Re: PROPOSAL to close ISSUE-39: RDFa term mapping triples

On 10/03/2010 02:18 PM, Richard Cyganiak wrote:
> On 3 Oct 2010, at 17:31, Manu Sporny wrote:
>> If there are no objections to this proposal in 7 days, we will close
>> ISSUE-39: rdfa term mapping triples.
> 
> I object.

Thanks Richard. The RDFa WG will need to re-analyze the issue in detail,
including your assertions below.

Just to clarify, if the group ends up not supporting the single property
mechanism that you describe, is that something that you can live with?
Or do you believe that it will do a great amount of harm? In other
words, how strongly do you feel about this?

Are there any other "RDF Experts" that support your viewpoint? What
about folks on this mailing list?

>> The group has found that while Richard's proposal
>> does ease RDFa Profile authoring, it simultaneously raises concerns that
>> the mappings are not being modeled in a way that is acceptable by the
>> broader semantic web community.
> 
> I don't think the WG can presume an opinion of the “broader community.”
> 
> The concerns are Mark's own, not those of the broader community.

To be fair, the concerns are both Ivan and Mark's.

> Mark's concerns about “semantical correctness” are unfounded. The
> proposal to use a single triple for expressing term mappings is
> absolutely in line with RDF Semantics -- see [1], which answers all
> points raised.
> 
> If Mark disagrees with this, then I would like the WG to present a
> rejection that is based on what's actually stated in the RDF Semantics
> document [2], rather than based on some imaginary RDF Semantics boogeyman.

Perhaps Mark and/or Ivan can discuss this issue with you on-list before
we push the discussion back to the RDFa WG. I would certainly like to
see a rebuttal to the points that you raised here:

http://lists.w3.org/Archives/Public/public-rdfa-wg/2010Aug/0076.html

I do support the single-property mechanism you describe, not because I'm
an RDF modeling expert (I'm definitely not), but because it's simpler
and easier to understand for authors and it doesn't require reading the
term/uri triple-pairs from a triple store to implement the feature.

Having to query a triple store is an issue for stream-based parsers,
such as my librdfa SAX-based implementation.

>> While the mechanism is a bit more
>> complicated, it does not risk polluting the modeling data for the
>> vocabulary term. In other words, it keeps a strict separation between
>> CURIE processing instructions and vocabulary term expression.
> 
> There was a proposal that addresses this risk, by defining the notion of
> a "profile graph" [3].

That's true... the "profile graph" should never become part of the
"default graph".

-- manu

-- 
Manu Sporny (skype: msporny, twitter: manusporny)
President/CEO - Digital Bazaar, Inc.
blog: Saving Journalism - The PaySwarm Developer API
http://digitalbazaar.com/2010/09/12/payswarm-api/

Received on Sunday, 3 October 2010 22:17:46 UTC