W3C home > Mailing lists > Public > www-rdf-interest@w3.org > November 2002

Re: [xml-dev] RDF for unstructured databases, RDF for axiomatic

From: Uche Ogbuji <uche.ogbuji@fourthought.com>
Date: Tue, 19 Nov 2002 07:51:05 -0700
To: "Danny Ayers" <danny666@virgilio.it>
Cc: "RDF-Interest" <www-rdf-interest@w3.org>, xml-dev@lists.xml.org
Message-Id: <E18E9io-0005NL-00@malatesta.local>

[Wow.  The to/cc headers on this thread seem to be growing without upper 
bound.  Trimmed a bit for now]

Shelley Powers:

> >> Anyone know of an [RDF] vocabulary using reification?

Moi-meme:

> >I'm glad they're hard to find.  If I saw an RDF vocab using
> >reification, I'd
> >steer very clear of it.  I have a very firm policy of avoiding reification
> >like the plague.

> >I either use explicit blank nodes or (if I can relax interop
> >requirements),
> >4RDF statement context/scope annotations.

Danny Ayers:

> Now that's exactly the thing - apart from being of a nice toy for the
> theoreticians, reification should give us context/scope capability without
> having to relax any requirements. I've still not got far enough through the
> latest spec suite to see how any recent changes might have impacted on the
> potential for contexts, but some of the n3 stuff and Graham Klyne's work on
> contexts [1] and scoping [2] suggest that it should be possible with little
> (or no?) extension to the current specs.
> 
> As I'm sure is the case in practice with 4RDF, contexts aren't particularly
> complicated in principle. They should be simple to do within the shared
> model. Very simple. Reification shouldn't be the 'big ugly' (as Shelley
> nicely put it) to be avoided by sensible developers, it should be a big
> friend.


I agree that proper context/scopes would be little work to add to core RDF, so 
I have always wondered: why the complete avoidance of the issue in the specs?  
This is a matter that almost every RDF implementor has had to rediscover 
and/or reinvent.  I would think a "standard" in development should take note 
of this fact.

As for reification in RDF 1.0, it is B.A.D., to use Tim Bray's cheeky term.  
It's B.A.D. from a model-theoretic point of view -- the lack of enforceable 
relationship between the four statements generated by its reification and the 
original statement itself makes it useless for a lot of the cases at which it 
is aimed.  It's B.A.D. from the implementation POV -- each reification bloats 
the data store, or the implementation has to add a big-O-rder of magnitude to 
typical query algorithmic complexity, unless one performs very dodgy tricks 
with statement hashes and the like (dodgy because of the model-theoretic 
failings of reification).

Thanks for the links, though, I'll reference them from my Akara page on 4Suite 
RDF scopes.


-- 
Uche Ogbuji                                    Fourthought, Inc.
http://uche.ogbuji.net    http://4Suite.org    http://fourthought.com
Python&XML column: 2. Introducing PyXML - http://www.xml.com/pub/a/2002/09/25/p
y.html
The Past, Present and Future of Web Services 1 - http://www.webservices.org/ind
ex.php/article/articleview/663/1/24/
The Past, Present and Future of Web Services 2 - 'http://www.webservices.org/in
dex.php/article/articleview/679/1/24/
Serenity through markup - http://adtmag.com/article.asp?id=6807
Tip: Using generators for XML processing - http://www-106.ibm.com/developerwork
s/xml/library/x-tipgenr.html
Received on Tuesday, 19 November 2002 09:48:47 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:57 GMT