parseType="daml:collection"

From: Frank van Harmelen (Frank.van.Harmelen@cs.vu.nl)
Date: 12/27/00 

Dear joint-everyone,

The current spec, example and walkthrough of DAML+OIL all rely on a non-existent RDF extension to deal with collections of items (unordered lists), in the form of a non-standard parseType directive:

<Disjoint parseType="daml:collection">
  <rdfs:Class rdf:about="#Car">
  <rdfs:Class rdf:about="#Person">
  <rdfs:Class rdf:about="#Plant">
</Disjoint>

At the same time, the spec introduces a whole terminology to deal with lists (first, rest, empty, item). This would make the above look like:

<daml:Disjoint>
 <daml:first><rdfs:Class rdf:about="#car"/></daml:first>
 <daml:rest>
  <daml:Disjoint>
   <daml:first><rdfs:Class rdf:about="#person"/></daml:first>
    <daml:rest>
     <daml:Disjoint>
      <daml:first><rdfs:Class rdf:about="#plant"/></daml:first>
       <daml:rest rdf:resource="daml:Empty">
       </daml:rest>
     </daml:Disjoint>
    </daml:rest>
  </daml:Disjoint>
 </daml:rest>
</daml:Disjoint>

Clearly, the first is preferable as syntax, but is currently unsupported, so tool builders will have to resort to the second (the above was taken from the output of Oiled). 

Any suggestions on how we will resolve this?

[1] remove the daml:collection parseType, and rely solely on the list construction
[2] solely rely on the daml:collection parseType 
    (then we might as well remove the list vocabulary from the spec?)
[3] live with the current situation (but add an explanatory note to the walkthrough)

Suggestions? Alternatives? Votes?

Frank.
    ---

This archive was generated by hypermail 2.1.4 : 04/02/02 EST