Re: ISSUE-204: aria-hidden - Chairs Solicit Proposals (was: Chair review of "Keep Longdesc Deprecated")

Quoting Matthew Turvey <mcturvey@gmail.com>:
>
> However, the HTML-A11Y-TF felt compelled to formally object to HTML5
> Last Call unless longdesc was included as a conforming feature, and
> WAI are now apparently working on a new aria-describedat attribute to
> provide the same function.

It is truly unfortunate that you cannot get to the point of  
understanding that "Universal Design" never is universal, and at some  
point accommodation is required. Numerous people have attempted to  
explain this to you, but you remain unconvinced. So be it.

>
> Is anyone planning to write a change proposal arguing against allowing
> ARIA to reference @hidden content? Or can the HTMLWG agree amicable
> consensus on this issue and move on?

I will. Your proposal does nothing to address the 3 key requirements  
that have been previously outlined, including the one of  
discover-ability and the ability to provide the end user with the  
option to consume or not consume the longer textual description.

Despite your assertion of "Universal Design" as the best path forward,  
you seek to have a Change Proposal that shifts the entire  
responsibility of access to the Accessibility API, failing to  
acknowledge that the majority of disabled users don't actually use  
tools that take advantage of the AAPIs. The proposed "solution" is a  
non-solution for sighted users with cognitive disabilities, or others  
who might require a longer textual description for whatever reason.

Finally, and most critically, your proposal completely runs afoul of  
tab-focus, and the impact that has on keyboard only users, screen  
magnification users, etc.  You cannot put focus on a tab-focusable  
item (derived by HTML rich content being "preserved" in some hidden  
container) that is not on screen and hope to not introduce complexity  
and confusion to sighted users.  For example, you cannot explain how a  
screen magnifier user would examine each cell of table data supplied  
in replacement to a pie-chart or bar-chart.

The real answer is that it is the User Agents, and not the mark-up  
code that is lacking here. Since it is beyond the scope of the HTML WG  
to prescribe UA behavior, your introduction of a Change Proposal that  
destroys existing backward compatibility by promoting an untried and  
completely unsupported "Hocus-Pocus" solution is actively harmful to  
end users.

JF

Received on Monday, 12 March 2012 22:22:52 UTC