The probe wasn't lost because of a human error in units. It was lost
because the Goldin-style faster-better-cheaper mantra required
a process in which humans were perfect, and didn't need
checking. THAT was the cause of the disaster, not the fact
that the project was implemented by normal human beings.

Good points here but still I think it should not have happened at all. Orbit
insertion has been accomplished so many times that the logic surrounding
the parameters should be COTS. I could realy imagine someone make a
sourceforge project surrounding it. I guess many students would love the
promise of real implementation.

Sincerely
Bjørn Ove