Re: [office-metadata] Export / Import of metadata

["Bruce D'Arcus" <bdarcus@gmail.com>]

On Mar 1, 2007, at 11:04 AM, Svante Schubert wrote:

>> I don't understand how xml:id has anything to do with this  
>> question? The subtree is specified as the content of the node  
>> which contains the metadata attributes; right? And there won't  
>> always be an xml:id in that context anyway.
>>
> The element with the xml:id specifies not only the element, but as  
> well it's embraced XML nodeset, the XML subtree.
> For example, if you give a table a xml:id and use the xml:id to  
> create a RDF subject representing the table in RDF/XML, you can  
> give RDF statements in RDF/XML about the whole table. The table  
> with all it's sub-elements is described as one, not only the string  
> of it's containing text nodes as it would occur by default - what I  
> suggest - when using RDFa.

I think you're confusing subjects and triples again Svante.

When we give an xml:id to a node, it identifies the node as a  
subject. The content of the node is really irrelevant.

When we add metadata attributes to a node, it defines the content of  
the node as a triple statement (a subject with a property).

E.g. subject identification and a triples are not the same thing, and  
so it makes no sense logically to compare them as a basis for  
evaluation.

That said, as I mentioned earlier, I am coming around to agree with  
the outcome (if not the logic) of your suggestion ;-)

Bruce