W3C home > Mailing lists > Public > public-xg-lld@w3.org > September 2011

Re: First mention of RDF - IMPORTANT ISSUE

From: Antoine Isaac <aisaac@few.vu.nl>
Date: Fri, 02 Sep 2011 09:13:31 +0200
Message-ID: <4E60821B.3020105@few.vu.nl>
To: public-xg-lld@w3.org
By chance, the cup has not yet any formal rec status within W3C :-)
And I agree that Tom's wording is ok. Linked Data is more a set of best practices, not a technology that has been standardized so far as to rely on one data model / syntax. Even though there's one stack of technical proposals (RDF and other SW recs) that is a very good fit, and is indeed strongly recommended by the main pushers of the technology in the community.


> This is tricky, in part because we have decided that Linked Data is a defined standard and Linked Data (based on the coffee cup [1]) must use RDF. Yet, we aren't tying library linked data to RDF in our report. So I like Tom's wording, but it implies that our decision on using Linked Data (rather than "linked data") may not be accurate. That said, I'm prepared to go with Tom's wording, to use Linked Data, and not worry overly about the RDF/not subtleties, since I think most readers of the report won't really care.
> kc
> [1] http://www.w3.org/DesignIssues/LinkedData.html
> Quoting Tom Baker <tbaker@tbaker.de>:
>> A point that had escaped me when reading the sections separately and out of
>> sequential order: The first mention of RDF currently comes in the Benefits
>> section, where it is mentioned in a way that assumes it has been introduced
>> earlier.
>> Where to introduce RDF? The logical place is in the Scope. In Scope, we
>> currently cite the TimBL's principles [1], and those principles say that where
>> links are relationships anchors in hypertext documents written in HTML, for
>> data they are links between arbitrary things described by RDF (and further
>> down: "When someone looks up a URI, provide useful information, using the
>> standards (RDF, SPARQL)"). As in the diff [2], I propose:
>> Linked Data uses Web addresses (URIs) as globally unique identifiers for
>> dataset items, elements, and value concepts -- analogously to the library
>> world's identifiers for authority control -- and provides data using
>> standards such as the Resource Description Framework (RDF).
>> I hesitate between "such as the Resource Description Framework" and other
>> possible wordings, e.g., "based on" or the stronger "using RDF", but this could
>> potentially create confusion re: references to OWL. It has been argued on our
>> lists that Linked Data does not require RDF, but we have grounded our usage in
>> [1] so I think it helps to make this clear.
>> Tom
>> [1] http://www.w3.org/DesignIssues/LinkedData.html
>> [2] http://www.w3.org/2005/Incubator/lld/wiki/index.php?title=Scope&diff=6052&oldid=6048
Received on Friday, 2 September 2011 11:01:24 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:35:58 UTC