Computing the Eclipse | Hacker News Reader

Computing the Eclipse

writings.stephenwolfram.com

96 pointslawrencechen25 comments

Putting aside the Wolfram Language related content, this is a really interesting introduction to the mechanics of solar eclipses. The notion of a "synodic" vs "draconic" really clears up a lot of my questions about why there are not eclipses during every new moon.

One thing I was wondering from looking at this is whether there is a good map projection that preserves the shape of the "solar disc" (my term) that represents the side of the earth illuminated by the sun, weighted by the solid angle from the view of the sun, so that eclipse paths would look like straight lines of constant width.

Such a projection would be possible-ish, but it would need a different centering depending on the eclipse. The NASA solar eclipse maps (these are the ones that you’ll see in the Wikipedia articles on the eclipses) kind of hint at in that direction.

Are all names in Mathematica really all together in a single namespace? Seems odd not to have such a specific function like SolarEclipse behind some kind of importable module.

And how do I discover what functions I want?

https://reference.wolfram.com/language/ref/SolarEclipse

I mean when you run it it downloads it.

> Are all names in Mathematica really all together in a single namespace?

No, there are package namespaces (Contexts) like most languages. When you import a package (call Needs) it adds that package’s namespace to the namespace path ($ContextPath).

I heard somewhere that the Spanish managed to wow the people they met in the new world with an eclipse.

Is this true? What calculations did they do to figure this out with 16th century math?

Christopher Columbus did it to terrorize people into giving him food (lunar eclipse, not solar): https://en.wikipedia.org/wiki/March_1504_lunar_eclipse

Geez, what an ass hole.

He did alright.

If they were familiar with Ptoloemy's almagest, they would be able to predict eclipses using geometric methods. Not perfect but surprisingly accurate according to Jayant Shah in the paper "Accuracy of Ptolemy’s Almagest in predicting solar eclipses"

More accurate methods using Newton's theories were yet to come.

A bit tangential, but this was one of the main ways that the Jesuits were able to gain access to the Imperial Court in Ming Era China. Predicting solar eclipses was one of the most important functions of the royal astronomers. However from the late Song Dynasty and into the Ming Dynasty there was some bureaucratic decay in the Astronomical Bureau and the quality of eclipse predictions declined. When the Jesuits arrived they were able to make superior eclipse predictions and leveraged that to gain appointments to the astronomical bureau. (And they were so successful at this that the Bureau was run by a Jesuit for a century and a half.)

Incidentally, this didn't work everywhere Europeans went. When Guillaume Le Gentil arrived in southern India in the 18th century he found that Tamil eclipse prediction techniques were superior to the Western methods. (His own prediction was off by 68 seconds whereas the Tamil prediction was off by 41 seconds.)

Lunar, not solar. No one worked out how to accurately predict the time and, more importantly, location of solar eclipses until Halley in 1715. [1]

1. https://www.quantamagazine.org/how-the-ancient-art-of-eclips...

Weird coincidence: while driving home from the eclipse yesterday, at one point I was right next to the Wolfram building in Champaign.

I wonder if it would be easy to figure out, given a set of coordinates, if there would be a totality in the next 100 years there and when. I cant figure it from the examples but that would be a neat website.

The UI for Wolfram is SO much prettier than Jupyter notebooks.

Stephen Wolfram is a national treasure.

This would be really cool if it wasn't a thinly disguised Wolfram sales pitch.

Why care? It's his life's project, and he's giving useful information to boot. And Mathematica is a wonder.

When in Rome do as romans do. That means keep it in the neutral language of the subject, which is not wolfram language despite his ego thinking otherwise.

What would be the "neutral language of the subject" for the material he covers in the article?

Considering that "Rome" here is Wolfram's personal blog, I think he is doing exactly as Romans do.

Reply on news.ycombinator.com