W3C home > Mailing lists > Public > public-html@w3.org > March 2008

RE: Exploring new vocabularies for HTML

From: Justin James <j_james@mindspring.com>
Date: Mon, 31 Mar 2008 14:25:42 -0400
To: "'Bruce Miller'" <bruce.miller@nist.gov>
Cc: "'Henri Sivonen'" <hsivonen@iki.fi>, "'David Carlisle'" <davidc@nag.co.uk>, <ian@hixie.ch>, <public-html@w3.org>, <www-math@w3.org>
Message-ID: <037a01c8935c$a0ddbda0$e29938e0$@com>

> Hmmm.... On the one hand, I very much sympathize with your
> point of view.  However, one can already do some of that now,
> modulo interoperability (or lack thereof) of <object>, etal.

Which is really my stance. :) If people really want to be able to embed
MathML (or any other format), they can write browser plugins that do this.
But by the time I finished writing this, I realized just how wrong I was.

> Sorry, but Math _is_ special....

Only because of HTML's original usage to exchange documents within the
scientific community. Just because a high concentration of the people on
this list are math oriented does not mean it needs to have a special
exception. Once you make one exception for MathML, everyone will want an
exception for their own pet format. Why can't Microsoft demand that OpenXML
be properly handled by HTML5 in an unobtrusive way? They could quite
convincingly argue that OpenXML is in the same spirit of HTML (document
exchange), and that it is XML, and certainly a lot more people generate
OpenXML compared to MathML documents. But we don't allow it, for the simple
reason that if someone wants to view OpenXML, they should use the
appropriate OpenXML parser (Microsoft Word, Excel, etc.), not an HTML parser
(Internet Explorer, Firefox, Opera, Safari, etc.).

I know, this is a "slippery slope" style argument which is always weak as
far as a formal debate would go. But I honestly believe that 90% of the
problems that users experience on the Internet are caused by authors (with
the spec and aggressive browser and tool vendors to enable them) create HTML
documents that completely deviate from the original intention of HTML.

The really, truly sad irony here, is that the problem is the very existence
of MathML. This should have been a full subset of HTML from the beginning!
Looking at the original intent of HTML, the people it was designed by and
for, and the types of documents that originally dominated HTML, including
mechanisms for including mathematics would have been a perfect fit. In a
way, you are right. Math *is* a special case. In fact, math is also *not* a
special case.

Why? Because it is only reasonable that to make a "document", one be able to
handle "numbers" and not merely treat them as a range in a regex of
allowable characters for input. Things like the full OpenXML spec, that is
much more than the "lowest common denominator" that HTML aims for. The same
for, say, a source code formatting spec. But being able to display numbers
is a critical part of HTML, and should be there.

I know, I am doing a 180 here, within my very own post.

Let's try thinking about this like so:

The need for proper formatting of math is no different, conceptually, from
the need for proper prose formatting. This means that just as HTML supports
certain concepts in prose such as paragraphs, quotations, abbreviations, and
so on, it should treat the world of numbers as in equal partner. In fact,
many of the constructs and assumptions in HTML regarding text are not
applicable in all languages, but math formatting is universal. If anything,
math is more deserving of full, standardized inclusion within HTML than text
itself. In other words, I fully support a unilateral folding of MathML into
HTML as a full "child format". HTML does not have these yet. But it should.
SVG should be the same way (you are right that drawing is special too).
Essentially, a "child format" is a subset spec that can stand on its own,
but it is fully embeddable and compatible with HTML. Basically, we need to
accept the concept that HTML as it currently stands is 100% designed to be a
nearly 100% text standard, but authors also need full drawing and number
support. These are the three elements of any document: images, text, and
numbers. By properly wrangling all three of these to be in line with each
other and working together, we can enable HTML to be truly useful as far as
its original purpose goes. We spend way too much time on things like making
AJAX work better, which have nothing to do with HTML's try purpose, while
ignoring things like MathML and SVG, which have everything to do with HTML's
purpose.

I know, it would be a political nightmare to do this, but I see it as the
best route available, compared to trying to incorporate standards without
cooperating with them, or continuing to allow HTML to be a sub-par format
for anything but basic text and imitation desktop applications.

J.Ja
Received on Monday, 31 March 2008 18:26:42 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:13 GMT