On 9 Sep 2003, at 23:57, Daniel L. Katz wrote:

> well, just in case this wasn't a joke, note that the problem at hand isn't
> simply a math problem. instead, the arithmetic relates to something physical,
> and units are attached to the numbers. whence:
>
> mpg = miles driven/gallons of fuel used = 198 mi/11 gal = 18 mi/gal.
>
> further, we usually have some sense of the expected magnitude for the
> result of a physical calculation, and the original sender was obviously
> suspicious that 35 mpg wasn't reasonable.
>
> as another example, suppose we wanted the average speed over the 198
> miles, and knew that the trip took 4 hours and 15 minutes = 4.25 hours.
> whence, mph = 198 mi/4.25 h = 46.6 mi/h. note that the entire time was
> converted to hours before doing the actual mpg calculation.
>
> keeping the units straight and having some idea beforehand as to the
> likely magnitude of the result are powerful tools in the business of
> physical calculations.
>

Yup.  Keeping the units in the equations can really keep the engineer out of
a lot of doodo by avoiding silly errors (like at JPL?).  Estimating the
expected magnitude is also good practice.  We were forced to do that in the
days of slide rules since the slide rule knows nothing of decimal points.

Vanagon content: Mine also averages roughly 20 mpg.  It has never seen 30
for a whole tank.

Jack Cook
1985 GL Weekender