Summer Dungeons started out as an ambisious summer project, but after one week it was already way to big.
The main focus had to be procedural content generation, since I was dying to try to implement PCG techniques
in a game. I ended up trying to generate dungeons with all their content and items with a Diablo-like stats.
This game is not finished and I doubt it will pretty soon. Still there is some amazing content which I'm quite font of.
- Project: Summer Dungeons
- My Role: solo project
- Development Time: July 2016 - Ongoing
- Current Status: ongoing
- Tools Used: Unity3D Engine, Visual Studio
- Language: C#
My item system worked amazing, but it was made for different player classes, had all these different stats and required different types of combat. But I only had a dungeon and items. I started implementing three player classes: hunter, warrior and mage. Each class has its own special weapon and has a stat thats better than others. I implemented a character creation and selection menu, save options and the different stats. The combat is all very basic for now: there is only one enemy, a silly spider, and all classes have the same combat. Future plans include skills, more monsters, leveling and class based combat.
When I figured a 3D dungeon was a bit much for me and my skills, I allocated all my time to a huge item system. It is completely based on the item system from Diablo 3, with names and stats and different items and all of that. It was amazing to make a system that was able to generate so much items. I ended up with 62 different items, 125 affixes (words that can be added to the name of the item that determine extra stats) and 4 qualities. The quality and level of a generated item and the item itself determine how its stats look and how strong it is. I have no clue how much different items this system can generate, but it is quite a lot.
My goal was to generate a dungeon. But it was quite harder than I expected. I ended up
using a pretty simple algorithm used in the Unity RogueLike tutorial. It was a major challange
to generate a 3D dungeon, so I figured that it would be easier to start of with an algorithm
I already implemented in 2D.
I'm planning on using another algorithm, since this algorithm doens't allow design elements to be tweaked and it's performance is pretty bad.