TallTed
commented
1 year ago RDF-protostar is enough for interoperability with labelled property graphs and in the end that's all what RDF-star is about, isn't it?
Short answer: No.
Very long answer: Read the entirety of the email threads from the RDF-star Focus Group of the RDF-DEV Community Group, and related threads on other mailing lists, from the past few years, since the initial RDF* paper was published.
Medium-length answer: I regret that I do not have the spare cycles to predigest the latter into a medium-length answer for you.
Specification of RDF-protostar
An RDF-protostar triple is a 3-tuple defined as follows:
By definition, every RDF-protostar triple is a RDF-star triple as well.
An RDF-protostar graph is a set of RDF-protostar triples. Every RDF-protostar graph is a RDF-star graph.
Background and Motivation
Recursively nested quoted RDF-star triples are nice in theory but complex both in theory and practice. Data models should be comprehensible. Statements about statements are almost comprehensible but statements about statements about statements are not. Recursively quoted triples might have their use-case in theory but for the average data-wrangler they likely just add another pain point to nasty aspects of RDF such as blank nodes, collections, reification, inference, http-303-redirects, and an endless number of serialization forms.
RDF-protostar is enough for interoperability with labelled property graphs and in the end that's all what RDF-star is about, isn't it? Use cases of deeply nested quoted triples are surely as detached from practice as the famous Pizza and Wine ontologies. By the way all examples in the Final Community Group Report are non-recursive RDF-protostar triples. Nested quoted triples are only exemplified at one place.