Re: Deconstructing WCAG: FWAP 0.1 Straw Man

Kynn,

Thank you for investing the time and effort needed to write up a
proposal. This provides the material necessary to allow the working
group to consider the merits of the "policy framework" analysis.

There was much discussion of the relationship between WCAG and policy
at last week's face to face meeting. It was also made quite clear at
the interest group meeting, by developers themselves, that they need
clear guidance regarding how to formulate an implementation plan and
how to determine the order and priority with which to address various
checkpoints. It was also argued that organizations should be provided
with some flexibility in defining policies, but that entities with
fewer resources would benefit significantly from a limited range of
pre-defined policy options' being laid down in advance.

There was also disagreement over whether the working group should, or
should not, prioritize the checkpoints of the 2.0 guidelines.
Opposition to the establishment of priorities arose from two distinct
but complementary arguments. First, as noted above, it was asserted
that policy makers, whether at a governmental or organisational level,
would inevitably determine priorities based on a variety of pragmatic
considerations, which depend on context and circumstances that the
WCAG working group is in no position to foresee or take into account.
At an organisational level, for example, implementation priorities
will be determined by a variety of factors, including the authoring
tools and implementation technologies in use, whether the content is
being created anew or "retrofitted", etc. Given the perceived
inevitability of policy setting, it was argued that the guidelines
should support these efforts by providing some flexibility, but
without allowing unconstrained subsetting or modification of the
requirements.

It was correspondingly argued from a user's point of view that any
prioritization within the guidelines would necessarily be
discriminatory, since almost every checkpoint represents an absolute
barrier to accessibility for some identifiable group, and by
establishing priorities, the working group would not only be
legitimising, but also engaging in the practice of determining whose
access needs should be treated as of foremost importance, and whose,
by contrast, carry less weight. It was recognised that policy setters
will inevitably make such decisions, but these, it was suggested,
should be required to be transparent as to whose needs will and will
not be satisfied by any particular policy formulation. Metadata was
favoured as the preferred mechanism through which to match users' needs
with the policy implemented by a particular site. An analysis of the
interdependencies among the checkpoints of the 2.0 draft indicated
that most could be implemented independently of one another, but that
the benefits of some checkpoints could not be realised unless certain
other checkpoints were regarded as prerequisites. It was acknowledged
that any implementation and conformance framework provided by WCAG
must take cognisance of these interdependencies, as well as the more
specific dependencies arising at a technology-specific level.

My impression of the meeting is that proponents of priorities were
willing to consider alternative proposals that did not involve the
prioritisation of checkpoints as in WCAG 1.0. The underlying concern
was that content developers be provided with firm policy guidance, and
there were suggestions that this could be achieved by offering a
well-defined framework for decision making, including one or more
sample policies.

Some participants were also concerned to ensure that the
prioritization and policy-setting process carried out by organisations
would not be completely open-ended, so that content which was tailored
for access by a specific group defined according to disability,
could not be regarded as conforming to the guidelines.

The upshot of these deliberations appears to be that the guidelines
should impose constraints upon, and offer guidance in, the formulation
of policies, whether by governments, organisations or individual developers.

It was agreed that there should be clear and explicit "success
criteria" with which to determine whether web content conforms to each
checkpoint. There was no consensus as to whether the guidelines should
have an over-arching conformance scheme requiring anything more than
that developers be able to assert (via metadata or otherwise) which
checkpoints they claim to have implemented.

A possible solution not discussed at the meeting would be to regard
web content as conforming to the guidelines if it satisfies those
requirements defined in a policy that accords with the framework
provided by WCAG. This would be broadly consistent with Kynn's
proposal and generally reflective of the ideas arising from the face
to face meeting. It should be pointed out that I am not necessarily
advocating this idea, but merely raising it as a possibility which
accords with the arguments reported above.

Full details of last week's deliberations will appear in the minutes
of the meeting, which will be published as soon as Wendy, Charles and
Cynthia have compiled their notes and made them available on the web
site.

Received on Sunday, 18 November 2001 04:05:32 UTC