On 2012 Feb 20, at 02:09 , zengyiping wrote:
Irv replies: Zengyiping has quoted an old version of my web page. I was last changed one month ago.
But I can't agree any leap week calendar,because it breaks the 365d year and insolation seasons.
Irv replies: I intentionally avoided having any equinox or solstice start the year or any month because the 7-day jitter of a leap week calendar causes those events to wobble ±3.5 days (at best) from the target.
5, Leap rule need improve. there are two choices :
a.leap rule is the same as Gregorian but add if the year number divided by 3200 deduce a leap year .
b.year number bigger than 2000 and divided by 128 deduce a leap year.
Irv replies: I don't understand Zengyiping's proposal (English grammar problem). I believe that he wants to introduce "corrections" at very long intervals, but these make no sense due to long-term astronomical changes that are ignored. I have intentionally avoided using such long calendar cycles. The 293-year cycle yields the desired mean year and its short cycle length is practical and useful. It is not clear what is Zengyiping's target calendar mean year, but I'm guessing it is the mean tropical year, which makes no sense if he wants the start of the insolating spring season to start the calendar year. It would be best to find a leap cycle that has its "primary calendar season" at February 4th, or more specifically at a solar longitude of 360° - 45° = 315°, as explained at:
The leap day cycles that bracket this target solar longitude are 108/445, with its primary calendar season at 315.6°, and 133/548, with its primary calendar season at 314.6°. Alternating these cycles gives the 241/993 cycle, with its calendar season very close to 315°.
Even if such a leap cycle is selected, there will still be problems with the equinoxes and solstices, because they are not evenly spaced and have mean years that differ from that of a near February 4th calendar season cycle. Using uniform month lengths of 30 days won't provide consistent positioning of equinoxes or solstices, and their positions will semi-independently drift as the centuries pass, because they all have different mean years that are changing at different rates, as explained at:
(I have not evaluated 5-day leap week cycles because Zengyiping seems to want to minimize equinox and solstice jitter. With 5-day weeks there would be 6 weeks = 5 x 6 = 30 days per month x 12 months = 360 days, requiring an extra week in common years and two extra weeks in leap years.)
-- Dr. Irv Bromberg, University of Toronto, Canada