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

Re: Change proposal for ISSUE-85

From: Maciej Stachowiak <mjs@apple.com>
Date: Tue, 15 Jun 2010 17:13:06 -0700
Cc: Sam Ruby <rubys@intertwingly.net>, Steven Faulkner <faulkner.steve@gmail.com>, public-html@w3.org
Message-id: <57BBD6FB-95CA-45CB-A441-0C011C8DBCCF@apple.com>
To: Ian Hickson <ian@hixie.ch>

Hi Ian,

On Jun 15, 2010, at 5:02 PM, Ian Hickson wrote:

> On Tue, 15 Jun 2010, Sam Ruby wrote:
>> 
>> Steven makes the case that this reasoning also applies to style 
>> attributes and href attributes with javascript: schemes.  See also bug 
>> 9872:
>> 
>> Care to update your proposal to address (either by incorporation or 
>> refutation) this?
> 
> I don't understand the relevance of that statement to this issue. Are you 
> saying that we should be encouraging the use of style="" attribute and 
> href="javascript:" attributes? That would seem to fly in the face of a 
> decade of accessibility advocacy. Could you elaborate on how you see these 
> points affecting this issue?

I think Sam's point was this: using href="javascript:" and the style="" attribute, one can make a link look and act like a button. The spec makes this valid, but makes it invalid to apply an ARIA role describing it as a button.

I the case of href="javascript:..." in particular, it seems to me that pretty much any time that is present on a link, the link is essentially acting as a button rather than as a link. However, the spec currently makes javascript: URLs conforming. On the surface, it seems inconsistent to allow a feature that can be used in almost no other way than to make a button act as a link, but at the same time forbid applying ARIA markup to tell AT that it is a button.

I must admit I personally hadn't thought about this issue in evaluating the ARIA roles allowed by the HTML5 spec. It seems like consistency would call for either disallowing links to javascript: URLs (or likewise links to href="#" with mouse event handlers), or allowing such links to carry role="button". I don't have a strong opinion on this issue, but there is a good argument to be made that the current spec is inconsistent.

Regards,
Maciej
Received on Wednesday, 16 June 2010 00:13:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:17:09 GMT