I’ve been thinking about building something like a Machinations-style tool in Defold and wanted to ask a simple question before I go too deep into it:
Would anyone here actually want this, and would anyone be interested in helping?
To me, the idea feels surprisingly doable, especially now.
At its core, this kind of tool is mostly about moving values through nodes, applying rules, and running simulation steps. That doesn’t sound like something that really needs to live on a server for a lot of money like Machinations.io. A lot of it should be perfectly fine running locally on the user’s own machine.
And the really nice part is that the minimum useful version does not need to be huge.
You could start with a small core:
- sources
- drains
- pools/resources
- converters
- gates
- links with rates
- simple conditions
- step-by-step simulation
- some basic graph inspection / charts
Another reason I think this is worth doing is that a lot of game designers still go looking for the old free Flash-era version / publicly available legacy code, because it used to do this job locally and without all the friction. To me that says there is still real demand for a simpler and more open version of this kind of tool.
And I think AI makes this idea even more interesting, but maybe not in the usual way people mean.
Right now, if I ask AI to generate Python code for a gamedev system for my game, it often confidently invents things, makes mistakes, and only admits them after you push back. That gets old very fast.
But if the engine already exists, and the core building blocks are already programmed and reliable, then AI does not need to invent the logic anymore.
It only needs to place components, connect them, and configure rules inside a system that already works.
That feels like a much more realistic use of AI:
not “please write the whole simulator for me,”
but “help me assemble a simulation out of valid parts.”
That’s a very different task, and probably a much easier one.
I also like the idea of having an open alternative in a space where game designers don’t really have many choices.
It feels like one of those tools that could become genuinely strong if built in the open by people who actually need it.
So I guess I’m trying to understand one thing first:
Is this something people would actually care about?
Because if the answer is “not really,” then I’ll probably just make it as an internal tool for myself and not bother anyone.
But if there is real interest, I think this could be a very good community project.
I’d love to hear:
- whether you would use something like this
- whether Defold sounds like a good fit for it
- which features would matter most in a first version
- whether anyone would want to contribute
