4 ms·
I think programmers should need to know complexity theory, because if there is a "programmer" type job that can be done without knowing it, then we should be wo
by moron4hire 12y ago
I think programmers should need to know complexity theory, because if there is a "programmer" type job that can be done without knowing it, then we should be working hard to eliminate that job through automation. In other words, dissolve the class of programmers that supposedly don't need to know complexity theory.
Also, big-O is not the be-all, end-all of complexity theory. What a ludicrous sentiment.
- j2kun 12y agoUnfortunately, nobody has yet figured out how to automate the process of encoding arbitrary sets of business rules written by non-programmers into code.
- moron4hire 12y agoThe problem isn't that the tools don't exist. The problem is that the work is organized to make someone not-the-manager think about the details. It's not just "arbitrary" business rules that need to be encoded, it's "poorly defined, poorly thought-out" business rules. That's why SQL didn't take off as a lay-person interface to databases; even though it was meant to be a system anyone could use, you can't get around the fact that you still need to know what you want and what comprises what you want. Your typical middle-manager type knows neither. That's also why UML can't be converted to a program: if it did, it would call managers out on their inconsistent logic and the manager would just give up and make someone else do it. The fact that it doesn't compile is a feature, one that enables UML to make managers feel like they are being a part of the technical process. It's not a technical problem, it's a cultural one.