So for the past X decades of my life or so I’ve had an idea for a gaming concept. However I take one look at any sort of code of any sort and my head swims. Some people just do not have a head for this sort of thing, and being born before computers were a thing doesn’t help much. With the advent of AI (I know, I know. Don’t hang me too hard from the nearest tree) I thought I could get an AI to help out – by using a free model’s limited number of context tokens, that’d at least cut down on the feature creep per concept or mechanic as I try to get it going.
So I came up with the thought that each game mechanic would be a “Crystal” that slots into a “Geode” in order to all come together. A couple of AIs suggested the Godot language, so here I am sniffing about before I commit to the few years of the development dive.
The game I’m noodling around in my head is just a typical first-person dungeon crawler (think of the Wizardry series) with a lot more interactables, a lot more race/class combos/skills, a lot more unusual mechanics (mostly inspired by DCSS, ADOM, and the like which tells you “I’m Old”) 40 years of D&D experience and homebrew, and far more levels than just the Wizardry series had (it had 10 floors to its dungeons; I vastly prefer 100 as it makes a win feel earned). Throw in roguelike mechanics on top of that to make it interesting… there’s a reason my design documents turned out to be 90 pages (cut down from 140, so don’t judge me too much).
So, vibe / feasibility check. I know at least not to vibe-code with an AI… we have enough eldritch horrors in the world already, but I have an AI currently writing an HTML document for reference with all 80 (soon to be 110) current “Crystals” and how they would “slot” into a “Geode” as a finished product… but would people want something like this, or am I chasing shadows and wasting time?
GDscript is not difficult to learn the basics of. Just start out trying implementing som simple features and learn as you go. For complex thijgs, you could use plugins.
AI can be helpful for learning the language and engine but i would advise against using it as anything else than a learning tool.
You’re clearly quite invested in this idea - I think it would be criminal not to at least give it a go
And creating a “crystal” system sounds like a good way to manage features - as a former producer, I’ve never worked on, or even heard of a game that was completed on time, on budget, with all the intended features. You can expect the same
I’d suggest going through a tutorial or two on the FPS basics in Godot, then using AI to build on your project. That’ll also introduce you to the Godot editor, which has a lot of “hidden” or unusual features.
You’ll also get a sense of the basic code that’s repeated throughout the project - if statements, loops, defining functions, data types and so on. Even with a little understanding, you’ll be able to roughly interpret every part of the code.
Disclaimer : I’m old (76 at next birthday…), and have fond memories of the Atari ST ‘Dungeon Master’. On the other hand, I’m not allergic to programming, nor coding, having spent a large part of my various careers in IT of one sort or another.
OK, having said that, I can completely understand the itch to create something inspired by one’s past experiences. I decided, a couple of years ago now, to make my own ‘version’ of Dungeon Master, and, having already some success with modelling with Blender, stumbled across Godot 3, and started to make my game. It’s in similar vein to that which you describe; currently I have half a dozen Levels working, with differing challenges, puzzles, obstacles and rewards as one ‘progresses’. Here’s a short video of the opening ‘Entry’ training hall, leading to Level One, with a whole labyrinth to get through…
All this to encourage you to take the plunge (come on in, the water’s lovely..!) and get cracking on advancing your dream. I’d be very willing to offer any modest help in getting such a project off to a good start (I’m far from expert, naturally, but I’ve managed to overcome quite a few hurdles along the way…). These Forums are an enormous help, too; I’ve not yet remains stuck for a response for any technical difficulty encountered, with advice, tips and explanations from many members here.
Give it a go, shove the boat out, get something very, very small off to a start. The first Level or two will pave the way to achieving your goal; what’s a hundred or so Levels got that’s so frightening, eh..? Get cracking; I can’t wait to try your first Demo.
My often quoted helpful encouragement : It’s the first forty or so that are the hardest, after which things sometimes tend to get slightly easier.
When i’ve started learning coding i chosen niche language and didn’t knew english, so no tutorials pretty much, but after years i don’t worry about coding when making any games
You need to just find fun in coding, or a way you find it fun, then learning will be possible, and with knowledge of english + modern tutorials it will take months, not years
Yeah, you are chasing shadows. Ain’t gonna happen unless you hire a couple of programmers and artists. Realize that you needed LLM crutches even for the thing you’re supposedly most competent with, i.e. game idea/design. How do you plan to proceed without “ai” if “one look at any sort of code makes your head swim”?
