>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 _______________________________________________ busybox mailing list [email protected] http://lists.busybox.net/mailman/listinfo/busybox
