Re: [humanmarkup-comment] [measurment] new version of type toolkit

[m batsis <mbatsis@netsmart.gr>]

Thanks for the feedback; it's much appreciated. BTW, how may I call you?

On Friday 27 September 2002 14:25, cognite@zianet.com wrote:

> As for value space, the section in the reference you give at W3C,  some of
> the system properties you might expect are surprisingly omitted.  Two are:

Well, it's not really surprising :-)

>         -  required-absence (prohibited membership, often definitive in
> linguistics) and

Would you be willing to provide some examples and discuss it further?


>         -   partial ordering (mathematical; also often appearing in natural
> language systems).

The commonly accepted way of building such types and relations is via complex 
types per the XML Schema terms. RDF containers for example.

I would appreciate the ability to map sets to inline constructs, for example 
the following container:

<rdf:Description about="http://mydomain.org/myplayroom";>
 <my:toys>
  <rdf:Bag>
   <rdf:li rdf:resource="http://mydomain.org/toys#hyperblaster"/>
   <rdf:li rdf:resource="http://mydomain.org/toys#spaceship"/>
   <rdf:li rdf:resource="http://mydomain.org/toys#chessboard"/>
  </rdf:Bag>
 </my:toys>
</rdf:Description>

to be expressed as 

<rdf:Description about="http://mydomain.org/myplayroom"; 
 my:toys="http://mydomain.org/toys#hyperblaster  
http://mydomain.org/toys#hyperblaster 
http://mydomain.org/toys#hyperblaster"/>

But RDF would interpret the above as a literal, not as resources. IMHO this 
would be acceptable in an application build on top of RDF and simplicity is 
always usefull.

Many of the limitations imposed by the RDF specs, would only be bypassed using 
a meta-graph. In other words, one could use an RDF processor to build the RDF 
graph from the markup and then process that to produce a more advanced 
interpretation, another graph. This would include extracting resources from 
literals (or rather, lexical representations of resources).

Anyway, I would be very interested to investigate use cases and pros/cons of 
possible approaches.

Manos