NavList:

Frank,

You wrote:

"... the whole world switched over [from the Julian to Gregorian calendar] in 1582. That's how it was spelled out in the Explanatory Supplement and other resources. But of course that just wasn't the case."

Indeed. That's one reason I also included a pop-up note that explains the way my spreadsheet handles dates. I even went so far as to display an error message (referencing the note) if the user were to input a date between the 4th and 15th of October, 1582. However, the note doesn't mention the difference in the handling of leap days between calendars. But I suspect that, in this day and age, anyone who understands what the Gregorian and Julian calendars are would probably also know the difference. And those who don't, but are using my spreadsheet for whatever reason, would be curious enough to find out for themselves. Maybe not ... IDK.

You continued ...

"Sean, your thinking --that you will now add calculations for every calendar on Earth ever invented in all of history-- will keep you busy until time itself comes to an end (ok, you didn't say you would add calculations for every calendar, but it's a slippery slope, and you're poised at the edge of the icy precipice!)."

Hahaha! Yes, you are certainly right that trying to add every calendar would be a Sisyphean task. I almost wrote in my last post that I was specifically thinking of the Islamic calendar because of the impact that Arabic astronomers had on the development and dissemination of astronomical knowledge. Beyond that, I might include a Buddhist lunar calendar just because I am Buddhist. But that's probably where I would draw the [arbitrary] line.

You also wrote:

"Sean, you mentioned implementing some polynomials for Delta-T. And they are NASA polynomials! OK, let's be clear here... They are Espenak polynomials, and sometimes Meeus polynomials, and sometimes Meeus-Espenak polynomials, but there's precious little "NASA" in them."

What I actually wrote was: "I used the polynomials from NASA's eclipse website." I didn't say that they were "NASA polynomials".  It's a minor difference in wording, but I think it's an important difference in meaning. The links I included in my original post point to the NASA website, which clearly states right at the top:

Five Millennium Canon of Solar Eclipses [Espenak and Meeus]

And I also wrote that:

"None of that really matters much, though. The problem only affected the years before AD 1, a time at which we can only guess what the value of Delta-T was based on dodgy eclipse reports. And the difference between years was only twenty-two seconds at the maximum, and that is only between the years 499 BC and 500 BC; the difference decreasing to ten seconds between the years 2 BC to 1 BC."

To be clear: I am not attempting to argue an appeal to authority. On the contrary, I'm trying to do the opposite: to make plain the shortcomings of my spreadsheet. But I agree with your point that the uncertainty should be made clear. And you can see the standard error of the tables on the NASA website if you click on the link to "table 1" ...

Past values of ΔT can be deduced from the historical records. In particular, hundreds of eclipse observations (both solar and lunar) were recorded in early European, Middle Eastern and Chinese annals, manuscripts, and canons. In spite of their relatively low precision, these data represent the only evidence for the value of ΔT prior to 1600 CE.

Anyway, all of this reminds me of a video I recently stumbled upon from Neil deGrasse Tyson. It was a light-hearted talk about leap days and their origins. Towards the end, Neil mentions the fact that most of us don't celebrate the "new year" at the exact moment of the completion of one full orbit of the Sun. But, for those of us in the Eastern U.S. time zone, we count down the moment to the [incorrect] second based on a time ball [which has no relation to its namesake]. This reminded me of several recent NavList discussions. Particularly, your very recent comment in the discussion about the BBC "pips" ...

In a few recent years, when I've been out at midnight on New Year's Eve, I have watched with studious interest as crowds celebrate the exact second of midnight (it's "3... 2... 1... Scream!!"). But of course, they don't celebrate the actual exact second. Despite all those near-perfect atomic clock repeaters that we all carry around, it's not the second of midnight, but rather the somewhat delayed broadcast --about 15 seconds delayed typically-- of the event in some public center, like Times Square in New York City. And that seems to be close to the level of "error" that mere mortals comfortably accept. The minutes matter. People care about getting the time right to the minute. They don't care about ten or fifteen seconds of net offset here and there.

Cheers!

Sean C.