W3C home > Mailing lists > Public > www-rdf-interest@w3.org > November 2000

Re: unreification (long)

From: Pierre-Antoine CHAMPIN <champin@bat710.univ-lyon1.fr>
Date: Mon, 20 Nov 2000 10:50:29 +0100
Message-ID: <3A18F3E5.B78EBDB@bat710.univ-lyon1.fr>
To: Bill dehOra <BdehOra@interx.com>
CC: "www-rdf-interest (E-mail)" <www-rdf-interest@w3.org>
Let's take the following example :
my model contains the 3 followin triples

	S1: [A, subClassOf, B]
	S2: [B, subClassOf, C]
	S3: [A, subClassOf, C]

and now I try to retract S3.
They are all in the model, so I should have the right to do that, shouldn't I ?
The answer is no, because the 3 triples, though stored in the same way in my model, are not independant.
Actually, it is a bad idea to store S1, S2 and S3 all together.
Or if I do, *then* I should consider that S3 has not the same nature as S1 or S2.
(note that the different kinds of triples come from an implementation choice, not from the model)

Ora stated "In essence you did not really add any information by reifying S'"
That is, a "reification quad" is not independant from the reified statement,
and though the figures in the spec, representing the statement and the quad all together, are right,
they are not a good *implementation* model.

But the spec does not adress implementation issues.
The Stanford and Jena APIs chose a sane way, making Statement inherit from Resource,
but as far as I know, no Parser rely on that feature, and this is a shame.


Sometimes I think the surest sign that intelligent life exists elsewhere in the
universe is that none of it has tried to contact us.
(Bill Watterson -- Calvin & Hobbes)
Received on Monday, 20 November 2000 04:51:14 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:07:33 UTC