Given the consensus that we reached in yesterday's call, here is a proposal on how to change the relevant definitions in RDF Concepts in order to capture Option 3.
Some remarks about this PR:
I am currently using the term "triple term" in this proposal while the term "triple descriptor" has been floating around as an alternative name for the same concept. Personally, I like "triple term" more but, of course, I am open to replace "triple term" by "triple descriptor" in this PR if there is consensus in the group towards the latter.
At the end of yesterday's call, there was a brief discussion whether the definition of triple terms should be recursive (i.e., whether a triple term may contain another triple term). The current definition in this PR is recursive. I defined it this way because of @pchampin's argument that, without recursion, it becomes impossible to talk about rdf:nameOf triples which have a triple term in their object position. Of course, if the group decides against such a recursive definition, this PR can easily be adapted to make the definition non-recursive.
I have separated the definitions of "RDF triple" and "triple term" from one another even if these two definitions are essentially the same (syntactically, an RDF triple and a triple term are exactly the same thing). My rationale for this separation is to make it more obvious that these are two different concepts. I have also added a brief 'Note' on this topic.
I have removed the definition of "asserted triple" because I think, by clearly separating the notion of an RDF triple from the notion of a triple term, it becomes obsolete to talk about asserted triples. Every triple (in a graph) is asserted; triple terms are not members of graphs.
This PR changes only the definitions. I explicitly decided not to touch Section 1.3, which (currently) provides an informal introduction of quoted triples. We will certainly have to change that section as well, but I think that should be a different PR. We should first make sure that we have the definitions right.
Given the consensus that we reached in yesterday's call, here is a proposal on how to change the relevant definitions in RDF Concepts in order to capture Option 3.
Some remarks about this PR:
rdf:nameOf
triples which have a triple term in their object position. Of course, if the group decides against such a recursive definition, this PR can easily be adapted to make the definition non-recursive.Preview | Diff