Work on my animation editor is going well. I’m well on track to my goal of having all major functionality in by the end of the month. All that remains are skin animation and the ability for animations to provide a spatial displacement. Once that’s in, it’s pretty much good to go.
This, of course, raises the question of “where do I go from there?” The first element of that is what I do with the animation editor and system. I’m inclined to put it in the public domain, and I plan to dig through my options in that regard. More than a few people have already come up to me asking for help implementing their own 2D animation systems, and being able to give this to them for their own uses or reference would be great.
Of course, I designed this with the intent of using it in a game, so I’ll have to start working on my game engine once more. Initial integration shouldn’t be too difficult, since it’s all been designed to port nicely, so the real difficulty will be managing a data driven way to drive these animations in a useful manner. That means that I need to have a working scheme for managing behaviour trees (my current editor may prove too unstable/unfriendly, though I certainly hope it holds up to extended usage), and the logic for them must work in-game.
After that we get to the fun stuff.
Looking back at the John Smith prototype from before, I want to ask myself the question: what was fun about it? What stood out? To me, these were the elements that stood out as enjoyable:
- Manoeuvring about the environment for tactical advantage
- Ebb and flow of combat, constraints that have to be taken into account (recharging health, limited ammo, etc)
- Exploiting or creating a flaw in an opponent’s position to do them in
Now, how can I use the animation system I’m building to get the most out of that?
I’m thinking that taking the environmental interaction to the next level is really the way to go. Focus on really cementing the player character into the world and giving him options to exploit it for his benefit. In the prototype, the player felt pretty disconnected from everything… he was basically just an accelerating brick that could shoot bullets around and collide with the world geometry. What would it take to improve on that?
Here are some preliminary ideas:
- Previously, players would walk through cover like it wasn’t there, and it would only block bullets. So lets have players collide with the cover and:
- Hop, vault, or clamber over low obstacles without forcing the player to jump, so long as you’re standing and moving
- Sidle up against it when using it as cover. When supposed to be completely protected, huddle in close. When popping up to shoot, present a minimal target profile.
- Be able to lean a shoulder into it and push it forward, like the box puzzles in Legend of Zelda.
- Have more meaningful interactions with the static geometry than “this stops you from moving forward or falling downwards forever”
- Be able to treat low walls in the geometry the same as any other piece of cover, as above
- Being able to cling to a ledge and pull yourself up, or hang there. Possibly be able to shoot overtop while clinging
- Advanced manoeuvres
- Diving and rolling (hitting crouch while running) to avoid fire
- Wall jumping
Okay, not the least ambitious thing I’ve ever planned, I’ll give you that. But I think the behaviours listed above would make moving throughout the world more fun, more kinetic, and give you many more options!
It sounds like a barrel of fun to me, so that’s the direction I’ll be heading. 
