Best way to create a 2d top down race track procedurally - unity3d

I am attempting to create a 2d top-down car racing game. This game will have a random road map each time the player plays the game. I have thought about doing this in two different ways: A tilemap, or just generate the roads by placing different prefabs (straight roads, turns, etc). I have decided to go with the prefab route.
The way I believe it should work is to have prefab square "tiles" which have their own colliders set on the edges so I can tell if a player goes off the track in which case they blow up. I would have a MapGenerator Script which will generate an initial random map by keeping track of the last tile placed (including its location and road type: left turn, straight, right, etc). This script will then keep adding onto the road randomly as the player gets closer and closer to the end which makes it an infinite road.
I just want to know if this is just not efficient or if I am thinking of this completely wrong.
Here are a couple of images showing my road tiles which I made in photoshop and then one prefab for a straight road (take note of the colliders on its edges).
A similar game to one I want to make is Sling Drift which I can provide the link if you want. I don't know the policy on adding links to forum chat.
Also, here is my code for the map generator:
//Type of tyle, types are normal (straight road or horizontal road) and turns
public enum MapTileType
{
NORMAL,
N_E,
N_W,
S_E,
S_W
}
//structure for holding the last tile location and its type.
public struct TypedTileLocation
{
public TypedTileLocation(Vector2 pos, MapTileType tyleType)
{
m_tileType = tyleType;
m_position = pos;
}
public Vector2 m_position;
public MapTileType m_tileType;
}
public class MapGenerator : MonoBehaviour
{
//Map Tiles
public GameObject m_roadTile;
public GameObject m_turnNorthWestTile;
//holds all the tiles made in the game
private List<GameObject> m_allTiles;
//Map Tile Widths and Height
private float m_roadTileWidth, m_roadTileHeight;
//Used for generating next tile
TypedTileLocation m_lastTilePlaced;
private void Awake()
{
//store the initial beginning tile location (0,0)
m_lastTilePlaced = new TypedTileLocation(new Vector2(0,0), MapTileType.NORMAL);
//set height and width of tiles
m_roadTileWidth = m_roadTile.GetComponent<Renderer>().bounds.size.x;
m_roadTileHeight = m_roadTile.GetComponent<Renderer>().bounds.size.y;
m_allTiles = new List<GameObject>();
}
// Start is called before the first frame update
void Start()
{
SetupMap();
}
void SetupMap()
{
//starting at the beginning, just put a few tiles in straight before any turns occur
for (int i = 0; i < 6; ++i)
{
GameObject newTempTile = Instantiate(m_roadTile, new Vector2(0, m_roadTileHeight * i), Quaternion.identity);
m_lastTilePlaced.m_tileType = MapTileType.NORMAL;
m_lastTilePlaced.m_position.x = newTempTile.transform.position.x;
m_lastTilePlaced.m_position.y = newTempTile.transform.position.y;
m_allTiles.Add(newTempTile);
}
//now lets create a starter map of 100 road tiles (including turns and straigt-aways)
for (int i = 0; i < 100; ++i)
{
//first check if its time to create a turn. Maybe I'll randomly choose to either create a turn or not here
//draw either turn or straight road, if the tile was a turn decide which direction we are now going (N, W, E, S).
//this helps us determine which turns we can take next
//repeat this process.
}
}
void GenerateMoreMap()
{
//this will generate more map onto the already existing road and then will delete some of the others
}
// Update is called once per frame
void Update()
{
}
private void OnDrawGizmos()
{
}
}
Thanks!

Have you tried splines? They let you make curvy paths like race tracks easily. If not, here is a video that might help: https://www.youtube.com/watch?v=7j_BNf9s0jM.

Related

How to make Trajectory Predictor on my ball?

