Make your own solar system
April 7, 2014 1:43 PM   Subscribe

The Kepler mission has changed the way we think about extrasolar planets and their abundance. It turns out that nature produces a bewildering variety of planetary systems, each in their own infinite majesty. But maybe, just maybe, you can do better?

(Bonus: this time sink is brought to you by a grant from the National Science Foundation, and the codebase is open source. And of course, much previously omitted.)
posted by RedOrGreen (24 comments total) 19 users marked this as a favorite
why arent their orbits elliptical this is hurting me
posted by Apocryphon at 2:02 PM on April 7, 2014 [4 favorites]

1 million points over 68 years.

bigger planets = more points.
posted by garlic at 2:09 PM on April 7, 2014

why arent their orbits elliptical this is hurting me

Circles are degenerate ellipses . Circular orbits aren't impossible, they just make me feel dirty.

I'm mad there isn't a Ptolemaic epicycle option. Teach the controversy!
posted by justsomebodythatyouusedtoknow at 2:23 PM on April 7, 2014 [5 favorites]

So I tried to do the planet X bit of two heaving things in the same orbit so that the sun was always between them, and about three seconds after putting B into position A goes flying out of bounds on the opposite side of the sun. So my score was for shit, but I discovered negative gravity. That counts for something, right?
posted by Kid Charlemagne at 2:26 PM on April 7, 2014 [1 favorite]

The Exoplanet App is pretty nifty. The author, Hanno Rein, is also responsible for the Open Exoplanet Catalogue, which among other things, lets you generate xkcd style bubble charts of known planets.
posted by zamboni at 2:45 PM on April 7, 2014 [1 favorite]

4 million points over 40 years. One dwarf star very, very close to the central sun, and a lot of earth-like planets in the (fluctuating) habitable zone.
posted by librosegretti at 2:51 PM on April 7, 2014 [1 favorite]

If you think this little web app is cool, try Universe Sandbox.
posted by General Tonic at 3:15 PM on April 7, 2014 [2 favorites]

But maybe, just maybe, you can do better?

My first thought was "A Kerbal Space Program upgrade is out?"
posted by Brandon Blatcher at 3:48 PM on April 7, 2014

The Exoplanet App is pretty nifty.

I'm sure it is, but all I ever do with it is load it up every few months and watch the "Confirmed exoplanets" figure go up by a bunch.

...Wow, this time it went from 982 to 1772. We sure live in interesting times.
posted by neckro23 at 4:10 PM on April 7, 2014

The game in the original post is fun, but is anyone else finding it a bit buggy in terms of submitting and loading scores? I just got 17 million points, and I was thrilled because my previous best was about 4 million, but the stupid thing wouldn't let me submit my score.
posted by erlking at 4:19 PM on April 7, 2014

Circular orbits aren't impossible, they just make me feel dirty.

Circular orbits rock. You're always at the periapsis! Burns are easy. (And, of course, you're always at the apoapsis as well!)
posted by eriko at 4:25 PM on April 7, 2014

Klemperer Rosette!
posted by lucidium at 4:28 PM on April 7, 2014

The orbits only start circular. If you want to see some ellipses, drop a dwarf sun into the simulation and see what happens.

It would be more fun if you could let the simulation run even after planets are ejected. And what's with the 2 AU barrier anyway? All the important planets that hold our own solar system together are way beyond 2 AU.
posted by localroger at 4:49 PM on April 7, 2014 [1 favorite]

Is there any way to remove that stupid earth that starts the simulation? It keeps getting ejected from my beautiful systems of giants.
posted by Hactar at 5:49 PM on April 7, 2014

If you check the mefi page source for the syntax of that Klemperer Rosette URL, you'll see that the entire solar system is defined in the URL, with a relatively simple syntax to set m, x, y, vx, and vy for each planet. This makes it possible to exactly specify all the parameters for all the planets without going through the web interface, which is undoubtably how the rosette was created.
posted by localroger at 5:53 PM on April 7, 2014

The orange planet (planet 4) is a jerk.
posted by aubilenon at 6:06 PM on April 7, 2014

URL hack to remove one of the giants from the Rosette. It's actually a lot of fun to watch this fall apart. By 10 years the symmetry is all blown but it took my sim 30 years (and this may vary due to butterfly wing flapping) for one of the runts to get kicked past the barrier.
posted by localroger at 6:18 PM on April 7, 2014

Binary star system.
posted by localroger at 6:32 PM on April 7, 2014 [2 favorites]

Also, if I had the patience to let it run to 500 years, that binary star system should smash the high score with over 200,000,000 points. And it could be tweaked more; since the half-solar-mass "planet" isn't a "star" it gets points for approaches to the habitability zone. It would probably be possible to hack the URL so it spends most of its time there. I'm also not sure why the barrier limit drops to 1.44 AU; if I increase Y1 from 1.15 to 1.5, the barrier goes from 1.44 to 1.88 AU. I'm not sure what's up with that.
posted by localroger at 6:42 PM on April 7, 2014

OK, how to obliterate the high score: Starting with the binary star system, drop a dwarf star in VERY CLOSE ORBIT around the "central" star. Even though it's not at the center any more because of the binary companion, the autostart logic gives it the right VX and VY for a circular orbit around the central star. The resulting three-body system will then be stable. I got to 150,000,000 points in 75 years before I stopped it because it's getting late. You get crowdedness bonus plus habitability bonus because the companions "star" spends a lot of its time in the habitability zone, and the sim thinks it's a planet.
posted by localroger at 7:55 PM on April 7, 2014 [2 favorites]

Ran localroger's set-up out to 500 years: 1,102,866,828 points.
posted by worldswalker at 8:38 PM on April 7, 2014

Wow somebody did something because the three high scores are now zero. I wonder if it's because the score display formatter can't handle a tenth digit.
posted by localroger at 4:44 AM on April 8, 2014

Wasn't me! It doesn't look like you can submit highscores from a custom system in any case, for obvious reasons. I popped the code I used up here if you'd like to play with the variables, just hit "Run" to get a new link. You can get some pretty (unstable) things if you mess with the starting velocity and angle.
posted by lucidium at 5:08 AM on April 8, 2014 [1 favorite]

« Older Most important product announcement that this...   |   Dangerous Newer »

This thread has been archived and is closed to new comments