I wrote:
> In an experiment with only 4 main parameters -- input power, inlet > temperature, outlet temperature and flow rate -- the number of potential > significant errors will BE small, and so will the number of ways > deliberately fake data can be surreptitiously introduced. > This is a bit like playing chess with only 4 pieces, a king and queen on both sides, and nothing else. The number of lines of attack are strictly limited. The problem is simplified. This is a key concept to computer security and program testing as well, as I expect everyone here knows. Limited, well defined inputs and outputs, and modularity. I did not mean there are only 4 ways to fake this experiment! I did not mean that every method of attack (introducing fake data) corresponds exactly to a potential error, although most of them are congruent. - Jed