Could you point which part of the spec specifies this (am looking at draft 10)? 
In any case, I would expect the auth server to include the scopes granted in 
the access token response to avoid any ambiguity.

On Nov 29, 2010, at 8:40 AM, Eran Hammer-Lahav wrote:

> #2. Asking for scope on the access token call can only reduce the already 
> approved scope.
>  
> EHL
>  
> From: [email protected] [mailto:[email protected]] On Behalf Of 
> Anton Panasenko
> Sent: Friday, November 26, 2010 10:54 AM
> To: [email protected]
> Subject: [OAUTH-WG] OAuth 2.0 server behavior
>  
> Hi,
>  
> What behavior is expected from the server, if in the query on access_token 
> without "scope" 
> (grant_type=authorization_code&client_id=s6BhdRkqt3&client_secret=gX1fBat3bV&code=i1WsRn1uB1&redirect_uri=https%3A%2F%2Fclient%2Eexample%2Ecom%2Fc)?
>  
> 1. The server must generate access_token for an empty scope.
> 2. The server must generate access_token for scope, which was approved for 
> access_code.
>  
> --
> Sincerely yours
> Anton Panasenko
> Skype: anton.panasenko
> Phone: +79179838291
> Email: [email protected], [email protected]
>  
> _______________________________________________
> OAuth mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/oauth

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

Reply via email to