Re: Mouse Drift

[Frank Sheeran <Frank.Sheeran@ska.com>]

> Mouse drift.

My favorite feature!

Every few seconds, TP's (and probably most other nipplemouse machines)
check for a steady voltage from the nipple.  If it is completely steady,
they assume that as the machine heats up that is the "centered" position
and call that value "zero".

The good news is that as your machine heats up, etc., and the mouse voltage
at center gets a bit higher, say, the machine will notice in a few quiet
seconds, and recalibrate that voltage as center.

The bad news is that if you press the nipple in one direction very steadily,
it assumes that the nipple is centered, and recalibrates that voltage to
zero.  You press a bit harder and get where you want, but when you let go,
the real center is off from the "recalibrated" center, and so the computer
thinks you are pressing the other way.

The harder you fight the drift, the longer it lasts.

On the other hand, if you see it drifting, just let go; it will stop in a
few seconds.

I killed many minutes "training" the mouse to do tricks, just to watch it.
Fun stuff!

I found that often when I was thinking, I'd rest my finger on the nipple
just a bit without noticing it.  That caused the problem.  Now that I've
"trained" myself not to do that, I never see the problem.  (This is on
a 760CD, 40MB, 1GB, 133MHz, but they surely use the same algorithm across
their range.)

Frank