Re: [office-metadata] Groups - MetaData-Examples_07-06-21(MetaData-Examples_07-06-21.odt) uploaded

[Svante Schubert <Svante.Schubert@Sun.COM>]

Bruce D'Arcus wrote:
>
> On Jun 22, 2007, at 8:48 AM, Svante Schubert wrote:
>
>>> Yes, but this document is not for tools; it's for humans.
>> I am human and see no obfuscation. ;-)
>> Bruce, all I want to say, it's a matter of personal taste and habit 
>> but we can change this, as it seems important for you as both ways 
>> are valid.
>
> But the point is the way you are doing it in the example is not a 
> standard serialization. So if an RDF newbie goes and looks on the web 
> for information, the examples they will see will use the more standard 
> "rdf:resource" approach. For example, the RDF Primer only uses this 
> serialization (though smartly uses N3 more than RDF/XML for 
> demonstration).
You nailed the problem of RDF/XML. There is NO standard for a 
serialization to RDF/XML. One RDF graph might have multiple RDF/XML 
representations.
If there would be a standard way of serialization, XML processing (e.g. 
XSLT) might easily work with RDF/XML.
There is no normalization of RDF/XML, resolving the IRIs to a compact tree.
>
> So I think it will he helpful for your TC colleagues for our examples 
> to look similar to examples they would find in other documentation. 
Sure.
> The inclusion of N3, BTW, is nice. Do we reference this?
>
> <http://www.w3.org/2000/10/swap/Primer>
Yes in a footnote, we might as well add the converter we used to create 
it http://www.mindswap.org/2002/rdfconvert/ and the validator 
http://www.w3.org/RDF/Validator/.
Newbies might have all most of the important URLs quickly at hand.

Svante