Leon wrote:


----- Original Message -----
From: "Mark & Juanita"

And errrr uh.... � The kilomete /mile calculation that shot millions
into the Mars surface.

THAT was Bush's fault.

Or was it Clinton's fault......?

Think that one occurred during the 90's under Clinton's watch.

Not that there's any correlation -- stuff happens despite rigorous
testing. Could just as easily have been a m / km unit problem and they'd
have still been off by 3 orders of magnitude. Systems Engineering is a
discipline that requires both top-level thinking and tedious attention to
detail.


Which is exactly why one would think that this kind of mistake should not
have been made. I think NASA would fall under the group of Systems
Engineers that would absolutely require top-level thinking and tedious
attention to detail.

IMHO it seems only common sense that one unit of measure should be adopted
and all others NEVER be allowed.

Doesn't always work that way. As I pointed out above, the same thing
could have occurred with a meter/kilometer interface error (or cm / meter
interface discrepancy). This isn't just a refusal on the part of one group
to adapt to someone else's units. In real-time embedded systems, it is
often the case that various order of magnitude selections are made in order
to preserve numerical precision in calculations. So, if a certain control
requires cm or other small unit precision, it is not practical to use that
order of magnitude for guidance that is working in km due to numeric
overflow issues.

I haven't seen the failure review report, but it is possible that the
units chosen made sense because of available data and the units in those
databases. Usually decisions made have some rational basis, particularly in
projects of this magnitude. The problem is that sometimes those decisions
don't get appropriately communicated between groups working on subsystems of
those very large projects.

--

There is never a situation where having more rounds is a disadvantage

Rob Leatham