Re: RDF-ISSUE-103 (dereferenceable-iris): Make dereferenceable IRIs a SHOULD in RDF Concepts [RDF Concepts]

While I agree with Richard re RDF concepts, I think it would be good if the
primer did contain such guidance...

  Pa


On Tue, Nov 6, 2012 at 4:02 PM, Richard Cyganiak <richard@cyganiak.de>wrote:

> Markus,
>
> On 6 Nov 2012, at 12:57, Markus Lanthaler wrote:
> > I would be even happier
> > if the sentence "A good way of communicating the intended referent to the
> > world is to set up the IRI so that it dereferences [WEBARCH] to such a
> > document" could be reformulated to a non-norminative recommendation. I
> find
> > it important to give developers clear recommendations they can make use
> of
> > without having to dig really deep.
>
> I think you misunderstand the purpose of RDF Concepts.
>
> It defines a data model that is the shared foundation for a number of
> other specifications, including RDF syntaxes, query languages, formal
> semantics, vocabulary definition languages, and so forth.
>
> Its purpose is to make it so that the mechanisms defined in all of these
> specifications can operate on the same data.
>
> For most of these specifications, it doesn't matter whether URIs
> dereference. The things that matter are things like whether the URIs are
> absolute or relative, whether they may or may not contain characters
> outside of US-ASCII, whether language tags are case-sensitive, whether
> blank nodes can be shared between graphs, whether URI normalization happens
> before URIs are compared for equality, and so on. That's the stuff that RDF
> Concepts defines.
>
> There's some 20 or so W3C specs that depend on RDF Concepts, and dozens of
> further specs outside of W3C. Most of them don't need to care whether the
> URIs in RDF graphs dereference; it's an orthogonal concern. Best practice
> for using the data model doesn't belong into the definition of the data
> model, because it just adds confusion for all the other specifications that
> rely on that foundation.
>
> If you want to give clear recommendations to developers, then write a
> tutorial! A data model definition isn't the place for advocacy.
>
> Best,
> Richard
>