Re: Engineering & Software Engineering

[Richard D Riehle <laoXhai@ix.netcom.com>]

In article <MpqV3.79$4D5.48229@ratbert.tds.net>,
	"M." <me@home.xx> wrote:

>Richard D Riehle wrote in message

>>Every engineering practice
>>includes some kind of design metrics. Software practice may proceed
>>oblivious to any notion of design metrics.
>
>I'm in trouble then, because I'm an engineering intern and still oblivous
to
>"design metrics".  Is there another term for the concept?  

Let's take mechanical engineering as an example.  When one designs
a piece of machinery, it is always designed to some tolerances. You
could say that you design to "snugness of fit."  Even a simple bolt
and nut are specified at the same TPI. A camshaft, a piston, the
compression ratio of your engine, all are measurable.  They all behave
according to the laws of physics.  Coefficient of friction is real.

We can measure the viscosity of liquids and plan for their behavior in
a design.  We can regulate the flow of electricity according to some
published specifications.  Over and over in physical engineering we
can apply what I call "design metrics".  The reason you never see the
term in conventional engineering is that it is inherent in the design
process.  The reason you never see a term like it in software engineering
is that it is almost non-existent.  I say almost because we do have some
opportunities for design metrics but they are rarely used.

So, as an engineering intern, you are using design metrics. It's just
that no one feels it necessary to explicitly use the term.  Who was
it that said, "I was writing prose all my life and didn't even know it."

Richard Riehle
http://www.adaworks.com