I agree Liddy. 

I also feel that some users *may* trust resources that have been verified by
an independent third party, 'self-labelled' sites are not trusted by
everyone all of the time. Similar to what you said; trust is like beauty;
it's in the eye of the beholder - you should see my slide for this statement
;)

Search engines and browsers *may* want to make use of trusted metadata.
Segala is co-editor of the MWI mobileOK document with Google and ICRA -
there is a use case for metadata that has been verified by an independent
'certification provider'. Users will make up their own minds as to whether
they trust CAs over self-labelled sites. The idea is to support both and not
mandate either. 

Let's not beat the CAs who have built trust slowly within their communities
- I know this wasn't/isn't intended.

Paul

      -----Original Message-----
      From: DCMI Accessibility Group 
      [mailto:[log in to unmask]] On Behalf Of Liddy Nevile
      Sent: 27 May 2006 08:08
      To: [log in to unmask]
      Subject: Re: certification program
      
      Sadly, I think we will not 'get it right' no matter what 
      we do, but we should be very careful.
      
      1. The 'test-the-tester' approach has to be flawed in 
      exactly the same way as the simple 'test-the-resource' 
      approach is flawed - either way it is just as easy to cheat.
      
      2. We know from experience that people decide for 
      themselves who and what to trust. It is very important to 
      be able to say by whom, using what tests, and when 
      something was tested.
      
      3. It is also important to say exactly what was tested - 
      the components of the resource, the resource as a 
      composite object, this or that version or, if you like, 
      which instantiation of the resource, etc..
      
      4. We have enough research to show that testing does not 
      and cannot guarantee accessibility to everyone.
      
      and, what is compelling for me,
      
      5. any individual user only wants to know if the resource 
      will be accessible to them at the time of delivery.
      
      We have recommended the AccessForAll approach in favour 
      of generalised testing for certification because that 
      way, the description of the resource is available for 
      decision-making.
      
      This list, beyond all others, should have many on it who 
      understand why we use metadata, and they should be able 
      to contribute to how and why metadata is useful in this 
      context. The credibility of the metadata is important but 
      so is the content. Specific details are often needed for 
      accessibility decision-making. If there are too few 
      details, people who could use resources will miss out on 
      them and if there is too much demanded, nobody will 
      bother to supply it.  These are the sorts of problems 
      that have been at the heart of our work for the last few years.
      
      I'd like to think that we can do, as we have always tried 
      to do,  and combine the use of metadata to convey 
      information ABOUT the resource, basing the statements 
      that make up that metadata on the tests that W3C so 
      carefully develops. In my humble opinion, this does not 
      lead to a 'let's certify' approach so much as a 'let's 
      describe' approach.
      
      Surely we want to empower the user to make the important 
      decisions about what they can/will tolerate just as, when 
      we choose a hard-to- get-into restaurant, we let our 
      colleagues in wheel-chairs decide to join us or not: it 
      should be their decision, not ours.
      
      Liddy