Re: Draft Partitioning

Graham Klyne wrote:
> 
> At 06:06 PM 6/15/01 +0100, Brian McBride wrote:
> >RDFCore: A base abstract syntax and a semantics for it.  The abstract syntax
> >is equivalent to n-triple (can n-triple be that abstract syntax).  Nothing
> >more - does not include type, containers, reification.
> 
> [...]
> 
> >Vocabularies (aka standard library): Reification and containers
> 
> I think that, to serve its intended purpose, that "reification" will need
> to be handled in the core.  A message I have taken from the RDF-logic
> discussion that it is really an abstract syntactic device to allow us to
> express statement "nesting" (i.e. use without assertion).

I'm of two (maybe more!) minds about this.  The message *I* took from
the RDF-logic discussion was that RDF would be improved if reification
(as currently provided, which is what I mean here by "reification") were
deleted.  I agree there was something of a conclusion that reification
provided a way to express a statement without asserting it (essentially,
by providing the more primitive statements out of which the original
statement could be constructed by some processor), and that this could
be used to express nesting, statement attribution (provenance), and
maybe some other things [I've noted before my opinion that the M&S
suggests at least two distinct uses for reification, which I don't think
are necessarily consistent].  However, I don't think anyone can read the
RDF-logic archives and conclude that anyone thought that reification was
a very good way of expressing any of these things.  For nesting, I'm
inclined toward Brian's S-expression suggestion [some straightforward
nesting syntax].  Quoting can be done in a similarly straightforward
way.  

On the other hand, for *RDF* (not reification) to serve [what *I* think
is] its intended purpose, I think we need to provide some guidance to
people as to how they should express provenance information, using
whatever facilities remain in RDF 1.0 when we get through with it.  It
seems to me the default is to think of this (as with other "uses") as
being at the Vocabulary level, work out the details on how to do it,
and, if necessary, move really critical and generally useful things into
the Core.  

--Frank

-- 
Frank Manola                   The MITRE Corporation
202 Burlington Road, MS A345   Bedford, MA 01730-1420
mailto:fmanola@mitre.org       voice: 781-271-8147   FAX: 781-271-8752

Received on Sunday, 17 June 2001 20:03:07 UTC