RE: XHTML comment form National Geographic

Curtis,

    (I am speaking for myself, not any Working Group or Corporation.)

    While I appreciate your enthusiasm for XHTML, I would like to caution
you in regard to statements you make below.

-----Original Message-----
From: Curtis C. Hovey [mailto:chovey@nationalgeographic.com]
Sent: Wednesday, April 07, 1999 2:21 PM
To: www-html@w3.org
Subject: XHTML comment form National Geographic


>I love it.  As the webmaster of nationalgeographic.com, I will be mandating
it's implimentation on our >site in the next two months.   

What applications do you expect to support it? 
 
>I think this will help ensure that we are producing valid documents by
simply letting us or the >developer attempt to open the document as XML.  

	I only wish it was that simple. The concepts of structural
validation are not changed by XHTML; both you and your developers can open a
file 'as XML' without any attempt at validation. The only real constraint is
that XML documents must be well-formed; this is a small improvement over
current HTML practice but not a major one. Document validity is generally of
interest only to document authors, and so is unlikely to occur on the client
side. (This is not the same thing as ignoring the DTD, btw.) In short, XHTML
does nothing whatsoever to improve the level of structural error checking
authors attempt.
Nor is there any reason to believe that authors will suddenly embrace
validation when it has proved of little value so far on the web.

>I believe it will also improve the portability of our content by letting us
use and XML/XSL processor >on it.  

	I am unclear on the import of this statement; XSL will work on any
XML document not just XHTML. XSL's transformational qualities may improve
portablity by allowing for run-time transformation to a different document
type that is acceptable to a specific client. But it has nothing to do with
XHTML.


>Of course the name-space option will open an opportunity to extend
functionality without worrying >about plugins and their distribution. 

	This is perhaps the most insidious error that is currently held in
the popular imagination with regard to XML/XHTML. In fact, nothing of the
sort if true. Namespaces will allow controlled extensibility of XML web
pages, but it cannot ensure that the user's browser posseses the needed
functionality to properly present the content to the user. If you send me an
XHTML document with foobar extensions, and my browser doesn't have a foobar
content handler, your page will fail. This doesn't mean that the page is
wrongly constructed; it may well be conformant and well-formed according to
the specification; it merely doesn't work. Plugins and their distribution
will continue to play the role of functionality extenders on the web, XML or
no.

>We intend to use this as a part of our internationalization effort later
this year.

	Well, this is certainly one area where XML can be of use. XML
requires the use of Unicode, a standard for character representation on
computers designed specifically to allow for ease of internatinalization in
documents. Once again, it doesn't have anything specifically to do with
XHTML.

	I am not writing all of this in order to rain on your parade Curtis.
:) I just don't want to see people act on their mistaken impressions of
XHTML's value. XHTML will be very valuable and worthwhile, but it is not a
panacea for all the Net's problems. And I like the Nat. Geog. and don't want
to see them  embark on a problematic journey.

Regards,

D-

Received on Friday, 9 April 1999 18:54:54 UTC