The GrabDragInteraction has markup like this:
When pressed, that element is destroyed and becomes:
<div tabindex="0" role="application">Drag me</div>
This means that there is a focus change when switching between two elements. This causes thte screen reader to always read the name of the new element BEFORE any new grab/release responses.
In phetsims/membrane-transport#491 @terracoda said it would be beneficial to hear the grab/release response first sometimes.
With the current markup, that is not possible. HOwever, if we changed it to use one element (like a div) that changed its role between the grab/drag state, we would have more control over the output.
This issue is to investigate if that can work, and if it indeed gives us the output that we want.
The GrabDragInteraction has markup like this:
When pressed, that element is destroyed and becomes:
This means that there is a focus change when switching between two elements. This causes thte screen reader to always read the name of the new element BEFORE any new grab/release responses.
In phetsims/membrane-transport#491 @terracoda said it would be beneficial to hear the grab/release response first sometimes.
With the current markup, that is not possible. HOwever, if we changed it to use one element (like a div) that changed its role between the grab/drag state, we would have more control over the output.
This issue is to investigate if that can work, and if it indeed gives us the output that we want.