Re: HTML-A11Y Task Force Consensus on Issue-204 (Updated)

On Mon, May 14, 2012 at 10:07 PM, Sam Ruby <rubys@intertwingly.net> wrote:
> So before proceeding, can I get confirmation that:
>
> 1) Benjamin: that you do not object to AllowAriaReferHidden?
>
> Should it be the case that you object to both proposals and do not bring
> forward a proposal of your own, then your input is likely to get discounted.

I believe the WG cannot make an informed decision because:

1. Use cases for referring to elements designated as @hidden have not
been provided, so it's not clear what problem we'd be solving by
making referring to @hidden content conforming.

2. The claim that more authors will understand @hidden to mean
invisible but exposed to accessibility clients when referred to by
visible elements, than will understand it to mean simply hidden, is
unsubstantiated, so it's not clear why we should require user agents
to expose @hidden content to ARIA processing even if we don't make
such references conforming.

3. Possible user agent behaviours, such as bringing @hidden content
into view, have only been superficially discussed.

In the absence of such information, I'm inclined to produce two Change
Proposals.

The first CP would strengthen the use of @hidden as a technique for
deactivating parts of the DOM, for example by defining the text
alternative of @hidden content as the empty string to effectively
exclude them from ARIA processing.

The second CP would require user agents to use @hidden references for
ARIA processing, but forbid authors to make such references on the
basis that such references tend to harm users. This would of course
leave open the option of making such references conforming in
HTML-Next.

What would be the deadline for these if I were to try to produce them?

--
Benjamin Hawkes-Lewis

Received on Thursday, 17 May 2012 00:25:38 UTC