W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > January to March 2001

Deprecated was: Layout tables

From: Matt May <mcmay@bestkungfu.com>
Date: Fri, 9 Mar 2001 09:16:16 -0800
Message-ID: <067a01c0a8bc$b6856f50$6501a8c0@sttln1.wa.home.com>
To: <w3c-wai-gl@w3.org>, "William Loughborough" <love26@gorge.net>
----- Original Message -----
From: "William Loughborough" <love26@gorge.net>
> Let's get specific.
> Do we go back to "until user agents" and detail how to do stuff without
> using CSS?

No, and yes.

I think "until user agents" is no longer an option, since according to the
F2F, checkpoints are to be normative as well. (Right?) Something like
"Evaluate (future technology X), because this is what we're expecting to
specify," or, conversely, "use of (old technology Y) is deprecated: use (new
technology Y) where appropriate," may work.

The issue with layout tables, however, is something I'm afraid we're stuck
with as long as we have HTML or a derivative. The utilization of tables for
layout is basically all everyone knows for formatting on pages (thanks,
David Siegel!). There are terabytes of HTML out there with layout tables,
and redesigning them all with CSS simply isn't going to happen. (This is one
of the issues I think of which suggest to me that a separate "legacy HTML
techniques" document would be helpful.)

I'll bet the number of people who know HTML is an order of magnitude larger
than those who know CSS. Telling the 90% who aren't aware of CSS to unlearn
tables for layout and do this other thing is not something the guidelines
can force successfully. The bulk of HTML books published up to this point,
the authoring tools that design pages this way, and the inertia of designers
guarantee that layout tables will continue to be generated, and if this
group can come up with a way to accommodate that (while still saying "you
shouldn't do that"), I think that's a necessary evil, and something where
"deprecated" may be of greater assistance in getting people to change their
behavior than "just say no".

> Do we ignore SMIL, SVG, and other probably-long-time-a-coming stuff and
> have them as notions informing checkpoints?

Well, we can't require them if they can't be used. SVG has a long way to go
before it's considered viable by the web at large. If we say "use SVG and
not Flash" as a requirement for a certain level of compliance, the average
Flash-using site will respond by not trying for that level, and the baby
goes out with the bathwater. Until the authoring tools offer parity between
accessible and non-accessible technologies, anyway.

> If as is claimed 96% of all systems in use have capability of displaying
> Flash or delivering RealMedia are the other 4% (of which it is likely that
> an inordinately high proportion are our clients) to stare hopelessly
> that particular arm of the digital divide?


Most technologies, HTML 4 included, require increased processing power to
provide additional functionality. That's not something that can be fixed as
long as Moore's Law holds up. I don't see a reasonable way to state in the
guidelines that compliance to (some moving baseline of) legacy hardware is a
core checkpoint, especially where the W3C is trying to ensure that future
multimedia technologies are "accessible." Would they have to tell them, "oh,
and your new product has to run on a 386, too"?

Received on Friday, 9 March 2001 12:16:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:47:09 GMT