I would like to thank the SGML Open group for these recommendations.
I share concerns about incompatibility with SGML, the procedural
semantics, and the overriding concern about backward compatibility.
These concerns force me to continue to consider HTML documents an
"end product", not a method for coding "working documents". This
difference, when magnified by large quantities, can cost web site
developers significantly, thereby reducing our ability to
"give back to the web".
I believe that addressing the procedural semantics will take time. So
I would suggest that the release of version 2.1 be reconsidered,
even if it is just advisory. 2.1 would provide a basis for browser
developers to advance (or justify their claims for advances) and test
some of the style sheet functions. Meanwhile 2.1 would give us the time
to reconsider the structural implications hierarchy containers and
eliminating some or most of the procedural elements.
> Therefore, we strongly recommend that HTML parsers (such as are
> embedded within WWW clients) be built to interpret HTML documents in
> the same way SGML parsers would (though of course they need not support
> features not required by HTML).
Perhaps, following the example of WWWLibrary_2, the W3C could would
with SGML Open's Parser Compatibility subcommittee and provide a
common, true, SGML-subset parser.
Dave Hollander Hewlett-Packard
Internet Technology Program Manager 3404 East Harmony Road, MS. 6U10
Electronic Sales Promotion Fort Collins, Colorado 80525
[log in to unmask] 303-229-3192
Access HP - http://www.hp.com