Brandon Rhodes mentions Nicolai Kopernik, a famous Polish scientist. He “lifted Earth into Heaven” by his book where he put the sun in the center of our universe instead of Earth. We’re no longer at the bottom.
The near-earth environment was pretty well mapped out. 300BC, the size of the spherical earth was already known. Around 100BC the distance to the moon was known.
But what about those planets? They were harder. Stars did move around the sky more or less linearly. But those planets. They seemed to move back and forth a bit. Did they need to have a different model? The sun-centric model was already known, but it didn’t catch on. The reason? Medieval science was too emperic: the earth cannot be moving, it seems to stay in place. Throw a ball and it falls down again towards the earth, it doesn’t career off into space. The church wasn’t totally idiotic by later convicting Galileo: there was just no solid emperical evidence :-)
One of the pieces of evidence that was missing was that there was no observable stellar parralax; visible movement between stars because of earth movement. It was only in 1838 that the instruments were good enough to actually observe parralax. Only then did we figure out how far away the stars were. So it took 200-300 year to get real evidence.
So why did Kopernik trust his theory despite the lack of evidence? Because of beauty. Brandon showed Kopernik’s theory as Python code; an algorithm for planet movements. And showed his theory as a code refactoring. The end result was quite simple and nice once you factored out Earth’s motion. The code looks nicely ordered.
There’s truth in beauty.
Brandon proposes a new term: Kopernican refactoring: making a system simpler by moving something new into the center.
It is so easy to behave according to a rule that’s not even really there. You look at a problem in a certain way and you’re stuck. Once you put something else in the center, you look at the problem in a different way.
For instance, you can use argparse to generate a help message when you pass
--help. You can also do it the other way around: using docopt which generates a parser from your help message! You
get a much nicer help message.
So if you’re stuck, think of Copernicus and his refactoring.
My name is Reinout van Rees and I work a lot with Python (programming language) and Django (website framework). I live in The Netherlands and I'm happily married to Annie van Rees-Kooiman.
Most of my website content is in my weblog. You can keep up to date by subscribing to the automatic feeds (for instance with Google reader):