W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 2000

Proposed checkpoints for WCAG

From: Jason White <jasonw@ariel.ucs.unimelb.edu.au>
Date: Thu, 27 Apr 2000 10:34:46 +1000 (EST)
To: Web Content Accessibility Guidelines <w3c-wai-gl@w3.org>
Message-ID: <Pine.SOL.4.10.10004270952380.7753-100000@ariel.ucs.unimelb.edu.au>
Please note: the following comments are written in my personal capacity
and do not represent consensus of the Chairs, editors, or the working
group as a whole.

Following our recent discussion of the importance of generalising the WCAG
to accommodate emerging standards and technologies, I would like to put
forward several proposals which may help to address some of the issues
that have been raised. Each suggested checkpoint is followed by
explanatory examples and clarification. The exact text of these proposals
could doubtless be improved; they are presented here only in incipient

Under guideline 3:

A. Provide content in a data format which best represents its meaning and
structure, independently of presentation.

Illustrations: SVG would thus be preferable to bitmap image formats, in
that it permits logical grouping of the components of an image and allows
them to be individually labelled and described. Similarly, a structured
markup language would be preferable to the transmission of formatting
objects directly to a user agent, as the former explicitly preserves the
logical structure and composition of a document whereas the latter do not.

Important qualification: this proposed checkpoint does not preclude the
provision of content in a different, presentationally-based format, where
this is requested in virtue of a user's preferences (E.G. via
client/server negotiation or as a result of a deliberate choice on the
user's part).

B. Use markup languages appropriately, according to specification. So far
as the available markup languages allow, ensure that the meaning and
structure of the content are explicitly captured in the markup.

This checkpoint requires little explanation as its rationale and purpose
have been discussed in detail throughout the guideline development
process, and in particular during recent meetings.

C. Provide headings, titles, labels etc., for all significant structures
within a document or user interface.

This would take the place of separate requirements in the 1.0 guidelines
related to table headers, proper use of headings in documents, labels in
form controls, titles, etc. It would also encompass the labelling of SVG
groups (portions of images), etc.

D. Ensure that labels, titles, headers etc., are associated explicitly
(E.G. in markup) with the content to which they apply, and that the
structural boundaries of the latter are properly delimited.

This somewhat convolutedcheckpoint, I admit, needs refinement. However its
basic intent is, I hope, apparent. The second part of the requirement is
intended to require not only that labels, headings etc., be marked up
properly, but that structural markup be used to indicate the portion of
document content to which the heading, label etc., is relevant. Sometimes,
as when the label or title is given in an attribute, additional support
for the latter requirement is not needed, as an attribute value naturally
applies to the element to which it is attached. However, headings are a
different matter. In HTML, for example, a DIV element can be used to
enclose that part of a document to which the heading appertains:
<h1> Heading </h1>
<div> <h2> Second-level Heading </h2>
[A form, heading or some other content, still within the scope of the H1]

E. Use style sheets to control layout and presentation (checkpoint 3.3).
Where feasible, provide style sheets for a variety of media types (E.G.
small and large screens, print, audio, etc.). This would be of a higher
priority where most user agents do not, by default, support the markup
language in which the document is written. It may also address some
cognitive issues.

There should be a general requirement (and I am not sure how best to
express it) that rich visual, auditory and tactile presentations be
readily derivable from the content, and that they accurately reflect the
structural and semantic distinctions captured by the markup. To some
extent, this will require the assitance of style sheets (and this concern
is a partial motivation for the checkpoint proposed above regarding style
sheets); but the basic intention here is that an effective and consistent
presentation be available in whatever happens to be the user's preferred
medium (or combination of media).
Received on Wednesday, 26 April 2000 20:35:27 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:59:14 UTC