Re: [External] : Re: one RDF1.2 "stated" 4-tuple per LPG edge

Hi James,

I was trying to say that "one LPG edge -> one RDF1.2 tuple" is easier to grasp than "one LPG edge -> two RDF 1.2 tuples (unless the asserted triple is already present)". Also, since with the use of the latter, n > 1 parallel (s)-[:p]-(o) edges of LPG will map to a single s-p-o triple (and n rdf:reifies triples), deletion of an rdf:reifies triple may require use of reference count to check if the s-p-o triple should be deleted as well (when ref count goes down to 0) or should stay on (otherwise).

So, the main point I was trying to convey is that one-to-one mapping is generally easier to grasp than one-to-multiple.

Thanks,
Souri.

________________________________
From: James Anderson <anderson.james.1955@gmail.com>
Sent: Wednesday, August 7, 2024 12:39 PM
To: RDF-star Working Group <public-rdf-star-wg@w3.org>
Subject: [External] : Re: one RDF1.2 "stated" 4-tuple per LPG edge

good afternoon;

> On 7. Aug 2024, at 17:38, Souripriya Das <souripriya.das@oracle.com> wrote:
>
> Converting an LPG edge to RDF1.2 will be simpler to think about, IMO, if we go with use of only a single "stated" id-s-p-o tuple compared to the alternative of using a "reified" id-s-p-o tuple along with an (asserted) s-p-o triple:

"think about", from which perspective.
my first thoughts are that neither their representation nor its implementation would necessarily be easier, but i would have to think about them some more.

>
> LPG edge:
>         (s) -[id:p]-> (o)
>
> RDF1.2:
>         id rdf:states <<( :s :p :o )>> .
> vs.
>         :s :p :o .
>         id rdf:reifies <<( :s :p :o )>> .
>  Thanks,
> Souri.

bet regards, from berlin,
---
james anderson | james@dydra.com | https://urldefense.com/v3/__https://dydra.com__;!!ACWV5N9M2RV99hQ!PNopp9WYcX3KBH2gCzgGACiccBYthWyuv8vtST_gWGHAH0dsNxb6nGv9D-PiWvd04oDrSXOdA-x69E7mGhcCH153d3idOw$