W3C home > Mailing lists > Public > public-rdf-star@w3.org > December 2020

Re: Agenda for 2020-12-11

From: Gregg Kellogg <gregg@greggkellogg.net>
Date: Fri, 11 Dec 2020 13:06:45 -0800
Cc: public-rdf-star@w3.org
Message-Id: <313CFBE0-1E12-45D7-8313-B33CCE8CDB86@greggkellogg.net>
To: David Booth <david@dbooth.org>
> On Dec 11, 2020, at 9:47 AM, David Booth <david@dbooth.org> wrote:
> 
> On 12/11/20 11:59 AM, Lassila, Ora wrote:
> > I am with Peter on this. Particularly, discussions of syntax
> > are hard for me when I don't know exactly what's "behind the
> > syntax" (and I get these horrible flashbacks to RDF M+S WG).
> 
> +1.  I don't see how this can sensibly go forward without first having agreement about whether bracketed triples are asserted.  As yet it looks indistinguishable from being a special syntax for an unnamed "named graph" that is limited to a single triple.

My understanding is that << … >> triples are _not_ automatically asserted.  They are asserted using {! … |}.

An N-Triples* syntax would only have << … >>, and graph isomorphism would need to be updated to consider that, so we can express the results of parsing Turtle* (or whatever*) in terms of N-Triples*/Quads*.

I believe Andy was going to create some Evaluation tests, based on what EYE does (which should be the same as my tools). In particular, I consider the following equivalent:

:a :b :c {! :d :e |} .

and

:a :b :c .
<<:a :b :c>> :d :e .

A hypothetical reification of this graph using the reification vocabulary might be the following:

:a :b :c .
_:st a rdf:Statement
  rdf:subject :a;
  rdf:predicate :b;
  rdf:object :c;
  :d :e .

And, that might be adequate for isomorphism, at least in test evaluation.

Gregg

> David Booth
> 
Received on Friday, 11 December 2020 21:07:03 UTC

This archive was generated by hypermail 2.4.0 : Friday, 11 December 2020 21:07:04 UTC