1137146 - Find better name rule for role::checkbutton and role::switch

Status: RESOLVED WORKSFORME

(Core :: Disability Access APIs, defect)

Description

[Marco Zehe (:MarcoZ)]

Spun off bug 1136563 comment #6:

> @@ +1061,5 @@
> > +     ATK_ROLE_TOGGLE_BUTTON,
> > +     NSAccessibilityCheckBoxRole,
> > +     ROLE_SYSTEM_PUSHBUTTON,
> > +     IA2_ROLE_TOGGLE_BUTTON,
> > +     eNameFromSubtreeRule)
> 
> I see we do that for checkbox as well but it's a bit weird since they don't
> really have name valueable subtrees.

Alex, what would be a better fit for these?

Comment 1

[alexander :surkov (:asurkov)]

You know that goes from the spec [1], they said name is computed from contents. I think we should ask the group about it before changing.

[1] http://www.w3.org/WAI/PF/aria/roles#checkbox

Comment 2

[alexander :surkov (:asurkov)]

here's ARIA thread https://lists.w3.org/Archives/Public/public-pfwg/2015Feb/0140.html

Comment 3

[Marco Zehe (:MarcoZ)]

I also gave this some more thought. In ARIA, widgets are often divs that contain some text, or spans that contain some text, in which case that inner text is the label. So for ARIA, the name from sub tree is definitely appropriate. Only for traditional inputs it isn't.

Comment 4

[alexander :surkov (:asurkov)]

That's true but I don't think I saw any real example doing this. Do you have any example at hands?

Comment 5

[Marco Zehe (:MarcoZ)]

(In reply to alexander :surkov from comment #4)
> That's true but I don't think I saw any real example doing this. Do you have
> any example at hands?

No. I hardly ever see ARIA checkboxes in the wild. Maybe Google has something in Mail or Calendar?

Comment 6

[alexander :surkov (:asurkov)]

ok, let's see if anybody comes with example at that thread

Comment 7

[Marco Zehe (:MarcoZ)]

Looks like the third example of the acc name calculation spec actually depends on this behaviour: file:///C:/devel/Testing/a11yNameComputationExamples.html

The ARIA checkbox depends on its inner text *and* the value of an input. We currently pass this test just fine. Alex, I think we should just close this. Agree?

Comment 8

[alexander :surkov (:asurkov)]

(In reply to Marco Zehe (:MarcoZ) from comment #7)
> Looks like the third example of the acc name calculation spec actually
> depends on this behaviour:
> file:///C:/devel/Testing/a11yNameComputationExamples.html
> 
> The ARIA checkbox depends on its inner text *and* the value of an input. We
> currently pass this test just fine. Alex, I think we should just close this.
> Agree?

if we do, then yes, let's mark it worksforme, however I'm not sure what /a11yNameComputationExamples.html is, since it refers to your local machine :)

worksforme per comment #7