RE: Conformance Claims and Logo

Gregg Vanderheiden writes:
 > A few questions.
 > 1.	Would this be accessible if the person couldn't handle RDF or RSS?

Probably not.
 > 2.	Are we requiring that users have RDF or RSS savvy access?  Even on
 > public systems?

This is similar to the issue underlying checkpoint 5.2, though it
isn't covered in that checkpoint, and I think it belongs in the
checkpoint 5.3/5.4 discussion. Specifically, the question is: what
flexibility should the developer have in requiring certain features of
the user agent in order for the content to be accessible (to a given
category of users). For example, for purposes of level 1 conformance
may I assert that my implementation of certain requirements of the
guidelines only works if the user agent supports particular
technologies or features thereof? RDF would be just one example of
this more general problem. The details of precisely what technologies
are available and implemented in user agents w(not to mention
proxy-based transformation services and so on) will of course vary
over time. For each conformance level, then, what types of assumptions
should the developer be able to make regarding the technologies used
"up-stream" (i.e., after the content leaves the originating server) in
order to make it accessible to the user?
 > 3.	Is RDF / RSS savvy user agents our baseline for the guidelines?
My proposed answer is in line with checkpoint 5.2: the baseline will
change over time and we need to decide, in writing checkpoints and
success criteria, what flexibility we want to offer at each
conformance level in determining the baseline. Ultimately this is the
entire problematique that Ian Jacobs was raising in his contribution
to this discussion. Currently I don't have a concrete answer to offer.
One solution would simply parallel checkpoint 5.2 with respect to the
technologies and features used in implementing the guidelines
themselves. At level 1 the required level of support would merely have
to be declared, as in checkpoint 5.2; at level 2 there would be a
stronger requirement to show actual implementation by client-side or
"up-stream" software and services, including user agent/assistive
technology combinations where pertinent. In the same or a separate
checkpoint one could then preclude the use of technologies in
non-standard ways that circumvent the accessibility support assumed in
one's user agent baseline. This would deal with cases of "custom" user
interface components, "custom" markup etc., described in my recent
message about checkpoints 5.3 and 5.4.

Received on Wednesday, 5 February 2003 03:20:34 UTC