hey guys , So as you can see i made a robot arm grab a slingshot's objectholder with a ball in it. My arm pulls it any direction I want it but I wanted the user to know which box is going to be shot at.
If you're applying an impulse force (or velocity) to your ball and there is gravity in your world, your item will follow the Projectile motion;
Here you can find details about it:
https://en.wikipedia.org/wiki/Projectile_motion
There are basically two main options
calculating it yourself
This is probably way better for performance especially if you want a really simple trajectory preview without accounting for any collision etc
refer to linked article. But it basically comes down to
and would need slightly rework from 2D to 3D physics, should be trivial though since the important part about the Y axis basically stays the same.
You would
Call this simulation with according supposed shoot direction and velocity
Visualize the tracked positions e.g. in a LineRenderer
Physics.Simulate
This allows you to run physics updates manually all within a single frame and actually let the Physics engine handle it all for you
This costs of course a lot of performance but you get all collisions etc accounted for automatically without getting a headache
You would
Make a snapshot of all Rigid bodies in your scene - in order to reset after the simulation
Simulate the desired amount of physics steps (XY seconds ahead) while keeping track of the simulated data
reset everything to the state tracked in step 1
use the simulated data from step 2 to visualize e.g. with a LineRenderer
This might look somewhat like e.g.
public class Prediction : MonoBehaviour
{
public LineRenderer line;
public Rigidbody tracked;
private Rigidbody[] allRigidbodies;
private void Awake()
{
allRigidbodies = FindObjectsOfType<Rigidbody>();
}
private void LateUpdate()
{
// Wherever you would get this from
Vector3 wouldApplyForce;
// Step 1 - snapshot
// For simplicity reasons for now just the positions
// using some Linq magic
var originalPositions = allRigidbodies.ToDictionary(item => item, item => item.position);
// Step 2 - Simulate e.g. 2 seconds ahead
var trackedPositions = new Vector3 [(int) (2 / Time.fixedDeltaTime)];
Physics.autoSimulation = false;
tracked.AddForce(wouldApplyForce);
for(var i = 0; i < trackedPositions.Length; i++)
{
Physics.Simulate(Time.fixedDeltaTime);
trackedPositions[i] = tracked.position;
}
// Step 3 - reset
foreach (var kvp in originalPositions)
{
kvp.Key.position = kvp.Value;
}
Physics.autoSimulate = true;
// Step 4 - Visualize
line.positionCount = trackedPositions.Length;
line.SetPositions(trackedPositions);
}
}
Of course we won't talk about performance here ^^

How do i get the the tile in front of the player in unity 2d

I'm trying to get all tiles around the player in unity 2d tileset.
What I would like to happen when player presses U I Would like to get all tiles surrounding the player including the one underneath the player or simply the one 1 tile in front of them.
I'm trying to make a farming Game where when the player pulls out an item it will highlight on the tilemap where they are about to place it.
Please note I'm not asking for full code, I just want what ever solution allows me to get tiles near the players position
Edit, found a solution by editing someone elses code but im not sure if there's a better way, if not I would also like this to work based on player rotation currently it places a tile above the player. Code Source https://lukashermann.dev/writing/unity-highlight-tile-in-tilemap-on-mousever/
using System.Collections;
using System.Collections.Generic;
using Unity.Mathematics;
using UnityEngine;
using UnityEngine.Tilemaps;
public class PlayerGrid : MonoBehaviour
{
private Grid grid;
[SerializeField] private Tilemap interactiveMap = null;
[SerializeField] private Tilemap pathMap = null;
[SerializeField] private Tile hoverTile = null;
[SerializeField] private Tile pathTile = null;
public GameObject player;
private Vector3Int previousMousePos = new Vector3Int();
// Start is called before the first frame update
void Start()
{
grid = gameObject.GetComponent<Grid>();
}
// Update is called once per frame
void Update()
{
var px = Mathf.RoundToInt(player.transform.position.x);
var py = Mathf.RoundToInt(player.transform.position.y);
var tilePos = new Vector3Int(px, py, 0);
if (!tilePos.Equals(previousMousePos))
{
interactiveMap.SetTile(previousMousePos, null); // Remove old hoverTile
interactiveMap.SetTile(tilePos, hoverTile);
previousMousePos = tilePos;
}
// Left mouse click -> add path tile
if (Input.GetMouseButton(0))
{
pathMap.SetTile(tilePos, pathTile);
}
// Right mouse click -> remove path tile
if (Input.GetMouseButton(1))
{
pathMap.SetTile(tilePos, null);
}
}
}
try to use the layer field inside the inspector, you can create new Layers inside the Project Manager and there you can easily create Layers like "Background" "Scenery" "Player" "Foreground".. after this you can assign them individually to your Tiles also its important to have different grids otherwise you will not be able to assign different layers to it i hope it worked already

