Re: superset and subset specification

po 10. 7. 2023 v 20:15 odesílatel Nathan Rixham <nathan@webr3.org> napsal:

> On Mon, Jul 10, 2023 at 6:07 PM Nathan Rixham <nathan@webr3.org> wrote:
>
>> ... unless either:
>> a) a superset specification was designed, which essentially says any
>> <uri> which dereference to an RDF response that asserts <uri> a :Agent
>> (where :Agent is a well specified class in a published ontology) w/ note
>> MAY 303 to handle range-14, or
>> b) a subset specification was designed, as above but constrained to a
>> single media type json(-ld)
>>
>> My personal opinion would be either to let it just move to solid and kill
>> this group, or take some consensus to scrap the current specification, and
>> produce (a) + (b) above, where (a) is deferred to by solid and anything
>> else implementing webid, and (b) is a subset which allows parties to
>> produce a very specific set of tooling, webid implementations that are only
>> json-ld.
>>
>
> This could be merged to one specification, which also defined an open
> ended list of sub specifications, as such:
>
> WebID broadly defines a <uri> which dereferences to an RDF response that
> asserts <uri> a webid:Agent, is a webid. (note about 303)
>
> WebID also defines an open ended list of sub specifications, where for
> each valid rdf response type, webid-{type} is an implementation which is
> constrained to require only that specific type.
>
> With that, we'd cover all bases, and webid-turtle, webid-jsonld, and many
> more, would automatically fall out.
>
> The specification would likely never need to be updated, be quite concise,
> and require only the publication of a simple vocabulary to cover
> webid:Agent, or some such universal term.
>
> The current webid specification, would be superseded by both WebID, and
> its inferred subspecification WebID-Turtle.
>

+1 everyone gets what they want, and it's future proofed for many years