applets wish: beep and sfdisk
Cathey, Jim
jcathey at ciena.com
Thu Sep 3 16:00:23 UTC 2009
>Perhaps an insane idea of using floating point for this is not
>totally insane after all... one day molecular/atomic storage may
>overshoot even 128-bit integers 8)
Where inexact answers are acceptable, so is floating point.
The analogy that always worked for me is that using floating
point for financial applications (my background) is like
determining the amount of money in your wallet with a
pair of calipers, rather than counting it. Not very appealing
when put that way. The temptation to use FP for money is
probably just because there's a '.' in the string. Not much
justification, really.
(I'm a floating-point hater from way back. Actually I don't hate
it at all, but I hate seeing it used where it's not necessary.
Depending on what/where you run on, there can be a significant
speed and/or size penalty when using it.)
Here it looks like it's just the bigness of numbers that's
glazing over the eyes. One could make a slight case for
3.235 TB as a printed result, or perhaps just some commas
in a big integer...
-- Jim
More information about the busybox
mailing list