Watching users at the HyperDesk the most common reaction was to place their
hand on the mouse and explore the space on the screen. Users would try clicking
on the images of keys on the keyboard, and the images of the desklamp's
lighbulb. Some users would even experiment by pressing the keys on the real
keyboard. Eventually users would find that clicking the mouse on the small
black lightswitch would have the effect of turning on a light.
The users then quickly realized that by clicking on the switch in the image of
the desk does not turn on that light, but rather turns on the light in the
image of the image of the desk; i.e. one level down. Users would then click the
second switch, turning on the third light and continue down until the tiniest
switch which would have the effect, not of turning on an even tinier lightbulb,
but turning on the physical desklight sitting next to the user. This would
cause the user, who moments before was leaned in close to the screen trying to
click a tiny switch, to lean back in surprise.
After
a minute or so most users lit all lights apart from one. That light was the
first light in the image on the monitor. To light that light required, as it
does for all the lights, to press a switch on the level up from that one. It's
impossible to click that switch with the mouse in the case of the first light,
as the level up from the first image is the physical desk itself. Most users
would not move their hand from the mouse the few inches to press the physical
lightswitch and thus lighting the remaining light.
The nature of the HyperDesk channels our vision progressively downwards into
repeating displays. This tunnel vision is a common side-effect of modern
computer terminals; gradually our peripheral vision is lessened as we stare
further into the screen. In the HyperDesk all we see is our peripheral vision,
it's just repeated many times.