- From: Seth Russell <seth@robustai.net>
- Date: Sun, 23 Jun 2002 12:59:57 -0700
- To: "Danny Ayers" <danny666@virgilio.it>, "Graham Klyne" <GK@ninebynine.org>
- Cc: <www-rdf-logic@w3.org>
From: "Danny Ayers" <danny666@virgilio.it>
> So in principle at least then, it may be possible to make statements about
> the arc instance, if this were identified by means of the attached nodes -
> an odd sort of {URIa & URIb} identifier. Hmm - but what happens if there
are
> two arcs of the same type between node a and b (which I *think* would be
> legal). bArcs anyone?
Within one context (model, graph, formulae, whatever), there can be only one
arc between <A> and <B> of a given type (ie with the same property). Of
course if you remove the restriction that you are in a particular context,
well then there can be as many ~different~ arcs of the same type between <A>
and <B> as there are instances of people declaring that triple. We beat
this to death a couple of years ago.
If you think about it, this is the primary reason that the semantic web will
not work unless we implement contexts. Thanks for giving me the
oppurtuinity to get that off my chest :)
Seth Russell
Received on Sunday, 23 June 2002 16:06:15 UTC