W3C home > Mailing lists > Public > public-rdf-wg@w3.org > February 2011

Re: Graphs and Being and Time

From: Nathan <nathan@webr3.org>
Date: Thu, 24 Feb 2011 23:45:57 +0000
Message-ID: <4D66EDB5.5090409@webr3.org>
To: public-rdf-wg@w3.org
CC: Guus Schreiber <guus.schreiber@vu.nl>
Pat Hayes wrote:
> On Feb 24, 2011, at 4:19 AM, Guus Schreiber wrote:
> 
>> Pat,
>>
>> I 'm trying to translate this into a simple story for outsiders:
>>
>>  RDF abstract graph: triples
>>  RDF graph: quads
>>
>> with a rationale along the following lines: the revised RDF spec does not *replace* triples with quads; it just *adds* the quad notion, to handle the established practice of how abstract graphs materialize on the Web.
>>
>> I hope this makes some sense...
> 
> Unfortunately it does not to me. Um... what have quads got to do with the issue? I just don't get it, sorry. 
> 
> It is much simpler: it is just wanting the WG to acknowledge that "an RDF graph" can either be a mathematical set, or it can be some kind of document or data structure or file than can be transmitted over a computer network. But it can't be both. 

Pat, do you see a graph-token as referring to:

A: a representation of a set of triples, a concrete realization of a 
Platonic abstraction in some distinct lexical form
(non temporally varying, blank node identifiers scoped at representation 
level, where an RDF/XML representation of those triples would be a 
different graph-token to a Turtle representation of those triples).

B: a data structure which holds a set of triples, and which can be 
serialized in to various lexical formats.
(non temporally varying, blank node identifiers scoped at data structure 
level, RDF/XML and Turtle representations both encode the same graph-token)

C: a triple container whose contents can be modified over time, where 
the value at a specific instant is a set of triples representing that 
containers current state, and where the value (set of triples) can be 
serialized in a representation, or held in some in memory data structure.
(blank nodes scoped where?)

D: something else

if C, then that is /one of/ the things which people name with a URI and 
refer to as a named graph and store in quads in a quad store, and would 
explain the distinction Guus made above (afaict)

aside wrt C: there is much confusion over this, as many also take the 
set of triples which represent the containers current state, and then 
give it a URI and call that a named graph too, referring to the abstract 
graph rather than temporally varying container of triples.

Best,

Nathan
Received on Thursday, 24 February 2011 23:48:09 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:25:38 GMT