W3C home > Mailing lists > Public > www-rdf-interest@w3.org > September 2004

Re: Upcoming wave of quad/namedgraph implementations ,was: Reification - whats best practice?

From: Graham Klyne <GK@ninebynine.org>
Date: Fri, 10 Sep 2004 11:23:15 +0100
Message-Id: <5.1.0.14.2.20040910111219.031c1d00@127.0.0.1>
To: "Chris Bizer" <chris@bizer.de>, <www-rdf-interest@w3.org>

At 11:14 10/09/04 +0200, Chris Bizer wrote:

>Hi all,
>
>it really looks like this reification discussion make some people sit down
>and work :-)
>
>We will see a wave of quad/namedgraph-based implementations being released
>in the next weeks. Up to my knowledge, the following people are currently
>working on implementations:
>
>1. We are working on a NamedGraph API extending Jena
>http://www.wiwiss.fu-berlin.de/suhl/bizer/ng4j/
>2. Andy Seaborne and Rave Reynolds work on a Jena Multimodel
>http://www.swed.org.uk/swed/doc/javadoc/com/hp/hpl/jena/util/MultiModel.html
>3. Andrew Newman said that Kowari 1.0.5 will somehow support
>quads/namedgraphs (??),
>http://kowari.sourceforge.net/
>4. Phil Dawes implements a store in python which will support
>quads/namedgraphs
>5. RDF Gateway already supports quads
>http://www.intellidimension.com/default.rsp?topic=/pages/site/products/rdfgateway.rsp
>
>Anybody else?

FWIW, both my RDF model implementations (one in Python and one in Haskell) 
have supported some variant of named graphs, though not using quads per-se 
[*].  I also understand that CWM uses quads internally, to support the 
formulae and contexts.

For me, it has always been important in principle to be able to deal with 
provenance, so allowing graphs or names-of-graphs to appear (somehow) in a 
graph seems to be a fairly minimal way to achieve this.  I use N3 syntax to 
represent these structures externally, ala:

    uri :- { formula }

where (I think) Trix uses:

    uri { formula }

(My use of N3 is debatable, but I originally deduced it by experimenting 
with an early version of CWM, and decided that I like the idiom.  Since N3 
isn't a standard as such, I don't feel to badly about bending it a little.)

#g
--

[*] My python code allows a set of statements (i.e. a raw graph) to be 
associated with any node.  My Haskell code uses a context-like structure 
which maintains a set of named graphs alongside any graph.



------------
Graham Klyne
For email:
http://www.ninebynine.org/#Contact
Received on Friday, 10 September 2004 11:00:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:09 GMT