In this installment of Building Something Cool, I will be going over the first few months of development of our project, and I will discuss the challenges we faced while learning and growing accustomed to the Unreal Engine interface.
After getting our design fleshed out, it was time to get started with the real work for the project. Knowing the sheer amount of stuff we wanted to have in our game, we decided to start early. Before the semester had come to a close, we had already tested our peripheral, the Oculus Rift, and had a github set up for our project. There was an uphill battle ahead of us, but we took early steps to ensure that we were up to speed before we started our climb.
Having made a repository for Angelis, set up a virtual reality lab on campus (with some badass new graphics cards), and hammered out our design, it was time to write some code.
Cue the classic Scooby-Doo line, “let’s split up, gang.” Sam, Kurt, and I began work on the C++ classes. Julia started doing some level design. Kurt also began work on AI research. And, Isaac began hammering out some code for the character animations.
I was tasked with creating a usable actor class. For this, I needed to make it so that we could form actions around specific objects in the game world. If you clicked on a spawner node, for example, a weapon would spawn close by (or in your inventory). If you walked in front of an elevator, the up or down arrows would light up when the reticle was over, and the elevatory would move correspondingly. This was also important for several character interactions, such as picking up a weapon or handing a weapon to an ally.
Kurt was working on AI classes. For this, he needed to create a diagram of ways the AI interacted with the world. This involved creating a state diagram, with four super-states – rest, patrol, alert, and attack – and many sub-states and nodes. Also, he had to manually create cover nodes for the AI to interact with.
Sam had reign over all of the character and weapon classes. He would gather attributes from our shared Trello page, then file them accordingly into classes. After this, he would tie them in with character or weapon blueprints and get them all working in one of our test maps.
Julia, after reading up on several design books, was working diligently on creating a captivating level. After creating one based on her original design, she made the revelation that it was entirely too large for use in the game mode we had in mind, and so she made up a new, more relevant level model.
Isaac was making slow-and-steady strides in the animations department. Finding assets and applying them to our current characters and weapons was his first task, most of which were readily available in the Unreal Engine asset store.
A couple weeks in, and we had already made strides that some groups weren’t even thinking about yet.
Barring some early midterms, most of us were putting hefty hours into the project every week. With the semester beginning to fly by, we knew that persistence was paramount to finishing on time.
I had compiled a comprehensive list of game sounds, as well as finalized my economy formulas; I was falling behind on the usable actor task, but I made some real headway on sound assets. Kurt and Sam ground out several working test levels and characters. Isaac found some useful information about an upcoming VR mode for Unreal. And, Julia finished her level entrance.
Things were shaping up fast and we were ahead… but that wouldn’t last.