Re: [OAUTH-WG] A token review of draft-parecki-oauth-v2-1-01

[Aaron Parecki <[e-mail]>]

Thanks for the detailed review Brian! I've made many of these edits to the
draft, but there are still a few things that need some discussion on the
list. My notes are inline below.


On Tue, Apr 21, 2020 at 3:49 PM Brian Campbell <bcampbell=
[e-mail]> wrote:

> Been working on this on and off for a while now (it's not exactly short at
> 80+ pages, various other priorities, etc.) but wanted to share my thoughts
> from an initial review of the OAuth 2.1 draft before the interim next week
> where it is on the agenda
> https://datatracker.ietf.org/doc/agenda-interim-2020-oauth-06-oauth-01/.
> So for better or worse, here's that review:
>
> Abstract:
> "replaces and obsoletes the OAuth 2..0 Authorization Framework described
> in RFC 6749."
> I think "replaces" is probably unnecessary here and, to be pedantic, is
> arguably inaccurate because published RFCs don't ever go away or get
> replaced.
> Probably should also consider using the official "obsoletes" attribute
> marker too for 6749 https://tools.ietf.org/html/rfc7991#section-2.45.8
> and probably also "updates"/"obsoletes" for others based on the scope of
> the rest of the document (not sure how this even works with a BCP or if you
> can or would want to update or obsolete a BCP) if this work proceeds. That
> scope could be better described in the abstract too as discussed somewhat
> in the thead
> https://mailarchive.ietf.org/arch/msg/oauth/Ne4Q9erPP7SpC5051sSy6XnLDv0/
> and also the 1st paragraph of
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-12.
> What is and isn't in scope is another larger question that I"m not even
> sure how to ask. What's included vs. what's referenced? Should this doc be
> incorporating bits of BCP 212 - OAuth 2.0 for Native Apps? Seems kinda
> antithetical to what a BCP is supposed to be but it's also a bit hard to
> tell how much was used. I mean, what happens if/when the BCP is updated?
> And that kinda begs the question of if it should also incorporate parts of
> or even replace the browser based apps draft? I guess that's a TBD circa
> page 68. There was talk about the device grant being in scope but I'm not
> seeing it (not saying it should or shouldn't be there but it was talked
> about). I dunno exactly but those are some scope questions that come to
> mind.
> Speaking of obsoleting, I do want to ensure that existing extensions and
> profiles of RFC 6749 that use legitimate extension points still present and
> unchanged in OAuth 2.1 aren't inadvertently impacted by this effort. I'm
> not sure how that should work in practice but want to be aware of it as/if
> this work progresses.
>
>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-1
> "is designed for use with HTTP ([RFC2616])."
> I was momentarily proud of myself for knowing that RFC 2616 is obsolete
> but then remembered that the nits tool can automate the check for other
> obsolete or problematic references
> https://tools.ietf.org/idnits?url=https://tools.ietf.org/id/draft-parecki-oauth-v2-1-01.txt
> so take a look at those issues.
> Probably should also check the errata of any documents this one intends to
> update/obsolete or just borrowed a lot of text from to see if anything
> should be applied.
>
>
I've fixed a bunch of these references. It was "fun" to track down where
all the stuff in 2616 ended up as that was split into like 5 RFCs.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-1.1
> "The interaction between the authorization server and resource server is
> beyond the scope of this specification."
> Don't want to try and change the scope but perhaps a mention that things
> like RFC 7662 Token Introspection and self-contained tokens a la JWT exist
> here (or sec 1.4 or 7 or...) would be worthwhile.
>
>
I'm not sure how best to phrase this here since it's so early on in the
draft, but I'm generally in favor of something like this.


>
> https://tools.ietf..org/html/draft-parecki-oauth-v2-1-01#section-1.5
> <https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-1.5>
> "   Steps (3), (4), (5), and (6) are outside the scope of this
>    specification, as described in Section 7."
> But Section 7 incorporates parts of RFC 6750 so being out of scope isn't
> really true.
>
>
Agreed, I dropped that "out of scope" mention.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-1.8
> Not too sure about this section in this document but seems that it should
> at least reflect the fact that some things like RFC 8414  Authorization
> Server Metadata do now exist.
>
>
This section was a last-minute add to the OAuth 2.0 spec and probably needs
to be entirely rethought for this update.


