RE: 6.1 - priority

I agree that there is a chicken and egg problem between authoring and
supporting features in user agents. My take is that we set priorities based
on the impact on people with disabilities, and those priorities are applied
to requirements for authoring (Web Content and Authoring Tool Guidelines) and
User Agents. The strength of this approach is that it provides encouragement
to conform, and do the right thing, without waiting for the other side to
make the first move. The cost of this approach is that our current priority
system says that any feature which is beneficial, but is not a significant
hindrance or total barrier to access is only a Priority 3 requirement.

It seems to me that this group, as the Authoring Tools Group did, should not
try to decide the priority of content features, since that is the
responsibility of the Web Content Guidelines.

Personally, I think that anything that does not meet P1 requirements (as they
are for people, which I hope and expect we will capture in the P1
checkpoints of these guidelines) is only useful in the absence of another
choice. Double-A conformance is my rough yardstick for "this tool is worth
recommending". That is of course in a general case - in some cases I would
suggest trying to match more closely the needs of a given user, and may
suggest a tool that meets those needs best even if it is not, overall,
conformant.

I hope and seriously expect to see triple-A tools in a couple of
years. (Hacking away at Amaya actually reinforces that faith).

Charles McCN

On Mon, 1 Nov 1999, Denis Anson wrote:

  Charles,
  
  This brings up an interesting philosophical philosophical point.  There is
  not incentive for a web author to insert accessibility features that are not
  supported by browsers.  There is no incentive for browsers to support
  features that are never used in pages.  If support for content has the same
  lower priority as the implementation of those features, why would a UA
  writer ever support them?
  
  Ultimately, we are hoping that priority 2 and priority 3 features are
  supported.  This might be a serial raising of the bar for "conformance" over
  time.  Would that be incentive enough?
  
  Just wondering....
  
  Denis Anson

Received on Monday, 1 November 1999 12:00:36 UTC