Paul Novitski schreef:
documents be written according to the prose of the specification and
not just the machine readable components of it.
The DTD almost always errs towards the liberal, it is expected that
That's a very interesting assertion and gets right to the heart of
many of the debates on this list. It sounds counter-intuitive to me:
I would expect the prose to be more liberal than the machine-readable
DTD. Can you recall the source of that expectation? If we could nail
that one down it would certainly help clear up much of the apparent
tension between the very specific DTD and the comparatively loose
descriptive passages of the spec.
I read the HTML spec as an annotated DTD, using prose to discuss and
exemplify the element and attribute definitions for us mushy wetware
types. Every section of the spec begins by quoting the DTD and then
discussing those definitions. On a quick re-reading of the spec's
introductory sections I don't see where we're advised to place more
authority in the prose than in the DTD.
Just to maintain perspective let me add that I'm pursuing this aspect
of the discussion NOT as a campaign for fieldsets without form
controls (I feel that part of the debate has been laid to rest) but
rather because I want to better understand the DTD and its
relationship to the spec, especially in a case like this where they
appear to contradict.
I was just reading something about this on the HTML5 mailing list (so it
might not be applicable to the current HTML/XHTML versions). In the
HTML5 working draft spec it says:
"...
To put it another way, there are three types of conformance criteria:
Criteria that can be expressed in a DTD.
Criteria that cannot be expressed by a DTD, but can still be checked by
a machine.
Criteria that can only be checked by a human.
A conformance checker must check for the first two. A simple DTD-based
validator only checks for the first class of errors and is therefore not
a conforming conformance checker according to this specification."
So in HTML5 the spec is definitly not just proze version of its DTD, but
a lot more than that. Not all of which can be expressed in a DTD.
Although I'm not sure I guess something similar will be the case with
the current HTML and XHTML specs.
cheers,
Sander
*******************************************************************
List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm
Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm
Help: [EMAIL PROTECTED]
*******************************************************************