What is wrong with this float?
1230.00089F
There are nine decimal places of precision.
Data type float can't handle that.
(The compiler will round the number into a value that can fit
in a float).
You might wish to argue that there are only five places used in the above number: the places used by the digits 1, 2, 3, 8, and 9. However, the four 0's in the middle do count. It takes bits to represent them, even if they are zeros.
Primitive data type double uses 64 bits, and has a much greater
range, -1.7E+308 to +1.7E+308.
It also has a much greater precision: about 15 significant decimal digits.
Because of this, if you write a literal like 2.345 in a Java program, it will automatically be regarded as a double, even though a float might be good enough. The other numbers in the program might need to be double, so we might as well make them all double.
Do you suspect that characters are important enough to be one of the eight primitive data types?