Desktop or laptop testing: active with cursor

[Thomas Pfeiffer]

> We only need the mouse cursor when a mouse is plugged in, right? Or am I
> missing cases here... ?
> 
> In that case, I'd suggest we do what we planned to do a while ago
already,
> and switch on the cursor when a mouse is plugged in, and switch it off
> when the mouse is plugged out, that makes for natural behaviour to me...

+1. In case this mechanism fails for someone (e.g. because their
virtualization
software simulates a touchscreen instead of a mouse or something), they
can
still switch to a different theme anways, right?