Thread: c++ to C#

What's better? Well, I think it's largely a matter of taste. It has advantages and disadvantages.

* By limiting initialization to the constructor, you always have all initializations in one place. That's a good thing.
* By allowing inline initialization, you reduce the number of initializations in the constructor. If you have multiple constructors, not only do you save multiple initializations, you also don't need to ensure that they stay identical.
* In C++, objects embedded directly are automatically initialized with the default constructor. In C#, you have to instantiate every object explicitly. The higher number of necessary initializations in C# probably tip the balance in favour of the reduced-typing solution.

That is, in C++ you'd have

Code:

class A
{
private:
  StringBuilder s;
};

and it would have the same effect as your C# code.