Converting mouse coordinates for ui in world space

Im having problems with position convertions. The way im trying to solve it may be very wrong but thats due to inexperience in that case and im up for any suggestion on how to do it differently.
What im trying to do is a gui with a dot graph envelope that the user can change by draging the dots with the mouse.
This is what i would wan it to look like.
https://imgur.com/FP6f1Cz
First i did the UI like normal in overlay but i couldnt get the line renderer to work so i took the whole ui into world space. This makes the line renderer visible. With the UI in world space ive tried both to put the envelope line renderer in the canvas with the rest of the ui and outside the canvas UI.
Here is the code that renders the lines where the dots are and moves the dots when the mouse drags them :
public class Envelope : MonoBehaviour
{
LineRenderer lineRenderer;
// Start is called before the first frame update
void Start()
{
lineRenderer = GetComponentInChildren<LineRenderer>();
}
// Update is called once per frame
void Update()
{
var points = GetComponentsInChildren<EnvelopePoint>().Select(ep => ep.transform.localPosition).ToArray();
lineRenderer.positionCount = points.Length;
lineRenderer.SetPositions(points);
}
}
public class EnvelopePoint : MonoBehaviour
{
[SerializeField] bool isHeld = false;
// Start is called before the first frame update
void Start()
{
}
// Update is called once per frame
void Update()
{
if (isHeld)
{
// Vector3 mousePos = Camera.main.ScreenToWorldPoint(Input.mousePosition);
Vector3 mousePos = Input.mousePosition;
transform.position = mousePos;
}
}
private void OnMouseDown()
{
isHeld = true;
}
private void OnMouseUp()
{
isHeld = false;
}
}
The best result is to put the envelope outside of the canvas.
The lines render well to where the points are but im not able to convert the mouse position to correct coordinates for the dots. When i click on a dot to drag it the dot snaps to a position a bit lower and a bit to the left of the mouse. Like this:
https://imgur.com/3KK6VD3
But then i can drag the dots and the lines adjust perfectly.
I guess i have two questions:
How should i get the mouse position conversion correctly?
Is this a strange or over complicated way of doing this? Is there a more reasonable way?
Id love some tip as well on what i should read up on to better understand the different screen types and how to convert between them.
RectTransformUtility.ScreenPointToWorldPointInRectangle: Transform a screen space point to a position in world space that is on the plane of the given RectTransform.
There is also ScreenPointToLocalPointInRectangle but since you are modifying Line Renderer's points (which are in world space), I think ScreenPointToWorldPointInRectangle best suits your needs.

Spawn sprites on action

I'm trying to make a Pinata GameObject, that when clicked bursts and gives a variable number of Gift GameObjects with various images and behaviors in them.
I'm also not sure what the unity vocabulary for this is so as to look this up in unity docs.
Can anyone please lend me a hand here? Thanks!
There are several ways to handle this.
The simple way is to use Object.Instantiate, Object Instantiation is the vocab you're after.
This will create a copy of a predefined Unity object, this can be a gameobject or any other object derived from UnityEngine.Object, check the docs for more info https://docs.unity3d.com/ScriptReference/Object.Instantiate.html.
In your case, your Pinata would have an array, or list, of prefabs. These prefabs are created by you with a certain behaviour and sprite for each one. When the Pinata bursts, you instantiate random prefabs at random positions surrounding the Pinata, up to you how to position these objects.
Something along these lines should do the trick:
class Pinata : Monobehaviour
{
public GameObject[] pickupPrefabs;
public int numberOfItemsToSpawn; //This can be random
//any other variables that influence spawning
//Other methods
public void Burst()
{
for(int i = 0; i < numberOfItemsToSpawn; i++)
{
//Length - 1 because the range is inclusive, may return
//the length of the array otherwise, and throw exceptions
int randomItem = Random.Range(0, pickupPrefabs.Length - 1);
GameObject pickup = (GameObject)Instantiate(pickupPrefabs[randomItem]);
pickup.transform.position = transform.position;
//the position can be randomised, you can also do other cool effects like apply an explosive force or something
}
}
}
Bare in mind, if you want the game to be consistent, then each behaviour prefab would have there own predefined sprite, this would not be randomised. The only thing randomised would be the spawning and positioning.
If you did want to randomise the sprites for the behaviours then you'd have to add this to the Pinata class:
public class Pinata : Monobehaviour
{
//An array of all possible sprites
public Sprite[] objectSprites;
public void Burst()
{
//the stuff I mentioned earlier
int randomSprite = Random.Range(0, objectSprites.Length - 1);
SpriteRenderer renderer = pickup.GetComponent<SpriteRenderer>();
//Set the sprite of the renderer to a random one
renderer.sprite = objectSprites[randomSprite];
float flip = Random.value;
//not essential, but can make it more random
if(flip > 0.5)
{
renderer.flipX = true;
}
}
}
You can use Unity random for all your random needs, https://docs.unity3d.com/ScriptReference/Random.html
Hopefully this'll lead you in the right direction.

