Mach 4.0
Just one final thought.
Your logic is that you can step through your systems, bit by bit and arrive at a works/doesn't work conclusion for the whole. I believe that this is a rather naive view and it suggests a lack of exposure to testing complex systems, I'm not being insulting here, just commenting on what I see in the interests of trying to give you a different view to consider.
My experience in working with complex, interdependent, electronic based systems is that the only real proof (or should I say non proof of failure) is to test the whole system, since individual modules may pass a test, but when integrated, the system has a reasonable probability of failure due to unforeseen circumstances, what one may call a known/unknown scenario.
So we started with a deductive approach and then often fell back to an inductive approach to understand why the system did not work as we expected it to work. The causes of failure were often relatively trivial, but nonetheless the system, as a whole, failed.
And that was in conventional testing, not something as exotic as NEMP/DEW, where a system that is certified to operate with one set of parameters is then exposed to a different set of parameters.
That does not give you any hard data or logic, you may say and that is precisely the point.
One needs to test the system to have any degree of confidence in a hypothesis.
If I were a consultant to your boss, I would feel ethically obliged (after all it is his neck potentially on the block) to point out the severe limitations of this deductive methodology and also to point out that a full systems test is not feasible, so the confidence level in the hypothesis is adjusted accordingly.
Last edited by Final 3 Greens; 3rd Jun 2006 at 06:20.