- From: Frank Manola <fmanola@acm.org>
- Date: Fri, 29 Apr 2005 10:06:53 -0400
- To: public-swbp-wg@w3.org
http://www.w3.org/TR/swbp-xsch-datatypes/ is a useful document, and I hope to have a more complete set of comments later on. However, I wanted to make one observation (and suggestion) right away concerning material such as Section 5 (The Use of Numeric Types), since the point applies to certain other general "best practices" material as well. Section 5 provides generally-useful advice about representing data for engineering and science applications. I'd like to suggest that in conveying such advice, it should be made clear that the issues described are not necessarily peculiar to the Semantic Web (or Semantic Web languages), but apply generally to the representation of such data (as opposed, for example, to issues that arise specifically when XML Schema datatypes are used in RDF or OWL). In this case, for example, similar issues would arise in representing engineering data in Java, C++, or other languages (I would agree that such issues assume particular *importance* in the Semantic Web, due to its presumbed increased interoperabilty requirements, but the issues are not peculiar to Semantic Web *technology*). The motivation for making this distinction is simply to try to forstall comments of the general form "Semantic Web languages have problems representing engineering data" (or whatever the particular issue covered happens to be), and citing SWBP documents (I've already heard comments to this effect about the Duration issue). --Frank
Received on Friday, 29 April 2005 13:56:55 UTC