bug#36918: 26.2.90; calendar: the chinese sexagesimal name for month is wrong

[Basil L. Contovounesios]

Zhang Haijun <ccsmile2008@outlook.com> writes:

> The sexagesimal name for month today should be Xin-Wei not
> Ren-Shen. It will becomes Ren-Shen after 4 days(on 8/8. which is the
> beginning of autumn). As far as I known, It always changes at the
> monment when the ecliptic longitude is 15, 45, 75 … (for every 30
> degrees).

Sorry, I am not very familiar with either the Chinese calendar or the
Calendar app, but I think the following points might be of interest.

The cal-china.el package contains the following disclaimer in its
commentary:

  The rules used for the Chinese calendar are those of Baolin Liu
  (see L. E. Doggett's article "Calendars" in the Explanatory
  Supplement to the Astronomical Almanac, second edition, 1992) for
  the calendar as revised at the beginning of the Qing dynasty in
  1644.  The nature of the astronomical calculations is such that
  precise calculations cannot be made without great expense in time,
  so that the calendars produced may not agree perfectly with
  published tables--but no two pairs of published tables agree
  perfectly either!  Liu's rules produce a calendar for 2033 which is
  not accepted by all authorities.  The date of Chinese New Year is
  correct from 1644-2051.

p C (calendar-chinese-print-date) describes 2019-08-05 as follows:

  Cycle 78, year 36 (Ji-Hai), month 7 (Ren-Shen), day 5 (Jia-Xu)

and a quick online search revealed two websites[1][2] which give a
similar description:

  Ji Hai Year (4717), Ren Shen Month (7), Jia Xu Day (5)

[1]: https://www.travelchinaguide.com/intro/focus/calendar.htm
[2]: https://www.prokerala.com/general/calendar/chinese-year-converter.php

So to an ignorant person like me, there is nothing obviously wrong with
Emacs' calculations.  Perhaps someone else can shine some more light on
this.

Thanks,

-- 
Basil