[Tfug] Virtual Keyboard

[Bexley Hall]

Hi, Joe,

--- jblais <joe.blais at pti-instruments.com> wrote:

> Well - I was thinking, because there are a number of
> scan windows, and it
> doesn't seem that it's just a matter of being in the
> path of the "h" key at
> the right instance, that there seems to be more
> going on. It may be more the
> intersection of the projected image and what appears
> to be the 2 lines of
> sight towards that image. ?????

I'm sure the only way it knows "distance" is by
simple trig -- knowing the orientation of the
emitter(s)/detectors and watching where your
finger appears "projected" onto them.

This could make my guess about scaling the image
incorrect -- things might only line-up with a
single geometry...

<shrug>  As I said, it would be amusing to play
with... (like how do they get the guts inside a
lightbulb?  :> )

> I may be completely wrong, I just haven't played
> with it enough, but the
> fact that you can bring your hnd in and press a "y"
> which is in the scan
> path of h, b and perhaps n, and only 'y' is
> triggered (not counting the
> number row and the function key row, it seems like
> there's more
> "triangulation" -the old surveyor that I was- going
> on.
> 
> > > > Does it deal with multiple key presses? 
> (e.g.,
> > > > Ctl-Alt-DEL, Shift-J, etc.)
> > >
> > > ???? I noticed it didn't have all the Function
> keys.
> > >  Insetad it had some
> > > funny "icon" like keys for the top row. I jst
> didn't
> > > think about the
> > > three-fingered salute.
> >
> > Or any *other* combinations where one key is
> > "in the shadow" of another!
> 
> That's just it, that's why I think there is a
> "triangulation" going on, and
> that implies a position specifically  defined in
> space rather than just
> where it happens to fall on some arbitrary surface,
> like my fat finger on a
> '7' being hit by the images for y,u,h,j,b,n,m and of
> course 'space-bar'

But this is different... being able to resolve two
dimensions is one problem ("triangulating" as you
call it).  OTOH, being able to *see* a second
keypress *behind* an existing keypress wouldn't
work (?)
 
E.g., a regular keyswitch matrix can only reliably
decode a single keypress -- you need to play other
games to be able to detect *multiple* closures
(e.g., diodes to inhibit "sneak paths")

Fun to think about, though!  Took me a fair bit
of time to sort out how the SpaceBall worked...
(sure beats wasting time on crossword puzzles!)

--don


      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page. 
http://www.yahoo.com/r/hs