Re: DIV/CLASS: Mike Wexler: Re: HTML 3.2

Joe English (joe@trystero.art.com)
Tue, 14 May 1996 15:53:41 PDT


Message-Id: <9605142253.AA07051@trystero.art.com>
To: www-html@w3.org
Cc: dsr@w3.org
Subject: Re: DIV/CLASS: Mike Wexler: Re: HTML 3.2 
In-Reply-To: <199605142148.AA026890508@w3.org> 
Date: Tue, 14 May 1996 15:53:41 PDT
From: Joe English <joe@trystero.art.com>


Dave Raggett <dsr@w3.org> wrote:
> [attribution lost]
> > How about:
> >    DIV document divisions
> >       Groups related elements together. Can be used with the CLASS
> >       attribute to create meaningful document divisions such as
> >       Chapter, Section, and Sub-section. 
> >       Can be used with ALIGN attribute to set the text alignment of the
> >       block of elements it conains. ALIGN can be one of LEFT, CENTER, or
> >       RIGHT. Requires start and end tags.
> > 
> > > I think the issue of the CLASS attribute could be opened up.
> 
> This is fine for the forthcoming version of HTML which supports
> style sheets. HTML 3.2 on the other hand captures the state of
> HTML as deployed in early `96.  ID and CLASS are therefore not
> part of HTML 3.2.


But CLASS is useful -- or at least could be -- independant of
any stylesheet mechanisms.  Adding user-defined semantic information to
HTML documents is still considered by some to be a worthwhile thing
to do, even if _browsers_ don't do anything with that information.

At the very least, it would allow authors to use CLASS in anticipation
of stylesheet deployment; the attribute is completely backward- and
forward- compatible.  At best, it would serve as a foundation
for site- and community-specific experimentation with HTML as a richer 
medium for information delivery.

Paul Prescod put it best in
<URL: http://www.acl.lanl.gov/HTML_WG/html-wg-95q4.messages/0795.html >,
included below for reference.


--joe@art.com


|Message-id:     <199512071847.NAA18993@undergrad.math.uwaterloo.ca>
|From:           Paul Prescod <papresco@calum.csclub.uwaterloo.ca>
|To:             Multiple recipients of list <html-wg@oclc.org>
|Date:           Tue, 12 Dec 95 11:46:22 EST
|Subject:        Re: draft-ietf-html-style-00.txt
|Reply-To:       papresco@calum.csclub.uwaterloo.ca
|X-Listprocessor-Version: 6.0c -- ListProcessor by Anastasios Kotsikonas
|X-Comment:      HTML Working Group
|
|At 10:17 AM 12/7/95 +0700, you wrote:
|
|>2) Is there some markup purpose OTHER than style for which the HTML
|>   working group wants to introduce the new (not defined in RFC 1866)
|>   attributes of CLASS and ID?
|
|Yes! 
|
|CLASSes can be used for searching.  ADDRESS is a wide open category, but
|<ADDRESS CLASS="email"> is much smaller.  Maybe you don't care if email
|classes are _formatted_ differently, but you want to be able to search for them
|.
|
|CLASSes can be used to recover the "original form" of a document where the
|source format is lost or unavailable.
|
|CLASSes can be used to convert HTML documents into some other form of
|document that doesn't have a 1 to 1 mapping from HTML elements.
|
|CLASSes can be used for addressing.  "Give me the last ADDRESS of type EMAIL
|in the document."
|
|CLASSes are supposed to have semantic names so that someone without access
|to a supported style-sheet platform (e.g. visually impaired people) can
|reverse-engineer the structure of the document and build their own
|style-sheet specific to their "platform."
|
|IDs are a way of referring to an individual element in a document, for
|searching, formatting, hyperlinking, or referencing.  Even today, they are
|desperately needed just for hyperlinking without a bunch of extra <A> tags.
|Furthermore, I believe that Netscape and Spyglass already use it for their
|MAP implementation.
|
|It worries me that this discussion has become so style-centric.  Have we
|forgotten that HTML documents can be used for things other than "surfing?"
|
| Paul Prescod