[COFF] Early programs that were run more than once?

[Dave Horsfall <dave@horsfall.org>]

That was the title of a sidebar in Australia's "Silicon Chip" electronics 
magazine this month, and referred to the alleged practice of running 
scientific and engineering programs many times to ensure consistent 
output, as hardware error checks weren't the best in those days (bit-flips 
due to electrical noise etc).

Anyway, the mag is seeking corroboration on this (credit given where due, 
of course); I find it a bit hard to believe that machines capable of 
running complex programs did not have e.g. parity checking...

Thanks.

-- Dave