README.md

🏓 pong-online

Online multi-player classic pong game.

How to Play

1. visit https://www.play-pong-online.site/
2. click PLAY
3. send the invitation link to your friend
4. when your friend joins, the match starts🔥
5. play till you win😤

Motivation

I wanted to play pong with my partner in long-distance. We didn't find any on the Internet. So I decided to build one.

I also wanted to try things further with my Full-stack SAM application pattern blueprint (https://github.com/yozibak/serverless-chat)

Development

# Install deps
pnpm i

# Deploy backend resources to dev environment
pnpm deploy:backend # -> prints API endpoint url

# Create .env
cd frontend/app
cp .env.default .env # <- write API endpoint url & key

# Run frontend
pnpm dev

# Optionally, you can deploy your frontend build onto dev s3 bucket
pnpm deploy:frontend

Design

Match

sequenceDiagram
	actor player1
	participant server
	actor player2
	
	player1 ->> player1: initiate game as a host
	player1 ->> server: subscribe
	player1 ->> player2: send invitation link or show QR
	player2 ->> player2: access the game
	player2 ->> server: subscribe
	player2 ->> server: initial message
	server -->> player1: receive initial message
	player1 ->> player1: locally set the start time
	player1 ->> server: start sending starting time constantly
	server -->> player2: receive starting time
	
	rect rgb(80, 80, 80)
		player1 --> player2: waiting, set up initial state and stuff
	end
	rect rgb(100,100,100)
		player1 --> player2: start game at time, counting frames
	end

Loading

Process

sequenceDiagram
	actor player1 as player1(local)
	participant server
	actor player2 as player2(opponent)
	

	loop frame at n-x
	player2 ->> server: send state payload
	server -->> player1: receive opponent's state as frame[n-x]
	note right of player1: save the payload
	end
	loop frame at n
		player1 ->> player1: import input & get snapshot
		note right of player1: combine network state & local
		alt if ball is on p1 side 
			player1 ->> player1: calc next ball position based on local data
		else
			player1 ->> player1: calc next ball position based on network data
		end
		player1 ->> server: send final state as frame[n]
		server -->>player2: receive state
		player1 ->> player1: render state
	end

Loading

Frame resolution

	player bar	opponent’s bar	ball
when ball is on player's side	once	recursive	once
when ball is on opponent's side	once	recursive	recursive

flowchart LR
	localInput --> merged
	localSnapshot --> receiverState
	networkPayload --> receiverState
	receiverState --> merged
	merged --recursively resolve frames--> finalSnapshot

Loading

Deployment / CICD

You can basically deploy to your AWS environment by selecting test or prod as sam deploy --config-env just as same as dev environment.

To setup CICD, follow steps below.

Prerequisite: Create Github Connection

Go to AWS console -> Pipeline -> Settings -> Create connection

Select your repository & create connection. Copy the connection ARN.

Deploy Pipeline and other resources

Once Github connection is created, deploy pipeline by running:

sam deploy 
    \--config-env pipeline 
    \--parameter-overrides ConnectionArn=GITHUB_CONN_ARN,FullRepositoryId=<username>/<repository_name>

Pipeline overview

The entire stages and actions are as follows:

flowchart TD
  subgraph Source
    ConnectionArn
  end
  subgraph Build
    BuildStack
  end
  subgraph Test
    deployStackTest(DeployStack)
    buildFrontTest(BuildFrontEnd)
    test(TestApplication)
    deployTest(DeployFrontEnd)
    
    BuildArtifactAsZip --> deployStackTest
    deployStackTest --stack info --> buildFrontTest
    buildFrontTest --build--> test
    buildFrontTest --build--> deployTest
  end
  subgraph Deploy
    deployStackProd(DeployStack)
    buildFrontProd(BuildFrontEnd)
    deployProd(DeployFrontEnd)
    deployStackProd --stack info --> buildFrontProd
    buildFrontProd --build--> deployProd
  end
  Source --✅--> Build
  Build --✅--> Test
  Test --✅--> Deploy

Loading