W3C home > Mailing lists > Public > www-html@w3.org > January 2001

RE: WWW: Interoperability Crisis?

From: Edward Barrow <edward@platopress.co.uk>
Date: Sun, 21 Jan 2001 11:48:15 -0000
Message-ID: <01C083A1.AAB5FA40.edward@platopress.co.uk>
To: "'David Meadows'" <david@heroes.force9.co.uk>, "www-html@w3.org" <www-html@w3.org>
On Sunday, January 21, 2001 9:47 AM, David Meadows 
[SMTP:david@heroes.force9.co.uk] wrote:
> I'll just jump in here if nobody objects...
> "Sean B. Palmer" <sean@mysterylights.com> wrote:
> > [...] It was originally a scientific document
> > format, but got evolved into Hyper*Media* Markup Language. But no-one 
> > any thought with how best to integrate media and HyperText, so they 
> > thought "oh, we'll embed it in the doucment"... and there you have it.
> [...]
> > Personally, I'd ban <img> and <script> from XHTML, but there you go.
> > Doesn't top you using <a> or <object>, does it?
> First define the objective that HTML is trying to achieve (to make
> scientific documents available on line). Then look at what a scientific
> document *is* and decide how you will achieve that objective.
> I have a few technical journals here. I'll see what they contain:
> (1) Text.
> (2) Images (charts and pictures) *embedded in* the text.

With respect, this assumes a print model for scientific documents (a.k.a. 
"papers"). It should be possible to transcend the limitations of  paper, 
using the new media better to explain science (or more generally, 

> To properly model the documents,

Why do you want to model something that is limited by its own obsolescent 

I would need the <img> tag that you want to
> ban. You want me to use <a> instead? But my technical articles integrate
> images and text on the same page. The text refers to the pictures. They 
> an integrated whole. Yet in effect you want me to say "now close this
> journal and open the accompanying picture book. Memorise the picture.

Who  says [*]  that links must be rendered this way? The link is semantic, 
not presentational.

> HTML, right from its inception, was, according to its stated purpose, a 
> for displaying multimedia content. It accomplished this task very poorly
> until graphical browsers and the <img> tag came along.

The <img> tag enables mainly flat 2D illustrations. Hardly multimedia.  The 
<img> tag is of its nature more presentational than semantic.

Hypertext is a powerful tool for explaining scholarship, since it can be 
used both to put scholarship into its wider context and to link it to 
greater detail - down to raw data. Multimedia extends it further by 
enabling non-verbal abstractions which may often be better than words for 
conveying scientific concepts. The <a> approach would link these 
abstractions semantically. Unfortunately, since user agents [* OK, the 
builders of user agents say that <a> must be rendered in a new 
environment.... ] work the way they do, it isn't a serious contender at the 
moment but I can certainly see the logic behind Sean's argument.

Strict separation of semantics and presentation imposes a useful discipline 
upon the new media and requires an ascetic approach to authorship which may 
at first seem at odds with the  potential of the technology, but it is no 
different to the disciplines of orthography and typography which are needed 
fully to exploit the potential of paper.

Edward Barrow
new media copyright consultant
Received on Sunday, 21 January 2001 06:59:56 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:05:56 UTC