[commonbullet]

I guess this is the root of the 'problem' I'm getting to convert floats to strings.

In this crap example, I fail to understand why fl3 is 0.199951. Shouldn't it be 0.2 ?

PHP Code:

float fl1 = 1234.2f;
float fl2 = 1234.0f;
float fl3 = fl2 - fl1; 


[SweatyBanana]

Originally Posted by commonbullet I guess this is the root of the 'problem' I'm getting to convert floats to strings. In this crap example, I fail to understand why fl3 is 0.199951. Shouldn't it be 0.2 ? PHP Code: float fl1 = 1234.2f; float fl2 = 1234.0f; float fl3 = fl2 - fl1;

It should be -0.2

[commonbullet]

Ooops, anyways it's -0.199951.

[schnitzelmaker]

Anywhere exist an thread where i have explained it a little better but i hope this helps also.

The Problem is with converting hex into floats. Some float numbers dont exist as hex.
As Example 1234.2 is in hex 0x449A4666, but 0x449A4666 is in float 1234.199951.

A trick to fix this is multiply the float before caculate it and divide it after caculate.
((1234.0*10.0)-(1234.2*10.0))/10.0

[PM]

This is possibly a bit unrelated, but you might be interested in reading this:
http://docs.sun.com/source/806-3568/ncg_goldberg.html

[commonbullet]

Thanks schnitzelmaker and PM (very interesting reading, looks like I'll take years trying to digest all that).

It seems that I came across an old issue which is not exclusively related to C/C++ as I had believed.