I’ve been giving this some thought, and I will provide a few simple rules here, but I may add to them in the future. I never wrote these things down. So, please be patient as I recall them over time.
I will start with two rules from two Rays, Ray Horstman and Ray Trush, neither of whom knew each other. I can’t say with assurance that the rules were originated by the Rays, but they certainly could have been. I am still in contact with RH, but I have not seen RT in nearly thirty years. And although these two rules are both possibly in the category of humor, there is quite a bit of serious truth in them.
1. RH: “Thumb times it works, and thumb times it doesn’t!” Depending on your own view of the world, you could interpret this to mean that either success is inevitable or failure is inevitable. At any rate, you have to be ready for either.
2. RT: “An once of image is worth a pound of performance!” My memory of RT is of his positive humor. He used this phrase to softly mock those who followed the path of image before performance. I include it here to remind you that good engineering is sometimes an uphill battle of attitudes.
And here is a third rule of thumb. It is a serious comment, and I am sorry to say that I do not remember the name of the person who penned it. It was in an article about the engineering of systems from many years back.
3. “Remember, a model is not reality!” Mathematical modeling of physical and electrical systems has become common place with the introduction of computers on every desk. And it is both a wonder and a danger. When you print out the results of computers on graphs and other media (especially in color), it adds a sense of reality that all too often is not quite true, sometimes not true at all. Computer output may be impressive, but it should always be looked at with caution. Don’t read too much into it. It is the easy way out at times, and very tempting.
Here’s an example: computational fluid dynamics, CFD, depends on things called “turbulence models”. And to quote a phrase, thumb times they work and thumb times they don’t. The reason for this is that no one truly understands turbulence, and therefore can not truly model turbulence. And the answers produced by these various models can be quite different from actual test data. Now don’t throw CFD out the window because I just said that, but as with all mathematical models, be careful how you use it.
And finally, here is an old one that has been around the block many times.
4. “We never have time to do it right, but we always have time to do it over!” (origin unknown, at least to me) Unfortunately, this is more than the ordinary rule of thumb. In many cases it is the “Law of Thumb.” All I can say is that when you run across it, do the best you can.
There is a category of rules that is worth a short discussion.
Murphy’s Law, or Laws, consists primarily of one statement, “If something can go wrong, it will,” or words to that effect. You can find various lists of Murphy’s Laws, some being rather long. Many items on these lists probably did not come from Captain Murphy. Some are entertaining, and some are true; however, the original law is not a central point in design. Rather the point is that everything has a probability of failure, and the point of good design is to reduce that probability to an acceptable level.