Re: marked sections
Case #2 is easily the most readable and useful. Like Michael, I've tried some of
the others and switched.
As far as I can glean, James Clark's objection to marked sections apply only to
their use in mixed content. He has proposed to solve this (and other RE related
problems) by abolishing mixed content for XML. I think that is an excellent
I would propose going a step further and dropping all declared content from XML.
not just CDATA and RCDATA, but also EMPTY. (I would use "(#PCDATA)" instead of
EMPTY. I realize that it doesn't allow an XML parser to enforce emptiness, but
there are many application-specific constraints that even an SGML parser can't
enforce. Inability to enforce emptiness would be a small price for XML to pay to
achieve simple, DTD-free, instance parsing.)
Charles F. Goldfarb * Information Management Consulting * +1(408)867-5553
13075 Paramount Drive * Saratoga CA 95070 * USA
International Standards Editor * ISO 8879 SGML * ISO/IEC 10744 HyTime
Prentice-Hall Series Editor * CFG Series on Open Information Management