Fourside
This is the largest game that Magnet Hill attempted. It never got a name, we still only refer to it by its code-name, Fourside. Ari and I had discussed several game ideas and inspirations and time and again we kept returning to some central themes. We've always been a fan of genre mashup games, and a big influence on both of us are the works of the now defunct studio Quintet, particularly their titles Act Raiser and Soul Blazer. Fourside was based off those core concepts, a genre mashup of a dungeon crawler and city building game.
Premise
Long ago, humanity existed in a golden age of technology that allowed it to colonize the galaxy. Little did people know that the very fabric of space is shaped by humanity's powerful collective will, and when humans began to fight each other over control of the portal network that linked their worlds, it created a violent rebound that shut down the portal network, and corrupted the worlds that used to be connected to it. Humanity was isolated, losing its connection between the worlds. Thousands of years later, a robot reawakens at the space station that served as the hub for the portal network, and is tasked with reuniting the remaining human worlds. This is a game that is ultimately about positive vibes, we wanted strong themes of friendship and cooperation to be the driving focus of the story. Gameplay wise, the player alternated between moving through procedurally generated dungeons (different worlds), and building out the city (the space station). The dungeon generation had a planned layer of player influence as well, with the idea that the player would be able to modify input parameters of the dungeon generation to create a succession of dungeons that would spawn things the player needed (specific loot, specific environments, specific enemies, etc).
We worked for nearly a year on Fourside, producing a prototype that had most of the core game in place. It was an ambitious project, and faced many challenges. Unfortunately, with only Ari and myself working on it, we didn't have much available manpower. I did the programming and original 3D art for the game, which was all placeholder, while Ari contributed to the design. Other art came from the Asset Store, mostly to use as placeholder and test rendering styles. As of right now, Fourside's development is on hold.
We took a breadth-first approach to building out features, but started on combat. There was a lot of mocking up we did, everything ranging from strict top-down approach to a more Dark Souls like ARPG combat system. Eventually we settled on something reminiscent of older Zelda games like Link to the Past, albeit with a deeper combat system.
Eventually we added in a variety of traps, environmental hazards, and interactive elements like switches, pressure plates and locked doors.
Once combat was to this point, we transitioned to working on the town development. We wanted town building to be the level progression, working in a way similar to the sphere grid from FFX, with buildings (and their decorations/upgrades) being the nodes that would provide you with stat bonuses and unlock new skills. The idea is that you get resources from the dungeons, which are then spent upgrading the town.
With narrative a big part of the planned game, we started working on our dialog solution. We started off with YarnSpinner as a base, but I added a more advanced run-time parser that was compatible with TextMeshPro. It offered richer text rendering with lots of markup options to have very flexible text displays. The runtime parser also supported binding function calls to coincide with the text cursor reaching specific , giving content designer have a pretty rich set of animation features without having to leave the dialog editor.
Additionally, I created some early versions of the skill editor, and a replacement for Unity's animation event system that lived on the Mecanim layer. This had some strong advantages, like being able to have different mecanim states that shared the same animation, but had different event structures, and the ability to duplicate event packages between different animation states.
Fourside further developed our 3D pipeline from the time I worked on Sploot. I developed a more advanced animation armature, creating my own humanoid IK driven rig in Blender complete with bone drivers for common animation tasks like finger curling.
We'd chosen to focus on combat prior to dungeon generation, feeling that how combat worked was going to impact dungeon generation a lot. Unfortunately, this was near the end of the project before we decided to put it on hold.