>
> https://tools.ietf.org/html/rfc6749?#section-1.9
> All the cool drafts are doing BCP 14 [RFC2119] [RFC8174] these days.
>
>
Fixed.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-2
> Mentioning the existence of RFC 7591 Dynamic Client Registration here
> seems appropriate. And I think it's be super useful to say that even when
> RFC 7591 isn't being used directly, it (and registered extensions to it
> https://www.iana.org/assignments/oauth-parameters/oauth-parameters.xhtml#client-metadata)
> imply a common general data model for clients that's pretty widely used and
> useful even with so called static registration that "typically involve
> end-user interaction with an HTML registration form."
>
>
I've added some text mentioning the optional dynamic client registration.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-2.2
> "Authorization servers SHOULD NOT allow clients to influence their
>    "client_id" or "sub" value or any other claim if that can cause
>    confusion with a genuine resource owner."
> This text taken from draft-ietf-oauth-security-topic is out of context and
> somewhere between meaningless and confusing here in this document. Removing
> it or maybe something like this might be more appropriate:
>   "Authorization servers SHOULD NOT allow clients to influence their
> "client_id" value in such a way that it might cause confusion with the
> identifier of a genuine resource owner during subsequent protocol
> interactions."
>
>
I like it, I've incorporated this.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-2.3.1
> "The client MAY omit
>       the [client_secret] parameter if the client secret is an empty
> string."
> Let's remove that sentence. As Michael Peck noted in his review, it
> doesn't even make sense in the context of this section describing
> authentication of confidential clients using a password/client secret. And
> that sentence has been the source of some other problems. More detail is in
> this thread
> https://mailarchive.ietf.org/arch/msg/oauth/D1-FrLrSeWrg9M_ca9dbkYeruc4/
>
>
Agreed.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-2.3.2
> "Other Authorization Methods" should be "Other *Authentication* Methods"
> It's probably worthwhile to note or reference the other client
> authentication methods that have been specified since this text was written
> in 6749 and/or point to the OAuth Token Endpoint Authentication Methods
> registry
> https://www.iana.org/assignments/oauth-parameters/oauth-parameters.xhtml#token-endpoint-auth-method
> for them and also maybe mention that, despite the Token Endpoint in the
> name, they are general methods of client auth to the AS when making direct
> requests.
>
>
Sounds reasonable to refer to that registry. I don't know if there is an
"official" way, but I've just mentioned it in the text for now.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-3.1
> "The means through which the client obtains the location of the
>    authorization endpoint are beyond the scope of this specification,
>    but the location is typically provided in the service documentation."
> Maybe add something like "or in the authorization server's metadata
> document [RFC8414]."
>

Sounds good.


> "Request and response parameters
>    MUST NOT be included more than once."
> please clarify that sentence to say:
> "Request and response parameters
> defined by this specification MUST NOT be included more than once."
>

Done.


>
>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-3.1.2
> "   The authorization server MUST compare the two URIs using simple
>    string comparison as defined in [RFC3986], Section 6.2.1."
> There's absolutely no context here for understanding what two URIs are
> being compared.
> Mandating full redirect_uri comparison is maybe more appropriate in 4.1.1
> with the redirect_uri parameter somewhere. And 3.1.2.3. and 3.1.2.2 needs
> some attention with respect to this too.
> But also an exception (for the port part) to full redirect_uri comparison
> is needed for loopback redirect_uris on native clients as in
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-9.6.1 and
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-9.2 and
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-10.3.3
>
>
This is a good point, but is too much for me to try to tackle right now.
Let's revisit this.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-3.1.2.1
> As Michael Peck noted in his review, it's probably okay now to just
> mandate TLS for HTTP redirect URIs.. Although custom scheme or loopback
> redirect URIs for native apps wouldn't use or require TLS.
>
>
This is probably worth discussing on its own.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-3.1.2.2
> <https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-3.1.2..2>
> and
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-3.1.2.3
> Seem to still allow for registration of partial redirection URIs. Which
> isn't gonna work with an exact match requirement.
>
>
After making the edits suggested by Michael Peck, I don't think this
section suggests partial registration anymore.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-3.2
> "Request and response parameters
>    MUST NOT be included more than once."
> please clarify that sentence to say:
> "Request and response parameters
> defined by this specification MUST NOT be included more than once."
>

Done.


> "   The means through which the client obtains the location of the token
>    endpoint are beyond the scope of this specification, but the location
>    is typically provided in the service documentation."
> Maybe add something like "or in the authorization server's metadata
> document [RFC8414]."
>

Done.



>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-3.2.1
> "Client authentication is critical
>       when an authorization code is transmitted to the redirection
>       endpoint over an insecure channel or when the redirection URI has
>       not been registered in full."
> Isn't full registration of redirection URI now required (other than maybe
> the port for native apps) by virtue of requiring a full comparison be done
> when validating the authz request?  And other than a native app using the
> loopback, maybe it's time to move to require that redirection URIs be
> accessed via secure channel?
>

I've taken off the "not been registered in full" part for now.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-4.1.1.2
> "Clients are
>    permitted to use "plain" only if they cannot support "S256" for some
>    technical reason and know via out-of-band configuration that the
>    server supports "plain"."
> With code_challenge_methods_supported from Authorization Server Metadata /
> RFC 8414 it doesn't have to be out-of-band anymore.
>

I mentioned RFC8414 here.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-4.1.2
>  " Typically, the "code_challenge" and "code_challenge_method" values
>    are stored in encrypted form in the "code" itself but could
>    alternatively "
> 'Typically' - really?
>

