| > The point of declarative programming is to declare what you want done. The runtime may turn it into greatness or crap, that's not really relevant. I didn't say the quality of the runtime was relevant. The point is that the runtime is free to achieve the goal you give it in any way it feels fit. In Python, the runtime isn't; it needs to ensure specific runtime guarantees. >> In your example, if you replace max() with a function that prints something, you'll see it's executed for each value in 'table', which is extremely inefficient. > It's also not relevant and an implementation detail, with a known max() on a known type the implementation would be free to lift the computation out since this expression doesn't have side-effects. No, it can't. The Python reference specifies the semantics of generator expressions, and it says "Only the outermost for-expression is evaluated immediately, the other expressions are deferred until the generator is run"[1]. A implementation that lifted the computation would not be implementing Python, but a derivative language. Furthermore, your example didn't specify neither the type of neither 'table' nor 'max', so I still think it's deceptive. > The runtime can know that all types and functions involved are their native side-effect-less versions and is free to optimize them if it wishes, that it does not is irrelevant. See above. According to the language spec, it can't. [1]: http://www.python.org/dev/peps/pep-0289/#the-details |