In article ,
Ned Simmons wrote:

On Wed, 25 Nov 2009 15:56:40 -0500, Joseph Gwinn
wrote:

I think it's optimistic to assume that number drills follow a rational
sequence. The drill gauge is more or less equivalent to the Stubs'
Steel Wire Gauge which, according to Wikipedia, was derived by
averaging the diameters of samples from a large collection of wires
belonging to a Mr Stubs in the 18th century.

The averaging could well be to overcome run-of-manufacturing variations
in what wire diameters they could easily make. Mr Stubbs had to be
important if they rummaged around in his shop to set the standard.
Perhaps he was a leading maker of hard or hardenable steel wire in the
day.

Whatever the article in question, all the manufacturers had their own
gauge or number series, usually generated by some rule, no two exactly
the same. The standard was written after one or a few of these
manufacturers had won in the marketplace.


This reminds me of a case of "standardization" I ran across when doing
some design work for GE's steam turbine division about 20 years ago.
One of the things I worked on was adapting punch press tooling that
was made in Lynn MA to fit a different machine in Bangor ME. The punch
mounting plate had a hole pattern for the fasteners that was almost
symmetrical, but not quite. The difference was small enough that it
was not apparent when looking at the part or a drawing, and I was
constantly checking to make sure that the mounting holes in the parts
I was designing were offset in the proper direction. I finally had to
go to Lynn to get some information from the old toolmaker who
maintained the die sets and asked what the reason was for the odd hole
spacing. I expected to hear something about avoiding interferences
with their press or the like. What he told me was that the guy he was
apprenticed to made the first prototype die set back in the 20s or so
and, since it was only supposed to be a prototype, simply eyballed the
holes. All I could think of was how many man-hours had been wasted
over 60 years keeping track of those odd locations and reworking the
resulting screw-ups.

Sounds like the story of rail gauges in the age of steam.

We all get burned by that only-a-prototype scam, once.

In the programming world, the standard defense was to do the hack work
in a language forbidden for the delivered system, so some desperate
manager couldn't declare the hackwork to be "good enough - we need only
a chevy, not a caddilac, ...". We all know this song.

Joe Gwinn