W3C home > Mailing lists > Public > public-vocabs@w3.org > April 2014

Re: How to avoid that collections "break" relationships

From: Peter F. Patel-Schneider <pfpschneider@gmail.com>
Date: Sun, 13 Apr 2014 19:26:43 -0700
Message-ID: <534B4763.3050200@gmail.com>
To: Luca Matteis <lmatteis@gmail.com>
CC: Ruben Verborgh <ruben.verborgh@ugent.be>, Markus Lanthaler <markus.lanthaler@gmx.net>, Pat Hayes <phayes@ihmc.us>, Linked Data community <public-lod@w3.org>, W3C Web Schemas Task Force <public-vocabs@w3.org>, Dan Brickley <danbri@danbri.org>

On 04/13/2014 04:10 AM, Luca Matteis wrote:
> On Sun, Apr 13, 2014 at 3:07 AM, Peter F. Patel-Schneider
> <pfpschneider@gmail.com> wrote:
>> RDF  includes a meaning for triples.
> I think this is your biggest misconception. RDF does not include any
> meaning of anything. It's a framework that is entirely unbiased about
> what you define. The meaning of statements is only inferred by the
> tools that interpret the RDF.

RDF includes lots of meaning, and RDF tools are supposed to abide by that 
meaning.  Of course, some tools don't fully abide by RDF's meaning, and quite 
a few users of RDF don't.

Some aspects of RDF's meaning include that IRIs are global identifiers, that 
the meaning of triples in an RDF graph are conjoined, that using a 
dereferenceable IRI does not commit you to include the RDF triples at that 
document, and that different IRIs do not necessarily identify different entities.

Some aspects of RDF's meaning are not even in any W3C document.  For example, 
using RDFS vocabulary in an RDF graph implies that the author of the document 
almost certainly meant them to be used as defined in the RDF semantics.  One 
could argue that certain other IRIs carry a similar import, particularly 

Received on Monday, 14 April 2014 02:27:18 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:29:39 UTC