W3C home > Mailing lists > Public > www-html@w3.org > February 1996

RE: Initial Draft --Cascaded Speech Style Sheets

From: David Seibert <seibert@hep.physics.mcgill.ca>
Date: Thu, 15 Feb 1996 21:54:14 -0500 (EST)
To: Gavin Nicol <gtn@ebt.com>
Cc: cwilso@microsoft.com, www-style@w3.org, www-html@w3.org
Message-Id: <Pine.ULT.3.91.960215204811.16514A-100000@prism.physics.mcgill.ca>
On Thu, 15 Feb 1996, Gavin Nicol wrote:

> Chris Wilson wrote:
> >1. I don't feel that this is the only major application of stylistic
> >properties to text; I think many authors are the hack 'n' slash kind,
> >who will think "I want this word to be blue," without considering what
> >it is about that text that makes them want it blue.  

Supporting "hack 'n' slash" authors should not be a major goal of  
stylesheet designers.

> >                                                     Also, sometimes it
> >*IS* purely presentational - in writing poetry, for example, the author
> >often wants a particular format of presentation - the goal is to express
> >or convey a feeling, not to describe the content model of each piece of
> >text on the screen.
> >

You can convey the feeling at least as well through markup, even for a 
poem, as markup allows you to control both visual and aural presentation 
in a consistent manner.

> >                      Especially, it would seem, in creating
> >advertisements (inarguably a major faction of Web publishing), designers
> >often want a particular font face, size, color or whatever, applied to a
> >section of text for purely presentational (vs. content-based)
> >reasons.

OK, here you can justify uniqueness for its own sake, for business reasons.

> For such cases, people might be *far* better off using PDF or RTF, or
> defining a tag set all of their own, with associated stylesheets. They
> could define: <FONT>, <COLOR>, <MARQUEE>, or whatever they want.
> >2. An explosion of tags to allow for tagged content (e.g. a particular
> >tag solely for tagging names of people, like <PERSON> from an old HTML
> >3.0 draft) would be great, if everyone could agree on the list of tags,
> >and keep their UAs relatively up-to-date. 
> Alternatively, they could use a language which allows them to specify
> an arbitrary set of tags.
> >The abstraction through CLASS allows unique tagging without having to
> >extend the DTD.  
> Sure, and supporting this is *at least* as complicated as supporting
> unlimited tags. That is what the point of my posting with the "spam"
> element in it was.

I think that there is a psychological advantage to encouraging use of 
the class attribute to generate new tags rather than just having people 
make up their own tags purely as they wish.  People will tend to think 
of element.class as a refinement of the basic element (if not, they can 
be taught this interpretation relatively easily), so the intended 
meaning of element.class will probably not stray too far from the 
meaning of element.  Thus, UAs will at least be able to make an  
intelligent guess as to how the author wants them to render  
element.class as long as the element is familiar to them, while this  
won't be true if authors construct arbitrary element names.  As time 
goes on, UAs will pick up the most popular refinements of the elements, 
which are probably more important to most authors (except advertisers)
than the ability to describe totally new elements.

<Lots of speculation removed>

> >Wrapping a more-or-less-stable DTD and the stylistic properties
> >applied to elements in that DTD into one (easier-to-implement)
> >package has allowed HTML the glory of driving Internet publishing to
> >incredible heights. 
> I do not consider this incredible heights. I consider this incredible
> chaos, and incredible noise to signal ratios. 

Gavin and Chris are both right - internet publishing has been driven to 
incredible heights by commercial publishing standards, where the money 
you make is all that counts.  Too many web pages are reminiscent of the 
world's most prolific author; he was quoted in the paper here last  
Saturday as saying that he had never read most of his own work.  I don't 
remember his name, and I'm *sure* I've never read anything he's written, 
but he certainly could be called very successful by commercial  

Do we want the web to be the best source of information, or just to make 
money?  I realize that TV is the opiate of the masses, and that the 
web could easily develop into just such a form of cheap entertainment, 
and web developers could make a lot of money that way.  However, I think 
it also has a lot of potential for being incredibly useful, and that 
capabilities for transmission of high-quality information should be 
developed in parallel with capabilities for high-tech entertainment.


Work: seibert@hep.physics.mcgill.ca         Home: 6420 36th Ave.
Physics Department, McGill University       Montreal, PQ, H1T 2Z5 
3600 Univ. St., Mtl., PQ, H3A 2T8, Canada   Canada
(514) 398-6496; FAX: (514) 398-3733         (514) 255-5965
Received on Thursday, 15 February 1996 21:54:20 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 30 April 2020 16:20:17 UTC