> From: "Erik E. Fair" <fair@clock.org>
>
> ::: [ pentium bug ]

The pentium bug wasn't about the last digit of a computation that is probably
about to underflow to zero anyway. It was an error in the 5'th significant
figure of an operation that was supposed to have 17 significant figures.

> :::
> I just want the right answer from dc and bc when I do simple arithmetic.

The bc(1) and dc(1) systems do not use floating point arithmetic at all. Hardly
anything in NetBSD does.

	Ross.Harvey@Computer.Org