RE: Compliance and html validation - how to interpret?

Thank you so much Joe

The templates for NN4 compliant layouts look great in all the browsers I have
tested (including NN4) and all achieved with style control (no tables)! These
are the exemplars I needed. I believe that the site does now meet triple-a
compliance requirements(based on my assessment of manual checks that is) and it
does now pass html validation and CSS validation tests also.

Your help is certainly appreciated.


Denise



-----Original Message-----
From: Joe Clark [mailto:joeclark@joeclark.org]
Sent: Sunday, April 28, 2002 6:56 AM
To: WAI-IG; Denise.Wood@unisa.edu.au
Subject: Compliance and html validation - how to interpret?


> used lay out tables. The original site I developed using only CSS2 for
> positioning looked great in later browsers but of course is not really very
> aesthetic in broken browsers such as NS4 (I need to accommodate all browsers
> particularly NS 4 as this is still commonly used by many staff).

There are several NN4-compliant all-CSS layouts available: Mark
Newhouse's is the one I use <http://theBRML.org>,
<http://joeclark.org/axxlog/>. Craig Saila has one.

It *can* be done. It is not easy, but it is possible.

<http://homepage.mac.com/iblog/nn4_3col_header.html>
<http://homepage.mac.com/realworldstyle/>

(Mark does an excellent job of hiding his discussion of the layouts
on his blog. I can never find them and I give up looking.)

I also do not know of a single convenient page that lists every
NN4-compatible CSS layout. I will ask the css-discuss list if
somebody wants to make one.


> sure how to interpret the issue for compliance level 2 regarding "3.3 Use
style
> sheets to control layout and presentation". Does this mean that if I have
> chosen to support lower end browsers and have used layout tables that the
site
> is not triple-A compliant (ie I have not used CSS2 for positioning)? I have
> used CSS for all styles it's only layout that is the issue here.

I would simply look the other way. As in so many other cases, the
WAI WCAG 1.0 requirements are pie-in-the-sky and reflect the
authors' lack of use of real-world Web sites. CSS exists, and it
works just fine in theory, so every other practice that works fine
in the real world must immediately be put to an end. In reality,
all-CSS layouts are very difficult to make work cross-browser, while
table-based layouts work in everything and in nearly every screen
reader save for OutSpoken for Macintosh.

There is no such thing as a working Web designer with a wide
clientele and a range of topics and audiences on their respective
sites who uses *nothing* but CSS-only layouts. I expect that will
still be true five long years from now. And even then, a billion old
HTML pages laid out in tables will still be online and must still be
read by then-new equipment.

> Secondly, with regard to HTML validation - the site reports validation errors
> (using W3C html validator) with regard to attributes that are no longer
> supported in HTML 4 (ie align, bkcolour, border and hspace). These are
required

Your only choice is to use a browser fork and serve different HTML
to different browsers, only some of which will validate. Unless you
can get the NN4-safe layouts to work for your application.

This, I believe, is the substantive answer you were actually looking
for rather than multi-screen digressions on how to train your
students.

You seem to be doing things quite the right way even if a few
corners must be cut here and there for real-world usability.

-- 

    Joe Clark              joeclark@joeclark.org
    Accessibility:        <http://joeclark.org/access/>
    Weblogs and articles: <http://joeclark.org/weblogs/>
    <http://joeclark.org/writing/> | <http://fawny.org>

Received on Sunday, 28 April 2002 12:29:32 UTC