[Tfug] Speaking of zero...

Paul Scott waterhorse at ultrasw.com
Mon May 14 14:51:13 MST 2007

William Stott wrote:
> Recently, a conversion on zero was brought up in a class I am taking. The conversation focused on positive and negative values for zero, and what they meant in terms of a computer. Anyone care to throw their ideas into TFUG on this? The instructor said that negative and positive values of zero can cause calculation issues, but he did not elaborate. I am not exactly sure how since the end result seems to be the same. I did see the problem when doing signed magnitude calculations, but only due to the rules in which the book we use gave to us.
The beginning is that there are two kinds of hardware even though there
aren't many examples of one of them (one's complement).

1.  twos complement:  0 (all bits off) is zero, all bits on is -1. 
Lowest possible negative integer is 2**n (n is number of bits) highest
possible positive integer is 2**n - 1

2. ones complement:  all bits off is positive zero, all bits on is
negative zero.  2**n - 1 is largest possible integer, - ( 2**n -1 ) is
most negative number.

(** is exponentiation)

The only example I remember of ones complement computers is the CDC 6000

Paul Scott

More information about the tfug mailing list