W3C home > Mailing lists > Public > www-rdf-interest@w3.org > February 2000

RE: A certain difficulty

From: Mark Birbeck <Mark.Birbeck@iedigital.net>
Date: Thu, 24 Feb 2000 10:20:53 -0000
Message-ID: <A26F84C9D8EDD111A102006097C4CD0D0E92F5@SOHOS002>
To: xml-dev@xml.org, www-rdf-interest@w3.org
Mark Grossman wrote:
> I sympathize.  It took me much much longer than a week to 
> "get" RDF.   I
> started at a disadvantage, having very little background in knowledge
> representation and scant experience with XML.  I really did 
> think it was I who
> was stupid for a while.

Knowledge management is a great application for RDF, but it is not a
precursor to understanding it. I'd suggest that anyone who has good
relational database skills already has all the concepts they need to get
a lot out of RDF. Obviously that's no coincidence - RDF can be
represented as entity relationships, as can RDBMSs. So if you imagine
that the entire internet is an RDBMS, and the URIs of the data you are
interested in are the primary keys, then RDF is just a load of join
statements. Theoretically, this is the wrong way round to understand it
- you must spend five years studying knowledge management, learn to draw
entity diagrams in your sleep and then you can understand RDF (and
RDBMS). However, most people understand relational databases a long time
before they even hear of ER, nodes and arcs, and the like.

So, I'd suggest starting with the practice and working back to the
theory:

1. The internet is a massive relational database.
2. The brilliant invention of URIs allows us to give everything a
   primary key.
3. We have no control over someone else's 'tables' (resources) so
   we can't add to their stuff, so we have to add to our own tables
   and do joins to theirs (in other words, make statements).

RDF is just like those many-to-many tables you've been doing for years
in your databases.

One last point, I must rehearse an old argument from the days of the
namespaces debate (remember that one?!) - that it is not the
responsibility of the spec writers to make their inventions accessible.
A spec must be rigorous and avoid ambiguity so that others can implement
their software in a way that they know will be compatible with others.
Contributions to the list on how obtuse the RDF and RDFS specs are, are
pointless (as are existential debates on whether, I, the reader, am
stupid, and philosophical debates on whether it is possible for a spec
to be stupid). What we need are good illustrations and articles and this
thing will turn round fast.

Best regards,

Mark
Received on Thursday, 24 February 2000 05:21:20 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:42 GMT