Re: [core] Opsdir last call review of draft-ietf-core-problem-details-04

[Carsten Bormann <[e-mail]>]

Hi Joel,

> On 2022-06-06, at 19:44, Joel Jaeggli via Datatracker <[e-mail]> wrote:
> 
> Reviewer: Joel Jaeggli
> Review result: Ready
> 
> I reviewed draft-ietf-core-problem-details on behalf of the  ops directorate. I
> nsummar y this draft is largely ready. I have one perhaps clarifying question.
> 
> in the regards to the following statement:
> 
>   Consumers of a Concise Problem Details data item MUST ignore any
>   Custom Problem Detail entries, or keys inside the Custom Problem
>   Detail entries, that they do not recognize; this allows Custom
>   Problem Detail entries to evolve and include additional information
>   in the future.  The assumption is that this is done in a backward and
>   forward compatible way.
> 
> This seems like less of a gesture at compatibility as opposed to simply
> ignoring conditions that would otherwise produce errors by the receiving
> parties. it would see likely that coap problem detail collectors may collect
> such data for processing by other systems since the whole collection pipline
> may not move in lock step or doesn't it?

Indeed, the text only describes that an unrecognized entry should not cause an error in a “Consumer”, not how to deal with it in further processing/forwarding.

I have attempted to clarify (and refactor) this in

https://github.com/core-wg/core-problem-details/pull/27

I think the SHOULD (RECOMMENDED) is justified because of systems/protocols that have their own problem details formats.

Grüße, Carsten