- From: Richard Cyganiak <richard@cyganiak.de>
- Date: Wed, 10 Aug 2011 19:49:06 +0100
- To: Souripriya Das <souripriya.das@oracle.com>
- Cc: Ivan Herman <ivan@w3.org>, public-rdb2rdf-wg@w3.org, "Sundara,Seema" <seema.sundara@oracle.com>, Ashok Malhotra <ashok.malhotra@oracle.com>
Hi Souri, On 10 Aug 2011, at 19:29, Souripriya Das wrote: > >From our point of view > • R2RML is an RDF vocabulary (that introduces all those rr:... RDF terms and explains their semantics) [We made a choice to use RDF, instead of say XML] Well, we made a choice to use *Turtle* instead of, say, XML. That choice entails that the spec has to define an RDF vocabulary. Just like we would have produced an XML schema if we had used XML. I see the vocabulary as peripheral. It's only a means to an end. > • Use of this vocabulary in an RDF document allows us to express a Relational-to-RDF mapping Sure. > • The syntax in which the RDF document is expressed is peripheral to this task I disagree with that. Syntax is *not* peripheral at all. Exchange requires syntax. Teaching requires syntax. I gave examples for the problems that arise from not mandating support for a standard syntax in [1]. > It seems that we are trying to create R2RML as a package: R2RML vocabulary (chosen to be in RDF) + Turtle first WD syntax. > An alternative for our WG that avoids this bias towards a particular exchange format could be as follows: Why is biasing the R2RML specification towards a particular exchange syntax a problem? > • Create an R2RML vocabulary document (illustrations can be in Turtle). > • Create a separate "R2RML using Turtle (first Working Draft) syntax" note > • This allows future addition of notes such as > • "R2RML using RDF/XML (Recommendation) syntax" > • "R2RML using N-Triple syntax" > • ... This strikes me as a completely impractical proposal. Try writing an R2RML document in RDF/XML or in N-Triples. Why would anyone want to do that? Consequently, why should anyone be interested in writing these additional notes? > The advantage (of this modular structuring, instead of packaging) is that the R2RML vocabulary document remains independent of any exchange format (or textual syntax for RDF). Why do you consider such independence an advantage? Best, Richard [1] http://lists.w3.org/Archives/Public/public-rdb2rdf-wg/2011Jun/0165.html
Received on Wednesday, 10 August 2011 18:49:35 UTC