[Top] [All Lists]

[TowerTalk] Great Conspiracy method of evaluating Antenna Software

To: <>
Subject: [TowerTalk] Great Conspiracy method of evaluating Antenna Software
From: (Guy Olinger, K2AV)
Date: Sat, 2 Feb 2002 12:25:48 -0500
Working for a large software house (SAS Institute) I am often struck
that the meanest critic of our stuff is ourselves, who wrote it. I
think that's really how customers want it. My own view of quality
software is that the customers never consider the software to be an
issue, it just works. We really take quality issues personally. I
think most in the business do. I really have never known anybody that
went to work to cheat, code junk and make sure that customers got
wrong answers. From what I read and see on TV there seem to be those
here and there, but they don't speak for the population.

It has become a fixture of the post-Watergate world to presume
cheapness, lack of attention to quality, carelessness of design, mean
or criminal intent on the part of designers, manufacturers, etc. On a
given day it seems a firehose of skepticism played out on anything
except our own personal observations.

*> OF COURSE <*, there is some aspect of a mathematical or idealized
solution to antenna modeling. But there also is an entire aspect of
mathematics devoted to nailing formulaic descriptions to measured
points, and correcting the formulas to generate the observed data
directly. And there is hardly anything you drive or ride that didn't
depend on some kind of computer modeling to get a product at a
commercially feasible price.

It is the DREAM of any model writer to predict actual data accurately.
It's what makes them drool, dream, and stay up late nights dinking
with the methods. The instinctive motivation of these guys is to make
the models better, not generate cheap excuses for cutting corners.

The one who knows MOST the weaknesses of making something map to real
data is the software writer.

What we appear to miss so often here is that the inconsistencies that
would invalidate NEC2 or Minninec solutions are well known and
publicly documented for all to see. If nowhere else, these are on
W4RNL's excellent web site. The collective user mass for NEC style
programs has been testing and refining them for a long time now. These
are in the public domain.

There really are no more great mysteries about a half-wavelength piece
of wire up in the air. No more than about a wheel. If you want to know
where NEC-2 breaks down, go read it.

A lot of it is solved in NEC-4, but NEC-4 is tied up in an expensive
licensing issue. You can buy an MP for less than the total licensing
cost for a NEC-4 based solution. Interesting, the license for NEC-4
includes a contract that all improvements to the source code
(included) will be shared back to the university for propagation to
the user community, peer review, etc.

One of these days when the computers gain about another 1000 x factor
in calculating speed, the weather will cease to have the mystery it
does today.

I suppose then we will be complaining that it rained at 8:32 instead
of 9:06 as predicted, and grumbling about a conspiracy among weather
forecasters to make sure we get wet going to work.

73, y'all.


-- Sometimes the only thing behind a bush is the back of a bush --

(apologies to N4ZR)

Where do you get ICE bandpass filters & beverage matching boxes?  The
same place that pays for the hosting of this list:  The eHam Store.
Order online at

FAQ on WWW:     
Administrative requests:

<Prev in Thread] Current Thread [Next in Thread>