Open gkellogg opened 9 months ago
Do serialization syntax documents need to distinguish non-Basic constructs? I don't think so - it should suffice to have a section in Concepts on Basic and state that the behaviour of Basic-compliant systems on syntax that would produce quoted triples is undefined. Alternatively Basic-compliant systems might be required to reject syntax that would produce quoted triples.
I agree and would prefer to not have to touch every serialization spec. If the processing mode is RDF 1.2 Basic it might be better if a graph containing quoted triples is detected so that we can test this case. Note that, depending on the outcome of the blank graphs (or whatever) discussion, this could be moot, if the use of graphs in place of quoted triples becomes a purely semantic distinction and does not impact the abstract syntax.
From https://github.com/w3c/rdf-star-wg/issues/19#issuecomment-1749041899, the group resolved the following:
Serialization syntaxes need to distinguish the portions of the grammar that are dependent on this profile.