Date: Mon, 10 Oct 1994 12:30:54 -0400 (EDT)
From: Kimberley Burchett <OKRA AT max DOT tiac DOT net>
Subject: Re: keyboard messups!!!! :(
To: Stephen Turnbull <turnbull AT shako DOT sk DOT tsukuba DOT ac DOT jp>
Cc: djgpp AT sun DOT soe DOT clarkson DOT edu


> Given "fixed point" computation, this sounds like some sort of round
> off error to me.  Did you really have the patience to do 5 x 10 plus
> several iterations in the form where it waits for a keypress to see if
> it was *really* not decaying in that form?  Have you printed out all
> the coordinates and checked the math?  (ha ha; well, you could write a
> program to do that, I guess.)

  I have done all of that.  5*10 iterations is not necessary - after only 
10 iterations, the radius of the circle has gone to less than two.  That 
means it is knocking about 35% of the radius off every rotation.  With 16 
bits of fraction, that is ludicrous.  I have checked the math, I have run 
17 different tests on my trig and fixed point functions.  All work 
extremely well - accuracy of .001 or so.
  The point of showing you the code was for you to realize that the 
problem was _not_ with the trig or fixed point routines, but the way I 
used kbhit() and getch().
							Kim