> overestimating the unique complexity of their particular situation. Then we prove it by forcing solutions on them that artificially reduce that complexity. A

> overestimating the unique complexity of their particular situation. Then we prove it by forcing solutions on them that artificially reduce that complexity. As an electrical engineer I have the habit of asking people to explain to me the function of the part they want to remove. Just because it worked briefly without that part does not mean that part did not serve some vital function. How you can tell experts from beginners is that beginners will be pleased with the thing working. Experts will think about weirder edge conditions, thermal cycles, maintainability, safety, security, testability, sustainability, you name it. Don't get me wrong, I love simple, elegant solutions with little (or no) moving parts — they often win out on all of the above. But sometimes the people with the complexity of the domain in mind are right, when they insist that certain elements of a system are needed.