In MacBSD, the command `cal 9 1752` shows the shortening of that month in the British Empire. If I reinstall MacOS and choose Italian as its default language, will the shift show up instead in October 1582?

Each month shall have 30 days, except within a lune spanning 157°15′57″ (5.242199 × 30°) of longitude, wherein the month shall be extended by one day which shall not affect the cycle of the week. The lune so affected shall shift westward(?) each month by its own width. The phase of the lune shall be such that, in every longitude, the northward equinox shall fall on the last day of March. (During a transition of roughly two years, each month shall have 30 days without leaps, to shift the equinox from March 21 to March 30/31.)

Friday thirteen come on Friday this month

Somewhere or other I recently mentioned having heard that, because the Gregorian calendar cycle of 400 years is a multiple of 7 days, the 13th of the month is not evenly distributed and falls more often on Friday than on any other day of the week; but I had not done the math myself and did not have the numbers. Now I’ve done it but can’t remember where to post the followup!

Average illumination near the Moon’s south pole, showing which crater floors never (or almost never) see sunlight. Unfortunately the text doesn’t quantify what the whitest pixel means, i.e., how much time the most-illuminated point spends in shadow.

Wobbling time exposure of Regulus and Mars, showing ‘twinkle’ in a novel way.

another thought that I’m never likely to have occasion to apply

If one has the luxury of designing a calendar from scratch, it might be good to put leap day at aphelion, where its angular value is least.

Dan Alderson once made a map of nearby stars by mounting little colored spheres on threads strung between holes in two sheets of heavy clear plastic.

It occurs to me that, taking the stars in pairs, he could use half as many threads; each would be oblique and therefore longer, but none would be twice as long as the straight threads.

Such a design would be error-prone in execution, and thread is cheap. But I think Dan would chuckle at the suggestion.

I got yet another wacky idea for a Martian calendar. Start with 24 months of 28 days each. Drop one day from every seventh month (so that a given month is short in one year out of seven), and add one day every 48 years. The result is longer than the mean tropical year by one day in 6176 years.

An analogous calendar for Earth: start with 12 months of 30 days, add 3 days to every 7 months (so the cycle is 30 30 31 30 31 30 31), and add one day every ten years; this is long by one day in 219130 years.