"Sigfrid Lundberg, Lub NetLab" wrote:

> Yes, Eric and Tod work, while the rest of the community worries about
> the uniqueness of triples in RDF interest group ;)

lol .. don't get me started.

> > I know somebody is trying something similar using XSLT/Schematron
> > approach but it would be really nice to have a coordinate effort on
> > it.
>
> I suppose that would be me... And the reason I'm trying it that way is
> that I've been unable to see how to serialize RDF metadata given the
> RDFS of its semantics.

Hmmm ... it seems to me that we can never ever get any meaningful semantics without
emerging up into the discourse level of dialogue, which rdfs refuses to do.  So
maybe that's why nobody seems satisfied with the rdfs semantics ... taint there.

> I'll give up that effort when I see how to
> achieve that, or if I find some software that does it for
> me.

Put the semantic memory in the client ... we call that the sembrowser.  Of course
that idea would have fought with the whole dot com craze ... thank god it died.  I
think semantic webbing is going to have an entirely different feel, culture and
texture than the www .... people just aren't use to it yet ... and i guess its a
good thing too cause there just are no popular tools out there.

> Schematron might be a good idea anyway, however, if an
> applications profile have requirements as regards obligation.

I almost understand what you're saying .. sounds like you're at the dialogue level.
Got any schematron tools that i can easily run on a Win32 ?

--
Seth Russell
http://RobustAI.net/MyNetwork/index.html
http://robustAI.net/MyNetwork/StickeyCyberMolecules.html
Http://RobustAi.net/Ai/Conjecture.htm