- From: Jason White <jasonw@ariel.ucs.unimelb.edu.au>
- Date: Sun, 5 Aug 2001 17:49:51 +1000
- To: Charles McCathieNevile <charles@w3.org>
- Cc: Gregg Vanderheiden <gv@trace.wisc.edu>, "GLWAI Guidelines WG (GL - WAI Guidelines WG)" <w3c-wai-gl@w3.org>
I would argue that, while the success criteria for a checkpoint must, taken as a whole, be both necessary and sufficient to satisfy it, individual criteria can be offered as alternatives (so long as this is clearly indicated). I would argue that effort should be devoted to refining, expanding and improving success criteria to ensure that they genuinely encapsulate the necessary and sufficient conditions, instead of hastily dropping them when an initial analysis indicates that they are inadequate. That is, we should endeavour to improve the criteria and only drop them in those cases where it is clear that only examples can meaningfully be given and that "success criteria", stricto sensu, are unattainable. I am sure this is the aim of Gregg's investigation of this point. There are various forms which success criteria can take, for example: You will have satisfied this checkpoint if you have done a, or b, or c, or ... You will have successfully satisfied the checkpoint if you have done either a and b, or c. and so forth.
Received on Sunday, 5 August 2001 03:50:05 UTC