W3C home > Mailing lists > Public > public-html@w3.org > October 2009

RE: ARIA roles added to the a element should be conforming in HTML5.

From: John Foliot <jfoliot@stanford.edu>
Date: Wed, 21 Oct 2009 16:23:20 -0700 (PDT)
To: "'HTMLWG WG'" <public-html@w3.org>
Cc: "'W3C WAI-XTECH'" <wai-xtech@w3.org>
Message-ID: <011401ca52a5$7a36e5a0$6ea4b0e0$@edu>
Thoughts on this thread:

Thomas Broyer wrote:
>
> The fact that the developer can technically turn an <a> into a button
> isn't a justification for making it conforming. If it's not a link but
> a button, you should use <button> or <span role=button>.
>

The fact that we are seeing this in the wild, and that non-conformant pages 
still render in all browsers (and will continue to do so) is justification 
enough that ARIA added here should not 'add' to the non-conformance.  ARIA 
is an attempt to provide real solutions to real problems, and if a developer 
can turn an <a> into a button and have it render on screen, that is a real 
problem.



Maciej Stachowiak wrote:
>
> ... a funky custom role
> on <h1>. But it does seem fairly common to use an <a> element with
> styling and a click event listener or javascript: URL as a button,
> instead of as a link. Is it worthwhile for the spec to tell people
> doing such things that they are wrong?

1) ARIA is no more 'funky' than microdata - and in fact is much more mature. 
Bad choice of description.

2) Having the spec introduce or take advantage of a teachable moment is good

3) Why *can't* any element take an ARIA role if it is appropriate?  Given 
the desire to have as much accessibility baked in as possible, this seems 
like a trivial thing to add to the spec - any element can take an ARIA role 
if/when required.  Why limit it to a subset of the entire tool-box?



Henri Sivonen wrote:
>
> Styling h1 to be a button probably isn't a cowpath.
>

Right, but it *is* a potential out-lyer, and more importantly, what *harm* 
is inflicted by allowing the <h_> element to take an ARIA role?



Ian Hickson wrote:
>
> Conformance is about what developers _should_ do, however.
>

There is however no real penalty for non-conformance, so that really doesn't 
mean a whole bunch in the grand scheme of things.  If however you truly 
believe that conformance is "...what developers _should_ do...", then they 
_should_ add ARIA role information to any element that they have 
'repurposed' via scripting and CSS, to, as Steven states "Make sense out of 
non-sense" for users of AT.  And to do this 'legally' the spec must say they 
can...

JF
Received on Wednesday, 21 October 2009 23:24:00 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:50 GMT