Planet Diversification

The primary issue I can see with rimmed craters is pathfinding. Currently I believe units will completely refuse to move past a 45+ degree angle, however this may simply be part of the way the costmaps work.

It'd be interesting if Bots could just trundle over the rim into the crater while Tanks would either be entirely incapable of entering the crater or would have to slowly crawl over the top, flattening the rim as it went.

Edit: Even though the only erosion the moon experiences is solar radiation/solar wind, it'd b e nice if craters were smoothed a bit after having a rim applied to them.

There is also the issue of what happens when you stop using spheres. For example if you used a rectangle and did the same process the 'long' sides of the rectangle would have less of a ridge than the short sides. and of course the more complex the shape well....it gets more complex.

It's not hopeless mind you, all that it requires is 2 handcrafted shapes, one to create the uplift and a second to create the crater but given the constraints Uber has to work under that doesn't seem likely for release.

Mike

I... Don't get what you're trying to say mike. I presume you watched the livestream where Uber explained how CSG brushes are made and how they are implemented and designed.

Yes, but I was more so commenting on this idea.

Good point Mars, that was actually another thing I was thinking about, having some variability/stretch to additive and subtractive meshes. So one canyon will be shallow, while another deep because their boolean meshes would be different. Craters would only have variance in the Z (depth) axis, maybe a tiny bit in the X and Y, while canyons and that sort of thing will have more variance in the X and Y directions (because of the way canyons form).

There appears to be some mesh scaling already, with some craters larger than others. This would just be partial scaling in certain directions, instead of all 3. They'll be lower and upper limits to this, so you don't have canyons 5 ft deep, or 5 miles deep. With craters of varying depth, and crater rims of varying steepness, bots and tanks may find it easier to climb some than others.

As far as having a more smooth crater rim, I think it depends on the subtractive mesh. If the crater mesh tapered outward near the top, then it wouldn't be so jagged I think. With erosion, I think it would be cool if canyons and sinkholes and that sort of thing got filled in over time with the passing of the seasons, while others get gradually deepened. It may be hard to do this, but just a mesh that fills in the bottom of a canyon as time passes is like having water or lava fill a crater, which is something I think they said they were going to do. It might even be easier because it would just be a solid mesh moving upwards, instead of anything flowing. It's hard to explain, I'll post a diagram later, but I don't think it would be all that difficult to do.

Are you referring to the livestream when they showed the metal planet? I'm not really sure what you mean by them having less of a ridge on the long side. I'm trying to picture a brick shaped mesh carved out of a spherical mesh, and yes, if it were arranged so the long side was parallel to the equator, the shorter sides would have less of a ridge, but I don't think that's how it works.

I remember seeing a blog post a few weeks ago where a dev explained how canyons and that sort of thing can wrap around planets of varying sizes. If I remember correctly, a canyon is made up of many rectangular and triangular prisms that are angled accordingly to contour to the surface of the planet. Does anyone else know what I'm talking about? If you remember where it came from, please post a link.

So, rectangles aren't really single rectangular meshes, but many smaller meshes that contour around the circumference of a planet, almost like an integration. I think, lol.

Both the hill elevation and the crater boolean mesh would be round/spherical, so I don't think they would run into this issue. I mean, in the planet maker, when you adjust the 'elevation' of a planet to make it lumpy, you're creating the same kind of hill you would need for a rimmed crater. A spherical subtraction would also have omnidirectional rim depth if it were carved out of a sphere of any size larger than itself.

Ahh, I see what you guys are talking about. Scaling on a specific axis is often not present in games because of the issues it can cause. As mike said, you can't scale a specific part of the geometry, you have to scale the ENTIRE MESH, which would result in a gigantic crater lip if a crater was scaled on the X axis only.

In games, scale is generally uniform across all axes. Games are often architectured this way because of how infrequently you will actually have to scale on a specific axis.

It happens a lot when MAKING models, but not during a game.

Knight, I found this on Mavor's Rants: http://www.mavorsrants.com/2013/02/planetary-annihilation-engine.html

I think this is what you were talking about, like in this picture:

The corner's of the square subtraction have shallower depth than the middle of each wall. I can see how this would be problematic if the subtracting mesh weren't round, or elongated some way to make this depth difference more apparent, but that's not how it works.

