Update #11

    Alright, I'm still a little sick, but I feel well enough to write something for this post. I'm not entirely sure "Progress Update" is the right thing to call these posts, so I'm just going to start calling them "Update" since for the time being I'm not actually able to work on the project. I probably shouldn't actually even be allocating the time to make these posts, but it's important to me that I don't forget about this project since I really want to make this game.

Learning to Art:

    My drawings are starting to get somewhat alright, I've still got along way to go before I'm able to make anything fap worthy, but it's getting to the point where I'm seriously considering buying a scanner and sharing them.

Game Ideas:

    Ok, so new section that I'm not entirely sure what to call it, but this is going to be where I talk about things I want to do for my game. For this post I was hoping to delve a little bit into math, but I seem to have misplaced my notes, so we're just going to cover the basic idea. For this post I would like to continue talking about sky boxes, or more specifically that thing hello games promised for No Man's Sky but apparently couldn't figure out how to do, realistic orbiting celestial bodies.

    There are two ways to have realistic orbiting celestial bodies, the first of which is to make it part of the physics simulation like the way universe sandbox did. This will produce accurate results and will even allow bodies to interact with each other and affect one another's orbits, but is very cpu intensive, must be run continuously even when no one is there to see it, and can't handle large skips in time.

    The other way to do it (and the way I intend to) is use mathematical models (I don't remember which ones or I would specifically name them) to predict where an orbiting body will be at any given time, like how Kerbal Space Program dose. How this basically works is if you know the mass of the object being orbited, the mass of the orbiting object, and the initial position and vector of the orbiting object then you can predict where that object should be in it's orbit at any given time. The advantages of this is it requires very little cpu time compared to the physics based method, it only needs to run when it's needed (such as when the player is looking up or the game wants to know what season it is), and it can easily handle large skips in time both forward and back. The only real disadvantage of this is that celestial bodies won't really be able to interact with each other, but that shouldn't really be an issue as I don't intend to place them close enough to one another for them to interact.