"Jeff Findley" wrote in message
...

In article ,
says...

I was always taught that to change more than one thing at a time was
unwise. If you do that in electronic design, or software, yu deserve to
get
your fingers burned. I suppose its OK if you can test all the changes
individually, but I suspect in this case cost and time are probably
driving
the path here.
Unfortunately, quality control can be a nightmare when you are relying
on
contractors who, on paper perform to the same spec, but humans make
errors.

True, but in this case the failed helium tank strut was a clear single
point of failure which was easily solved. That and it was not really a
design variable; it was much more of a quality control problem.

Also, changing more than one thing on a launcher is quite common.
Saturn V evolved quite a bit over its short lifetime, to the point that
no one Saturn V was quite the same as any other.

Jeff

You know I've been thinking about that a lot lately with programming.

The difference is, I can change a line of code and recompile and test in a
matter of minutes for the cost of a few dollars.

We can't do the same with rockets.

And the reason we made it to the Moon on time was because in part of the
switch to the all-up testing.
Risky, yeah, but sometimes the choices are limited.


--
Greg D. Moore http://greenmountainsoftware.wordpress.com/
CEO QuiCR: Quick, Crowdsourced Responses. http://www.quicr.net