The process of adding an subtracting is explained in this blog post: http://allenchou.net/2013/07/bending-solid-geometry-in-planetary-annihilation/

An additive/subtractive mesh is arched around the sphere according to it's size. This allows for uniform depth rectangular trenches to be cut out of metal worlds. What I am proposing is simply to elevate or deepen the terrain mesh around additive/subtractive meshes for better blending, quite different than adding an additive mesh with a subtractive mesh, but the same principle. Also, as you can see from the picture above, there's a huge amount of tesselation around the subtracted area perimeter, particularly around the corners. I do not believe my suggestion would add more to the poly count than this process already does, in fact it may better use tris that area already a part of the mesh. In the above pictures, the tris are there, but are flush with the surface of the sphere (providing no added dimension), when they could be elevated like in my diagram, giving some added dimension.

Ohhhh, I see too, lol.

Well, I was hoping the terrain elevation would match with the dimensions of the additive/subtractive meshes.

In my initial proposal, the terrain elevation would match the dimensions in the x, y, and z directions already, so if the mesh were distorted in some way too, than the terrain deformation would be scaled accordingly. I also suggested this elevation for craters because they have uniform depth, and do not need to be arched. Also, I said that there should. only be a slight scaling in the x and y directions. For canyons and mountains and that sort of thing, I think it would still be possible. It's hard to explain, I make diagram.

ethannio, the point I'm making is that you can't just uniformly scale something that isn't uniform to do your suggested process.

Mike

I do like the idea of easily adjustable colours allow for creating a much more Alien planet than the current standard range of earth like tones.

Are we all talking to each other from different universes or something? I feel like I missed half the conversation.

MushrooMars, preface posts with @ethannino, or @knight to avoid confusion. For a round boolean mesh elongated in the x and y, arching can be added just like it's already in use for canyons. I definitely think it's possible.

@knight: In most map makers with deformable terrain, you can use paint brush tool to 'paint' elevation on the playing field. This is similar to what I suggest, except the brush would be scaled along with the boolean mesh. EX, for a round crater, the elevation brush would be round, for a square crater, the brush would be square, for an elongated square, the brush would also be elongated the same amount. The brush would be applied normal to the sphere, this differs from classic map makers which have flat terrain. I /think/ the number of subdivisions in a mesh (for arching), depends on the size of the planet vs the size of the mesh. So a large additive/subtractive mesh on a large planet will have fewer subdivisions than a large mesh on a small planet. In the diagram I made, I only showed arching planes along one axis, but I'm pretty sure meshes are arched in the X and Y directions.

At any rate, I said there shouldn't be that much X & Y distortion for craters, so I don't see what the problem would be. Terrain elevation is normal to the surface, just like arching, so they should match up no matter what shape the mesh is.

The problem is that Non-Uniform shapes don't scale nicely which in turn requires 2 unique shapes/brushes/whatever that have to act as 2 parts of a whole, you couldn't mix and match.

Mike

Actually if I remember correctly, the "arching" of a brush is perpendicular to the radius of the planet at the angle the brush is built at. Not the terrain's normal.

If it was based on the terrain normals we would probably get Frankenstein craters.

"Biome: Escher"

In any case, even some minor color palette variations would spice things up quite a lot.

Very cool thread.

Colour changes -> Sounds great and readily implementable

Strategically relevant differences in planets -> Simple 'multiply a certain thing by a factor' adjustments like gravity depending on size of planet and affecting range, unit speed etc would add a lot for not too much effort. Different yield from solar collectors depending on distance from the sun would make them the primary choice near to the sun, and almost irrelevant further out.

Seasons etc -> Sound great but I'd doubt there would be time for that type of thing. Having the temperature of biomes depend on distance from the sun would be clever, though I'm sure if we can create and save systems, the good ones will keep that type of thing in mind and put desert planets near the sun and icy ones further out.

I would like to see some of the Biome from Total Annihilation to come over. I know that Water, Metal, Earth, Lava, Ice, Lush (Tropical) and Moon have been transfered. But I would like to see Acid, Crystal, Mars (Red Planet) and Urban style Biome Planets as well. See the link below for the Tilesets TA used.

http://www.rakrent.com/rtsc/html/ta3-01.htm

There are non uniform shapes that scale already in game, like canyons. They scale the same in the x, y, & z directions, so they're either larger or smaller. My suggestion is to have the scaling in every direction, but also randomly slightly more in some directions. Nothing too extreme, maybe a minimum of 75% the width/height as a normal one to make it look like a different canyon. The elevation painter would be made from the custom made cutting mesh I think.

However arching is done, if it can apply to complex geometric meshes like branching canyons, it should also apply to elonged spheres. I'm thinking the scaling would just to make them look like different brushes, to break up the uniformity a little.