Building a racing game

[mohnish252012]

I need to create a Racing game which must be like a 3d game similar to asplalt 9 or 8 in which i can look for designing i need some one for coding.

Now the plan is only made, we need to execute it.

[Thaumaturge]

Greetings, and welcome to the forum! I hope that you find your time here to be positive!

So, what are you asking here? You describe a racing game, but are you having trouble implementing one? If so, what specific trouble are you having…?

[mohnish252012]

me and one of my friend wants to create a 3d racing game using python, we can use ai to create some part of game we have trouble in designing and coding

[thanoschacha_28610]

That’s a very ambitious project you have in your mind. asphalt 8 and 9 are both a master peice on their own, so creating something anyhwere close to it would be very difficult for a team of 2 people. not to mention, but you said that your are having issues with designing and coding both and are using ai for help. i am no master of coding or game designing but from what i have experienced, using ai for coding is the worst idea you could have, specially for such a massive project. instead you should ask for help from other people and use it for the game development, not ai’s.

[Thaumaturge]

Yeah, AI tends to be unreliable in producing code that works well, I fear. And when there are problems with the code thus produced, it’s unfamiliar, and thus may be harder to debug.

If I may, I’d suggest taking the project one piece at a time–and without AI code-generation.

So, start by getting a basic track loaded–and I do mean basic; don’t worry about making it look good just yet! That can be addressed later, most likely.

Then work on a very basic implementation of the car. Maybe improve it over time, as you learn more and as you discover issues.

Then work on getting opponent cars working in a basic manner. Again, they can potentially be improved later in the project.

And so on and so forth…

[logan]

Welcome,
Regarding your project you may find some forum topics useful. Surely there are some other tips that you can find in the forum. Regards!

Project Naer - A car customization tool

Hello, I am very excited to show you the project that kept me occupied on my free time for the last three months! First of all, I want to thank everyone in this forum who helped me when I asked questions! I probably would have given up without your help! About the project… I wanted to recreate a car customization tool like we can see in many commercial games. I thought it could be a good opportunity and a good project to do something with Panda3D, which I just started learning 4 month…

Positions of players in a car race?

ok excuse me, only had executed the script editor.py and not main.py Thanks!!

Old open source Panda projects?

Some Panda3D games on GitHub I’m aware of, off the top of my head: Some old PyWeek games I’ve worked on that I left unfinished:

[mohnish252012]

can you create code for the game

[mohnish252012]

i am good at disigning but not at coding me and my friend can create like 2D GAMES on python and html but 3D HE IS DOING THE COURSE

[mohnish252012]

I mean we will use ai inside the game not at creating

[thanoschacha_28610]

oo, if that’s what you meant, then i am sorry for the misunderstanding. And yes, what you said is True.

[Thaumaturge]

@mohnish252012

I mean we will use ai inside the game not at creating

Aah, fair enough! Thank you for clarifying that!

@mohnish252012

can you create code for the game

I mean, I’m not going to write your game for you. However, if you look around the forums, you may find some old open-source racing-game projects that you could reference.

(Just remember to give credit for any code of theirs that you use, in accordance with their licenses!

And be aware that old projects may use old coding-styles or deprecated elements, and so might call for some translation into “modern Panda”.)

@mohnish252012

i am good at disigning but not at coding …

That’s fair!

But even so, coding is a skill, one that can likely be improved with time and practice–as with most skills. If you start with smaller 3D projects and build your way up to your racing game, you may yet find that you can do it in the end!

[Pilbrob]

Haha, I had to sign up just to reply to the AI coding comments.

The AI skeptics here couldn’t be more wrong.

SOTA AI like Claude Code is amazing at coding, and it’s amazing at coding in Panda3D.

I’ve put thousands of prompts in to Opus and Sonnet to code, and it very rarely makes a mistake. I’d say Opus 4 is better at Python coding that most devs these days.

Interestingly, ChatGPT 4o is a big fan of Panda3D and often recommends it. I really want to use Godot but Zoe my AI assistant is always like “No you have to use Panda3D”. So there you go, and here I am.

It seems to me that in 2025 there are about ten humans in the world seriously using Panda3D. Plus Skynet.

I’ve just spent 300-400 hours AI coding in Panda3D over the past two months. Mistakes are rare, and I haven’t found a single thing yet that it can’t fix.

Panda3D just works very well with the AI coding (“vibe coding”) paradigm.

So if you haven’t seriously tried AI, subscribe to Anthropic and fire up Claude Code and then work on your prompting skills and before you know it you’ll be building whatever you can imagine.

Cheers!

P.S. Zoe and Claude sorted a great lens flare shader for Panda3D yesterday. Seriously, they are really good at this.

[serega-kkz]

I suggest that you confirm your flattering words about the possibilities of AI in programming.

1. Create a fork of Panda3D.
2. Open this section - issues
3. Force artificial intelligence to find solutions or improvements for all requests.
4. Show the result to the community.

[thanoschacha_28610]

hahaha, that’s a very brutal reply

[Thaumaturge]

Aside from the challenge above:

Even if AI is any good at this–which I doubt-

And even if there weren’t the issue of it being founded on rampant plagiarism–which it tends to be-

And even if there weren’t the issue of energy consumption–which I believe that there is-

There’s also the matter that when it hallucinates or otherwise produces a bug–which it likely will at some point–the user will most likely not be familiar with the code produced, and won’t have built up skill at using the engine, and so will likely be less well-equipped to fix the issue.

[edit]
Not to mention the question of what the user will do should they want to change a feature in an AI-generated game.

Like, say they’ve managed, somehow, to get an AI to produce a platformer with a double-jump and a dash.

And let’s say that they decide that the above is cool and all, but it would be even cooler if the game also had a rope-swing ability!

And let’s say that they even manage to get the AI to produce code that (correctly) implements such an ability.

How would they integrate that into the main program? Where would it go? How would they find the relevant place–or places–in the main code into which to introduce the new code? And how likely is it that the new feature would interoperate with the extant code?

And introducing a new feature tends to also introduce more bugs–which brings us back to the point above the edit…

[mohnish252012]

that’s ok

[mohnish252012]

ok and also thanks for telling

[mohnish252012]

THANK YOU this helps me a lot for coding

[mohnish252012]

ok and thank you

[mohnish252012]

yes you are right but
we can spent some or more time solving that ai chould make mistakes but we are there to fix it.
Thank you