RE: More On the Semantic Web (or: is RDF any good?)

Date: Wed, 8 Nov 2000 11:51:20 -0500 (EST)
From: Sean Luke <seanl@cs.umd.edu>
To: Gordon Joly <gordo@dircon.co.uk>
cc: Matt Jensen <mattj@newsblip.com>, William Loughborough <love26@gorge.net>,
        Craig Pugsley <craig.pugsley@mimesweeper.com>,
        "'www-rdf-interest@w3.org'" <www-rdf-interest@w3.org>,
        "'semantic-web@w3.org'" <semantic-web@w3.org>
In-Reply-To: <Pine.BSF.4.21.0011081647160.6852-100000@tdc.dircon.co.uk>
Message-ID: <Pine.GSO.4.21.0011081149580.11520-100000@jifsan.cs.umd.edu>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

On Wed, 8 Nov 2000, Matt Jensen wrote:

> On Wed, 8 Nov 2000, Gordon Joly wrote:
> 
> > Yes, to convince them. But like HTML (the well know computer virus),
> > inertia rules.
> > 
> > Did PNG take off like a rocket? Answer - no. And that was a very
> > simple in comparison.
>
> There is no network effect in PNG adoption, but there would be in Semantic
> Web adoption.  If ABCNews.com were to support a semantic standard, I now
> have more of an incentive to support the standard on my own site; I'll be
> able to make inferences with ABCNews.com's content.

Let me propose the glass-is-half-empty scenario, which I think to be much
more likely: network effects also prevent systems from being adopted.  If
ABCNews.com were to *not* support a semantic standard, then I now have
more of an excuse to *not* spend all that time supporting this standard on
my own site. I think the right way to view the PNG situation is that it
didn't take off like a rocket *despite* having no nework inertia to hold
it down.  This causes Luke's First Law :-), namely quality is inversely
proportional to popularity.

I imagine the only realistic way to get a semantic standard accepted is
through standards hegemony.  It's nice to see W3C behind it, but I don't
think it's a sufficient hegemon.  Microsoft probably is, assuming they
decide to actually go full-bore with RDF (or whatnot).  This is made even
tougher by the fact that I have yet to see a killer application for RDF
with regard to the average joe (or Joe Business) on the web.  Lots of
nifty applications.  But no *killer* application, without which it may be
difficult to justify the infrastructure expenditure.

Sean

Received on Wednesday, 8 November 2000 12:06:52 UTC