Re: Accessibility of sites driven by databases.

Perhaps Ian's requirement could be reformulated as follows:

In order for a web site to be considered accessible, all of its content
must be available in a format that complies with the guidelines. This does
not preclude the availability of other formats which may be selected by
the user, either through explicit action or by content negotiation in
pursuance of user-specified preferences. Stated differently, provided that
the structure and semantics of the content are available in appropriate
markup, equivalents are provided as mandated by the guidelines,
device independence is ensured, etc., in one version of the content, this
will suffice to satisfy the conformance requirement.

One difficulty attendant upon this approach, however, is to define exactly
what is meant by "the same" content, for purposes of ensuring compliance.
For instance, if the author of a complex sociological monograph were to
provide two versions: (1) a highly mathematical version, containing
detailed treatment of the relevant statistical parameters and the
findings of the study, supplied in a highly presentational file format
that can not be rendered effectively in non-visual media; and (2) a
simplified exposition of the findings of the study, intended for a general
(non-specialist) audience, and in a standard, highly accessible markup
language, we would not consider them to be equivalent in the relevant
sense, and hence the site as a whole should be regarded as non-compliant.
Stated differently, the differences in markup technique (in the file
format used), between the two versions, would characterise one as being
more accessible than the other, but would not, of itself, lead to the
conclusion that they were "different content" for purposes of conformance.
However, the simplification of one version, by contrast with the
mathematical detail provided in the other, would justify the view that
the two versions were not, relevantly, the same, and hence that the site
was inaccessible, as there was no accessible version available of the
mathematically sophisticated edition of the study. Suppose, on the other
hand, that the detailed mathematics were included in both editions, but
that the version marked up in a presentational file format made extensive
use of graphs and charts that were not reproduced in what the author might
regard as the "accessible" edition. Suppose further that, in the latter
version, the graphs were replaced by tables purporting to convey the same
information. Again, the question arises of whether the two versions have
"equivalent" content for conformance purposes.

I am concerned that in allowing site-wide compliance on the basis that Ian
has suggested, a move which has considerable merit, we need to be careful
in defining what is, and what is not, "the same" content, or amounts to
two versions of "the same" content, for purposes of judging conformance.

Disclaimer: the above are personal comments and have not been made in my
capacity as co-chair.

Received on Thursday, 4 May 2000 19:10:25 UTC