html 5 and accessibility issue

Hi,

I will try to resume the accessibility open issue currently in html 5.

the basic fact that currently can have impact on accessibility ( from
http://dev.w3.org/cvsweb/~checkout~/html5/html4-differences/Overview.html) 
:

- |accesskey| attribute on |a|, |area|, |button|, |input|, |label|, 
|legend| and |textarea| is removed from spec.
- longdesc| attribute on |img|| |is removed from spec.|
- summary| attribute on |table |is removed from spec.|
- headers| and |abbr| attributes on |td| and |th| is removed from spec.
- |scope| attribute on |td| is removed from spec.

Actual argument for removing them :
- they have not been used often
- in some case new parsing mechanism can achieve the same result

Actual argument for keeping them :
- even if they have not been used often, they are used
- they are useful for users
- they are supported by assistive technology
- in some case, information can't be retrieve without them even with new 
parsing mechanism
- they already are supported by wcag 1.0, 2.0 spec and by a lot of 
country specific policy
- they currently are teaching in some educational program
- current HTML 5 design principle clearly say : Support Existing 
Content, Evolution Not Revolution, Universal Access

So we need to improve them not to removed them :
- work on a reel usable accesskey API, where users can reach the current 
accesskey of the page and modify them of even add his own accesskey to 
element in webpage like user style sheet
- Since longdesc can be useful for everybody their is a willing to have 
a way to make it available to every users. We need too a way to have 
access to longdesc content and to come back to the image in a visual or 
keyboard navigation mechanism and maybe have a fallback for this too 
like replace the longdesc by a reel link  when image is turned off or 
when the browser can't render it.
- Same problem for summary, maybe summary attribut can support #idvalue 
like the longdesc attribut, this way we can have a relationship between 
some visual content and and the summary
- Actually for the headers issue i don't see any other solution than to 
find a new table parsing algorithm or improving the existing one. 
Without it we still needs headers

Finally here is others issues i have found :
- dt/dd need a reel relationship like the for attribut on label mechanism
- there is huge need to have access to fallback content of  embedded 
content even if the element is well rendered (for example people can 
have the flash plugin but can have reel pain to use it or access it 
content with assistive technology, so they need a way to access to the 
fallback content)
- actually their is no fallback content for embed element
- the only fallback content for iframe elements is pure text, i think we 
need to authorize an "a" element to link to the content of the iframe
- Drag and drop API is great but what about people without mouse, is 
there a keyboard usable drag and drop API, this statement :
"On media without a pointing device, the user would probably have to 
explicitly indicate his intention to perform a drag-and-drop operation, 
stating what he wishes to drag and what he wishes to drop, 
respectively." need much more explanation i think
- there is no way to have synchronized caption and audiodescription on 
video element or is it source element role, why did not you take the 
SMIL audio and text element ?

-- 
Aurélien Levy
http://www.fairytells.net

Received on Saturday, 30 June 2007 13:54:05 UTC