Agreed, I rephrased this. Though I am curious to do a proper survey of
implementations to see if anyone is doing this at all.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-4.1.3
> "   *  ensure that the "redirect_uri" parameter is present if the
>       "redirect_uri" parameter was included in the initial authorization
>       request as described in Section 4.1.1, and if included ensure that
>       their values are identical."
> The reference should be to 4.1.1.3.
>
> Fixed.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-4.1.4
> "   An example successful response:
>    HTTP/1.1 200 OK
>    Content-Type: application/json;charset=UTF-8
>    Cache-Control: no-store
>    Pragma: no-cache"
> Pretty sure application/json shouldn't have a charset (see note at end of
> section https://tools.ietf.org/html/rfc8259#section-11)
> and I've long thought that "Pragma: no-cache" shouldn't be there (see
> https://mailarchive.ietf.org/arch/msg/oauth/9DdkE2P0RrUZMeZAbdf3NrMfy0w/
> )
> Note that these apply to most of the example responses in the draft.
>
>
I've removed the charset from the examples. I'm not sure about no-cache.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-4.3
> " specifying the grant type
>    using an absolute URI (defined by the authorization server) as the
>    value of the "grant_type" parameter"
> The words in the parenthetical have led to questions in AD review of
> documents making use of the grant type extension point (see
> https://mozphab-ietf.devsvcdev.mozaws.net/D4278#inline-1581) and I think
> "understood by the authorization server" might be better phrasing or even
> removing the parenthetical all together.
> RFC7522 is mentioned here as an example extension grant but maybe worth
> also including mention of others like RFC7523, RFC8628, RFC8693, and maybe
> even non IETF ones.
>

Suggestions welcome here.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-6
>  "  *  _Sender-constrained refresh tokens:_ the authorization server
>       cryptographically binds the refresh token to a certain client
>       instance by utilizing [I-D.ietf-oauth-token-binding] or [RFC8705]."
> Given the relative immaturity of ways to do this, maybe something more
> open ended would be appropriate? This reads like token-binding or MTLS are
> the only ways allowed. I'd think wording that would allow for DPoP or some
> yet-to-be-defined method would be better here. Also maybe drop the
> token-binding reference all together (it's long expired and doesn't look
> like that's gonna change).
>

I'm generally supportive of something more open ended here. Would that also
suggest that the corresponding text in the Security BCP be updated as well?



> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-7.4.4
> The SHOULDs/RECOMMENDEDs in this section seem a little overzealous and/or
> too specific.
>
>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-9.4
> Seems to be a lot of overlap and duplication between 9.4. Access Tokens
> (under 9. Security Considerations) and section 7.4. Access Token Security
> Considerations, which could/should maybe be reconciled.
> "(#pop_tokens)" hints that some text was copied from elsewhere but the
> markdown references weren't fixed/updated
>
>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-9.5
> "(#refresh_token_protection)" same thing about markdown references not
> fixed/updated
>
>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-9.6
> "(#insufficient_uri_validation)", "(#mix_up)",
> "(#open_redirector_on_client)","(#csrf_countermeasures)", again
> "(#mix_up)", and "(#redirect_307)"
>

I'm not sure how this happened but it'll take a bit to track these down.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-9.7
> "Attacker A4 in (#secmodel)"
> "The use of PKCE is RECOMMENDED to this end."
> PKCE is required elsewhere so this doesn't seem quite right. Similar
> comments about text ijn
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-9.14 that
> talks as though PKCE might not be there.
>

I believe this text was included since an extension such as OpenID Connect
can define other methods of authorization code injection. Also note that
this same sentence is in the Security BCP so if we update anything here
they should probably match.


>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-9.17.1
> "(#redir_uri_open_redir)"
>
>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-9.20
> "(#client_impersonating)"
>
>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#section-12
> "   *  Redirect URIs must be compared using exact string matching as per
>       Section 4.1.3 of [I-D.ietf-oauth-security-topics]"
> Should that maybe be qualified to cover dynamic ports on ephemeral local
> http servers used for redirect URI with native clients?
> BTW, does [I-D.ietf-oauth-security-topics] need to make a similar
> allowance?
>

Sounds like there are a few things around redirect URIs that need to be
clarified in both documents.


>
>
> https://tools.ietf.org/html/draft-parecki-oauth-v2-1-01#appendix-C
> "TBD"
> Given the potentially high visibility of an OAuth 2.1 effort, I think it'd
> be worthwhile to list organizational affiliations of individuals here in
> the acknowledgements along with their names. Something like what was done
> in the first part of
> https://tools.ietf.org/html/draft-ietf-oauth-access-token-jwt-06#appendix-A.
> This can help with visibility and justification of (sometimes not
> insignificant) time spent on the work by non-authors/editors.
>
>
>
>
>
> *CONFIDENTIALITY NOTICE: This email may contain confidential and
> privileged material for the sole use of the intended recipient(s). Any
> review, use, distribution or disclosure by others is strictly prohibited..
> If you have received this communication in error, please notify the sender
> immediately by e-mail and delete the message and any file attachments from
> your computer. Thank you.*_______________________________________________
> OAuth mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/oauth
>