Floaating point systems using IEEE convention
take care of the problem of infinity:
a special pattern of bits represent the infinity
a number greater than any other floating point value.
It has a sign too.
IEEE compliant floating point library / processor
observe the following rules
infty + x = infty
x / 0 -> infty
etc
and 1/(1+x) = 1 if x==infty
---------------------------
Every elementary programming lectures
talk about numeration, number representation and
the inifinite problem.



Daniel Frerejacque
------------------
dan### [at] acmorg

Post a reply to this message

frerejacque <dan### [at] freefr> wrote:
> x / 0 -> infty

  Well, that's just plain wrong. In the n/x function there's a discontinuity
at x=0, which makes n/0 undefined (if you approach x->0 from the right,
the limit will be +infinity, while if you approach from the left, it
will be -infinity; it can't be both at x=0; it's a discontinuity and thus
not defined).

> and 1/(1+x) = 1 if x==infty

  If 1+infinity = infinity, then 1/(1+infinity) = 1/infinity = 0

-- 
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -

Post a reply to this message

These dicussions have always been out of my league, but I'll make a
fool and a nuisance of myself briefly to add that assigning a +/-
to infinity has always seemed to me about as sensible as the
following conversation:

Person 1: "Look straight ahead. What do you see?"

Person 2: "Nothing, sir, I'm blind."

Person 1: "Oh, in that case, just tell me what color it is."


:)

Post a reply to this message