- From: Smith, Michael K <michael.smith@eds.com>
- Date: Wed, 30 Jul 2003 18:13:17 -0500
- To: Jeremy Carroll <jjc@hpl.hp.com>, www-webont-wg@w3.org
Thanks very much Jeremy. And thanks for the call. Once I have uploaded the changed docs I will let you know. I hope that will be within a day or so. I have been working on a proposal and have had no time. - Mike -----Original Message----- From: Jeremy Carroll [mailto:jjc@hpl.hp.com] Sent: Friday, July 25, 2003 7:00 AM To: www-webont-wg@w3.org Subject: review of wine.owl and food.owl Mike please can you acknowledge receipt of this, otherwise I'll give you a call later. Note: due to the tardiness of this review (which makes points that I don't recall having previously made) I am happy for the editors to totally ignore the suggestions - although on the first two I would defer to DanC and Sandro ... 1: Base location It is an improvement that xml:base is used. I think it is probably more appropriate to use the base location that includes the publication URI so e.g. for the current WD it would be http://www.w3.org/TR/2003/WD-owl-guide-20030331/wine# and say the CR gets published 10 August it maybe http://www.w3.org/TR/2003/CR-owl-guide-20030810/wine# This is a bit of pain at publication time, but is probably better than the implicit link off to the changing location "http://www.w3.org/2001/sw/WebOnt/guide-src/wine#" 2: Changing suffix to .rdf Given that we decided not to register a mime type I think we should be using the .rdf suffix rather than a .owl suffix. 3: (if doing 1) If you decide to follow the suggestion in point 1, then it may be helpful to replace the namespace declarations with entity refs to minimize the points of change. e.g. <rdf:RDF xmlns = "&vin;" xmlns:vin = "&vin;" xml:base = "&vin;" xmlns:food= "&food;" xmlns:owl = "&owl;" xmlns:rdf = "http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:rdfs= "http://www.w3.org/2000/01/rdf-schema#" > 4: making food.owl conform with OWL DL (I can't remember if we have already discussed this - if you have already rejected this suggestion apologies for repeating myself) food owl is not in OWL DL because it refers to the objects in the wine ontology without giving them types e.g. <owl:Class rdf:ID="Wine"> <owl:equivalentClass rdf:resource="&vin;Wine"/> </owl:Class> This can be rectified in three different ways: a) easy way import wine.owl from food.owl b) harder way include specifc type information in food.owl, e.g. above fragment becomes <owl:Class rdf:ID="Wine"> <owl:equivalentClass> <owl:Class rdf:about="&vin;Wine"/> </owl:equivalentClass> </owl:Class> c) very hard way make a new file which just has the type information for both wine.owl and food.owl and import them. 5: delete xmlns:xsd="...." from wine.owl This is not used 6. Keeping wine.owl in OWL DL - imports object Food.owl declares its own URI to be of type owl:Ontology using the xml:base and the idiom <owl:Ontology rdf:about=""> The xml:base ends in a # which gets ignored when resolving the empty same document reference, so that the subject of the rdf:type owl:Ontology triple is the URIref http://www.w3.org/2001/sw/WebOnt/guide-src/food However the object of the owl:imports triple is http://www.w3.org/2001/sw/WebOnt/guide-src/food.owl While the imports mechanism works, this uriref remains untyped and so the document is in OWL Full. Deleting the .owl suffix may work (it definitely will work if you use .rdf as your suffix). Including the suffix in both places will also work. (I am still working on this review - there may be one or two more points) Jeremy
Received on Wednesday, 30 July 2003 19:13:35 UTC