Simplify gmtime_r() day-of-year calculation, correct comment
In the refactored gmtime_r()
(!309 (merged)), I made a wrong assumption that the day-of-year calculation for dates in Jan/Feb was sensitive to the previous year being a leap year (it is not).
Only the comment text was wrong as such, the implementation was functionally correct. Even so, it can be simplified a bit as submitted here.