Re: [Json] is it possible to re-register application/json with a profile media type parameter?

["Martin J. Dürst" <[e-mail]>]

Hello Erik,

I think that in terms of process lawyering, I wouldn't worry too much.

What matters is 1) whether this profile parameter really makes sense 
(see separate discussion), and 2) what happens with actual 
implementations when they see a profile parameter.

Regards,   Martin.

On 2013/11/07 8:11, Erik Wilde wrote:
> hello.
>
> during the JSON WG meeting, there was a brief exchange about the
> question whether it would be even possible to use 4627bis to introduce a
> "profile" media type parameter. regardless of whether that's a goal or
> not, i simply wanted to check if there is a problem with it.
>
> http://tools.ietf.org/html/rfc4288#section-4.3 says:
>
> "[...] the names, values, and meanings of any parameters MUST be fully
> specified when a media type is registered in the standards tree, [...]"
>
> as well as:
>
> "New parameters SHOULD NOT be defined as a way to introduce new
> functionality in types registered in the standards tree, although new
> parameters MAY be added to convey additional information that does not
> otherwise change existing functionality."
>
> i believe that this would apply to a "profile" parameter if that
> parameter was based on the RFC6096 model of a profile, meaning that
> anything that represents a profiled media type representation, also
> represents the media type in its unprofiled way.
>
> this means that 4627bis could add a profile media type parameter, and
> the I-JSON draft could use this parameter to expose the I-JSONness of
> representations on the media type. this indication would be optional,
> and old implementations not knowing anything about I-JSON or the profile
> media type parameter would keep working. one possible source of friction
> is that some implementations might get confused by this parameter
> showing up, even though the proper behavior would be to ignore
> unrecognized parameters.
>
> cheers,
>
> dret.
>