Re: This calculation is just wrong / computer can't count!

Tech Tip: Click here to run a free scan for Windows Errors and optimize PC performance

It is the programmer's JOB to choose an internal representation that
fits the problem. You wouldn't use a sledge hammer to pound in a
finishing nail. Likewise, you shouldn't use a float where more
precision than it can offer is required. More precision requires more
computer resources. The programmer making the choice has to weigh the
costs and benefits and select the best option. Does the carpenter
LOVE his sledge hammer or his claw hammer? Professional programmers
don't love their FPU's either, but a good one learns to use the
correct tool for the job.

The OP stubbornly refuses to acknowledge the nature of the (float)
beast and continues to insist that it is wrong, wrong, wrong. It
isn't wrong. It is simply limited. The sledge hammer is not "wrong"
when it destroys some delicate carpentry. What is wrong is the choice
of tools.


On Wed, 10 Oct 2007 13:53:00 -0700, "David Ching"
<dc@xxxxxxxxxxxxxxxxxxxxxx> wrote:


I agree there is no bug in the floating point implementation, and it's
getting as close as it can to the ideal answer. But it is still "wrong,"
much like early digital audio systems that used only 10 bits to encode audio
and sounded like crap were "wrong" even though theoretically their
implementation lived up to the potential. Yet back then people were saying
that digital audio perfectly captured the sound, which is laughably wrong
today. I'm just saying it's easy to fall in love with the technology and be
overly sensitive to criticisms of its flaws, and that's what I'm seeing
here.

-- David

.

Relevant Pages