You are right Bob. The xml schema is very loose and permits any of the id terms in any class.
If I recall correctly this was done in the early days to avoid discussions on the exact class relationships and leave that to the ontology. To get to an agreement on a full class model is much harder than agreeing on a flat list of properties alone.
We also felt that for various uses a different degree on complexity or granularity is needed and we didn't want to force everyone to use the most complex one - sth you would do when you start representing the world with an exact and restrictive model. Like Chuck has said there are the majority of users right now living very well with a completely flat dwc. If you force them to provide ids for every class - even if they are entirely transient - its much harder to generate and consume dwc. I guess we wished that users *could* build that complexity with the dec, but never were forced to. If they wanted a more rigid schema they would have to simply design one for their purpose themselves.
Markus
On Sep 12, 2011, at 11:38 PM, Bob Morris wrote:
Ummm, In the example in http://rs.tdwg.org/dwc/terms/guides/xml/index.htm#classes, the solution leaves me still confused on several grounds:
0: Maybe what's below shows an insufficient understanding of Substitution Groups and is nonsense on my part.
- In tdwg_dwc_classes.xsd It looks like the Substitution Group for
anyIdentifier admits multiple taxonIDs and multiple occurrenceIDs, so there would then still no way to know which Occurrence is being identified to which Taxon, except to assume the element order gives the semantics of association, and that the software got it right. That doesn't work for RDF even if it does for XML-Schema. (And I'm suspicious of its robustness for XML-Schema, but I suppose that is a generic problem about xs:sequence. At the least, the solution seems to require careful attention to the serialization order on the part of a programmer, and an informal semantics that says that the order determines what's associated with what).
- To a certain extent, the Substitution Groups around there seem to
permit the association of objects in any of the eight classes with any of the other eight, in ways that would harass, oh, say Rich Pyle :-) . For example, given the example and schema, it seems as though the method permits an Identification to associate a Taxon with an Event. Since I'm a firm believer in resisting design questions that start with "Why would anybody ever", I'd better shut up on this point....
Bob
On Mon, Sep 12, 2011 at 3:49 PM, "Markus Döring (GBIF)" mdoering@gbif.org wrote:
Bob, this is a very late reply, but hope its still useful.
The dwc xml guidelines already give examples on how to relate Identification with Taxon and Occurrence instances for class usages as xml: http://rs.tdwg.org/dwc/terms/guides/xml/index.htm#classes
Those examples simple use the ID terms (occurrenceID,taxonID) on the Identification instance to link out to the other classes and there is no further relationship class involved.
Markus
On Sep 7, 2011, at 10:46 PM, Bob Morris wrote:
As far as I understand for DwC, in order to associate a dwc:Identification with a dwc:Occurrence we need to define a dwc:ResourceRelation analogous to http://rs.tdwg.org/ontology/voc/OccurrenceRecord#identifiedTo (The context is the building of examples of data annotations for the data extensions to the W3 incubation group document Annotation Ontology http://code.google.com/p/annotation-ontology/). The examples will be RDF).
a. Am I right? b. Is there one already blessed by TAG? c. If b. is "No", does it make sense to put one on the table? -- Robert A. Morris
Emeritus Professor of Computer Science UMASS-Boston 100 Morrissey Blvd Boston, MA 02125-3390 IT Staff Filtered Push Project Department of Organismal and Evolutionary Biology Harvard University
email: morris.bob@gmail.com web: http://efg.cs.umb.edu/ web: http://etaxonomy.org/mw/FilteredPush http://www.cs.umb.edu/~ram phone (+1) 857 222 7992 (mobile) _______________________________________________ tdwg-content mailing list tdwg-content@lists.tdwg.org http://lists.tdwg.org/mailman/listinfo/tdwg-content
-- Robert A. Morris
Emeritus Professor of Computer Science UMASS-Boston 100 Morrissey Blvd Boston, MA 02125-3390 IT Staff Filtered Push Project Department of Organismal and Evolutionary Biology Harvard University
email: morris.bob@gmail.com web: http://efg.cs.umb.edu/ web: http://etaxonomy.org/mw/FilteredPush http://www.cs.umb.edu/~ram phone (+1) 857 222 7992 (mobile)