a possibly contentious suggestion

It seems to me that although the guidelines refer to several other
documents which treat particular subjects more fully (for example the Web
Content Accessibility Guidelines, Notes on SMIL and CSS, and various
documents dealing with the accesibility of software applications) we
should summarise the most important points of those documents, as well as
referring to them as 'authorative sources'.

This opens a Pandora's box of opportunities, but one into which I feel we
should look. As points for discussion, rather than a specific proposal, I
think we should make mention of the following principles:

In section 2: 

Content produced must be device independent, or should have accompanying
information (eg alternative and/or descriptive text) which can be used to
ensure the information or function provided by that content is accessible.
For example, where a Java applet is used to provide a navigation system,
an alternative system which relies on ordinary HTML must also be offered.
This could be done in the content of the OBJECT or APPLET element.

Content produced must seperate, or allow the seperation of, presentation
from structure. For example, where headings are indicated with a certain
colour and font-style, it must be done by applying style to the
appropriate level of Heading element (H1, H2, etc in HTML) rather than
styling 'ordinary' text (content of a P in HTML)

Content produced should have simple and consistent orientation and
navigation. For example, navigation bars should use icons and ALT text
consistently (see appropriate guidelines/checkpoints). Features which
disorient users (such as the unexpected opening of new windows) should be
avoided.

In Section 4:

Access to functions provided by the tool should be device independent. In
many instances this can be achieved by providing keyboard access to every
function, according to operating system conventions, since in nearly all
operating systems any device can control the keyboard commands. Although
this may in principle be true of a mouse, the use of a mouse is
exceptionally difficult for many users, as it requires much more complex
feedback than keboard-based functions.

The representation of content must be configurable. For example, font
sizes and colours used in editing the content must be configurable
independently of the intended representation of them by a user agent.

Representation of the content being produced must be accessible through an
interoperable accessiblity interface. Complex elements (such as tables),
or those which can only be rendered in certain media (such as graphics)
may not be otherwise accessible to users who are relying on a
representation other than a common-sized colour monitor. For example,
the position within a table, and the relationship of the 'current' cell to
other cells may not be readily apparent via a speech interface, and
requires a further exposure of the document structure. As dynamic content
becomes more wdiespread, this need is likely to increase significantly.

I think that this effectively covers the issues, although the reader must,
of course, be referred to the various full treatments which are included
as references in the document. I feel that something along these lines
justifies the volume it adds to the document. What do others think?

Charles

--Charles McCathieNevile            mailto:charles@w3.org
phone: +1 617 258 0992   http://purl.oclc.org/net/charles
W3C Web Accessibility Initiative    http://www.w3.org/WAI
MIT/LCS  -  545 Technology sq., Cambridge MA, 02139,  USA

Received on Tuesday, 16 February 1999 15:36:45 UTC