Re: RFC: a generic pointer proposal

Elliot Lee (sopwith@cuc.edu)
Fri, 15 Nov 1996 01:14:36 -0500 (EST)


On Thu, 14 Nov 1996, Doug Ridgway wrote:

> I don't know whether things like the SGI bank of knobs should be included
> under this, but they have at least a dozen knobs. I can't remember exactly
> how many. The ones I saw were 3x4 or 4x4, but I think they are available
> larger. (The SGI bank of knobs is a panel with an array of knobs on it.
> It's used by various graphics programs for rotating matrices or adjusting
> things in 3d models. I can certainly see how I'd want thirty or forty
> physical knobs to tweak if I was trying to do the modelling for ``Toy
> Story''. A complex Toy Story model had something like three hundred
> adjustable parameters.)
>
> Along the same lines, one might imagine Linux running a more specialized
> system, such as a chemical plant process control console, a professional
> audio mixing console with many inputs or a realistic flight simulator. How
> many knobs does a 747 cockpit have?

I'm sure there is a mouse driver for a 747 cockpit available from Boeing
*somewhere* (getting MS Flight Simulator to use it would be another matter
altogether :-)

> Then there are also the various VR input devices. A boom-mounted helmet, for
> example, is going to need at least six dimensions (three space plus three
> rotation) and it's unlikely to be the only thing hooked up. I tried a
> setup at the San Diego Supercomputer Center with a boom helmet and a
> wheelchair motion input device. Two more dimensions, one for each wheel.
> Now imagine a glove. You'd need dozens of dimensions just for an accurate
> hand position.

I don't think there is (or I hope there isn't) any limit on the number of
"mouse" devices you can have hooked up to your system. The 255 dimensions
proposed(?) would be more than enough for the glove. Then you'd have other
separate devices for each particular input device. You don't need to
combine them all into one.

Sounds good to me otherwise.
-- Elliot

http://www.redhat.com/