W3C home > Mailing lists > Public > www-html@w3.org > January 2003

Re: XHTML 2.0 considered harmful

From: David Woolley <david@djwhome.demon.co.uk>
Date: Tue, 14 Jan 2003 21:49:57 +0000 (GMT)
Message-Id: <200301142149.h0ELnw101634@djwhome.demon.co.uk>
To: www-html@w3.org

>      "xHTML 5.0" language. Clearly a successor to HTML 4, feature-oriented,
>      made for the _web_.

I think this rather depends on how you define _web_.   I suspect your
definition has a rather tortuous etymology in that the sorts of site you
are talking about don't actually form a world wide web of interconnected
links (the web never was the physical communication links; it was always
the information links, even though some popularist glossaries confuse them).
What you are really talking about is the incrementally loaded marketing
documents and thin client applications designed to be accessed by people
using Internet Explorer and similar.

The non-webbed view of the "web" certainly accounts for most of the popular
books written on the subject and most of the specific employment.  However,
there are people, even some doing so for commercial reasons, who want something
closer to the original concept.  I'm not sure if they are under-represented in
the market for books because they are few in number, or simply because they
don't need the books because they are at home with markup langugages, and
don't need to learn tricks to make browsers do things that HTML was not
intended to do.

A consequence of this is that there are are actually several different sorts
of thing that people want to do with "web" browsers, that would probably 
justify specific markup languages, although there will always be a the problem
that having a single universal solution tends to be easier to sell to 
organisations.

One application would be marketing documents.  This might be the most 
difficult as the nature of the people who create such documents is that
they want to be different from everyone else (or rather everyone else in
the previous round, as they often are all different in the same way!).
Being able to specify in a standard way might well not be attractive.  Another
problem in this case is that people do not want to reveal the true meaning
of a document, because it is often about the psychology, rather than the
literal content.

Most personal web pages try to emulate this sort of site, but authors are
probably even more WYSIWYG oriented.

Another area is thin client applications, where forms, client side
scripting, and multiple views can be important (ignoring accessibilty).
This is complicated in that there is often an desire to impose a corporate
image, which again introduces a desire to be different, and blurs the
distinction from the previous category.

Then there is the application for which XHTML 2 is designed, which
tends to be librarianship and simple structured textual documents.
This is also the target, I believe, of ISO HTML.  Oddly, the sort of
document that fits most with this, and the original spirit of HTML,
tend now to be produced as PDF, possibly because HTML, as a relatively
new technology, tends to be a marketing plaything.

I'd suggest that HTML 4 is not well designed for your concept of a "web",
but has been compromised from the requirements of the last sort of use.
I tend to think that writers of marketing documents will never really be
happy with structural languages, and really want page descriptions 
languages, and it is only an accident of history that PDF didn't become
the language of that "web".

(I've ignored the factor that existence of HTML has raised the possibility
of accessible sites and therefore removing structure in the markup may
no longer be possible.)
Received on Wednesday, 15 January 2003 01:50:57 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:15:54 GMT