Talk:Last Friday of each month: Difference between revisions

I don't have a dog in that fight (as pertaining to Java code and how it interprets/presents/discombobulates/converts Gregorian and Julian dates). But, as far as I can interpret from the quoted text from the documentation about the Object that the Java code is using, it states that the Gregorian calendar is a hybrid calendar. It isn't. (Possibly, it meant the way Java treats dates is a hybrid system?) Possibly, the Java code treats dates after a "switchover" (there were many switchovers, depending on the state/country) as Gregorian, and Julian before that.

If the Java code supports Julian '''and''' Gregorian as an option (as separate function/subroutine options), that would seem to be OK. But the quoted statement says it supports a single discontinuity with an assumed date of when the Gregorian calendar was adopted by most of the (Holy Roman) Catholic world back in 1582, then it appears that the Java code writer(s) think that the Gregorian calendar isn't proleptic [proleptic essentially means that it extends backwards in time with no discontinuity back to January 1st, year 1], and the Java code then (it appears to me) switches back to the Julian year method. If that is true, then the Java code would have to use the Julian leapyear rules, and also that some Julian new years start on March 25th (Lady Day) [in some countries], not January 1st. To check this, see what year George Washington (USA president) was born in the O.S.); if Java says 1731, then that's the correct year for the Old Style (Julian) format.