Re: New tags. (fwd) -Reply (fwd)

Joe English (joe@trystero.art.com)
Sat, 08 Feb 1997 20:40:17 PST


Message-Id: <9702090440.AA17426@trystero.art.com>
To: www-html@w3.org
Subject: Re: New tags. (fwd) -Reply (fwd)
In-Reply-To: <199702041216.EAA16787@server.livingston.com>
Date: Sat, 08 Feb 1997 20:40:17 PST
From: Joe English <joe@trystero.art.com>


MegaZone <megazone@livingston.com> wrote:

> The IETF has handed control of HTML to the W3C -fact of life, stop whining.
> And no other organization is interested in it.

IIRC, the IETF didn't want any part of the HTML standardization
effort to begin with either (since it's a file format, not a
protocol).  And the HTML-WG didn't exactly "hand control"
over to W3C -- it was more like the W3C principals said
"OK, this isn't working, we'll take over now."


> But it isn't.  The ONLY standardization process with any credibility
> is the W3C - that is the only one the major browser makers and authoring
> tool houses are going to listen to.

I have a different opinion about who's listening to whom here :-)
What MSIE and Netscape Navigator implement seems to have a far greater
influence on what gets into the "official" DTD than vice versa.

As to the RFC1866 vs. Wilbur vs. Cougar vs. HTML Pro debate,
it's really a toss-up:  Since HTML 2.0 and 3.2 can't seem
to make up their minds whether they are prescriptive (i.e.,
what authors should use) or descriptive (i.e., what browsers
can be expected to handle), they seem more useful to people writing
browsers than to people writing Web pages -- and only then
so that the browser writer can put a "Supports W3C-endorsed HTML 3.2"
icon on their splash screen.

HTML 3.2-conformant documents are likely to be rendered
correctly on most current browsers, but the DTD has too much
backward-compatibility (with old documents) and not enough
forward-compatibility (with future browsers and versions of the
DTD) to make it terribly useful for people who want to make sure
that their Web pages are going to work, not just today, but six
months from now.  For example: Cougar still doesn't require <HEAD>
and </HEAD> tags or ALT attributes for IMGs, and it still allows
the long-deprecated XMP and LISTING elements by default, just
to name two.

Conversely, neither HTML 3.2 nor Cougar include enough of
the bleeding-edge Kewl New Feechurs to be useful to authors who
*want* to be bleeding-edge.  Given the choice of using Netscape
du jour's whiz-bang new tags or running a validator, many
authors simply won't validate.  HTML Pro is very useful for
people in that camp.  While there is no guarantee that an HTML
Pro-conformant document will render correctly on _all_ current
browsers, validating against the Professional DTD will at least
ensure that all the tags are correctly nested, HREFs quoted,
t's crossed, and i's dotted; and ensuring that much can go a
_long_ way.

More than that: HTML Pro, unlike Cougar, *does* enforce some level
of "best common practice" (e.g., the distinction between phrase-level
and block-level elements, requiring ALT attributes, &c), and Peter has
done a very good job of cataloging which features are supported by
which versions of which browsers (including those supported by "fringe"
browsers like Lynx and Emacs-W3, which Cougar ignores completely); this
helps authors make an informed choice.

But like I said, it's a toss-up.  The principal tenet of the
SGML philosophy (and don't let anyone tell you any differently)
is: "they're *your* documents: you have the right (and the
responsibility) to choose how they should be encoded."  IMHO,
Peter is no less qualified than the W3C to write a DTD to which
I'd entrust my Web pages.  In many ways he's more qualified,
since he listens to people who actually write Web pages; I
I don't see much evidence of that from the W3C.



--Joe English

  jenglish@crl.com