From: owner-3dui@hitl.washington.edu on behalf of Doug Bowman
[bowman@vt.edu]
Sent: Tuesday, January 25, 2000 3:06 PM
To: 3dui List
Subject: RE: system control - techniques and guidelines

Let me just clarify something Ernst said in his last
message. I did indeed do a little research into "cross-
task" interaction techniques - basically using the same
technique for multiple tasks - and even found some
clues that performance might be improved in certain cases
with these. Unfortunately, the paper was never published
(maybe I'll resurrect it someday).

At any rate, I certainly didn't introduce this idea. The
most common type of cross-task technique is to use a
manipulation technique for a travel task. For example,
Ivan's go-go technique can be used to grab a faraway object,
or it can be used to grab the air far away from you, and
then pull yourself in that direction. Mapes used this
air-grabbing technique in PolyShop/SmartScene. Jeff Pierce
et al also showed this type of cross-task technique in
their paper on image-plane interaction techniques at I3D '97.

There are other ways to cross task boundaries. For example,
I experimented with using a travel technique to manipulate
an object - by having the user "become" the object when he
selects it. Then the object's motion can mirror the user's
motion until the object is released. Object selection can
be used for travel - select the object to move to its location.

The point is - people have been doing this for a while, and
I just added a little formalism (and a not-so-catchy name)
to it, plus doing a couple of simple experiments.

Any other ideas for cross-task techniques? Hypotheses for when
this might be useful?

--Doug

--
Doug A. Bowman, Ph.D.          (540) 231-7537
Assistant Professor             bowman@vt.edu
Computer Science        www.cs.vt.edu/~bowman/
Virginia Tech