W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > July to September 2002

Re: 5.4: proposal re level 2 success criteria

From: Jason White <jasonw@ariel.ucs.unimelb.edu.au>
Date: Thu, 5 Sep 2002 11:54:46 +1000
Message-ID: <15734.47462.86706.674175@jdc.local>
To: Web Content Guidelines <w3c-wai-gl@w3.org>

john_slatin writes:
 > Proposed wording:
 > 	1.	The interface has been tested using a variety of assistive
 > technologies, and an accurate summary of the test results is provided.
 > 	2.	The interface has been tested by a number of people with
 > different disabilities using assistive technologies familiar to them, and a
 > summary of the results is provided.

In an earlier discussion of this checkpoint, it was argued, notably by
Gregg, that testing with actual users should not belong as a requirement
in the success criteria. Essentially, it was suggested that the
technical requirement could be satisfied without conducting any
testing, and, conversely, that despite carrying out tests with actual
users, there could still exist incompatibilities (with assistive
technologies that had not been tested, or at least not tested with
sufficient thoroughness). Thus it was concluded that the question of
whether the implementation meets the checkpoint or not is separate
from the issue of what type of testing has been carried out, in that
there is no necessary connection between the two.

Also, I think this checkpoint, more than most others, should be
amenable to automated verification (if documented API's have been
followed correctly, and these are known to have been properly
implemented by assistive technologies, then compatibility is assured).
Also, I think this checkpoint will be superseded by checkpoint 5.3.

If checkpoint 5.4 is retained, I suggest that the level 2 success
criteria be rewritten to correspond to the review requirements
elsewhere in the guidelines, viz.:
The content has been reviewed and is believed to be compatible with a
range of assistive technologies.

In Appendix B where review requirements are discussed, we already
recommend testing with actual users. This change would thus bring
checkpoint 5.4 into line with the remainder of the guidelines. Recall
that it was written prior to the introduction of "review" requirements
elsewhere in the document, and is now an anomaly in that its review
requirements are not consistent with the pattern followed throughout
the remainder of the guidelines.
Received on Wednesday, 4 September 2002 21:54:54 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:42 UTC