, and the information in it is confidential. Use of this email or
>>> of any information in it other than by the addressee is unauthorised and
>>> unlawful. Whilst reasonable efforts are made to ensure that any attachments
>>> are virus-free, it is the recipient's sole responsibility to scan all
>>> attachments for viruses. All calls and emails to and from this company may
>>> be monitored and recorded for legitimate purposes relating to this
>>> company's business. Any opinions expressed in this email (or in any
>>> attachments) are those of the author and do not necessarily represent the
>>> opinions of Moneyhub Financial Technology Limited or of any other group
>>> company.
>>>
>>>
>>> --
>>> TXAuth mailing list
>>> [e-mail]
>>> https://www.ietf.org/mailman/listinfo/txauth
>>>
>>
>>
>> --
>> Dave Tonge
>> CTO
>> [image: Moneyhub Enterprise]
>> <http://www.google.com/url?q=http%3A%2F%2Fmoneyhubenterprise.com%2F&sa=D&sntz=1&usg=AFQjCNGUnR5opJv5S1uZOVg8aISwPKAv3A>
>> Moneyhub Financial Technology, 5th Floor, 10 Temple Back, Bristol, BS1
>> 6FL
>> <https://www.google.com/maps/search/10+Temple+Back,+Bristol,+BS1+6FL?entry=gmail&source=g>
>> t: [phone]
>>
>> Moneyhub Enterprise is a trading style of Moneyhub Financial Technology
>> Limited which is authorised and regulated by the Financial Conduct
>> Authority ("FCA"). Moneyhub Financial Technology is entered on the
>> Financial Services Register (FRN 809360) at *https://register.fca.org.uk/
>> <https://register.fca.org.uk/>*. Moneyhub Financial Technology is
>> registered in England & Wales, company registration number  06909772 .
>> Moneyhub Financial Technology Limited 2019 ©
>>
>> DISCLAIMER: This email (including any attachments) is subject to
>> copyright, and the information in it is confidential. Use of this email or
>> of any information in it other than by the addressee is unauthorised and
>> unlawful. Whilst reasonable efforts are made to ensure that any attachments
>> are virus-free, it is the recipient's sole responsibility to scan all
>> attachments for viruses. All calls and emails to and from this company may
>> be monitored and recorded for legitimate purposes relating to this
>> company's business. Any opinions expressed in this email (or in any
>> attachments) are those of the author and do not necessarily represent the
>> opinions of Moneyhub Financial Technology Limited or of any other group
>> company.
>>
>> Moneyhub Enterprise is a trading style of Moneyhub Financial Technology
>> Limited which is authorised and regulated by the Financial Conduct
>> Authority ("FCA"). Moneyhub Financial Technology is entered on the
>> Financial Services Register (FRN 809360) at https://register.fca.org.uk/.
>> Moneyhub Financial Technology is registered in England & Wales, company
>> registration number 06909772. Moneyhub Financial Technology Limited 2020 ©
>> Moneyhub Enterprise, Regus Building, Temple Quay, 1 Friary, Bristol, BS1
>> 6EA
>> <https://www.google.com/maps/search/1+Friary,+Bristol,+BS1+6EA?entry=gmail&source=g>
>> .
>>
>> DISCLAIMER: This email (including any attachments) is subject to
>> copyright, and the information in it is confidential. Use of this email or
>> of any information in it other than by the addressee is unauthorised and
>> unlawful. Whilst reasonable efforts are made to ensure that any attachments
>> are virus-free, it is the recipient's sole responsibility to scan all
>> attachments for viruses. All calls and emails to and from this company may
>> be monitored and recorded for legitimate purposes relating to this
>> company's business. Any opinions expressed in this email (or in any
>> attachments) are those of the author and do not necessarily represent the
>> opinions of Moneyhub Financial Technology Limited or of any other group
>> company.
>>
>>
>>
>
> --
> Dave Tonge
> CTO
> [image: Moneyhub Enterprise]
> <http://www.google.com/url?q=http%3A%2F%2Fmoneyhubenterprise.com%2F&sa=D&sntz=1&usg=AFQjCNGUnR5opJv5S1uZOVg8aISwPKAv3A>
> Moneyhub Financial Technology, 5th Floor, 10 Temple Back, Bristol, BS1 6FL
> <https://www.google.com/maps/search/10+Temple+Back,+Bristol,+BS1+6FL?entry=gmail&source=g>
> t: [phone]
>
> Moneyhub Enterprise is a trading style of Moneyhub Financial Technology
> Limited which is authorised and regulated by the Financial Conduct
> Authority ("FCA"). Moneyhub Financial Technology is entered on the
> Financial Services Register (FRN 809360) at *https://register.fca.org.uk/
> <https://register.fca.org.uk/>*. Moneyhub Financial Technology is
> registered in England & Wales, company registration number  06909772 .
> Moneyhub Financial Technology Limited 2019 ©
>
> DISCLAIMER: This email (including any attachments) is subject to
> copyright, and the information in it is confidential. Use of this email or
> of any information in it other than by the addressee is unauthorised and
> unlawful. Whilst reasonable efforts are made to ensure that any attachments
> are virus-free, it is the recipient's sole responsibility to scan all
> attachments for viruses. All calls and emails to and from this company may
> be monitored and recorded for legitimate purposes relating to this
> company's business. Any opinions expressed in this email (or in any
> attachments) are those of the author and do not necessarily represent the
> opinions of Moneyhub Financial Technology Limited or of any other group
> company.
>
> Moneyhub Enterprise is a trading style of Moneyhub Financial Technology
> Limited which is authorised and regulated by the Financial Conduct
> Authority ("FCA"). Moneyhub Financial Technology is entered on the
> Financial Services Register (FRN 809360) at https://register.fca.org.uk/.
> Moneyhub Financial Technology is registered in England & Wales, company
> registration number 06909772. Moneyhub Financial Technology Limited 2020 ©
> Moneyhub Enterprise, Regus Building, Temple Quay, 1 Friary, Bristol, BS1
> 6EA
> <https://www.google.com/maps/search/1+Friary,+Bristol,+BS1+6EA?entry=gmail&source=g>
> .
>
> DISCLAIMER: This email (including any attachments) is subject to
> copyright, and the information in it is confidential. Use of this email or
> of any information in it other than by the addressee is unauthorised and
> unlawful. Whilst reasonable efforts are made to ensure that any attachments
> are virus-free, it is the recipient's sole responsibility to scan all
> attachments for viruses. All calls and emails to and from this company may
> be monitored and recorded for legitimate purposes relating to this
> company's business. Any opinions expressed in this email (or in any
> attachments) are those of the author and do not necessarily represent the
> opinions of Moneyhub Financial Technology Limited or of any other group
> company.
>
>