Re: [media-types] Media subtypes containing "+"

[Manu Sporny <[e-mail]>]

On 12/24/20 8:41 AM, Graham Klyne wrote:
> 1. In the discussion of suffix registrations, the draft says "each 
> subset of suffixes, starting from the right-most suffix, MUST be 
> individually registered as a subtype in its own right" - I think this
> is incorrect use of the term "subtype" as used by the MIME
> specification. RFC6838 uses the phrase "media type name suffix"
> and/or "Structured Syntax Suffix Registration Template".
> 
> https://www.rfc-editor.org/rfc/rfc6838.html#section-6
> 
> It's not clear to me what the best term would be, but I'd probably
> aim for something like "media type name suffix" or "subtype suffix"
> if that's clear from context.

Ok, can do... we'll probably go with the more precise "media type name
suffix".

> 2. I think an example would help to show what is being proposed (I
> think the description is clear enough, but it's always good to have
> something concrete to cross-check against).

Yes, agreed, we'll add an example and use a real one as an example --
application/did+ld+json

> In particular with reference to registration of multiple subtype
> suffixes used together.
> 
> E.g. in the case of refinements of application/ld+json (per 
> https://w3c.github.io/json-ld-syntax/#iana-considerations), one
> could imagine more specific types defined; e.g.
> 
> application/invoice+ld+json
> 
> Per the description, I think this would require the following
> suffixes have been registered:
> 
> +json +ld+json
> 
> But not:
> 
> +ld

Yes, that is correct. We may want to put additional language in there
that states that you can only build valid media type names from a
structured suffix from right to left (not left to right).

> And if this data may be offered with alternative RDF syntaxes, such
> as, e.g.:
> 
> application/invoice+ld+rdf+xml
> 
> then:
> 
> +xml +rdf+xml +ld+rdf+xml

Yes, agreed.

> 3. A nit:  the draft makes reference to and uses RFC2119 language.
> It is usual in RFC documents I've seen to explicitly call this out in
> the document.  E.g.
> 
> https://www.rfc-editor.org/rfc/rfc6838.html#section-1.2

Ok, will do, thanks Graham!

-- manu

-- 
Manu Sporny - https://www.linkedin.com/in/manusporny/
Founder/CEO - Digital Bazaar, Inc.
blog: Veres One Decentralized Identifier Blockchain Launches
https://tinyurl.com/veres-one-launches