Use Case: Concept of Graph/Dataset without quoted triples

[gkellogg]

As an specification editor,
I want to be able to refer to the concept of a graph or dataset not including quoted triples,
So that algorithms such as that defined in RDF Dataset Canonicalization can be written to so as to not depend on the concept of triples as a resource (quoted triples).

This concept has been described as an "RDF 1.1 Dataset", but that may not include other desired components of an RDF 1.2 Dataset, hypothetically including support for text direction or other concepts other than embedded/quoted triples.

[TallTed]

Perhaps "Simple {RDF, RDF Dataset, etc.}" and "Complex {RDF, RDF Dataset, etc.}" might serve the purpose?

[afs]

See #19.

[afs]

The reason there is "RDF-star triple" is because the CG report is separate from RDF 1.1 and was speaking about RDF 1.1.

RDF 1.2 , SPARQL 1.2, has "triple". The definition is a triple and "triple" is used through this and other specs.

algorithms such as that defined in RDF Dataset Canonicalization

rdf-canon is already explaining it is URDNA2015 which implies classic RDF. That is the right place to explain that it targets RDF 1.1 while taking the canonicalizaed N-Quads. It helps the rdf-canon reader.

[gkellogg]

Because of “full” and “simple” conformance levels, we need a way to talk about triples, graphs, and datasets without quoted triples. Other terminology is defined as a consequence. The suggestion on terminology has been around for a month without further comment, until w3c/rdf-concepts#32. Happy to bikeshed the names, which is why the PR was labeled needs discussion.

[afs]

"RDF-star triple" is due to the CG report.

We haven't fixed on “full” and “simple” yet. We are waiting on action #19.

It is not clear we need a name of "full" at all - it just is RDF 1.2 as covered by concepts and semantics so at best it is emphasis, not defined terminology.

The restricted form is whether quoted triples are in the set of RDF terms. Triples are not changed, only constitute parts.

around for a month

This issue may have been around; little discussion indicates not noticed - may be because of waiting for #19 (This is not a "use case" as the WG uses the phrase.) The WG has a backlog of things to discuss. We have to live with that.

The PR is says terminology from the CG report. It isn't.

[gkellogg]

Such concepts already exist with the definition of Full conformance and *Classic conformance.

We could consider terms such as "Fully conforming RDF Graph/Dataset" and "Classically conforming RDF Graph/Dataset".

[pchampin]

We could consider terms such as "Fully conforming RDF Graph/Dataset" and "Classically conforming RDF Graph/Dataset".

yes, that's how I used it in w3c/rdf-concepts#115

@gkellogg Can we close this issue, then? Or do you think that more specific terminology should be explicitly added?