RE: Conformance levels and core /extended

Lots here.  Let me see if I can shed some light.

 

Core and extended - is based on the premise that there will be difficulty
passing laws that affect people's freedom to express themselves.   This is
another civil rights issue held by people with and without disabilities.

 

Within each checkpoint there is Required and Best Practice and also
Additional Information.    This is the equivalent of priority.   The most
important or highest priority things for each checkpoint are in the
required.     

 

If you want a   A, Double A and Triple A format you can use

 

A = Core

AA  =  Core plus Extended

AAA = Core plus Extended plus Best Practice (for core and extended
checkpoints)

 

No one ever gets AAA with WCAG 1.0  - and those that claim it are always
attacked by others saying that they didn't really do one thing or another.
In 2.0  it looks like we are heading toward having AAA (Best Practice) not
be normative.  But that could change.

 

Please note that there is one big difference with 1.0.   In 1.0 we decided
that some checkpoint were more important than others.  

In 2.0 we recognized that what was "helpful" to some was "critical" to
others.   Thus do specifically decided NOT to rate the checkpoints as P1, 2,
3.    All checkpoints have same "priority".  

They are divided into those that can be done without restricting expression
(CORE) and those that involve changing the expression to make it more
accessible without using AT (Extended).   The CORE require the person use AT
or some accessibility feature in their User Agent, in order to access the
content.     

WITHIN each checkpoint we then have different priority assigned to different
measures or techniques for addressing the checkpoint.  That is where our
priority rating lies.   Required success criteria are the highest priority.
Best practice is for those who want to go beyond the minimum for any
checkpoint.  There are also often a number of additional items of
information or advisories for a checkpoint.  

 

 Hope this is helpful in undestanding where we are.   I don't know where we
will be exactly when we end, but we didn't get here by accident.  There were
carefully thought out decisions made about the underlying rationale for
these.   This note is an attempt to quickly lay out what some of those were.


 

Those with questions - please come to the working group teleconferences.  It
is hard for me to type out all the rationale in a short memo - and respond
to questions.   Also, you need to address the working group as a whole to
get the full story.     Please come.


Gregg

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

-----Original Message-----
From: w3c-wai-gl-request@w3.org [mailto:w3c-wai-gl-request@w3.org] On Behalf
Of Sailesh Panchang
Sent: Monday, September 22, 2003 8:48 AM
To: Web Content Accessibility Guidelines
Subject: Re: Conformance levels and core /extended

 

1. Well then, the  proposed conformance system (claim to core / extended /
core+ etc) is no measure of accessibility at all. 
I believe that a system that measures conformance to accessibility
guidelines  must: 
- consider  impact on usability  for user population  being served by the
accessibility guidelines by  

- assessing compliance against checkpoints    (and accessibility guidelines)
applicable to a page / site / application.
   
I fully  agree with John's statement that the distinction between core and
extended  is  developer-centered" as   "Core do not make you change the
document  visually - just mark it up or add hidden text (or audio)
"equivalents"". (I recognize that some content that is meant to give a
"sensory experience" (recent list discussions) cannot really be made
"accessible".). If some "extended checkpoints are also important", then a
website that claims only "core"  might still have  many serious
accessibility-related flaws. How does it matter to me as a website user, if
a website claims "core" level? What does it tell me about the  accessibility
level of the site? For instance say  of the important-to-accessibility
checkpoints applicable to a site, only 25 percent   belong to the core group
while 75 percent belong to the extended group. 

 

2. So the core checkpoints might probably be relatively easy and quick in
some instances to make content accessible. 
  How is  the process used to make Web content accessible relevant to a
conformance to an accessibility claim? 
How does it matter if the developer needs to spend three minutes or three
hours to make some content accessible  as per the organizations adopted
accessibility policy?
If the developers realize that the technology or design adopted does not
support the  required accessibility level and need to  make drastic  changes
in order to gain compliance, then they may be required to do so.
An analogy: how is the time/process/cost  needed to build a bridge
relevant to a statement on the "safety level" of the bridge?

