I say go for it. Who cares if it's never complete, I love toying with stuff like this. DF will never be complete, but it's still an amazing piece of software. It might not even be possible to run it on today's computers; doesn't matter. If you or someone else doesn't start it now it'll never be finished, and if we do end up having computers that can run it down the line (which I think is actually very likely), we'll have to wait thirty years anyway for it to be written.
This is the kind of stuff that large software companies look at for big ideas of the future. We would have never had bullets penetrating flimsy wooden walls in our boring FPSes if someone with your spirit hadn't made a simulation that did it and proved the nay-sayers wrong
We're only -just now- breaking out of the god-awful PnP Final Fantasy RPG molds, and only tentatively. It's not too difficult to do better once you have an idea of the difficulties.
There are gamers, and then there are simulation gamers. Most of the latter go on to become some variety of engineer, at the very least as a hobby. I usually get bored and quit with games that try for a story pretty quickly (although I enjoy them while I can play with their other features), but games that have dynamic and unpredictable but meaningful results I can play with endlessly. Don't think for a second you won't have an audience
I can definitely see it being a great learning tool. Dump an elementary student in a world, give him some pointers on how to survive, and make it rewarding enough that they'll want to play more if they're that kind of person. If a community forms, people in it will inevitably do crazy things like try to find flaws in your chemical reaction for tanning, or build structures that can't exist in the real world. The kid will grow up learning this stuff as play, which is all you really need to get through any kind of schoolwork. If you actually enjoy studying the world around you, you're set for life as a scientist.
As for actual research... Start with what you know, start coding right away, and leave stubs where you think you might be missing mechanics. It's a lot easier to figure out what you need to add when you have something that you know needs work, because then you know where to look. Don't be afraid of making stuff that you might need to throw away later. That's a given, since you're new to coding. If you keep good notes on what you're putting in, like physical constants/formulas that you had to consider, it's going to be a lot easier down the line to rewrite your functions should you screw up something in program structure.
I personally would start with geology, making a natural landscape to explore using plate tectonics, erosion, soil creation (though this involves plants and animals being made that made it possible to eventually process rock into the crazy-complex nutrients that compose topsoil), and all the various processes in between. But that's because I had two geologist parents. Even that is a huge undertaking, but if you get it right you've got something you can sell to a niche and build on.
Author
Topic: In which Sonerohi tries to understand everything. (Read 2330 times)