1 Year of Game Engine Development (was it worth it ?)

Hey, guess who is still alive
my mic suck so sorry!!

Resources About 3D projection:

#FromNorth
#FromNorthProjekt

[rmt]

This is amazing, and restores some of my own motivation too. Glad you're one of the few ppl I have notifications on for, very rewarding!

I planned on making my own engine too, and still mostly do. Godot 3 uses a lot of the same parts I was planning on using, so now I plan on turning it into my own engine. Biggest hurdle is I'm struggling with Git, and sm trying to learn how to build Godot from source before modifying that build pipeline.

Anywho, keep up the great work!

[GoldSrcAesthetic]

I wish you luck. Always admired people making their own game engines

[photon]

man, your sense of humor is great =D
But beside all of those triangles rendering funny haha I really respect a ton of work you have done to accomplish such a result. It's great.

[Dragon_Dev]

glad you're still alive 🙂
well, the work is amazing, can't lie
Keep it up buddy, we're waitng for the game

[DeamonTanki]

Man loved every minute of this video really motivating, nothing better then following what interest u and spark enjoyment in you, Thanks for the video !

[friendlyfox]

I plan on making my own game engine also for the sake of learning and it will help me make me optimize and improve my 3d games later on.

[charleshornby]

You're incredible, this is excellent ingenuity and workmanship

[platonvin]

Personally, almost all the time when people have to move computations to GPU, they should not have had them in the first place. Cool, grass blades are data driven! (or particles). But why is per-particle/blade data exist at all? Anyways, there is a lot of sequential optimizations that are effectively moving computations up in the pipeline, and one day you will reach the perfomance wall which can only be removed by utilizing unused CPU time.

Loved the video, thanks for your story. I envy you a little, i wasn't smart enough for school 42.

Opinion: Vulkan abstraction layers are always either OpenGL perfomance (which defeats the point) or Vulkan complexity (which also defeats the point)
Major perfomance benefits from using Vulkan are subpasses, low drawcall cost (so gpu-driven loses main benefit), synchronization (semaphores / fences. Not layouts/memory barriers, they barely matter for modern drivers), actually available hw operations (e.g. 8/16-bit, reductions / quads / [just] subgroup operations, variable rate shading, etc.) / compression (arguable) and more explicit buffer's control (how memory is written, how (when) they are updated, atomics, some metainfo for driver like push constants, sets)

[breakdancerQ]

Thats awesome man! Definitely gave a me a good idea of how it is, may I ever want to suffer in this way

[vulnoryx]

This video inspires me to continue developing my own game enigne using c++ and opengl.
Its a 2d game engine, so its not as complicated as yours but still.
Anyways I wish you success during your engine development.

[danilafoxpro]

Haha, I like the squares joke (even if it's not technically correct, as those are rectangles).

I am making a replacement for the C standard library (with different API) and a compiled programming language, so I can kind'a relate to what you are doing. Though I suppose I wouldn't be using any libraries or frameworks eventually (currently still depending on the C standard library for some things). Still running on top of an operating system kernel though (technically a framework?). :D

Might make an engine at some point as well, though I would probably use CPU only. There is this game called King's Crook, written in C using only the CPU, only standard library and the system library. Renders everything to a pixel buffer internally. Might do something like that. GPUs just seem too roundabount. CPUs have an instruction set, and you can just give it the instructions, though still wrapped in ELF or PE. And I am still kind'a unhappy to have to deal with operating system stuff, like potentially having to call into DLLs on Windows instead of being able to use system calls directly. And having to dance around the blackbox GPU? Sounds like a nightmare. Though people tell me that Vulkan might get sufficiently close to the metal of the GPU, might see about that. But probably will still start with CPU only, and hopefully in my programming language.

Best of luck with your engine and future games. I am happy to see videos like this, just people passionate about making something technical. Though I hope it doesn't turn into a project that never produces anything, as such is a danger with those things sometimes. I had a GUI library I was writing for years, which turned out to be near useless in the end.

Good work, but in engineering, efficiency is key. You could spend the same amount of time to write a Godot 3D Engine plugin like godot-jolt, it could be more useful for the community.

[mostrealtutu]

try to load the deccer-cubes next :) also good luck with the rest of the engine

[epicDamnWin]

Honestly... i feel the same sometimes. I started questioning myself, if i just want to make an engine instead of the game and more often than not, imposter syndrome starts coming up. Rn back in unity and just trying to push out a first prototype of an idea, but even that is cumbersome and i really need to push myself to keep going...

[verendale]

Damn you got more dedication than I do. Honestly with Unity, *IF* they don't screw it up and stick to the plan their Roadmaps and such look amazing (ECS for All as a baseline but with simpler Gameobjects workflow, revamped Animation system, unified Renderer so theres no more URP and HDRP, yada yada). Thing is...that's IF they stick to the plan....or it doesn't take 6 years for any of those to reach "Production ready", if they even do and don't get do not have a good track record for it lol. Who knows maybe new CEO will fix that *copium*

[photon]

hey mate, are there any plans for future updates on your game engine?

[itodobien]

That intro was hillarious..no no no...wait. wait. wait.

[ourabig]

i can suggest also ogre rendering engine it's a bit easier i guess and has more documentation than the forge yes it gives you rendering controll too

[ArThur_hara]

10:10 Ho your school works like ecole42 :]

[tb_regular]

if u want to not struggle to create anything u want even game engine that is tap engine is lightweight and trivia not only tap engine or Godot engine every languages code ar Lua