If the goal is more modest than creation of a top-selling blockbuster, I say that it‘s perfectly possible to chase, and even catch up with, one’s shadows, and well worth the effort. A team of like-minded individuals, with compatible, non-commercial, objectives can achieve much, and have fun learning as one goes. For my part, I have no interest in hiving off the most rewarding parts of my Projects to an anonymous super-computer, and would rather keep to modest, achievable goals. Baby steps lead to toddling, then walking; running is now beyond me at my age, but it’s all good for exercising the old ‘grey matter’, anyway. I’d not discourage anyone willing to reach for the stars. No kittens get hurt. Just sayin’.
Some folk have big hand-writing. Or maybe lots of sketches..? Or a small notebook..? At least there’s been more than simply pipe-dreaming. I reckon it’s worth looking further into, and, to me, more interesting than a 2D platform jumper. Just my tuppence-worth, of course.
The OP basically asked for a reality check. To form an objective (or at least nuanced) view they should listen to as much varied opinion as they can, and then judge for themselves.
I agree, do chase your dreams. Never too old to learn.
That being said, there are some huge moral issues you are probably aware of when using one of those silicon valley token chatbots that mash open source code together. Besides the theft of intellectual property leaving bereft artists of all kinds there is the amount of energy it drains and the amount of coolant it needs.
On top of which you are now welcomed in one of the most helpful communities of IT folk I’ve ever seen… Of which some have been gaslit for four years now by Silicon Valley billionaires and their siccophantic main stream tech journalists that we might as well quit. But it is our open source work that they are misintroducing all over the place.
Programming is daunting for everyone. Just like translating, composing music and painting is. I put in my 100.000 hours and am more than willing to help. But not in exchange for being gaslit by big tech and their fanboys.
I love for you that you feel empowered, but finding a team of like minded humans will give you more joy than the dystopiabots. I promise.
About being negative: disillusionment is the most healthy growth experience imaginable. I wouldn’t call stating empirical facts negative either.
Dreaming is one thing, creating a working game is another.
To OP: I would say that you should stop writing design documents and start writing code. If you are not a programmer yet, do no rely on AI for creating the actual game. Use AI to explain stuff and review your code, but if you don’t understand the code, you don’t get a working game.
You may find out that programming is not for you, but writing a design document is. Then it is totally ok just to write fun stuff, but it is not going to evolve into a working without a lot of hard work.
Godot’s gdscript is quite beginner friendly. I don’t think it’s chasing shadows but for a project of this scope, realistically you will likely need to be more focused or else rely mostly on emergent gameplay. If you’re hoping for more scripted and planned combos that you explicitly code and most of the 100 levels to be unique or at least different, then the scope will be massive and you will likely take a few years.
Your 40 years of D&D experience and homebrew here is a big advantage you can use to simplify your game design to increase the chance of completion. The sheer content you’re planning is probably a bit too big because even if you finish coding, you’d have to start implementing the content, iterate on ui, balance with playtesting, do the art and music and refine the quests and so on.
I used to think somewhat like that too. Then I started playing with Unreal Engine’s Blueprints. You’ve probably heard about it, it’s a graphical interface for programming by connecting nodes instead of typing code. It seemed friendlier for non-coders like me, so I tried it, and I made a game prototype with only pre-built tools of the engine and blueprints, no actual coding whatsoever.
(this is the prototype, it’s short and not polished, but you can finish it, so I guess it’s technically a “full game”: Aero Ace (Prototype) by Alexandre Marcati)
The thing is, I eventually started realizing that the idea that coding in blueprints is easier than typing code is actually an illusion. I mean, it’s easier to follow the logic flow when looking at the nodes if you are a visual thinker, but it’s pretty much the same as typing code with an IDE, just with more clicks. So by using Blueprints, I actually tricked myself into learning the basics of coding.
After that I tried Unity, and tried to re-create the same game in Unity, in C#. I never finished it but it was a great learning experience. And now I think I’m pretty ok at coding in C# for gamedev. I’m no expert by any means, but when I set out to add a feature to a project, it may take a while but I eventually get there.
So, maybe try creating some stuff using Unreal Engine’s blueprints, I think it might help you eventually get to coding.
Normalized was a bit harsh, but if you’re not willing to learn how to code, I do agree it’s unlikely you will get a game done. But I do think you are able to learn it. Why wouldn’t you be?
If you find yourself chasing shadows and unable put in time and focus and energy to learn some of the hard stuff, you may not actually be as interested in it as you think you might be.
That said you don’t know until you know, you know? Start somewhere, learn basics of GDscript and , follow tutorials to get aquatinted with the software. Take some baby steps make some prototypes and test games. You’ll learn the most by doing.
I have a lot of friends who always talk about writing a novel yet never seem to be writing anything to help improve their skills. You can’t get better at something unless you actually do the thing.
