Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
Tags
(+1)

Pretty cool game, definitely a unique concept.

I’m a sucker for physics simulation, and n-body problems are tricky in general, although here I guess you’re getting away with it by keeping the number of planets small (as in, not thousands or millions). I’m not sure how physically accurate this implementation is but it doesn’t seem too bad.

Anyway, I had fun playing it and it’s an interesting take on the theme, could use a bit more polish in terms of the UI and other visuals, but it’s alright for a jam game.

The main actionable criticism I’d give, then, is that in the beginning it wasn’t really clear what the green and blue markers were and what the win conditions were. The guide does sort of explain it but it doesn’t really tie the rules together with the actual visuals on screen, and there’s a bit of a disconnect there.

It would also be nice being able to position the objects and set their velocities with more granularity, perhaps with some tools to aid with precise alignment (e.g. getting three equal mass objects to orbit each other a third of a full circle apart from each other at the right kind of tangential velocity to get them to orbit in a stable manner, not super easy to do with the current grid-based solution)

(+1)

Yeah one of my biggest regrets was not making checkpoints/goals clearer, especially since some require multiple bodies to hit them while others only need one. If i work on it in the future I'll definitely add a sort of start-of-level description like poly bridge has. 

I hadn't considered alternatives to grid/non grid position/velocity snapping but it would be super cool if once you placed a body the velocity could snap into place (maybe by holding control or some other button) to the nearest velocity that establishes a stable orbit. And then maybe an alternative grid that is radial rather than rectangular with the origin at the current center of mass of the system, or a given selected body or bodies.

And since you seemed interested in the implementation, I used a method i found described here: https://en.wikipedia.org/wiki/Numerical_model_of_the_Solar_System#Integration, where it seems better than a standard leapfrog integration, but not as expensive as Runge-Kutta. I also calculate all the bodies' updated positions first and then update them all after so that symmetrical setups don't break as quickly.

So it’s basically using the midpoint method? I’m not sure if RK4 is particularly expensive but it’s a bit more of a pain to implement than these explicit methods. Implicit methods are nice though because they handle numerically stiff problems without exploding, and can usually do much larger step sizes.

In any case, you guys did a pretty good job with it =D