Re (tilly) 1: Atoms as a concept for programming analysis

[tilly]

I like the idea of atoms in that it captures the point that functions should be small and simple.

But I really think it is key that a good programming model shows a good conceptual model which is going to be well-compressed. Among other details, that points out not only why you factor code, but also why you avoid repeating it.

[dragonchild]

I'm not sure why the 'but' ... an atom isn't going to be repeated because it already exists. You're going to reuse an atom every time you can. (But, then again, I'm assuming logic will be used here. *shrugs*)

------
We are the carpenters and bricklayers of the Information Age.

Don't go borrowing trouble. For programmers, this means Worry only about what you need to implement.

[tilly]

The reason for the but is that if you are trying to merely decompose into atoms, then you have no reason to aim to create atoms that are maximally reusable in creating other atoms. Indeed not creating them makes for more local simplicity.

When reusability conflicts with local simplicity (and they very often do), saying that reusability generally wins is based on some value system. Without a value system to reason from, your conclusion is not just logic. I think that the concept of intellectual compression captures the value system quite well. I think that the atomic hypothesis captures what you do, but misses why you do it, and cannot resolve that conflict.