- From: Jim Jewett <jimjjewett@gmail.com>
- Date: Tue, 15 Sep 2009 15:06:08 -0400
- To: HTML WG Public List <public-html@w3.org>
- Cc: Aryeh Gregor <Simetrical+w3c@gmail.com>
[This email is not Cced to any of the wai lists, but only because I suspect it would be either out of context or redundant.] Aryeh Gregor wrote: > I think a lot of the people who work on accessibility start out > with the attitude "Let's provide a mechanism that would be > helpful for accessibility if used correctly". Yes. That seems like a pretty low bar, but it is a necessary condition. In certain walled-garden situations, it may even be sufficient. The questions are: (a) How practical would it be to do even better? (b) If a "mostly better" solution came with trade-offs, how many such walled gardens would be hurt, and how badly? A fair amount of automated communication uses XML rather than HTML, because the extra effort of getting it right (for a program) isn't all that high, and the extra benefits (for automated processing) accrue quickly. My personal beliefs regarding summary are that: (a) It wouldn't be that hard to do better. This is largely because correct @summary usage has been so rare. (But also because the automated "even better" options are not as expensive or as unlikely as the advocates believe, so long as they can be clearly specified and implemented by the browser vendors instead of AT vendors or individual page authors.) (b) The number of sites (even walled gardens) using @summary correctly is fairly small, and probably limited to the same organizations that would be most willing to use an even better solution if one could be standardized at even the author level, let alone the browser level. I freely admit that I don't have any hard data to support those two beliefs. But I haven't seen any data to contradict them either, and I think accessibility advocates may have become too conservative in their aims. This leads to a reflexive defense of something that is less bad than nothing, rather than an exploration of what improvements are still plausible. -jJ
Received on Tuesday, 15 September 2009 19:07:07 UTC