There's also a hidden adventure side to it. It's not 'the game' but it's 'a game' - if you don't know about it, you're not affected. If you do, there's a new spin to the game to explore.
And just to throw a word out there that'll get the flames started: It'll (eventually) be MMOG. !
But honestly, I should qualify that a bit...The game won't be MMOG to begin with - it'll be AI-based and single player. But the scope is there to make it MMOG. I've already created (many years back, now) a small-ish scale MMORPG game that was browser based, and have since developed a pretty smart C++ server that should be able to stand up to a few thousand connections nicely - although looking at RakNet I'm starting to wonder if it was necessary! But, to overcome some of the fallbacks of MMOG's, I've worked out a few limitations already:
- Continuous space will probably be used, not 'zoned' - but in fact it will be zoned in the background. There will be VAST areas of empty space which you can, if so inclined, traverse - but using 'technologies' (like extremely powerful boosters) you'll be able to move quickly through these 'dead' areas.
- Each server will only control as much space as it can realistically handle. After a lot of looking into this, the only way (really) was to make the 'area of control' a square. If the server is under-used, it'll gobble up another line of coordinates that it controls, and visa versa.
- I toyed with the idea of releasing the server software (when developed!) into the wild for people to 'join in' to the server hosting, but it's too dangeous as the server will be authoritative.
- Each client will use auto-update code at the beginning of the game, so old clients will be quickly replaced with new ones.
That's all future stuff, however. The game itself will be developed as a single player game. Unfortunately, as a coder will little expertise in model development, some things are taking longer than hoped for. But my real emphasis on this game is game-play, as I'm not worried about graphics. It's 2D even though it's so easy to turn it into 3D with the same models - this is because I find 3D space-games tend to lose focus, and simply are more confusing to play! I want to remove that, and make the game the sort you could just fire up to play for 20 minutes instead of having to make a full session of it.
At present I don't need any help - I've settled on Irrlicht having used GDI, SDL, DirectX, and then Irrlicht. I'm happy with the superb featureset it has, and have already got the space-ship working (nearly) as I want it to. There's issues with texturing, lighting, and a tiny bug in the movement (which is a rounding problem!) but otherwise I'm happy.
I'm extremely tempted to create a client-server based model for the single player version itself - it adds compications, but it means when the game is eventually extended to work with more people, the architecture is already in place. That would mean that every movement needs validation from the 'server', and every AI move is done by the server also. Plus it makes abstracting data and display much easier.
I'd like to know what peoples initial thoughts are, and also what people think about abstracting client/server immediately even for a single player game is like.
And naturally - I have ideas flowing about what to put in to the game - but being realistic, getting an initial 'world' up and running is more important to begin with!