Making a modern SNES game from scratch

Why it was so common for programmers back during the 80's and early 90's to code their own game engines and tools from scratch and is looked down today to try to make something similar to those games?

Not talking about making unity today alone, talking about making a similar game to SNES and maybe early 3D games like genesis or maybe psx games today using modern libraries?

I mean, those niggas used to ship functioning 2D engines, even if primitive as well as the toolset to use the engine from scratch in like 6 weeks.

Attached: the-fuck-you-looking-at.jpg (623x469, 52.38K)

Other urls found in this thread:

assetstore.unity.com/packages/tools/physics/obi-softbody-130029#description
twitter.com/AnonBabble

The difference between 2d and 3d is massive
Especially when you add lighting
I think when games started to have lighting was around the time when studios decided to start making engines

The number of people who can make a decent 2D game engine is tiny. The number that can do it for 3D is almost nonexistent. Throw on top of this that every game is essentially a custom Quake map and it is nonsensical from a business standpoint to spend any time rewriting things.
The main effort in 3D games is the art, not the programming. Generating tons of good models to make modern AAA games is a fucking grind. The main job of modern game programmers is to help the artists make pretty pictures, possibly make some custom asset tools, read doco, and debug the engine. They don't actually do any engine programming.
The only exception to this is if your game isn't Yet Another Quake Map and you need certain behaviors to be core to the engine for it to work. A modernish example is Minecraft.

hm, so making your own engine might be pretty impressive to employers then?

Only if you want a job working on engines

that sounds like fun, im in. im pretty good with C, what else do i need?

Id assume yes.
But it mightvalso be overkill.
Still the most useful types of programming are graphics programming.
If you can make a game engine youve basically mastered graphics programming.

>talking about making a similar game to SNES and maybe early 3D games like genesis or maybe psx games today using modern libraries?
Because those tools aren't made for those things. Those games were a product of the hardware that was available.
As for why very few people can make games for that hardware or simulate it from scratch, it's because we don't teach programmers from the bottom up. We start with teaching them abstractions and never teach them why those abstractions were made in the first place unless if they specialize.

Making a relatively authentic SNES game is plenty doable. Check YouTube for some tuts and a really good breakdown of the assembly language the SNES employs. That said, it should be a small personal project. You're going to run up against some frustrating limitations, old consoles are not ergonomic. I hate to sound like that other guy, but if you don't need to build your own wheel, you probably shouldn't. Unity and Godot or Unreal would be good. Or Defold or Heaps if you wanna go off the beaten path.

If you're looking to have that primitive feeling, consider making a Gameboy game. You'll learn a lot and it's dead simple. Relatively. The best part of making a Gameboy game is that the console is really well documented and you can use other games as guides or models. If that's too archaic, check out the Playdate. You can build something very similar in C or Lua or both in a low level feeling way without implementing a lot of the square zero stuff like a draw call.

>making your own engine might be pretty impressive to employers then?
Yes but only because it means you have passion. It isn't the only thing.
Here's what I care about as an employer - are you able to teach yourself? Are you able to solve problems? Are you able to work with others? Will you do the not-so-fun parts of software development?
Most zoomers and millennials where I live don't have these basic skills. I am desperate for non-retards.

90s engines weren't as complicated as the ones used today. you aren't as smart as the people making these general-use engines - these people have decades worth of brilliance and optimization in code.
however if you were to make a game for SNES from scratch (knowing how hard it is to write in low level langs) you shouldn't worry that much. just know your shit enough to make an optimized engine for the game you want to build.
checking some quake builds might help you a little

Attached: 1649360394304453200.jpg (1417x1072, 167.74K)

Keep in mind that theres not getting around the math.
If you thought math was useless this is where youll find out you were wrong.
Youd have to use every single math subject found in the computer science curriculum. And then some cause im not sure if they teach linear algebra. Discrete math for the AI, trig for doing anything whatsoever, linear algebra for vectors, calculus for the physics, etc...

>Why it was so common for programmers back during the 80's and early 90's to code their own game engines and tools from scratch and is looked down today to try to make something similar to those games?
They didn't, lots of tools shipped with the dev kit, any code that was reusable was copied between games

I wouldn't say it's looked down upon, but people will definitely ask why you went out of your way to reinvent the wheel.

some people still do that sometimes, like cave story uses it's own engine, and is rather 4th-gen-like

that's because it was released during the 4th gen

it came out in 2004, during the 6th gen
while it's not very new now, it's new enough that it uses technologies still current today, meaning someone could make something similar to cave story with up to date libraries
there's also a foss engine implementation of cave story, which op could use for research or i guess probably even just making a new game on that engine

vulkan-tutorial.com
vkguide.dev

>Why it was so common for programmers back during the 80's and early 90's to code their own game engines and tools from scratch
Where were they gonna get a game engine from? Game Engine licensing only became a real thing later. In the early days you made your own or maybe there were random one off deals between game companies to share an engine. But the environment was nothing like it is today.

Also the hardware was super quirky back then and nothing was standardized. You had to code directly for a single platform. Console dev kits shipped with SDKs that were basically just interfaces to the hardware, and not very elegant ones I should add.

Studios did kind of start building their own in house engines and adding onto them from game to game. But this required you to be successful and stay in business first, and second know what the hell you were doing. Game programmers weren't exactly the cream of the crop for the most part.

Yeah if you were capable of doing this stuff youd just make shaders or some sort of asset for these big engines and sell them.
Itd be more productive and youd make money.
Some people but assets without even intending to make a game probably.
Like in unity a hot selling asset is for cutting jelatin objects. When was the last time you saw a game with a character chopping jelatin?
I cant think of any. People just buy the asset just to play with it i guess.
That must not have been easy to make though.

I think it was this one right here if i remember right
assetstore.unity.com/packages/tools/physics/obi-softbody-130029#description