Hi all,

This is a follow-up on the presentations last Monday (oauth) & Thursday (wimse) 
on Delegated Authorization 
(https://datatracker.ietf.org/doc/draft-li-oauth-delegated-authorization/, 
slides at 
https://datatracker.ietf.org/meeting/124/materials/slides-124-oauth-sessa-delegated-authorization-00).
 Thanks to all who provided valuable feedback!

We have created a GitHub repo for the draft at 
https://github.com/liuchunchi/li-oauth-delegated-authorization.

Below are some clarifications on how this draft differs from other delegation 
solutions.

This draft vs token exchange and its derivatives (Transaction Tokens, Identity 
Chaining, Identity Assertion, ...):

*         Both can support:

o    downscoping

o    txn binding

o    reducing lifespan / single-use

*         Advantages of this draft:

o    It enables local downscoping at the client or middle parties along the 
call chain -> can be useful for large scale delegation / constrained 
environments.

o    The delegator can obtain the delegatee's public key and bind it to the 
sub-token (dt/dat) before sending it to the delegatee, to prevent unauthorized 
use of the sub-token if leaked.

?  (side note: perhaps this can be applied to token exchange as well in a 
separate draft?)

o    Delegation can happen outside the trust boundary of the target resource 
(compared to txn tokens), no federation needed (compared to identity chaining).

*         Limitations:

o    It does not allow adding additional asserted values in the new token

Another alternative is passing access/refresh token around, which has similar 
properties as above.

Look forward to comments and collaborations!

Best Regards,
Ruochen
_______________________________________________
OAuth mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to