 

We should not drift away from the objective of measuring accessibility to
user population served by the guidelines. 
3. And this measure must  definitely  consider importance of  the
checkpoints  with respect to accessibility. For instance absence of  a text
descriptor for an  image   that is important to understanding content is a
serious violation. But if  every link begins with the words "Link to"
instead of the key/action word in  the link, it may make navigation
inefficient(but not inaccessible) for the   screen reader user trying to
navigate using a list of links pulled up by  the access technology used.
This will be a less serious accessibility  violationn.
  
So I am afraid I am again proposing the  conformance system used for WCAG
1.0. 

 

4. By the way, the philosophy underlying core / extended is radically
differennt but what are the serious weaknesses / limitations or
disadvantages of P1/P2/P3 system? Is it so vital to adopt a new conformance
system to measure accessibility? Why not proceed with a workable system
already in use, perhaps with some tweeking? The P1/P2/P3 system has been
used by other W3C guidelines including the  one that measures the QA process
for creating W3C guidelines.

 


Please do not interpret this as resistance to change, but I  want a good
reason to subscribe to the _new accessibility_ conformance system being
advocated by WCAG 2.0.
Thanks for your time and  indulgence,
Sailesh Panchang, Senior Accessibility Engineer
Deque Systems, Reston, Virginia 20191
Tel 703-225-0380 (ext 105)
  On Thursday, September 18, 2003 10:09 am Gregg Vanderheiden  wrote:
  In 1.0 it was  priority 1,2,3 based on user need

 

  In 2.0   the dividing line between core and extended is not user need but
whether it involved constraining the content or presentation.
  In 2.0   the dividing line between core and extended is not user need but
whether it involved constraining the content or presentation. 
  Core do not make you change the document visually - just mark it up or add
hidden text (or audio) equivalents.

 

  There are many extended checkpoints that are VERY important.

 

  It is true that all the core checkpoints are important.  But it is not
true that all the important checkpoints are core. 

 

  On  Thursday, September 18, 2003 11:53 AM  John M Slatin  wrote

 

  "I still have some fundamental concerns  about the idea that CORE
checkpoints are (or should be) those that don't require content developers
and designers to change the  efault visual presentation.
  ...
  I'm concerned that the current criteria for distinguishing "core" from
"extended" is *developer*-centered.  
  ...
  I would argue, then, that extended checkpoints that we agree are truly
important to meeting the needs of users with disabilities should go into the
"core.  

 

  John"
On Tuesday, September 16, 2003 12:07 AM
Jason White
wrote

 

Subject: Re: Conformance levels and best practices

 

"Others have argued, however, that precisely because the conceptions of
conformance levels have changed so radically between WCAG 1.0 and WCAG
2.0, the same terms (Priority 1, Priority 2 and Priority 3) should not be
carried forward into the 2.0 document, as this would create confusion for
those who are already acquainted with the 1.0 conformance structure and its
definitions.

 

In developing WCAG 2.0, the distinctions underpinning the 1.0 priority
definitions were subjected to detailed, critical scrutiny, with the result
that the working group decided they were no longer tenable for purposes of
the 2.0 document, and that a new set of distinctions was needed to classify
checkpoints and success criteria in constructing the 2.0
conformance scheme. The argument for new terminology, then, is that unless
different terms were used, readers acquainted with 1.0 would construe 2.0
as following the same, or a similar, system of priorities and conformance
levels as WCAG 1.0, and would make policy decisions and claims about
accessibility on that erroneous, but tempting, presupposition. The new
terminology is supposed to signal the fact that the 2.0 categories of core
and extended are very different from the 1.0 priorities, and that the
impossible/difficult/easier approach worked out in the 1.0 conformance
scheme does not apply to 2.0."

 

My purpose in stating these arguments is not to defend them, though I do
happen to find them persuasive, but to attempt to recapitulate the reasons
which led the working group to the conclusion that new terminology was
needed and that retention of the old terminology would do more harm than
good.

 

Comments, counter-arguments and proposals are, of course, most welcome. If
anyone is dissatisfied with the working group's prior decisions on this
matter, the issue should be re-opened in the context of our ongoing
conformance discussion.
"