WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Would pressing enter on a radiobutton in a <form > with onsubmit event be called under 3.2.2 ?

for

From: Mallory van Achterberg
Date: Sep 10, 2014 9:41PM

Hitting "Enter" has submitted forms since forever, with the
exception of Internet Explorer of old (not sure about recent ones)
which required first the user be focussed on the submit button.

While I've always liked that behaviour of IE, nobody else did it,
and on things like search forms, too many users expected to hit enter
from the text input (you'll find a lot of questions/complaints on
the internet from several years back about "why we can't submit
with Enter in IE" etc).

Users who keyboard for the first time (or had always used Opera
and now use anything else) will likely accidentally submit until
they learn the interface. Experienced keyboarders know better.

If you worry you have a lot of newbie keyboarders filling out your
form, you could always somewhat annoyingly intercept all submit()s
and present a confirm popup ("Are you sure you want to send this
form?") which lets users learn how to use radios/checkboxes without
constantly submitting forms... but this isn't something I personally
would do. :)

_mallory

On Wed, Sep 10, 2014 at 07:02:42PM -0400, Birkir R. Gunnarsson wrote:
> My conclusion after some thought is that this is not a WCAG violation.
> We always assme that users, be they users of assistive technology or
> not, understand the patterns of interaction with the web (i.e. that
> you use the spacebar to check radiobuttons).
> And I do not see a way to call this under 2.1.1 (the element is
> keyboard accessible), or 3.2.2 (as jOnathan pointed out, the user
> initiates this action).
> See also: http://tjvantoll.com/2013/01/01/enter-should-submit-forms-stop-messing-with-that/
> I do not necessarily think that screen readers are as bothered by
> people not using the form element as the author claims, they are
> pretty smart applications and know to work around that, but this
> article has at leat led me to the initial decisiontaht this is not an
> accessibility violation.
> I can be convinced otherwise if necessary.
> Cheers
> -Birkir