- From: Jason White <jasonw@ariel.its.unimelb.edu.au>
- Date: Mon, 25 Apr 2005 18:26:57 +1000
- To: boland@nist.gov
- Cc: w3c-wai-gl@w3.org
boland@nist.gov writes: > > To me, a paraphrasing of the intent following "ensure that information, > structure, and behavior are preserved when presentation of such information, > structure, and behavior is adapted to the needs of users with disabilities" The above formulation is inaccurate and misleading. In order to adapt content to meet the needs of people with disabilities, the structure, information and behaviour have to be represented in such a way that the user agent can perform the adaptation. The purpose in separating information and structure from presentation is to enable the adaptation process to be possible (without requiring the intervention of a human being who can derive structure from presentation in ways that software can't). Thus the current wording in the guidelines is preferable to the above, as it clearly states what is required. In addition, the above-quoted formulation could be misinterpreted as requiring content authors to know what would or would not be preserved by adaptation processes, which in turn depends on the nature of the software available to the user. The whole point of the guidelines is to set a standard that content authors can work to, without having to know in this degree of detail what is happening at the user's end. > seems more specific than the wording using "separable" ("separable" seems to > me more ambiguous from an eventual techniques/testability point of view) Another way of expressing this might be to say that information, structure and behaviour are represented in a conventional and supported manner that enables them to be separated from presentation. > > Questions: > > (1) Is such a guideline that does not include qualifying wording (as mentioned > in previous messages excerpted following) able to be satisfied in a widespread > way? Yes. > > (2) Would use of such a guideline be "raising the bar too high" for > developers, or is it reasonable and practical to expect developers to meet > such a guideline (as opposed to "leaving the bar too low" referenced > following)? I don't think this is a problem. Rather, the issue with 1.3 is that it fails to state how high the bar is raised at all: exactly what degree of separation is required, and what structure has to be represented explicitly in the content? I think there is an implication in 1.3 that it is relative to the technology baseline. The kind of structure that can be represented in SVG is vastly greater than that which can be represented in PNG (discounting the possibility of using RDF in the latter). Likewise, the structures representable explicitly in XHTML 2.0 will be significantly richer than those recognized by earlier versions of (X)HTML. The solution for 1.3 is to make it relative to baseline. The problem, then, is the temptation to choose baselines that minimize one's responsibilities by selecting semantically impoverished formats even where richer alternatives are available and supported. This comes back however to offering clear guidance on the selection of baselines and I don't think guideline 1.3 is the right place to attempt to prevent it. Rather, for the sake of specificity in the guidelines I think we'll have to make 1.3 relative to baseline, which must surely have been the implicit assumption underpinning 1.3 all along.
Received on Monday, 25 April 2005 08:27:43 UTC