I just noticed that, although they say that the attributes aria-label, aria-labelledbyand aria-describedbywork on each element (see. Http://www.w3.org/TR/wai-aria/states_and_properties#aria-describedby ), they seem to only work for multiple elements of a type a, not for, for example, divor pin NVDA and JAWS.
I created a little code to demonstrate the problem (view it using the view and focus mode):
http://codepen.io/jmuheim/pen/avWbPe
For example, in NVDA, in the element a, aria-labelthey aria-labelledbywork both in viewing mode and in focus mode. But it is aria-describedbydeclared only in focus mode, and not in view mode.
For an element, inputnone of the attributes work in view mode, but all work in focus mode.
For bare text elements such as por div, none of the attributes work.
In JAWS, this is a very similar behavior, but at least for the element p, when there is aria-describedby, it announces that the description can be read by pressing "JAWS + alt + r".
I really do not see a clear picture for this, so I wonder what are the general rules in the firmware regarding the use of these attributes? Or better: why don't they just work for each item, as the specification suggests?