My first contribution to https://exercism.org got accepted: An improvement of some tests which are now using `assert_in_delta` for floating point numbers instead of `assert_equals`.

I noticed that there is an issue because a student I was mentoring had to round the result of a multiplication to make the tests pass, while my one solution didn't need this. For that particular exercise no solution should need rounding.

Remarkable: The only difference was the order of the operands in a rather simple multiplication: `x * y * z` resulted in a slightly different number than `y * x * z`.

Take care when processing flotation point numbers.