Re: SGML confirming? and HTML conformance testing

Glenn Adams (glenn@stonehand.com)
Tue, 18 Apr 95 08:53:54 EDT

Date: Mon, 17 Apr 95 19:39:48 EDT
From: connolly@www10.w3.org (Daniel W. Connolly)

I think it would be more accurate to say "If the HTML implementations
had been stricter, there would have been much fewer illegal documents."

But if HTML implementations are being created according to the HTML
specification, and the specification leaves certain matters unspecified
or only implied, then who's to say that an HTML specification is not
strict? The HTML specification clearly needs to be more explicit in
its reference to the constraints imposed by ISO 8879 or it should replicate
the pertinent text from ISO 8879 into the HTML spec. For example,
the current draft 3 does not specify the behavior regarding:

1. internal declaration subset

- should a document be able to declare and reference internal
and/or external general entities?

- should a document be able to declare parameter entities in such
a way as to modify an externally referenced DTD? e.g., for including
and/or ignoring marked sections in the DTD? for extending/modifying
content models?

2. marked section declarations

- should an HTML implementation be able to parse these properly?

3. processing instructions

- should an HTML implementation be able to parse these properly?

4. minimized tags

- should an HTML implementation be able to parse these properly?

This list is probably not exhaustive. In any case, I think it would
be extremely useful to HTML implementers to at least be aware of these
components of SGML, which, if they only read the HTML spec, they may
fail to support properly.

Regards,
Glenn Adams