2 Sep
2006
2 Sep
'06
10:48 p.m.
Chuck Guzis wrote:
on 9/3/2006 at 12:39 AM der Mouse wrote:
No. The cheap calculator does fixed point math and moves the
point accordingly. It freely discards digits when the point
can not be shifted any further within the display (or, resorts
to displaying "0." or "FLASHING ERROR" to indicate it's
failure to deal with the operation requested).
Because it *is*. It differs from your other
example (which I cut) in
that the point can float over a much wider range, and it stores only
some small (compared to the size of that range) number of digits where
the point floats to.
I know I'm splitting hairs, but a cheap 4-banger calculator displays
numbers in floating point. A more expensive "scientific calculator"
displays in "scientific notation". It used to be that commercial business
types were paranoid about scientific notation and one HAD to implement
floating point the way I described.
The term for what everyone calls "floating point" is really "scientific
notation". In fact, outside of a rather narrow range, it's displayed as
such; e.g. x.xxxxEyy. Dig up any high-school science text where the
notion of scientific notation is introduced--it's never called "floating
point".
The distinction is small, but it's there.