Re: WWW Support Questions

Bert Bos <bert@let.rug.nl>
Errors-To: listmaster@www0.cern.ch
Date: Mon, 9 May 1994 21:27:59 +0200
Errors-To: listmaster@www0.cern.ch
Message-id: <9405091916.AA27729@freya.let.rug.nl>
Errors-To: listmaster@www0.cern.ch
Reply-To: bert@let.rug.nl
Originator: www-talk@info.cern.ch
Sender: www-talk@www0.cern.ch
Precedence: bulk
From: Bert Bos <bert@let.rug.nl>
To: Multiple recipients of list <www-talk@www0.cern.ch>
Subject: Re: WWW Support Questions
X-Listprocessor-Version: 6.0c -- ListProcessor by Anastasios Kotsikonas
Content-Type: text/plain; charset=US-ASCII
Content-Type: text/plain; charset=US-ASCII
Mime-Version: 1.0
Mime-Version: 1.0
X-Mailer: ELM [version 2.4 PL13]
X-Mailer: ELM [version 2.4 PL13]
Daniel W. Connolly writes:

[lots of good stuff deleted...]

 |	And if we are to evolve HTML technology from its current
 |	very-useful-but-far-from-sufficient state, we MUST apply
 |	more formal methods to abstract the essential techinques
 |	from the various applications. I suggest we take a serious
 |	look at the architectural-form techniques from the HyTime
 |	standard development, and develop (1) a set of WWW
 |	architectural forms for linking and navigation, and
 |	(2) a stylesheet mechanism so that the WWW linking and
 |	navigation techniques can be applied to a variety of SGML
 |	DTDs.
 |	
 |</SoapBox>
 |
 |Whew! I needed that!

Hear! Hear!

I support this view wholeheartedly. Indeed, this is the subject of the
<A HREF="http://tyr.let.rug.nl/~bert/SGML.html"> WWW'94 workshop</A>
that I propose to organize. We need to reach enough consensus on this
topic that we can start a coordinated development effort.

Architectural forms are nice, I even wrote an HTML meta-DTD some time
ago (on paper, and I've thrown it away), but eventually I decided that
they have too many disadvantages compared to stylesheets. Unless we
have access to so many resources that we have development capacity to
spare, I think we should focus on style sheets first.

HyTime itself is much too complex for us at the present time, maybe in
five years time it will be feasible. HyTime-like Architectural Forms
have the advantage that they are pure SGML, but the disadvantage that
they are difficult to understand and that SGML doesn't seem to be a
very convenient language for writing programs in. Stylesheets can be
written in better languages.

Although I hate Lisp, I think the Scheme-based DSSSL (currently an ISO
draft standard, I believe) is the best bet when it comes to defining
the tree-transformations that define the on-screen/on-paper
representation of an SGML element. I'm waiting for the first WWW
browser with an embedded Scheme interpreter!

<sidebar>
  Surprisingly, a stylesheet in Scheme would work in exactly the
  opposite way from previously proposed stylesheet mechanisms. Simple
  stylesheets are `document-driven': they define for each SGML element
  how it would be rendered. A formatter would read a document as a
  stream and apply a style to each element it encounters.

  A stylesheet in Scheme is very different: a formatter would read the
  style and *execute* it as a Scheme program. The document is treated
  as a database from which the Scheme program can retrieve elements,
  in any order, even more than once or not at all.
</sidebar>


Bert
-- 
                     __________________________________
                    / _   Bert Bos <bert@let.rug.nl>   |
           ()       |/ \  Alfa-informatica,            |
            \       |\_/  Rijksuniversiteit Groningen  |
             \_____/|     Postbus 716                  |
                    |     9700 AS GRONINGEN            |
                    |     Nederland                    |
                    |     http://tyr.let.rug.nl/~bert/ |
                    \__________________________________|