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

Re: change proposal: Allow the a element to have ARIA role attribute values other than link

From: Maciej Stachowiak <mjs@apple.com>
Date: Tue, 02 Mar 2010 00:11:51 -0800
Cc: HTMLWG WG <public-html@w3.org>, HTML Accessibility Task Force <public-html-a11y@w3.org>
Message-id: <56010B34-7180-411F-B1EA-A09065B34020@apple.com>
To: Steven Faulkner <faulkner.steve@gmail.com>

On Feb 26, 2010, at 4:20 AM, Steven Faulkner wrote:

> Have completed the change proposal for HTML WG tracker Issue 85 [http://www.w3.org/html/wg/tracker/issues/85 
> ]
>
> change proposal: Allow the a element to have ARIA role attribute  
> values other than link
> http://www.w3.org/html/wg/wiki/ChangeProposals/ARIAonanchor#Negative_Effects
>
> any feedback welcome

<chair hat off>

I'd like to check with you whether this proposal is really doing what  
you intend. In Details it says:

> Remove the first row of the first table in the section 3.2.6  
> Annotations for assistive technology products (ARIA).
>
> code/content to be removed:
>
> <tr><td><a href="text-level-semantics.html#the-a-element">a</a>  
> element that represents a <a href="interactive- 
> elements.html#hyperlink">hyperlink</a></td><td>link role</td></tr>

So that will have two effects:

1) <a> will not be considered to have any default role at all; it will  
no longer be considered role="link" by default.
2) Any role whatsoever is allowed on <a>, even dubious things like  
landmark roles, or role="textbox".

Was that really your intention? I think maybe not.

I think if you remove <a> from the "Strong native semantics" table,  
then you should add it to the second table, which defines "Default  
implied ARIA semantics". I suspect you probably intend for <a> to have  
at least a default role of "link", and perhaps you also may want to  
limit the set of roles allowed. I understand from your Change Proposal  
and your bug that you definitely want the "button", "tab" and  
"menuitem" roles allowed. But I suspect roles like "document" or  
"grid" would indicate an author error.

Note: I'm not necessarily taking a position on whether this should be  
changed at all. I saw what seemed like decent arguments for both sides  
of the issue in the bug. I'm just trying to help you improve your  
proposal.

Regards,
Maciej
Received on Tuesday, 2 March 2010 08:12:27 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 04:42:03 GMT