Re: GitHub issue on checkbox and radio button labels

Andrew wrote:" If you use the title attribute, I believe that you are
making the relationship available in text and that is sufficient to
meet 1.3.1".
Sailesh: The title only provides an accessible name that is PD. The
relationship is not available in text. I have been given to understand
that suffixing (or prefixing)  words like "form label" or "heading"
next to a form label or section heading respectively, for instance,
are examples of conveying relationships in text. And this is seldom
desirable for most UIs.

With regard to Steve's comment about checkbox in an inbox email
folders: The sender or subject serves as a row identifier and a label
for the checkbox I believe Not the column header ... often there is no
column header for that column.

Here  is my take:
ARIA16 (aria-labelledby) notes: "Clicking on a label focuses the
associated form field. This does not occur with aria-labelledby. If
this behaviour is required then use label or implement this
functionality using scripting".
I do note that ARIA16 is called 'Using aria-labelledby to provide a
name for user interface controls' i.e. for 4.1.2  but it is listed as
sufficient for 1.3.1 too.
So a div role=checkbox and aria-labelledby= xxx> type of control, will
pass WCAG2 (including 1.3.1)  even if  clicking label does not
activate the checkbox on that platform.
It was clear in HTML4: When a  LABEL  element receives  focus, it
passes the focus on to its associated control. This is a UA function.
WCAG2 does not mandate this label functionality.
If I am interpreting this right, in HTML5 it says: the UA should match
the platform's label behavior. The default activation  behavior for a
label is to do nothing.
And even on a platform where the label moves focus to the control, no
WCAG2 SC requires this.
An accessibility feature built in to browser functionality that helps
some user groups is not a criterion mandated by WCAG 2.0.
Therefore,  one cannot call a 'fail' when a form control with a
visible label relies on title / aria-label / aria-labelledby
to convey purpose of control instead of the HTML LABEL element
implicitly or with explicit
association.

Best regards,
Sailesh Panchang


On 11/19/15, Jonathan Avila <jon.avila@ssbbartgroup.com> wrote:
> Ø  If in that pattern the row header is the label for the check box then the
> label should be clickable to check the box, there’s nothing that prevents
> that from being possible in the code.
>
> This might not always be the case – for example, what if the label was link
> that did something else and the user had to check the checkbox to check the
> item rather than performing an action on the link.  While I think we agree
> such as thing is bad/poor design it’s used commonly and would pass WCAG
> level AA IMO.
>
> Jonathan
>
> --
> Jonathan Avila
> Chief Accessibility Officer
> SSB BART Group
> jon.avila@ssbbartgroup.com<mailto:jon.avila@ssbbartgroup.com>
>
> 703-637-8957 (o)
> Follow us: Facebook<http://www.facebook.com/#%21/ssbbartgroup> |
> Twitter<http://twitter.com/#%21/SSBBARTGroup> |
> LinkedIn<http://www.linkedin.com/company/355266?trk=tyah> |
> Blog<http://www.ssbbartgroup.com/blog> | Newsletter<http://eepurl.com/O5DP>
>
> From: Paul Adam [mailto:paul.adam@deque.com]
> Sent: Thursday, November 19, 2015 12:05 PM
> To: Steve Faulkner
> Cc: Andrew Kirkpatrick; WCAG
> Subject: Re: GitHub issue on checkbox and radio button labels
>
> If in that pattern the row header is the label for the check box then the
> label should be clickable to check the box, there’s nothing that prevents
> that from being possible in the code.
>
> If there is no visible label then there is nothing to click but visible
> labels are required under labels and instructions.
>
> [cid:image001.png@01D122E0.B2DCD290] Gmail example is different because if
> you clicked on the text it opens the email, there is no visible label for
> the checkbox. In my examples I’m referring to static text labels that are
> not links or buttons so they could easily be connected labels.
>
> Paul J. Adam
> Accessibility Evangelist
> www.deque.com<http://www.deque.com>
>
> On Nov 19, 2015, at 10:52 AM, Steve Faulkner
> <faulkner.steve@gmail.com<mailto:faulkner.steve@gmail.com>> wrote:
>
>
> On 19 November 2015 at 16:36, Andrew Kirkpatrick
> <akirkpat@adobe.com<mailto:akirkpat@adobe.com>> wrote:
> Any thoughts?
>
> It is a common UI pattern in web applications to have checkboxes in a column
> of a data table, the visible label is provided as a column header, it is not
> always the case that a visable label presents itself in each row. (have a
> look at the gmail inbox view for example. )
>
> --
>
> Regards
>
> SteveF
> Current Standards Work
> @W3C<http://www.paciellogroup.com/blog/2015/03/current-standards-work-at-w3c/>
>
>

Received on Thursday, 19 November 2015 21:05:48 UTC