Don Foreman wrote:
On Tue, 07 Apr 2009 17:31:51 -0400, Bob Engelhardt
wrote:

Pete C. wrote:
Actually writing good software *is* engineering, ...
I was a software "engineer" for 25 years before retiring in '92. Maybe
things have changed since then (although I doubt it), but in my career I
never did myself, or saw, or heard about, or read about someone using
math to analyze the behavior of software. That is the essence of
"engineering": the use of math models, based on science, to analyze &
predict the behavior of a design. Be it electrical, mechanical, civil,
chemical, metallurgical, whatever real engineering. All based on
science. Software is NOT included.

Don't get me started on "computer science".

Bob

Some fairly heavy-duty math is behind software that does:
secure computing
encryption and decryption
compression
image processing
some controls, particularly controls that must not fail like
flight controls, refineries, etc.

I'd agree that few of the engineers I knew that did this sort of work
were computer scientists.




I'm an "Electrical Engineer". I do 95% digital work and freely admit
that most of what I do is not engineering. My heavy lifting math skills
have degraded significantly since graduation in 1981 due to lack of use.
Most of the engineers I work with (mechanical, electrical, software,
systems, optical) lean very heavily on computer simulation of one sort
or another to do the math for them. Actually calculating, or even
estimating in your head and even on paper seems a lost art. I was
trying to get a couple of engineers to look at an alternative way to
calculate Standard Deviation. One guy found it on the web, but nobody
trusted it. It took me about 2 minutes to prove equivalence with the
standard form of the equation, pretty simple algebra, expand the square
of a difference, distributive property, definition of average, rearrange
terms. No one seemed to be able to follow it, or want to try it even
though it significantly reduced hardware memory requirements.

CarlBoyd