What would be the implications of using just one scene for everything? Would I get a lot of memory consumption or anything like that? My strategy have been to use Unity3D as a view engine for my game logic.
I already have a scene-management system in place in my framework for creating and destroying objects, lights, cameras, etc.
Would there be any problems considering memory and such?
Depending on how you build it, putting everything in one scene could cause some drawbacks.
In the pro version you can use occlusion to help with this, found here:
If you set up your code properly so that functions aren’t all running when they don’t need to be, and objects aren’t being rendered when they don’t need to be it could be fine. You just have to be careful about how much you let go on at once. Things like your example of removing lights would help.
Do you have more examples of the types of things you’d like to mash into the one scene so we know how far you’re thinking of going?
We all know how annoying it is when playing a game, and there is a 30-second wait between hitting Next Level and when it appears. That’s a major advantage. You can swing the camera from the menu to the game if you want, making really professional looking graphics easily.
However, if each level uses a lot of different geometry, it might be better to just use different scenes. See how long it takes to load one of your complex scenes from a different scene and decide whether that’s acceptable to you.
In my game, there is just one terrain, etc., which is the same throughout the game, but the missions are different with different AIs. So I’m going to stick with one scene.