About Me

Hi, I am River! I am a game programmer studying in Belgium, especially insterested in gameplay and AI programming. I love implementing fun game mechanics and systems and enjoy learning new things in the process.

Projects

Fluffy Engine

<aside>

• Tech: C++, SDL </aside>

I created a basic 2D game engine and implemented some Galaga levels in it. This project gave me a much deeper understanding of how game engines work. In the project I had to make some design choices for my engine, such as:

• Using multithreading for SDLSoundSystem:
  • As a requirement of the project, I had to use multithreading at least in one system, so I decided to use it in my sound system. When a game event requests a sound to be played, the request is queued and a new thread to process that request is created and then is detached, since we don’t need to wait for that thread to finish executing. Currently sound requests are made one at a time so the feature is not really used but it is still functional.

GitHub - RiverHillbug/FluffyEngine: My custom 2D game engine for Programming 4.

GOAP Implementation for Survival Game

<aside>

• Tech: C++, building on a provided dll library </aside>

I was given a starting code that had the basic layout for a survival game, consisting of different types of enemies, consumables, buildings, and a player character agent. My task was to implement the behaviour of the agent so that it survives as long as possible. I implemented GOAP for my agent because I was very interested in learning an alternative for state machines.

Challenges

• The provided library only provided information about the enemies in the agent’s field of view each frame without a unique identifier for each. For more appopriate decision making, I wanted to prevent the agents from reacting to the same enemies every frame so I wanted to implement a memory system for them. In the provided EnemyInfo data, we had an EnemyHash variable, which looked perfect for my use. However, after implementing the memory system using it, I realized that the hash value was generated based on the enemy’s current data (including position), rather than only once per enemy. Time being short, I had to prioritize other areas and did not have time to implement a better solution, but I was able to make good use of the memory to keep track of enemies no longer in sight, so the agent could keep running away from an enemy it could not see anymore.

GitHub - RiverHillbug/SurvivalGOAP: My GOAP implementation for survival game agents.