"bob urz" wrote in message
...
http://www.edn.com/article/CA6588368.html

bob

This type of calculation is good if you are a manufacturer and concerned
with the mean lifetime of a very large population.

If you, on the other hand, frequent this newsgroup because you are dealing
in ones and twos or even dozens then it is an interesting exercise but will
not have a lot of use in practice.

It is sort like saying the probability of incurring cancer is one in a
hundred, but if you are the one those odds are meaningless.

A long time ago I was doing this sort of stuff for solid state devices.

I learned about the cancer odds the hard way.


Charlie