6 ms·
Unity has built a very impressive engine and runtime. With that being said, I have never had the desire to use it (or any other engine). Perhaps I'm a minority
by consoomer 3y ago
Unity has built a very impressive engine and runtime.
With that being said, I have never had the desire to use it (or any other engine). Perhaps I'm a minority here, but I dislike engines and I dislike the idea of "building my own engine."
I think you should set out to make the game. You begin by creating a window. Then you draw some pixels or render a texture. You add events and controls. You make the game logic and states. Then you have a game. It can take as little as a hundred lines of code to have a basic game up and running.
From there, you make the thing you want to create. No more, no less.
- Sohcahtoa82 3y ago> You begin by creating a window. Then you draw some pixels or render a texture. You add events and controls. Yes, but without an engine, how do you make this all cross-platform? Using something like Unity, you can theoretically write your program once and it runs on damn near anything. Android, iOS, Linux, Windows, PlayStation 4/5, Switch, and more.
- consoomer 3y agoThat sounds nice, but it looks a lot like a "what if" scenario. What if I need to support a dozen systems? What if I need to do fancy 3D graphics? What if I need to do some other thing that isn't my problem right now (and likely ever)? It's like when a web developer starts talking about how they will handle the c10K problem, or how their todo app is deployed on a fancy kubernetes cluster across the planet running on edge networks and fault tolerant, but they have zero users and no good ideas. Just make the game you want to make. If it's fun and people love it, build it out more. Rewrite it if you have to. Hell, Minecraft was wrote in Java, then C++. Just make a game.
- Sohcahtoa82 3y agoI think, at the very least, it makes sense to want to support Windows, Mac, and Linux. Trying to support consoles adds a massive layer of complexity regarding their stores. "Just make a game" is fine if your game is going to be simple and can be done in a month, but there are plenty of indie games that have had only one or two developers for a couple years before even entering Early Access. A game with that kind of complexity helps a LOT to use an engine.
- dbrueck 3y agoTo each his own, but it sounds like you're overlooking the opportunity cost, which is potentially massive.
- matt3210 3y agoIt’s much harder than you make it out to be.
- andersa 3y agoAnd by the time you actually ship something, 10 years have passed.
- sovietmudkipz 3y agoAny advice on how to build a tender pipeline from scratch? Im struggling learning how to implement interesting shaders as a user. I couldn’t imagine building a render pipeline engine on top of writing shader code.
- peteforde 3y agoThat's fine for a hobby; specifically a hobby where you plan to spend 90% of your available time reimplimenting crappy equivalents to things you get on day one of using an engine. In other words, a hobby project where the point is the exercise and has no need to actually ship. Just pray that you're never employed to be the #2 developer to someone who built their own game without an engine. "Why would there be documentation? Just follow my inner voice."
- LarsDu88 3y agoAnd then you setup the forward rendering pipeline, and the programmable shaders, and the skeletal animation system, then the inverse kinematics for procedural animation, then the asset pipeline for publishing to different platforms, then the netcode, then the realtime physics system. Did I forget about reflection probes? How about a terrain editor? Maybe you need to bake lighting to make your game look good. I think this comment might wildly underestimate just how much modern game engines like Unity, Godot, and Unreal bring to the table. They may very well be the most complex pieces of software in the world today