W3C home > Mailing lists > Public > public-rdf-wg@w3.org > April 2012

Re: the term "named graphs"

From: Sandro Hawke <sandro@w3.org>
Date: Sat, 28 Apr 2012 00:49:17 -0400
To: Andy Seaborne <andy.seaborne@epimorphics.com>
Cc: public-rdf-wg@w3.org
Message-ID: <1335588557.9663.815.camel@waldron>
On Fri, 2012-04-27 at 19:27 +0100, Andy Seaborne wrote:
> 
> On 27/04/12 19:01, Sandro Hawke wrote:
> > On Fri, 2012-04-27 at 10:45 -0700, Jeremy Carroll wrote:
> >> On 4/27/2012 1:44 AM, Antoine Zimmermann wrote:
> >>> Remember that "named graph" is normative as it's part of a normative
> >>> section of the SPARQL recommendation.
> >>
> >> I just looked this up, and the definition seems to be in section 8
> >> http://www.w3.org/TR/2008/REC-rdf-sparql-query-20080115/#rdfDataset
> >> "An RDF Dataset comprises one graph, the default graph, which does not
> >> have a name, and zero or more named graphs, where each named graph is
> >> identified by an IRI."
> >>
> >> This wording seems to me to suggest that G is a named graph if and only
> >> if there is a pair
> >>
> >> (u, G)
> >>
> >> in the dataset.
> >
> > Thanks for looking that up.  I'm fine with that meaning for the term.
> >
> > Being pedantic, and using my terms from 6.3, the named graph is the RDF
> > Graph comprising the RDF Triples known to be contained by the graph
> > resource which has the given name.   But I don't think boiling that down
> > to "named graph" is too misleading.
> >
> >       -- Sandro
> >
> 
> Strictly, the definition is:
> 
> http://www.w3.org/TR/2008/REC-rdf-sparql-query-20080115/#sparqlDataset
> 
> and definition is a named graph is the pair.
> 
> A red thing has an attribute that it is read.
> A named thing has an attribute that it is named.
> 
> So it could be a set of { name , triples} , of { name , set of triples } 
> or with named attribute (name=u, graph=G) but we happen to write the 
> name aspect and graph value aspect as a pair (u,G)
> 
> G can have a life of it's own - and a different name elsewhere.
> 
> (u1, G) and (u2, G) are different named graphs.

I understand how it's formally defined.  My concern is with how people
use the term in practice, and whether that usage conflicts with the
formal definition.  I think it does, but I could certainly be wrong.  

Richard mentioned the idea of a survey, and that does seem like a way to
help me find out if I'm just being crazy on this subject or not.

I've put one together, as a bit of a lark.  Before I tweet the link, do
you want to take a look and see if you think it's fair?  I think
question 2 will draw out whether people's thinking tends to agree with
you (and the formal definition), that (u1, G) and (u2, G) are actually
different named graphs.   The rest of the questions are just to get some
other ideas about how people think about this stuff.

https://www.surveymonkey.com/s/named-graphs

    -- Sandro
Received on Saturday, 28 April 2012 04:49:27 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 22:02:04 UTC