New idea for planet shapes.

[fr0stbyte124]

I discussed it a bit on the main thread, but I want to get everyone else's oppinion.

The gist is that is might be possible for cube planets to work, even at the corners. The problem with the cube model has always been that gravity will behave funny at the corners and by conforming to an axis-aligned player orientation, whatever it may be, we pretty much guarantee that the change in orientation at the corners will be entirely nonintuitive except in the case where there is a perfect convex lip right on the diagonal.

The way we might be able to get around this is by mostly preserving orientation on an entity on the ground, but having the force of gravity be angular while the entity is in freefall. The stronger the gravity one way or the other, the more influenced your freefall orientation becomes the longer you are in the air. The ultimate goal is that you should be able to viscerally determine the rough orientation of gravity without it greatly affecting gameplay and be able to stand on the blockface of your choosing depending on how you move, preferably in an intuitive fashion, or at least with a little practice.


The benefits of this make it seriously work considering. For one, with a cube world, we don't have to worry about projections or distortions, and the rendering will look a lot cleaner. We'll also be able to do more with the interior of a planet, such as drill a hole from one side to the other. This type of gravity would also most certainly be good to work out about smaller vessels with artificial gravity. For the most part it will all be uniform, but how cool would it be to have artificial gravity produced via spinning torus?

[hyperlite]

Well, this is back to normal.

What?

[fr0stbyte124]

Okay, you remember back when we had that big debate over projected spheres vs. solid cubes for planet shapes? Spheres won because we couldn't come up with an intuitive ruleset for how to handle gravity at the edges. It can't be an instant flip at the corner because sometimes you will have concave corners and they will behave opposite of how you expect. Plus there is no indication as to how close you are to the flipping point until you cross it, which will feel arbitrary and obtuse to a player.

Though still very much a WIP, this might be the start of a working ruleset. If we can make it work, it will solve a lot of the really major obstacles regarding game mechanics. We have to try and make it work if we can.

Coppied from the main thread, here is what I have so far:

-Gravity is normal away from the edges.
-At the edges, so long as you keep walking on a flat surface, you can cross the diagonal an not change orientation for some threshold distance.
-Jumping near the diagonal will cause you to fall not straight down, but more towards the planet core. The effect becomes stronger the closer you get to the diagonal.
-As you fall, your body starts to orient to the direction you are falling. Reorientation to block alignment can be smoothed in the ticks before you land an adjustable amount.
-Falling while oriented naturally with the side of the diagonal you are on produces less orientation change. Falling opposite natural orientation produces much more change.
-Whether you stand or slide when you hit a surface depends on how strongly you were oriented before you landed.
-There will be some sort of visual cue near the edge indicating that the gravity is much more angular than normal. A slightly tilted horizon might be sufficient (you do roll with the camera, so any changes to the horizon should be pretty noticeable.)
-There needs to be some mechanism for voluntarily changing orientation near the diagonal. For instance jumping vs. walking off the edge, or how far you clear the edge when you jump. These rules can't be directly programmed; they have to be incidental from the previous rules and will help to determine how to tweak the settings.
-If falling is angular, then jumping must be angular too. This angle should not be directly opposite of the gravity vector, or the pull won't be noticeable. Instead the jumping vector should be angled to the same degree the horizon is,
-Stairs will make things awfully interesting.

So give it some thought. I am looking for input and I honestly don't know what the right answer is this time.

[Tunnelthunder]

That could work as long as you have it so I can't mess the whole thing by building a large wall at certain points or forming a mass larger than what is being rendered below me.

[҉]

That could work as long as you have it so I can't mess the whole thing by building a large wall at certain points or forming a mass larger than what is being rendered below me.

I very highly doubt that gravity will actually be tied to mass.

[Tunnelthunder]

I am thinking that it will need something to latch to and I was thinking that "Hey minecraft is rendered by proximity, so if gravity is attached to what is rendered, then all I need to do is create something larger then what is rendered below me and gravity will flip!" That is most likely a run-on sentence but grammar is not in my skill set, spelin isnit ether.

[҉]

I am thinking that it will need something to latch to and I was thinking that "Hey minecraft is rendered by proximity, so if gravity is attached to what is rendered, then all I need to do is create something larger then what is rendered below me and gravity will flip!" That is most likely a run-on sentence but grammar is not in my skill set, spelin isnit ether.

I don't think it will work like that. Gravity probably won't be attached to blocks. It will be attached to planets, which are magic and which will continue to pull you down with the same acceleration regardless of how much mass they have.

[Tunnelthunder]

YAY MAGIC! Good luck frost. I hope you understand magic.

[Shadowcatbot]

But then if you dug a tunnel straight through the very center of the planet, Would that mean you would fall half the way then get stuck in the middle?

I smell best trap ever...

[Tunnelthunder]

Well I think that momentum would carry you back and forth for a bit, and the chance of you going exact enough not to crash into a wall or be sucked to one side is slim to none.

[fr0stbyte124]

Yeah, pretty much magic is the way gravity is going to work. The scales are so different, there is no way real physics are going to work here.

But what does everyone think about the rules for crossing edges? If you walked off the edge of a block on the threshold, would you expect to fall a distance, or would you curve around and start walking on another face? What about jumping? Would it make more sense to change directions by jumping or would you expect to fall straight?

[Tunnelthunder]

It just sounds choppy. Its like "weeee" "ooof" "ow" I must be on the other side of the world...

[Iv121]

The nature of minecraft suggests square yet making a square will make it nightmare on the edges. I already suggested to have a plain minecraft world linked to a round shape, when you enter the round shape you are teleported to the plain normal minecraft world.

[fr0stbyte124]

Yes, that's been the plan since forever. However, there are some impossible to avoid issues with it, like distortion and seams which require teleporting around and concealing. It's much more complicated to render, and will necessarily warp as you change altitude. You won't be able to dig too deep or build too high without terrain looking significantly wider or narrower than it should, respectively.

This cube model can work in real space and therefore solve all those problems, which would be amazing if we can make it work.

[Prototype]

Cube model sounds good to me, while a sphere might look better it would be a nightmare to implement, and would probably end up more distorted than the cube