“后期约束力”与“控制权倒置”之间有什么关系？

Question

In his definition of OOP, Alan Kay points out he supports "the extreme late-binding of all things". Does his interest in late-binding share the same motivation as people's interest in IoC?

In particular, would it be correct to say that both are motivated by the concept "specify as little as possible, and leave implementation details to be filled in later"?

Answer 1

It depends what you mean by inversion of control - the term has been overloaded to include dependency injection, but they are really different concepts. IoC originally described a method of controlling program flow, whereas DI is specifically concerned with reducing coupling between types.

That said, it could be argued that all these methods/patterns/philosophies share the same fundamental principle: to lower the cost of change.