Closed pchampin closed 2 years ago
Suggestion: as mentioned in the 2021-10-15 telecon:
Split the vocabulary box into two - one box for the basics, rdf-star:triple, rdf-star:graph, rdf-star:source and one for TEP.
Then all TEP can be put in on place.
Given likely further discussion of TEP and its semantics, putting TEP separately helps put int in one place and does not lose sight of the the more basic concepts.
This was discussed during today's meeting: https://w3c.github.io/rdf-star/Minutes/2021-10-15.html#t05
I think that all comments above are now addressed.
I am generally happy with this PR. However, I would prefer a different order for the subsections of Sec.6.4. In particular, I would prefer to move Sec.6.4.6 (Alternatives to referential opacity) before Sec.6.4.5 (Selective referential transparency), in which case the following sentence may be added to the very end of Sec.6.4.6 (then Sec.6.4.5):
Another option is to allow users to enable referential transparency explicitly for selected cases while still using the current semantics for the general case. The next section introduces an approach to this end.
If we don't find consensus for swapping these subsections, then the first sentence of Sec.6.4.6 should be changed as follows: Instead of "..., as described in the previous section, ..." it should say "..., as described in § 6.4.4 Referential opacity, ..."
The order you suggest was actually the original order (hence the wrong "in the previous section" in "6.4.6 Alternatives"...
The reason I changed it was to have a decreasing "gradient of maturity" among those 3 sub-sections
For this reason I would rather keep the current order, fixing the introductory sentence of 6.4.6 -- and possibly renaming it to "Other alternatives to referencial opacity".
@pchampin thanks for explaining your rationale for putting the subsections in the order in which they are now. I am fine with it. So, the only thing that remains from my perspective is to change the first sentence of Sec.6.4.6 as mentioned in the last paragraph of my previous comment.
This was discussed during today's call: https://w3c.github.io/rdf-star/Minutes/2021-10-29.html#x117
@hartig I was about to change the introduction sentence of 6.4.6, but I realized why I kept it that way. Actually, Section 6.4.4 does not anymore discuss the drawbacks of ref-opacity. The paragraph that contained that discussion has been moved as the introduction of 6.4.5:
Referencial opacity is well suited for use-cases where the syntactical form of the annotated statements may be significant (like annotating statements from other graphs or annotating commit deltas). On the other hand, other use-cases would expect some degree of referential transparency. Consider the case of attributed/evidenced triples. The annotation is understood to be about the fact asserted by the triple rather than the triple itself. In this use case, one might then expect more inferences than provided by the base semantics.
Still, I made the reference clearer, hopefully.
Okay, thanks.
The semantics still needs to be adjusted to also support the semantics of rdf-star:Triple. (and renamed to something more generic)
Preview | Diff