I'll be starting on my two-month project, but it may actually take three or four. It's a sci-fi "horror" game called "Blue Bloom". You'll have to clean a planet of an infectious, parasitic, fungus while stationed upon an island. This means that most clean-up efforts will be long-range. Cures will have to be analyzed and built via a crafting and modular building system.
This project will use resources from my two previous projects, along with adding more resources. while it will draw from previous experience, I'll be learning new things as well. Currently I'll be setting up a basic player and testing distance. I'll be adding a LOD system to this game for optimization. I think large faces render slower when they are close to the player, so this should break that up.
There will be several different spawn caps for creatures, both soft and hard. The soft cap is just for normal spawning, while the hard cap controls "event" spawning. Since creatures may be spawning in/out regularly, I'll be using signals to coordinate the spawns; reducing lag. This entire project is me seeing how fast I can put a game together when I have numerous resources.
Little One will still be worked upon, and updated. Many ideas from that game will be adapted for Blue Bloom. Not only the vacucell, but the crafting materials as well. A few materials will be added in Blue Bloom, making the count about twelve or more. If I use preexisting resources, most of this project will just be coding.
The map for Blue Bloom will be comprised of smaller scenes, also comprised of smaller scenes. Each map piece will be made of cells. The idea is to make a handful of specific cells and use a lot of them to build the map pieces. These pieces will be controlled via scripts to change their mesh or visibility as needed. I think the map will run a bit faster if I set the cell visibility even outside the player's view. I'm not sure if the creatures will have navigation abilities, but I'll try to include that from the start.
The map will be 8x10 areas in size and the areas might be roughly 10x10 cells. That's 800 scripts running a loop of 100 each frame to handle cell visibility, or is it? If I only check one cell per frame, I eliminate all those loops. The script can get through all the cells in 1.5-3.4 seconds (depending on frame rate). This will cause some problems with low-frame rate or lag.
While a single script could handle all the map areas, iterating all the cells of all the map areas every frame or X-frames could cause lag. To reduce even more lag, I might just check the nine map areas around the player.
To do this, I'll probably surround each map area in an area node and give them all grid coordinates. When the player enters an area, the current grid coords will be changed and I can write a script that organizes and acquires the map areas.
Creatures will change their own animation progression and visibility since they'll all have scripts. They should spawn within a pair of ranges from the player, to be changed by other factors. Any critters spawned will be direct children of their spawner and should only travel a set distance from their parent before returning/despawning. To keep the pressure on the player, I'll be playing fast and loose with the spawns.
Major hard points will include the sky, navigation, creatures, collision shapes, and map handling. I'm in no way ready for this project, but I'll be doing it anyway.
No comments:
Post a Comment