A false midnight

I think that the main reason why other language doesn't do this is that reals have several possible representations: float32, float64, float intervals, rationals,etc which make choosing the 'real' type for the division's result quite arbitrary..