Why must you use floating-point division when converting Fahrenheit to Celsius in Java?

"Try typing and running the program as defined above and you’ll find that you don’t get correct answers. In fact, you’ll find that no matter what values you enter, you get zero. This is because of the calculation using (5 / 9): recall what happens with integer division: truncation! This will always end up being zero.\n\nOne way we could fix it would be to pull out our calculators and find that 5/9 = 0.55555... and replace (5 / 9) with 0.555555. But, how many fives? It may be difficult to tell how accurate we can make this floating point number by hardcoding it ourselves. A much better approach would be to let the compiler take care of the optimal computation for us by making at least one of the numbers a double to prevent integer truncation. That is, we should instead use 5.0 / 9.\n\nThe full program can be found in Code Sample 26.3.\n\npackage unl.cse;\n\nimport java.util.Scanner;\n\npublic class TemperatureConverter {\n\npublic static void main(String args[]) {\n\ndouble fahrenheit, celsius;\n\nScanner s = new Scanner(System.in);\n\n//1. Prompt the user for input in Fahrenheit\n\nSystem.out.println(\"Please enter degrees in Fahrenheit: \");\n\n//2. Read the Fahrenheit value from the standard input\n\nfahrenheit = s.nextDouble();\n\n//3. Compute the degrees Celsius\n\ncelsius = (fahrenheit - 32) * 5.0 / 9;\n\n//4. Print the result to the user\n\nSystem.out.printf(\"%f Fahrenheit is %f Celsius\n\n\", fahrenheit, celsius);\n\n}\n\n}\n\nCode Sample 26.3.: Fahrenheit-to-Celsius Conversion Program in Java"