No. Well… maybe fear of simplicity. That’s something I use personally to rationalize why people prefer super complex error prone approaches as opposed to appro

No. Well… maybe fear of simplicity. That’s something I use personally to rationalize why people prefer super complex error prone approaches as opposed to approaches that are more durable. The reason almost universally comes down to familiarity instead of measurement as qualified by a long list of selective biases. The best way out of this, that always works for me, is to go lower. That means one or more steps lower in the technology stack than what you are currently using. That could means eliminating a dependency, a more primitive language, or merely rolling back a shallow layer of abstraction. The reason why this works is because forcing a more primitive approach requires elimination of some convenience. This isn’t a technology problem. It’s a people (discipline) problem, so you have to impose the means to require a more disciplined participation. Yes, this will piss people off, because it’s a people problem, but it will solve the problem and result in a superior product.