GameObjected "sucked" into another GameObject effect while dragging

I am building a chemistry game where the user can drag around different atoms and connect them together to build molecules. I am able to get the atoms to "stick" to one another when they collide. However, the spheres are stuck just on the outside of each other. I want to get them to overlap slightly like the picture below.
Basically, I want the user to click and drag the mouse to drag atom spheres within the radius of other atom spheres. When that collision happens, I want the atom to be "sucked" into the other atom. However, the problem I am facing is that because the user is holding on to the atom with the mouse, any sort of movement is negated. I've tried each of the three lines of code below but none of them work in getting that atom to suck into the other one while the user is holding the mouse.
atom_entering.transform.position = Vector3.MoveTowards(atom_entering.transform.position, this_atom.transform.position, 5*Time.deltaTime);
atom_entering.GetComponent<Rigidbody>().MovePosition(this_atom.transform.position + (atom_entering.transform.position - this_atom.transform.position));
atom_entering.GetComponent<Rigidbody>().AddForce(this_atom.transform.position, ForceMode.Acceleration);
If I try to decrease the radius of each atom, it can work but then the atom starts clipping any other collisions like the table itself.
Edit: Asked to post complete code:
void OnCollisionEnter(Collision collision)
{
GameObject atom_entering = collision.gameObject;
GameObject this_atom = this.gameObject;
if(atom_entering.name == "HydrogenPrefab(Clone)" && num_hydrogens < max_hydrogens && !hydrogen_connected)
{
// Code to make the "suction" effect here
// One of the three lines above went here but none of them worked.
// First, make the hydrogen a child of this oxygen atom.
// Next, create a Fixed Joint component on the hydrogen and stick it to the oxygen
atom_entering.transform.parent = this_atom.transform;
atom_entering.AddComponent<FixedJoint>();
atom_entering.GetComponent<FixedJoint>().connectedBody = this_atom.GetComponent<Rigidbody>();
atom_entering.GetComponent<HydrogenCollider>().setConnectedStatus(true);
// Increment the number of hydrogens connected to the oxygen
num_hydrogens += 1;
}
}
In a separate C# script:
void OnMouseDrag()
{
this.GetComponent<Rigidbody>().isKinematic = true;
Ray ray = Camera.main.ScreenPointToRay(Input.mousePosition);
RaycastHit hit;
if(Physics.Raycast(ray, out hit, Camera.main.farClipPlane))
{
float oldY = this.transform.position.y;
this.transform.position = new Vector3(hit.point.x, oldY, hit.point.z);
}
this.GetComponent<Rigidbody>().isKinematic = false;
}
Edit: Here is a video of how it is currently behaving. https://vimeo.com/185896832
I want that same behavior expect I just want to get it "sucked into" the parent atom. If I do anything with the radius, the atoms sink into the floor a bit which is not what I want. It doesn't make sense to have an object with a normal radius when it interacts with the floor and a different radius when it interacts with an atom.
Well here's how I would do it :
I would create prefabs of my atoms with empty GameObjects as children serving as references (example : OxygenAtom => HydrogenPos1, HydrogenPos2, HydrogenPos3 and HydrogenPos4)
On atoms I would have a script managing the available position and occupied ones
When the atom you're carrying (let's say Atom2) collides with another atom (Atom1), you check the available positions for one that fits the Atom2 type.
If it fits, you tag the available position as occupied and parent Atom2 as child of Atom1 and finally place him at the position you just checked as occupied.
EDIT : here's a sample of code to bind atoms together :
protected void OnCollisionEnter(Collision collision)
{
if(gameObject.name.Equals("Oxygen") && collision.gameObject.name.Equals("Hydrogen"))
{
Debug.Log("Binding...");
StartCoroutine(BindAtomAfterFixedUpdate(collision.gameObject));
}
}
private IEnumerator BindAtomAfterFixedUpdate(GameObject atomToBind)
{
GetComponent<SphereCollider>().isTrigger = true;
// or GetComponent<SphereCollider>().enabled = false;
yield return new WaitForFixedUpdate();
atomToBind.transform.position = transform.FindChild("HydrogenSpot").position;
gameObject.AddComponent<FixedJoint>().connectedBody = atomToBind.GetComponent<Rigidbody>();
GetComponent<SphereCollider>().isTrigger = false;
// or GetComponent<SphereCollider>().enabled = true;
}