W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > July to September 2004

Re: First draft: proposal for guideline 4.2

From: Jason White <jasonw@ariel.ucs.unimelb.edu.au>
Date: Sat, 11 Sep 2004 15:25:54 +1000
Message-ID: <16706.35938.819687.98854@jdc.local>
To: Web Content Guidelines <w3c-wai-gl@w3.org>

Joe Clark writes:
 > This wording requires conformance to a set of guidelines, UAAG, that is 
 > almost entirely ignored (as ATAG also is). The Web author has no power 
 > whatsoever to make a user agent conform to UAAG. Make it a requirement and 
 > nobody at all, nobody, will be able to comply with WCAG Level 1 for an 
 > indeterminate number of years, probably five or more.

Some might say instead that such a requirement on the part of WCAG 2.0
would create a demand for the completion of UAAG 1.0 implementations
and thereby stimulate UAAG conformance. I leave that observation aside
in what follows.

In order for a content developer to be entitled to rely on a
technology, there must exist implementations that make the
resulting content accessible to a broad variety of users. Requiring
the content to be written using technologies supported by user agents
which conform to UAAG 1.0 is a way of ensuring that this is true, and
hence that the accessibility path from content author to end-user can
actually work.

It would be possible to require only conformance to a subset of UAAG
1.0 for this purpose, but in doing so one would (a) undermine the
conformance scheme of UAAG 1.0 by requiring something less than the
minimum; and (b) give the needs of some users priority over others in
selecting the subset of UAAG requirements that would have to be met by
a user agent as a prerequisite to allowing content authors to rely on
the supported technologies.

Another option would be to say that there must be a set of user agents
that collectively meet UAAG at Level A; but here we run the risk that
no single user agent may satisfy the needs of a given class of users,
which needs would in fact be met by a user agent conforming to Level A
of UAAG 1.0. Note that there are sets of user agents that collectively
conform to UAAG; UAAG 1.0 did, after all, meet its Candidate
Recommendation exit criteria.

At the more general level I do not think it is at all reasonable to
expect content authors to compensate for non-implementation, or
inadequate implementation, of UAAG. I also maintain that there is far
too much emphasis placed on making demands of the content developer,
instead of encouraging better implementation of both ATAG and UAAG.
The three sets of guidelines form a package which attempts to divide
responsibilities appropriately. They all need to be implemented for a
higher degree of accessibility to be attained, and to ensure that no
party is burdened with a disproportionate share of the responsibility,
or indeed of the work involved.
Received on Saturday, 11 September 2004 05:26:45 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:50 UTC