Re: HTML 3.2

Daniel W. Connolly (connolly@beach.w3.org)
Tue, 07 May 1996 23:53:10 -0400 

Message-Id: <m0uH0JL-0002U5C@beach.w3.org>
To: Charles Peyton Taylor <CTaylor@wposmtp.nps.navy.mil>
Cc: www-html@w3.org
Subject: Re: HTML 3.2 
In-Reply-To: Your message of "Tue, 07 May 1996 14:39:35 -0800."
             <s18f60b3.005@wposmtp.nps.navy.mil> 
Date: Tue, 07 May 1996 23:53:10 -0400
From: "Daniel W. Connolly" <connolly@beach.w3.org>

In message <s18f60b3.005@wposmtp.nps.navy.mil>, Charles Peyton Taylor writes:
>  It would be nice
>if the W3C put out standards for a logical, more
>functional HTML, and then the vendors coded to 
>that spec, rather than the other way around.

Wouldn't it? In fact, we _are_ headed in that
direction. The first step in that direction is the <OBJECT> spec:

	http://www.w3.org/pub/WWW/TR/WD-object


>No, but I think the effort put out in adding 
>presentation markup in HTML as opposed to 
>CSS might cause CSS to lose user focus and
>fade away.

This is a good point. I'm noodling on some different
approaches to this issue. Things like HTML 3.2-strict
DTDs that strip out all the stuff you could do
with stylesheets in stead of HTML markup.


>But WHY is there no CLASS attribute?
>The browsers don't have to do anything with it, 
>but it should be there for those of us who are 
>experimenting with Styles and Arena. 
>
>I feel I should be able to use the class attribute 
>and still have my document validate. (This is
>kinda important when you want to do web pages 
>for a living.)


Another good point. The ID/CLASS/LANG/STYLE attributes got thrown out
all at once, on the basis that they weren't supported: ID might make
folks think they could link to <p id=XXX>...</p>, when they can't
(YET!). LANG implied I18N support, STYLE implied stylesheets, etc.

I guess CLASS was a case of the baby going out with the bathwater.

I'll look into it.

>*My* disappointment is caused by not being
>able to use the neat stuff in HTML 3 (the 
>March '95 one).

Keep in mind that HTML 3.2 is descriptive: if an idiom
doesn't work in 3 or four major browsers, it doesn't
belong in HTML 3.2. On the other hand, we _are_ developing
new specs.

>For example: 
><FIG> ... I know there has been
>      a lot of discussion, but in my book, 
>      <Fig> rules. Ditto with <caption>.

Have a look at the <OBJECT> spec:

	http://www.w3.org/pub/WWW/TR/WD-object

You'll be pleased to see shaped anchors ala fig.

In fact, at one meeting, somebody observed that <fig> could be
trivially implemented in terms of <table> and <object> so it should go
in the draft. I think <fig> will go in soon, but I'm not sure when.

><UL src="%url"> *many* people use graphics for
>      Bullets, so why not make it easier?

We had a long talk about this. The counter-argument was: where
dou you stick height/width/hspace/vspace etc? What about using
<object> for a bullet? Basically, we never came up with a crisp
notion of something that would work.


><UL Plain> how hard could it be to just
>      *not* use bullets?

This is something you'll almost certainly see soon.


><HR src="%url">  makes your page look better
>      in lynx or any other browser.  Conversely,
>      how does "Dash, Dash, Dash, Dash, Dash..." etc.
>      sound out loud?  (From the alt text to a speech
>      synthesizer.)

I like this one too. But it has the same problems as <ul src=...>.

If somebody could do a crisp write-up of the syntax and semantics,
that would be nice.

>Why is <center> even in there when we have
><DIV align=center>?

We considered making <center> deprecated in favor of <div
align=center>, but we really haven't written things up to that level
of detail yet.

Dan