Re: RDF-star “baseline” document

On Thu, Jun 6, 2024 at 7:00 PM Franconi Enrico <franconi@inf.unibz.it> wrote:
>
>
>
> > On 6 Jun 2024, at 18:58, Peter F. Patel-Schneider <pfpschneider@gmail.com> wrote:
> >
> > There has been a "minimum" baseline around for quite some time.  Its current incarnation is at https://github.com/w3c/rdf-star-wg/wiki/RDF-star-profile-%22transparent%22
>
> I agree that the above could be considered really as the bare minimum baseline.

Yes, well it did appear to be the baseline some months ago. Then
objections were raised (on the list) that the reifiers were
many-to-many, and thus too flexible (an effect of this approach being
simpler).

Open issues (without numbers) seem to be:
1. Do reifiers capture the notion of statements about statements or
not? (Debated on the list; nothing resolved in a meeting.)
2. They might not be "isomorphic with LPG expectations'' (reifiers are
more powerful, LPGs look more like a kind of singleton property
pattern, from an RDF POV).
3. Reifiers are arguably a superset of classical RDF reifications (but
not fully isomorphic); the latter being "tokens of statements" (albeit
not always used/understood as such), and while informal, thus without
a defined well-formedness, they're used in the wild.

Resolving those appear to call for a functional restriction, i.e. a
stricter triple occurrence, being the annotation (or token!) of a
single triple. For that, opacity and functional well-formedness
appears necessary. Opacity being less "harmful" since equality of two
distinct triples is logically false(?) rather than entailing
unexpected identity, IIUC? It also comes with this feature to go
outside of interpretation into the terms for opaque data management (a
mechanism only seen in graph names before). (It appears much more
complex; but I'm trying to get to the bottom of it; especially whether
it is better or worse for users at large.)

But I could see the transparent profile as the baseline and view the
above as open issues on it. Or, conversely, accept viewing them as
requirements under criticism. Until those are resolved, everything's
at risk either way.

Best regards,
Niklas

> —e.

Received on Friday, 7 June 2024 13:04:38 UTC