Re: [Ace] [EXTERNAL] Francesca Palombini's Discuss on draft-ietf-ace-oauth-authz-38: (with DISCUSS and COMMENT)

On 2021-06-09, at 12:45, Ludwig Seitz <ludwig.seitz@combitech.com> wrote:
> 
> Hello Carsten,
> 
> Can you clarify what exactly you consider is broken and why?

What: attaching the representation choice to the protocol choice
Why: Because the protocol may transport data that were generated with another protocol (or no particular protocol) in mind.

So, recommending as a default choice is fine, but saying that the protocol choice dictates the representation choice is too limiting.  Maybe the MQTT people discover CBOR at some point :-)

Grüße, Carsten

> 
> I was indeed trying to attach the 'when' to both arms (token encoding and protocol message payload encoding), but I don't have a strong opinion on this. If the WG can decide how this should be I will implement.
> 
> /Ludwig
> 
> -----Original Message-----
> From: Carsten Bormann <cabo@tzi.org> 
> Sent: den 9 juni 2021 09:15
> To: Ludwig Seitz <ludwig.seitz@combitech.com>
> Cc: Francesca Palombini <francesca.palombini@ericsson.com>; Seitz Ludwig <ludwig.seitz@combitech.se>; The IESG <iesg@ietf.org>; art-ads@ietf.org; ace-chairs@ietf.org; draft-ietf-ace-oauth-authz@ietf.org; ace@ietf.org
> Subject: Re: [Ace] [EXTERNAL] Francesca Palombini's Discuss on draft-ietf-ace-oauth-authz-38: (with DISCUSS and COMMENT)
> 
> 
>> In 2021-06-09, at 08:42, Ludwig Seitz <ludwig.seitz@combitech.com> wrote:
>> 
>> " ... size.  Self-contained tokens and protocol message payloads are encoded in CBOR when CoAP is used.”
> 
> This is not what the old NEW text says.
> 
> (The new NEW text attaches the “when” to both arms.)
> 
> The whole idea of attaching the representation choice to the protocol choice is broken, but if we pursue it, we at least need to make the logic clear.
> 
> (1) If you use CoAP, you use CBOR for protocol message payloads.
> (2) Self-contained tokens use CBOR.
> (3) No other hard limitations are implied, but of course CBOR is the format of choice to maximize interoperability, so deviations from that need to be justified.
> 
> Grüße, Carsten
> 
> _______________________________________________
> Ace mailing list
> Ace@ietf.org
> https://www.ietf.org/mailman/listinfo/ace