RE: Level 3

Actually,

I wasn't talking about testability but applicability.  And effort. 

We have said that all SC have to be testable. 

 
Gregg

 -- ------------------------------ 
Gregg C Vanderheiden Ph.D. 
Professor - Ind. Engr. & BioMed Engr.
Director - Trace R & D Center 
University of Wisconsin-Madison 


-----Original Message-----
From: Matt May [mailto:mcmay@w3.org] 
Sent: Monday, June 06, 2005 11:04 AM
To: Gregg Vanderheiden
Cc: 'WAI-GL'
Subject: Level 3

Gregg Vanderheiden wrote:

>And at level 3 I guess I don't have a problem. 
>

I've seen and heard this comment so many times now that I have to say 
something.

What is to be inferred from this statement in most contexts is that the 
rigors of higher-priority success criteria (e.g., testability) are less 
applicable to Level 3 items, and therefore less worthy of further 
specification or extended discussion.

And I agree. I've been saying that all along. This is why I have 
advocated two levels of conformance (single-A and double-A) while 
keeping three priority levels in the document.

Take a look at the requirements set out in Level 3. Who in the world 
would be able to conform to all of these for any single site? And yet, 
we know from experience that people will claim triple-A conformance 
against all evidence to the contrary, simply because the conformance 
level exists. WCAG 2 raises the bar even further, and yet, we will have 
the same false claims in WCAG 2 if we offer a triple-A conformance 
level. This is damaging not only to the triple-A mark, but to all WCAG 
marks. In other words, it's bad for Web accessibility to set norms no 
one (or nearly no one) can achieve.

It appears that Level 3 is a place for success criteria that are 
ambiguously-worded, or are good ideas that lack adequate definition, or 
are not broadly applicable, or are an awful lot of work. If this is the 
case, we are more likely to gain adoption of these criteria in 
mainstream sites by presenting them as advanced accessibility needs, and 
suggesting them for those who can find a way to implement them, than by 
adding them to a conformance set that is burdensome, confusing and 
sometimes impossible to achieve.

Further, those SC that aren't well-defined and achievable are bound to 
hold us back when we try to advance the document. We can ignore these 
issues now, but eventually they will be called up, and we will have to 
have an answer when someone asks us why they deserve to be in the 
document. It's better to answer that now than be recycled back to 
Working Draft status later.

I renew my objection to the triple-A conformance level on these grounds. 
I would accept A+ or AA+ marks for those who conform to a given level, 
plus certain Level 3 success criteria, or some means of itemizing what 
SC a site conforms to. But I'd like us to stop dancing around what these 
SC really are.

-
m

Received on Monday, 6 June 2005 17:34:28 UTC