|
From: Hobern, D. <DH...@gb...> - 2002-10-29 20:28:41
|
(I need to think about this area much more than I have so far done. I may well do this in England over the next few days. I also understand that many of you have been thinking about this for far longer.) There are clearly other (possibly simpler) ways for us to define the relationships between schema elements than by using substitutable elements. I am certainly tempted by the idea of just defining the federation schema and then adding attributes from a DiGIR namespace, where these attributes define the essential characteristics of each element as it relates to DiGIR (e.g. whether it is queriable, and if so, what class of query operations is appropriate). Possibly a compiler could then be used to derive a more natural query schema from the federation schema. Secondary configuration documents could still be used to define the mapping to the physical data model. In any case I believe that (whether we stick with using substitutable elements or something more like attributes), we need to model more information within the protocol schema. In particular, can we lose Darwin Core bounding box in favour of identifying that Location is an element which is an instance of a 'CoordinateSubstitutableElement'? We could then simply allow software to deduce that a bounding box query construct can be applied when querying it. This would allow the concept to become a generalised part of the protocol rather than a special case for one particular coordinate element. Donald --------------------------------------------------------------- Donald Hobern Programme Officer for Data Access and Database Interoperability Global Biodiversity Information Facility Secretariat Zoological Museum - University of Copenhagen Universitetsparken 15, DK-2100 Copenhagen, Denmark Tel: +45-35321483 Fax: +45-35321480 E-mail: dh...@gb... --------------------------------------------------------------- |