From: Mike Dean ([email protected])
Date: 07/24/01
Dan, I've been reluctant to change [1] to drop the .daml suffices for several reasons: 1) I think it helps folks to be able to tell what the type of the file is (DAML vs. n3 vs. HTML, etc.). I don't think this by itself necessarily causes them to use the suffix in their references. 2) I have some concerns about the practice of omitting the type in ontology references: a) Content negotiation in the presence of multiple types doesn't currently work as desired (e.g. with [2], which is available in .daml, .xml, and .html formats) because DAML API's don't/can't set HTTP_ACCEPT and there isn't a standard DAML MIME type (we had planned to register one, but I'm now holding off until RDF Core decides what it wants to use as an RDF MIME type). b) You have to specify the suffix with file: URIs (which I use quite a bit for private and offline development), at least with the Java APIs under Windows NT/2000. c) At least [3] and one of the WWW servers used by the Horus team require the .daml suffix. I haven't yet tracked down which server they're using or other details. Mike [1] http://www.daml.org/2001/03/daml+oil-index [2] http://www.daml.org/tools/tools [3] http://www.daml.ri.cmu.edu/ont/AirportCodes.daml
This archive was generated by hypermail 2.1.4 : 04/02/02 EST