W3C home > Mailing lists > Public > www-xml-schema-comments@w3.org > January to March 1999

No subject given

From: <doconnor@bna.com>
Date: Fri, 26 Feb 1999 14:09:59 -0500
Message-Id: <9902269200.AA920056432@centaur.bna.com>
To: <www-xml-schema-comments@w3.org>
     Thanks to all involved in the creation of this document, and 
     especially to the editors for the excellent, well organized 
     presentation of these ideas.
     
     The first two comments below describe functionality provided by XML 
     1.0 DTDs that does not seem to be provided for in the requirements. 
     The third comment relates more to conformance, but may be out of scope 
     for this effort.
     
     1) Inheritance - Inheritance of element, attribute, and datatype 
     definitions is addressed, however, I find it often useful to inherit 
     PARTS of definitions. 
     
     For example I define "classes" of element types such as "citations", 
     "emphasis", and "display", using parameter entities. I then use these 
     "classes" in a several definitions. I may define that paragraph 
     contain data (#PCDATA), footnotes, citation elements, emphasis 
     elements, and display elements. I may then define heading to contain 
     data, citation elements, display elements. I may then define footnote 
     to contains data and citation elements. Only being able to inherit the 
     entire definition of an element may make this very obscure and 
     difficult to implement. I have also used this technique with "legal 
     lists" of attribute values.
     
     2) Redefinition of inheritance - DTDs allow for the redefinition of 
     parameter entities. I have used this functionality to prepare reusable 
     DTD fragments that are configurable by redefining parameter entities.
     
     For example a DTD fragment contains a definition for the element type 
     "footnote." This definition presents the allowed content of the 
     footnote as "data and citation elements". A parameter entity is used 
     to provide the list of the citation elements. The various DTDs that 
     call this fragment may then override the definition of the citation 
     elements parameter entity.
     
     3) "Backward compatibility" - I strongly suggest that the schema 
     language define how a schema should be converted to an XML 1.0 DTD. 
     This may be addressed by item #4 under the heading "Conformance" but 
     it's not clear to me. 
     
     For example, if an attribute is defined to contain a "date" (or some 
     other non-XML 1.0 data type), what should it be converted to in an XML 
     1.0 DTD? Should it become CDATA or NMTOKEN.
     
     Thanks very much for the opportunity to participate in this process. I 
     look forward to seeing further work in this area.
     
     Dennis J. O'Connor
     The Bureau of National Affairs, Inc. 
     doconnor@bna.com
Received on Friday, 26 February 1999 14:12:29 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 6 December 2009 18:12:45 GMT