Path: utzoo!mnetor!uunet!husc6!ut-sally!ut-emx!wca
From: wca@ut-emx.UUCP (William C. Anderson)
Newsgroups: comp.sys.sgi
Subject: Re: Mouse driven rotations
Message-ID: <1227@ut-emx.UUCP>
Date: 16 Mar 88 16:06:32 GMT
Organization: The University of Texas at Austin, Austin, Texas
Lines: 29
Summary: mouse rotation human interface

Interestingly, almost all workstation manufacturers have settled on the
optical mouse design originated by Mouse Systems, Inc.  This mouse is
interesting in that it has *two* LED/phototransistor "eyes" in its bottom.
Although usually only one eye is turned on (this is sufficient to report
relative position), both eyes can be enabled and will report on their
respective relative positions.  This is sufficient information to generate
(small) rotations.

I wrote just such a driver for a Sun 1 workstation (serial number 91 - my
age is showing) and a simple program to test the human factors related
to the idea that by rotating and translating the *mouse*, an object on
the screen would likewise rotate and translate.  It was a very intuitive
interface, in my humble opinion.

Those readers schooled in geometrical matters will realize the limitations
of the mouse in reporting the rotation.  Whereas the information which you
need to generate accurate rotations is sin(theta) or cosine(theta), the
two eyes in fact generated information related to tangent(theta).  Therefore
the rotation information became difficult to deal with as theta -> pi/2, or
tangent(theta) -> +infinity.  Still, with a "reset rotation" command/switch
and by rotating in small increments, the user interface seemed to me
to be unusually transparent.

I have never seen this implemented elsewhere.

William Anderson - University of Texas Computation Center - wca@emx.utexas.edu

Disclaimer:	No connection with Mouse Systems Inc. or Sun Microsystems Inc.,
		except as a satisfied user of their equipment.