Validation was Re: Non-conformance by W3C members

--- Jonas_Jørgensen <jonasj@jonasj.dk> wrote:
> 
> Karl Dubost wrote:
> > for example You can make a valid document, which
> > is completely incorrect for the semantic, like using
> > a "blockquote" element to indent un text and
> > not  for a citation.
> 
> IMO such a document cannot be said to be valid. That
> validator.w3.org doesn't find any errors does not mean
> that the document is valid, similar to how running a
> document through your word processor's spell 
> checker is no match for having it proof-read by a person.

The W3C page on document validation somewhat discusses this problem
at <http://www.w3.org/TR/html4/sgml/intro.html>, even though they're
focusing on technical mistakes like illegal (but 'valid') attribute
values.

Validation is simply a check against the referenced DTD, nothing
else. So any HTML document that conforms to the referenced DTD is
valid. Whether the document uses the right markup for the right
content, or whether the text makes any sense at all, is not part of
the validation process. That's why even valid HTML, just like
spell-checked documents, can be complete gibberish. See
<http://groups.google.com/groups?selm=35080%40sdcc12.ucsd.edu&output=gplain>.

But of course I do agree that semantic is very important. This aspect
of the standards sometimes gets lost in the recent coolness of
validation. Perhaps it gets lost because correct semantics require
that the author actually know the standards, while validation can be
done by a stupid machine.


Matthias


__________________________________________________
Do You Yahoo!?
Yahoo! Finance - Get real-time stock quotes
http://finance.yahoo.com

Received on Monday, 9 September 2002 08:20:29 UTC