There is a tradeoff with the readability of the diagram. It's already a busy one

the diagram itself does not perturb me. the notion that it adequately conveys a real problem does. the same concern applies to the rdf-star report.

I think RDF* attempts to solve the problem that software engineers find LPGs more appealing because they can quickly qualify edges without giving much thought to domain modeling. RDF* allows one to more easily punt on domain modeling in RDF too.

RDF* is advertised as supporting LPGs but the CG report goes into a totally different direction: versioning, explainable AI, syntactic fidelity, triple types instead of actual occurrences. LPGs is the last thing the proposed semantics supports. So it will be ignored. One more...

RDF-star already covers very well what developers are used to do with edge properties in LPG. I don't feel like judging whether that's good or bad thing to do. The CG dives deep into many semantic aspects - something people in the RDF camp us to do. I don't see this harmful

"harmful" is not the question. try "useful". - does anything more than "what developers are used to [doing] with edge properties" - that is the "deep ... semantic aspect", matter? - are the real things one does with rdf supported?

Yes, @Synaptica uses @OntotextGraphDB support for #RDF-star for access control management in their Graphite knowledge management suite. The benefits in terms of simplicity of modeling and smaller storage footprint are clear. ontotext.com/blog/rdf-star-i…

Again unasserted assertions. Really, before the RDF-star CG I had no idea that unasserted assertions are such a thing among practitioners. Why did the proposals for an RDF literal datatype didn't get more support? Much more straightforward encoding of unasserted assertions IMHO

The problem with modelling quoted triples as literals (with a dedicated datatype) would have been blank nodes. You would not have been able to model something like : <#alice> foaf:knows _:bob {| ex:since 2019 [}. _:bob foaf:name "Bob".

Or would you? Isn't the example rather a case for Quoted Graphs? "Blank nodes belong to _one_ surface, end of story". Then i.e. skolemize into RDF-typed literal, end of story. No trouble with semantics. Your example above has wiggle room for strangely differing interpretations.