[core] Separate formats for fetch and patch (was: WG Last Call of draft-ietf-core-senml-etch-03.txt)

[Christian Amsüss <[e-mail]>]

On Mon, Jun 24, 2019 at 01:21:31PM +0200, Klaus Hartke wrote:
> General questions:
> 
> * It seems the new data formats are supersets of the existing data
> formats. Also, it seems the same data format is used for fetching and
> patching data. Is this what we, as CoRE WG, recommend (1 media type
> for data, 1 media type for fetching and patching)? CoRAL is doing it
> differently at the moment.

I'd view this not as a general recommendation but as a way to go if for
some reason the "descriptive" format can not be extended within its own
constraints. In CoRAL, I don't see a need for separate media types.
Applications mint their own dictionaries anyway.

An application can make restrictions like in [1] for some operations
(eg. "an Information Provider must only use dictionary I on resource
representations returned from GET, and may accept dictionary F for
FETCH-like operations if supported").

Best regards
Christian

[1]: https://tools.ietf.org/html/draft-hartke-t2trg-coral-08#section-6.1.3

-- 
To use raw power is to make yourself infinitely vulnerable to greater powers.
  -- Bene Gesserit axiom