Re: Exploding the myth of automated accessibility checking

On Mon, 08 Aug 2005 22:57:08 +0200, Patrick H. Lauke  
<redux@splintered.co.uk> wrote:

> Wendy Chisholm wrote:
>>  The 30 June 2005 Working Draft of WCAG 2.0 says, "The Working Group  
>> believes that all success criteria should be testable. Tests can be  
>> done by computer programs or by people who understand this document.

>  I'd say the core problem here lies with the "or" in that last sentence.  
> It implies that *all* tests can be done by *either* a computer *or* a  
> person. You would probably want something more along the lines of:  
> "Carrying out the entire series of tests will require people who  
> understand this document. However, some (or even "a small subset of")  
> tests can be automated / performed by computer programs."

It is a noble goal to have as much as possible testable automatically (so  
long as this is not at the expense of actually ensuring that real people  
get content they can really use). But I believe that the working group has  
clearly acknowledged that not all tests will be amenable to automation -  
at least at their current level of understanding.

If that is indeed what the working group is aiming towards, then I suspect  
the analysis is more devastating to the claims that automated testing  
along is very useful - and in aprticular the practice of running an  
automated test over a large body of web content and then making some kind  
of claim about the results that treats the results as reliable without  
checking.

I have seen many proposals for academic research papers that do this, and  
for reasons that I hope are obivous in this audience I tend to reject them  
very fast.

And yes, clearing up the actual wording used by the WCAG group would be a  
good thing. Especially if their collective output can be so easily  
misinterpreted as to think that they are only interested in machine  
testability by someone who participates in the work of the group. (If I am  
wrong, as has happened, and their real aim is machine testability as a  
primary goal, then the problem is a little deeper... but not insoluble)

cheers

Chaals

-- 
Charles McCathieNevile                              chaals@opera.com
          hablo español - je parle français - jeg lærer norsk
   Here's one we prepared earlier:   http://www.opera.com/download

Received on Monday, 8 August 2005 23:44:21 UTC