But by reinventing the wheel, you can learn how it works first hand. Sure it may not be as good as the provided wheel in some places (or better in some) but at least you learned and appreciate the work put into it.
… Or you didn’t know the wheel existed, and “big round thing” didn’t yield any result on the internet
But what if I covered the wheel in spikes and then tie a reanimated skeleton to the spokes?
Take all my fuckin money!
(I’m broke)
But how will I spend 3 years making my 2D platformer thats as generic as possible except for one slight difference I implemented poorly?
You’re taking away all of the fun of posting my 2D platformer post mortem #204582910058 where I refuse to see reality and never understand why my game sucks or why it flopped!
I take issue with this. My slight difference was implemented well and my game has not flopped yet
I’m gonna do the thing, I just want everyone to know that I am aware I am doing the thing.
Fucking general recommened default set up for a character controller in Godot.
If you want to actually have responsive character physics and collisions, in 3d, the way most people tell you to do it is wrong.
You want to base if off of a RigidBody3D, not a CharacterBody3D.
CharacterBody3D works if you are doing a simpler, more arcady, platformer type game.
It does not work well if you want more complex/interactive/ragdoll-like physics between the character and their environment, such as dynamj inverse/reactive animations, where say a box is thrown at a character, or a character runs into a wall, or a character’s own momentum would dynamically affect their animation and movement.
Now sure, most people probably aren’t trying to create something approximating MGSV or SplinterCell or Euphoria style character controller systems.
But some people are.
… at least I’m not inventing my own entire game engine.
but… if i made my own wheel, it could run faster… i definitely know more about game engine development than game engine developers…
hmmm i reckon it wouldn’t be bad if i reinvented it… i’m special, i can do it :3
Look at the thing. There’s clearly room for experimentation and improvement at least. And since you’ve used various wheels and understand the pros and cons, you could probably invent a really interesting wheel if you started from first principles.
When learning a new domain, there will always be things you spend a lot of time and effort figuring out only to find out someone else has already spent a lot of time and effort figuring out. This is fine and unavoidable; if you wait to start building until you’re sure you know everything you will never build anything. IME this tension never goes away entirely.
So if you find you’ve reinvented the wheel, don’t beat yourself up about it. Now hopefully you know a little bit more than you did about what a wheel actually is, how it functions in a video game, and when to use off-the-shelf wheels vs building bespoke wheels (do not pardon the pun it was intentional and I can take the heat). This is how you go from being a novice game developer to a better game developer.
Occasionally you may even stumble across a new way to make wheels that has never been done before, and the state of the art moves forward a tiny bit. It’s rare, but it does happen.
This is my stance on wheel reinvention. And my pitch is similar. I offer it to juniors when it feels like they’re setting impossible expectations for themselves. But I just have to say.
If you wait to start building until you’re sure you know everything, you will never build anything.
Is excellent phrasing that I don’t think I’ve ever used. The truth of it is self-evident to any senior, but juniors should be able to grasp that “more” rarely increases certainty, so for training and discovery the best you can shoot for is “enough.”
If you see learning the mechanisms as part of the challenge and journey that’s fine. If you want to understand how something works and why, then doing it yourself is the best approach. If you just care about something that should ‘just work’ without wanting or needing to understand the machinery, then of course, reinventing the wheel is not the best approach.
let God strike me down before I download a “card system” package from the Godot repo. I am making these out of individual control nodes and no one will stop me
Jigsaw you don’t understand, it’s Balatro but with turtles instead of cards.
Keep cooking
Uhh checks note cards when you reveal a turtle its shell color rarity will range from common green to legendary orange.
like super auto pets? (fun fact balatro was inspired by SAP that’s why they share so many mechanics)
i don’t think i reinvented the wheel, but if there was a better way to implement authoring frome data with hitboxes, i would love to hear it
https://bsky.app/profile/thegrainreaper.bsky.social/post/3mmagpyhmls2q
Recent first time indie dev here. I did not reinvent the wheel, but building a functional platformer demo still took years. I’d probably still be working on it if I tried my own code base or reworking a genre from scratch.
Edit: The game is called No More Mages, if anyone is interested (on Steam).
Featuring the new ai powered wheel for optimal rotation and automatic orientation detection
You don’t AI for that, you just need quaternions.
dies
A for effort, but death mechanics have already been implemented a million times over so no, sorry, into the spike pit with you.
But does it rust?
Is there an implementation of wheels for Bevy?
I’m gonna do the thing, I just want everyone to know that I am aware I am doing the thing.
Fucking general recommened default set up for a character controller in Godot.
If you want to actually have responsive character physics and collisions, in 3d, the way most people tell you to do it is wrong.
You want to base it off of a RigidBody3D, not a CharacterBody3D.
CharacterBody3D works if you are doing a simpler, more arcady, platformer type game.
It does not work well if you want more complex/interactive/ragdoll-like physics between the character and their environment, such as dynamj inverse/reactive animations, where say a box is thrown at a character, or a character runs into a wall, or a character’s own momentum would dynamically affect their animation and movement.
Now sure, most people probably aren’t trying to create something approximating MGSV or SplinterCell or Euphoria style character controller systems.
But some people are.
… at least I’m not inventing my own entire game engine.
the deepest part of godot source code I’ve ever worked on is just compiling export template, managed to reduce android apk size from 34 mb to 11 mb. but it’s an UI (Controls node) only app
I thought people hated asset reuse?
#gamers in certain specific spaces hate asset reuse. certain devs and artists who are somewhat puritanical of making things hate asset reuse. but its a standard practice, there’s no possible way to not reuse anything, u arent building ur own computer architecture and creating a compiler and everything so u are using preexisting libraries. but certain people think its ‘cheating’ or ‘not real development’ if they use specific libraries, frameworks, assets, main example being game engines. but examples like asset libraries, addons, are also included in this
the reason is very similar to “you arent a real chef if u use a cake mix” bullshit. and it exists to some extent in every field from what i understand. its just about learning that building on existing things isnt demeaning or bad in any way, atleast as long as u acknowledge and credit
Don’t get me wrong I have no opposition to it. I don’t need artisanal hand crafted crates.
I say it as an aspiring indie dev looking at the landscape and seeing all the hate for it. Even though I’d love to use an asset pack or 2.
i love how super mario reused bushes as clouds
