W3C home > Mailing lists > Public > public-rdf-comments@w3.org > June 2013

Re: [RDF-CONCEPTS] Skolemization

From: Niklas Lindström <lindstream@gmail.com>
Date: Sun, 16 Jun 2013 11:52:16 +0200
Message-ID: <CADjV5jdbeEH2ORrEkSApCRSHFDzfsXUR2LqJ1vWRwr1OMMC8EA@mail.gmail.com>
To: Pat Hayes <phayes@ihmc.us>
Cc: David Booth <david@dbooth.org>, Ivan Herman <ivan@w3.org>, public-rdf-comments <public-rdf-comments@w3.org>
On Sun, Jun 16, 2013 at 5:25 AM, Pat Hayes <phayes@ihmc.us> wrote:

>
> On Jun 15, 2013, at 9:38 PM, David Booth wrote:
>
> >
> > similarly the definition of RDF skolemization needs to be normative,
> with conformance terms, for the exact same reason.
> >
> > if someone transformed
> >
> >  _:b :foo :bar .
> >
> > into
> >
> >  :bar :foo :bar .
> >
> > then they should be prohibited from calling it RDF skolemization, just
> as people are currently prohibited from calling the following RDF:
> >
> >  "literal" _:bnode  "oops" .
>
> Its not about calling things names.



> Its not about calling things names.

To me, that's exactly what it is. The sentence David has a problem with is:

    "Systems wishing to do this SHOULD mint a new, globally unique IRI (a
Skolem IRI) for each blank node so replaced."

Because, in his reading (IIUC), the first "this" is bound to
"skolemization". The question is only if this is the case. If it is, David
is right, because skolemization is, by definition, "mint a new, globally
unique IRI (a Skolem IRI) for each blank node so replaced". So to do
skolemization, you MUST do skolemization. However, the sentence preceding
the one above in the spec is:

    "In situations where stronger identification is needed, systems may
systematically replace some or all of the blank nodes in an RDF graph with
IRIs."

It can be argued that this binds the "this" in question to "replace some or
all of the blank nodes in an RDF graph with IRIs". And, as Pat has
elaborated on, this act in itself has no requirements which are relevant to
specify in this context.

In either case, as this discussion is evidence of, something should
probably be done to the spec text to clarify what is actually required, and
under which conditions.

Cheers,
Niklas



> They can *call* that RDF all they like, but a conforming RDF engine will
> spit it out, is the point. But there isn't anything about skolemization
> that engines are required to do or prohibited from doing. They can do it,
> or not; and they can do other things, or not. If someone says their engine
> is doing skolemization when it isn't, then they are wrong, which can be
> checked by looking at the definitions. But just because their own tray
> tables are not in the fully upright and locked position does not make their
> RDF engine nonconformant.
>
> Pat
>
>
> >
> > David
> >
> >
> >
>
> ------------------------------------------------------------
> IHMC                                     (850)434 8903 or (650)494 3973
> 40 South Alcaniz St.           (850)202 4416   office
> Pensacola                            (850)202 4440   fax
> FL 32502                              (850)291 0667   mobile
> phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes
>
>
>
>
>
>
>
Received on Sunday, 16 June 2013 09:53:14 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:29:57 UTC