Note: Accuracy in days is lost by about 7 AD backwards because of irregularity in leap years at the beginning of the Julian calendar, due to wrong decisions, and further back by other calendars such as the Roman, which is not uniform, if you can call it that. But the weekly cycle has never been changed. Excuses for some ancient inaccuracy. Additional note: The result of the remainder in days of the difference detailed in the last or fourth line of results may not make much sense, even on current dates due to the method used, and what one considers to make up a month at a particular time.
Updates and useful information (September 11, 2024):
You may receive some warnings from old dates on BC when searching for a february 29 date. Expect inaccuracies on calculations (difference between dates) that involve those dates, but the navigation through the months from 45 BC to 7 AD match the historic weekday, this was achieved now to my better understanding. It is not easy to avoid inaccuracies when trying to have a calendar with both calendar systems and with irregular events at the beginning.
This Calendar considers the following historic events or data on dates.
1. The absence of a Year 0 (from BC to AD)
2. The irregular Julian Leap years are the following: BC Leap Years = [45, 42, 39, 36, 33, 30, 27, 24, 21, 18, 15, 12, 9]; AD Leap Years = [8, 12]; and from 12 AD and on, it is mathematical calculation.
3. The change from Julian to Gregorian calendar from Thursday October 4, 1582 to Friday October 15, 1582
All of this was built over and around the script moment.min.js 2.29.1 that works only with real and proleptic Gregorian calendar.