Re: A case for a datatype for identifying RDF graphs

> On 25. May 2021, at 16:09, thomas lörtsch <tl@rat.io> wrote:
> Antoine Zimmerman wrote:
>> [snip explanation of graph datatypes]
>> Thoughts? Criticisms? Acclaims?
> 
> Acclaim! This seems like a very sensible way to cite triples and graphs for the purpose of documenting specific states in the life of some piece of RDF data, like e.g. after ingestion but before its deductive closure was computed and co-references were materialized, before versions were created, when it was signed etc.
> 
> I also like the user interface aspect: literals are easily understood as not interpreted and not part of the "active" data and the quotes make that syntactically very obvious.
> 
> 
> I don’t see the limitation w.r.t. bnodes as problematic. I would however think that being able to properly query such RDF literals could be very helpful. How crazy would an effort to that effect be?

I think querying should be possible using tools like SPARQL. :-)
One should start off with that assumption at least.

Given that a graph literal is (I think) very similar to N3 formulae
without variables, one could learn a lot from N3 too.


Henry


> 
> Thomas
>