Misleading introductory text in "Terminology" subsection

[pchampin]

The "Terminology" subsection starts with a sentence saying:

This document uses the following terms as defined in JSON [RFC8259]. Refer to the JSON Grammar section in [RFC8259] for formal definitions.

While this was true in the 1.0 spec, this section has been augmented with many terms that are not defined in JSON, but are specific to JSON-LD.

I would recommend rewriting common/terms.html to split it into two separate subsubsections, one with "borrowed" terms from RFC8259, and one with original terms.

NB: this also applies to other specs (API, Framing).

[pchampin]

Meta-question: is it proper etiquette if I tag this issue as spec:editorial, or is it the prerogative or chairs?

[gkellogg]

Its clearly editorial, and editors regularly make this determination. Of course, if contested, it can be determined not to be editorial.

[gkellogg]

I say we change the introduction to be more appropriate. The terms are already split into different parts, but not sections. We could use sub-section for these different bits:

• Primitive datatypes from JSON, INFRA, and ECMASCRIPT. The only thing in that first bit we defined is the internal representation, which seems appropriate here.
• Terms defined in other specs, such as RDF, IRI. We might include here our own terms, such as base IRI, compact IRI, relative IRI, and JSON Literal.
• JSON-LD specific terms.

Perhaps we could say something like "From ECMASCRIPT we use the following terms:", and so on for the specs we import from.