Hey everyone!
I’m trying to generate a Dungeon Keeper based map, filled with cubes. I’ve been reading up on the process though, and as I understand it the process of creating a grid map can (for some reason) be very resource intensive. Therefor, I’m trying to be conscious of performance right from the beginning. Would it be significantly better in performance if I allow the first visible row of cubes to be full cubes and the rest of the map be ceiling planes?
My approach (currently just theory) is to have a script that will store an array of tile GameObjects. Each tile object contains a floor plane, a ceiling plane, and a full cube. The visible walls will be full rendered cubes and the rest only render the ceiling tiles. As the player digs through these full cube tiles, the game stops rendering the full cube and renders a floor plane instead.
As I understand it from reading through these forums, the creation of collider boxes is what really eats up computational time. So if I make sure only the visible wall and the floor tiles have collision (so you don’t run into the wall and you can stand on the floor), would that make the initial load time significantly faster?
And finally, I am concerned about the load time involved in loading an entire tile object with three different objects inside, even if two are not actively rendered. Should I do that, or would a more effective option be to store the data on what type of tile I need in a simple array and then load each individual game object as I need it?
I’m probably going to have several questions as the process continues, so I’ll just update this thread as I think of them. I really hope someone answers; as you can tell, I’m intensely new at this.