NetNews Usenet Archive 1993 #1/NN_1993_1.iso/spool/sci/space/18711

Text File | 1993-01-04 | 3.1 KB | 61 lines

Newsgroups: sci.space
Path: sparky!uunet!zaphod.mps.ohio-state.edu!cs.utexas.edu!torn!utzoo!henry
From: [e-mail] (Henry Spencer)
Subject: Re: Fabrication (was fast track failures)
Message-ID: <[e-mail]>
Date: Tue, 5 Jan 1993 00:39:01 GMT
References: <[e-mail]> <[e-mail]> <[e-mail]>
Organization: U of Toronto Zoology
Lines: 50

In article <[e-mail]> [e-mail] (Edward V. Wright) writes:
>>> Most engineering *is* paperwork, or workstation work today. Otherwise
>>> it's just tinkering on a wing and a prayer. You have to bend metal to
>>> *test* your engineering, but bending metal *isn't* engineering. It's
>>> fabrication done by tradesmen.
>
>Wait a minute, I never wrote that! That was Gary Coffman!

Not everyone is anal-retentive enough to waste bandwidth on keeping
every single one of those tedious attribution lines; some of us expect
people to have the sense to look at the ">" nesting depth too! :-)

>>I think engineering must consider how something is to be made.  The
>>most elegant design is useless if it can't be manufactured.
>>Knowledge of what can be made is obtained by bending metal, or
>>at least by interacting with those who do.
>
>Indeed, G. Harry Stine has an article on this very subject in
>the current Analog.

Moreover, there is a deeper issue here.  Yes, having the engineers out
of tune with metal-bending is a bad thing when the objective is to
manufacture something.  It's worse, though, when they're trying to
*develop* something.

Gary's comments (the >>> above) are squarely in the NASA mold:  if you
do it right on paper, it will work the first time (although of course
you test it just in case).  Unfortunately, the real world doesn't work
that way, as witness any number of NASA projects that *didn't* work
the first time.  Real-world development involves *finding out* what
works and what doesn't... and you cannot do that on paper.  You have
to test things *during* the engineering, not just afterward.

Note, you have to put reasonable efforts into trying to build the best
you can, or you're just wasting time with your tests.  (One of the
problems found in investigation of the early Ranger program was that
if you expect the first attempts to fail, nobody tries very hard to
make them succeed.)  But sensible engineers -- which includes Gary,
actually, based on his accomplishments rather than on how he claims
to have done them :-) -- know that the first prototype is most unlikely
to be identical with the successful final product.  Successful
engineering includes bending metal and watching it break. :-)

Whether the bending is done by the engineer himself or by a technician
is irrelevant (except insofar as the engineer needs to understand how
it's done); what matters is that the engineer is involved, because
getting that metal bent and seeing how well it works is part of his job.
--
"God willing... we shall return."       | Henry Spencer @ U of Toronto Zoology
       -Gene Cernan, the Moon, Dec 1972 |  [e-mail]  utzoo!henry