Re: WAI-ARIA: Text Equivalent Computation

I was asked to clarify some points from this comment:

   http://lists.w3.org/Archives/Public/public-pfwg-comments/2009AprJun/0005.html


The reason the usage of DOM attributes is inconsistent with HTML5 is that  
there DOM attributes refer to attributes on the interface object. What you  
call DOM attributes here is what HTML5 calls content attributes. So e.g.  
http-equiv is a content attribute where as httpEquiv is a DOM attribute.  
WAI-ARIA does not introduce any DOM attributes.


The third bullet point of 2A is not clear because markup languages do not  
define which attributes and elements need to considered here. I.e. HTML5  
does not state that the value of the alt attribute on img elements is  
non-tooltip native markup. (The alt attribute on the img element is only  
given as example in WAI-ARIA.)

Furthermore, from this requirement it is not clear what the text  
equivalent of e.g.

   <input id=x>
   <label for=x> A <b> B </b> <script> alert("C") </script> </label>

is.


The overall point is that this algorithm is much too complicated for such  
a simple task. I'd suggest to start out very simple and maybe increase the  
scope of the specification later once there is more implementation  
experience.

Also, it seems the requirements here contradict the implementation guide.  
Why are there two separate specifications with user agent requirements? It  
would be more helpful if they were in the same document so that such  
discrepancies can be caught more easily.


-- 
Anne van Kesteren
http://annevankesteren.nl/

Received on Tuesday, 21 April 2009 18:33:57 UTC