Re: Why is RDF such a tough sell?

on 6/24/02 9:24 AM, MDaconta@aol.com at MDaconta@aol.com wrote:

> I am fairly new to this list but hope you can help me in defending RDF as an
> architectural direction. About six months ago, for a government integration
> project we proposed RDF as our registry data model but encountered significant
> resistance from developers, data architects and decision makers.  One
> suprising source of resistance was from an XML instructor who stated (to the
> class), "RDF is unnecessary, Schema can do everything RDF can."

I work on XMP for Adobe, which uses RDF. We have run into quite a bit of
similar feedback. I wasn't around when Adobe chose RDF, but from what I've
seen the advantages are subtle but real. At least for our usage.

Taking any one possible application in isolation, you can probably do an
equal or better job by designing your own XML. As long as RDF is adequate
though, it saves you from having to design all of the structural framework.
E.g. how to represent nested structures (nested rdf:Description) or
collections (rdf:Bag, rdf:Seq, rdf:Alt).

By using a standard form your metadata is potentially usable by others, even
without knowing all of the details of your property set. Those others need
not be everyone in the world. Just within related government agencies,
standardizing on a representation framework can improve interoperation.

A lot of it is perception and learning curve too. The RDF community has not
done a particularly good job of explaining or positioning RDF for practical
use. The Semantic Web emphasis drives a lot of folk away.

Alan Lillich
Adobe Systems Incorporated

Received on Monday, 24 June 2002 13:07:55 UTC