Re: reference needed

> or a document about the thing, or (b) to have syntactically
> distinguished URIs that are only used to denote things.  Speaking
> entirely personally, I prefer (b) as a more expressive solution that
> does not require excessive engineering - a new URI scheme (thing:...)
> would be sufficient.  However, before Pat and others round on me, I'm
> aware that this is considered a somewhat simplistic view :-).

I don't think it's simplistic, it's just that if you chose to use a
"thing" URI [1], you give up the benefits of the current HTTP web
infrastructure.  That infrastructure sure seems like it should be
useful for something, but we haven't yet reached a crisp consensus on
what.

       -- sandro

[1] as in my old proposal, http://www.taguri.org
[2] http://www.w3.org/2000/10/swap/doc/ontologies