Re: HTML working group does NOT dictate what ARIA can and can't do

On Wed, Mar 21, 2012 at 3:41 PM, Richard Schwerdtfeger
<schwer@us.ibm.com> wrote:
> Don't start breaking all of industry's implementations with hacks like this.

You're objecting to this CP?

    http://www.w3.org/html/wg/wiki/ChangeProposals/DeprecateLongdesc

What implementations would be broken by the proposed textual change
and how? Details would be useful information to add to John's CP.

> HTML5 does NOT control the ARIA spec.

Likewise ARIA does not control HTML5, except in so far as the HTML WG
is constrained its attempt to describe how to process the web corpus
(which includes features originating in ARIA) and through its attempt
to make HTML a host language for ARIA.

It's within HTML WG's remit to define the meaning and mapping of
@hidden, or indeed to stop trying to make HTML a host language for
ARIA and define the meanings of attributes like "role" and
"aria-hidden" independently, though that's best avoided if possible.

The current W3C HTML5 draft makes a normative provision: "Elements
that are not hidden should not link to or refer to elements that are
hidden." It gives an example: "It would similarly be incorrect to use
the ARIA aria-describedby attribute to refer to descriptions that are
themselves hidden. Hiding a section means that it is not applicable or
relevant to anyone at the current time, so clearly it cannot be a
valid description of content the user can interact with."

    http://dev.w3.org/html5/spec/editing.html#the-hidden-attribute

The ARIA Editor's Draft, even after the recent edits, includes
normative text implying authors should be allowed to expose hidden
content as text alternatives to end-users:

"Skip hidden elements unless the author specifies to use them via an
aria-labelledby or aria-describedby being used in the current
computation. By default, users of assistive technologies won't receive
the hidden information, but an author will be able to explicitly
override that and include the hidden text alternative as part of the
label string sent to the accessibility API."

    http://www.w3.org/WAI/PF/aria/roles#textalternativecomputation

Ignore the rationale and debate around the Change Proposal, and look
at the proposed textual changes:

    http://html5.org/tools/web-apps-tracker?from=6979&to=6980

The spec would be changed to say:

"Elements and scripts may, however, refer to elements that are hidden
in other contexts. … It would be fine, however, to use the ARIA
aria-describedby attribute to refer to descriptions that are
themselves hidden. While hiding the descriptions implies that they are
not useful alone, they could be written in such a way that they are
useful in the specific context of being referenced from the images
that they describe."

This verbiage would make it conforming to point @aria-describedby at
@hidden content. The proposed text does not does not imply that this
can be used to surface rich content to the user.

In other words, the actual result is to further harmonize ARIA and
HTML, not to have HTML5 "control" ARIA.

If your view is that it should be non-conforming HTML5 to point
@aria-describedby at content in @hidden, I suggest a first step would
be changing the ARIA spec to say that it is non-conforming to point
@aria-describedby at content that is "hidden".

It might be worth noting in the HTML spec that such descriptions would
be flattened in ARIA processing for text alternatives, but that seems
more like stepping on ARIA's turf than keeping off it.

--
Benjamin Hawkes-Lewis

Received on Wednesday, 21 March 2012 23:13:40 UTC