search

w3c / rdf-semantics

https://w3c.github.io/rdf-semantics/
Other
6 stars 2 forks source link

a possible extension of RDF's Simple semantics supporting quoted triples #14

Open pchampin opened 1 year ago

pchampin commented 1 year ago

mostly for discussion at this point

(I already discussed it briefly with @doerthe and @Antoine-Zimmermann )

Preview | Diff

pchampin commented 1 year ago

the only changes for the moment are in the 'Simple interpretation' section, and are highlighted to be easy to spot.

pfps commented 1 year ago

I don't think it is a good idea to have suggestions on technical changes be made in pull requests. My view is that technical changes should be suggested instead in issues or on the working group email (I prefer email) and discussion should happen there, not in a pull request.

If someone wants to create a version of a document to support their suggestion then this can be done without creating a pull request.

TallTed commented 1 year ago

I don't think it is a good idea to have suggestions on technical changes be made in pull requests.

PRs are a good way to collaborate on textual edits, technical or otherwise.

Such PRs should generally come after some discussion of the proposed change in an issue, but this need not delve deeply into the technicalities nor exact text changes.

It makes sense to me to use PRs to propose document revisions, in large part because the document markup is often important to full comprehension of those proposals, and the PR Preview is only available via PR -- not via email, issue, or other discussion medium.

pfps commented 1 year ago

I'm not sure in what respect this proposal is minimal. There is a considerable amount of machinery here. A truly minimal proposal might treat quoted triples just like a new datatype with a lexical to value mapping that maps into a triple of strings. A different minimal proposal might treat quoted triples as events. To determine what a minimal proposal should look like requires determining what quoted triples are supposed to be used for.

gkellogg commented 1 year ago

Discussing quoted triples in semantics really depends on the core concepts being defined. I added https://github.com/w3c/rdf-concepts/issues/23 to track this, although I think the group has a fair amount of work to do before settling on core concepts.

pchampin commented 1 year ago

I'm not sure in what respect this proposal is minimal.

Granted. I updated the title of the PR accordingly. The goal of this proposal is to be essentially equivalent to the semantics specified in the CG report through the unstar mapping (although in some corner cases, it might differ).

Discussing quoted triples in semantics really depends on the core concepts being defined.

Agreed. The underlying assumption here is that the WG would adopt the changes made by the CG to the abstract syntax.

pfps commented 1 year ago

An extract of PR showing only the parts relevant for the semantics would make it much easier to see what is going on and make it easier to compare with other proposals.

pfps commented 1 year ago

Suppose we have a class for triples. Then

:john :said :s . :s rdf:type rdf:Triple .

does not entail

:john :said << :a :b _:c >> .

because there might not be any ground triple that is mapped by IT to the domain element that John said.