Question

What is the correct way to implement the standard behaviour of __new__ in Python so that no functionality is broken?

I used

class Test:
    def __new__(cls, *args, **kwargs):
        return object.__new__(cls, *args, **kwargs)

t=Test()

which on some Python versions throws DepreciationWarnings. On the internet I had seen something with super() or with type(). What are the differences and which is prefered?

Answer 1

You should write

return super(Test, cls).__new__(cls, *args, **kwargs)

This is recommended by the documentation (and the same for 2.x).

The reason to use super is as always to cope with inheritance tree linearisation; you don’t know for certain that the appropriate superclass is object, so you should always use super.