Editorial comments
1. In section 1, it says "RDF is based on a graph syntax, which is typically serialized using XML.". I think it would be good to make clear here that the RDF graph syntax is not equivalent to the XML infoset, and that the XML serialization really constitutes a "recipe" for constructing the conceptual graph. Alot of folks seem to get confused when the RDF graph doesn't mirror the RDF/XML or visa versa, so making this clear up front may be helpful. 2. In section 2, it says "RDF builds on XML". I don't think this is accurate. RDF uses XML to serialize its graph structures, but is not an extension of XML, but only includes an application of XML. 3. In 2.2, a stated design goal is "Use XML schema datatypes". I think it would be better to say "Support the use of XML Schema datatypes". Otherwise, it may be taken to imply that XML Schema datatypes are integrated with or into RDF directly, which they are not (and IMO should not be). Ahh, yes, I see that this is clearer in the discussion of this bullet item, but still, perhaps the bullet item could be clearer. 4. In 2.4.3, perhaps "are not covered by this recommendation." could be expressed more strongly as "are not licensed by this recommendation for interchange of RDF graphs".
See: http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2002Aug/0074.html
Comments raised to WG prior to initial WD publication, but not folded in to the document.
[GK] Editorial fixes clearly in GK's text.
Item 1: added comment in section 1.
Item 2: instead: "RDF is a member of the family of languages that use XML, ..."
Item 3: instead: "Support use of XML schema datatypes"
Item 4: I think "not covered" says it all here. In connection with
the first part of the sentence I think its perfectly strong enough.
See: http://www.ninebynine.org/wip/RDF-concepts/2002-10-18/rdf-concepts.html
Changes folded into publicly accessible document.