> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On Behalf
> Of Toby White
> Sent: Sunday, February 20, 2011 8:12 AM
> To: OAuth Mailing List
> Subject: [OAUTH-WG] OAuth2 queries
> 
> Some more OAuth2 queries (based on reading -13)
> 
> * There's no direct indication of expected behaviour after a failed attempt to
> refresh a token. Since a refresh_token workflow is an example of issuing an
> access token, should I read the error behaviour describe in 5.2 as applying to
> section 6 as well?

" If the request
   failed verification or is invalid, the authorization server return an
   error response as described in Section 5.2."

> * Can I read throughout the spec, that wherever scope is an OPTIONAL
> parameter, I can read a request with an absent scope parameter as being
> equivalent to a request with a scope="" (empty string) parameter?

Yes, but an empty string doesn't necessarily means no scope. That's 
implementation specific. There is no definition of default scope and I would 
expect most services to define it as something other than nothing.

> * In 4.2.2 (implicit grant, response from the auth server), the response
> includes an OPTIONAL scope parameter, whose definition includes
> 
> " The authorization server SHOULD include the parameter if the requested
> scope is different from the one requested by the client."
> 
> I'm not quite sure I'm reading this correctly, but I think that implies that 
> if the
> client requests scope A, the server can actually grant scope B.
> 
>  - is it required/suggested that scope B be a subset of scope A?

No. The client can ask for oranges and you give it apples.

>  - is it true of any other workflows that a client requesting one scope can
> actually be granted another scope?

Yep. That's implementation specific.

EHL
_______________________________________________
OAuth mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/oauth

Reply via email to