W3C home > Mailing lists > Public > public-html-a11y@w3.org > January 2010

Re: HTML Chairs Solicit Proposals for ISSUE-85 - ARIA anchor-roles

From: Maciej Stachowiak <mjs@apple.com>
Date: Thu, 21 Jan 2010 15:14:48 -0800
Cc: Martin Kliehm <martin.kliehm@namics.com>, Steve Faulkner <sfaulkner@paciellogroup.com>, HTML Accessibility Task Force <public-html-a11y@w3.org>
Message-id: <F8941680-BC4F-4D10-AC42-171083EEE205@apple.com>
To: Laura Carlson <laura.lee.carlson@gmail.com>

On Jan 21, 2010, at 10:58 AM, Laura Carlson wrote:

> Hi Martin,
> 
>> I agree that the current spec draft [3] doesn't make sense and is
>> restricting the allowed roles too much. I have some ideas, but I
>> remember that Steve has created a matrix of roles in a Google doc,
>> didn't he?
> 
> Yes, I think it is at:
> 
> http://spreadsheets.google.com/ccc?key=0AlVP5_A996c5dG9RSE9GMy1JaVlBQ2dIWDliczJHckE&hl=en
> 
> Steve, is this the latest?
> 
>> I can contribute to such a document.
> 
> That would be great.
> 
>> The question to me is: is restriction and overriding with "strong native
>> semantics" the right approach?
>> Do we need such a table/matrix at all?
>> Shouldn't we trust that the author knows what he does?
> [snip]
>> if we need a restriction, is the HTML5 spec the right place?
> 
> All good questions for the task force to contemplate and discuss. I
> have a related question that I will send out shortly.

The Task Force can certainly consider these broader questions, however, I will note that they are beyond the scope of this particular issue, which is only about roles on the <a> element. It may be easier to deliver a timely Change Proposal by focusing on that specific case instead of the big picture.

As for "strong native semantics" generally, here is the background. There was much past discussion of this around the time ARIA was first integrated. I think everyone agreed that an example such as <input type=text role=checkbox> was nonsensical - it could never possibly be correct to present a text entry field as a checkbox to assistive technologies. It has to be assumed to be an authoring error of some kind. This was something that PFWG and HTML WG agreed on. I think the remaining issues to be worked out are which elements specifically should have "strong native semantics". If you look at the bugzilla bug on this, you will see that there is some disagreement over whether <a> belongs in this bucket. I think it makes sense to consider other elements on a case-by-case basis.

Regards,
Maciej
Received on Thursday, 21 January 2010 23:15:22 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 04:42:00 GMT