W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > April to June 2001

Reason Revisited

From: William Loughborough <love26@gorge.net>
Date: Sun, 10 Jun 2001 07:26:51 -0700
Message-Id: <>
To: <w3c-wai-gl@w3.org>
Cc: "Wendy Chisholm" <wendy@w3.org>
At 08:44 AM 6/10/01 -0400, Katie Haritos-Shea wrote:
>minutes updated

WL: How about [adding] the words "within reason" to the current guideline.

I think the full drill is "if/as/where/when appropriate/necessary" as well 
as "within reason".

As we got more into moving from WCAG 1.0 to 2.0 I changed viewpoints (I 
didn't participate materially in the details of 1.0) from thinking of this 
as a major overhaul to realizing just how effective the original document 
had been demonstrated to be. It has formed the basis for 
policies/laws/regulations and actually spawned an entire industry of 
services designed to enhance Web accessibility (and usability for that matter).

It has been much maligned for opacity and for "leaving out" major segments 
of PWD yet the changes proposed/made have been for the most part editorial. 
Generalization/abstraction have been accompanied by slight changes in 
emphasis but always an almost total mapping of the previous checkpoints 
into the new document.

What we find ourselves wrangling over are almost always the devilish 
details: the underlying principles are still rather solid.

The thread that is most recent in this process involves user control of 
various timed events and the idea of what would be "reasonable" to expect 
from the provider/content/user matrix. Put simply, we don't want to make it 
impossible for someone who cannot "just click here" within an allotted few 
seconds while at the same time we don't want to hogtie a huge enterprise 
from making server activity efficient by requiring that they allow 
unlimited unattended shopping carts ("deposits made after 5 PM will be 
credited on the next business day", etc.).

We pretty much understand this problem and even have rough ideas about how 
long certain instances should take, etc. On one hand we don't want to put 
"seconds/hours" into a checkpoint yet we must protect someone who needs a 
bit more time to send Morse code via a sip/puff straw to an interactive form.

I proposed that this and a great many (all?) checkpoints be considered to 
have a tacit "within reason" phrase for just about every term therein. E.g. 
"Provide a text equivalent for all non-text content" might have as 
a  "reasonable exception" making certain exceptions as is already done with 
alt="", "Synchronize text equivalents with multimedia presentations" might 
encounter discovery that there are instances where synchronization is 
meaningless/unimportant, "Synchronize a description of the essential visual 
information in multimedia presentations" clearly requires judgement as to 
what is "essential", "Use markup or a data model to provide the logical 
structure of content" makes no allowance for the possibility of there being 
such a thing as "unstructured content", and that's just the first few 

We will never find *all* the examples that require intervention of reason 
on the part of content providers and in order to maintain a "liftable" 
document we must restrict the potentially endless array of 
caveats/exceptions/examples we furnish.

As in the cited example, we can collect many of the ruminations into the 
techniques/examples section so that the harassed designer has practical 
guidance to accompany the ideals of the checkpoints. We must give a 
"flavor" that delineates the approach  needed to make content accessible, 
particularly for those who come clueless to our table. They want their 
stuff to be widely available in the "increased market share" sense as well 
as being required to conform to *something* - and as many detailed answers 
as possible should accompany (but not bloat) the document.

Gregg proposed that I try to locate "bright lines" that might exemplify our 
intent without burdening the checkpoints and I will keep trying, but I 
think we must all take one step back, a deep breath, and not try to include 
examples in the guidelines/checkpoints themselves but in their explanatory 
accompaniments, including the "lead-in" paragraphs and "post-checkpoint" 
summaries. To mix metaphors: our "Ten Commandments" will have an extensive 
"Federalist Papers" backup.

In summary, the introduction is so much friendlier, the organization so 
much "cleaner", and (at least the plan of) the techniques so much more 
practical/specific that we can expect much better use to be made of 2.0 if 
we keep on keeping on.

Received on Sunday, 10 June 2001 10:27:32 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:37 UTC