Path: chuka.playstation.co.uk!news
From: "Jon Prestidge (alias Moose)" <zmoose@premier.co.uk>
Newsgroups: scee.yaroze.programming.codewarrior
Subject: "No FP... no comment"
Date: 15 Aug 1997 14:21:17 GMT
Organization: PlayStation Net Yaroze (SCEE)
Lines: 37
Message-ID: <01bca987$927743a0$e6567ec2@default>
NNTP-Posting-Host: client86e6.globalnet.co.uk
X-Newsreader: Microsoft Internet News 4.70.1155

If Codewarrior does not support floating-point guff, how do I do a
square-root.   It's OK not supporting FP (I suppose it makes sure you use
more efficient integer stuff) provided there are a few integer function
equivelents to replace things like pow.    Does anyone know how to effect a
square-root using integer maths???

In particular I'm trying to get this ScaleUnitToVector routine to work.... 
the squared things are easy enough because I can just multiply them by
themselves but what do I do about the  'length = pow(square, 0.5);' line
... if anyones got any suggestions I'd be most greatful!

Sounds like Micheal in the above item "problems with sin()"  has a similar
problem with the sin function...... what is to be done???

Jon



void ScaleVectorToUnit (VECTOR* vector)
{
	int length, square;

	square = pow(vector->vx, 2) + pow(vector->vy, 2) 
					+ pow(vector->vz, 2); 

	length = pow(square, 0.5);
	assert(length != 0);	   // never allow division by zero

	vector->vx *= ONE;
	vector->vy *= ONE;
	vector->vz *= ONE;

	vector->vx /= length;
	vector->vy /= length;
	vector->vz /= length;
}