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

ISSUE-85 ARIA roles and button/links

From: Shelley Powers <shelley.just@gmail.com>
Date: Wed, 18 Nov 2009 09:14:58 -0600
Message-ID: <643cc0270911180714h6bfe5da1ufb6e8dae0e3f5c16@mail.gmail.com>
To: HTMLWG WG <public-html@w3.org>
I'm not sure what happened with the thread related to ARIA roles and
links that are used like buttons.

I did want to point out an Ajaxian post[1] today, pointing to a new
button library[2] that demonstrates the use of both button and
link-as-button.

Though there is some criticism of the person's code, and why the
person who developed the library didn't use a CSS solution rather than
JS, you'll notice in comments on both posts related to this library
that no one questions his use of links as buttons. This, even though
he also demonstrates how his library works with buttons, as well as
links.

Ajaxian has been one of HTML5's biggest fans, and yet Ajaxian points
to a library that demonstrates using links as buttons--and does so
without equivocation and without caveats to the effect of "One must
not use links as button..." Of course not: JavaScript developers use
any number of page elements as buttons, to suit the needs of our
applications. We'll use list elements, divs, even headers as
buttons--depending on the effect we want to create. The only time
people demur is when they see bad code.

I can see a general acceptance for ARIA roles, over time, in order to
facilitate accessibility. Especially since their addition won't
require any changes to the code, and only minor changes to the markup.
The use of the roles doesn't go against accepted practice. Telling
people they can't use something like links as buttons, will.

Like I said, I'm not sure what outcome occurred from the thread on
this topic, or if it the issue has been moved to the new
HTML/Accessibility work group. But throwing out warnings or errors
based on using links as buttons won't change what is generally
considered accepted practice. All the increased warnings and errors
will do, is add noise, and generate distrust in the validators.

Shelley

[1] http://ajaxian.com/archives/simply-buttons
[2] http://www.p51labs.com/simply-buttons-v2/
Received on Wednesday, 18 November 2009 15:15:32 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:54 UTC