thanks for the comment :)
yes, I know, there lot's of work on input and collision, the current one just suck xD hahaha
I'll try to do a live stream about how make then better today, you can follow it here ^^
https://www.youtube.com/channel/UCV6lLuuVfN0Agq2xfaWIPjg
If you're going to keep that scale, you might center it too - possibly put some image fluff on the edges. Consider the artwork on the actual cabinet - or use your imagination. :)
https://s-media-cache-ak0.pinimg.com/736x/90/50/6f/90506fb633777858dad68f1dbb09b61a.jpg
I really loved the "arcade frame" idea :D I'll try it ^^
like this one :D
https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcRi2lF3yRzeIqkwyX9sJx4YWtAqIRxLGM5gaZVQlKDcVpnwWZN5iQ
the pico-man is incredible beautiful! I played this one yesterday, UrbanMonk did a great work on it :D
but the idea from my project isn't to create another pac-port, my idea is create research material in my youtube channel, not just about pico-8, but about game and development, from C to Unity and Unreal...
another project of mine is the LEFT+START, where I'm developing and documenting everything about GameBoy development :)
oh! you can watch all the development here https://www.youtube.com/playlist?list=PLS_eTm_uMyblZ4bFeRlhWXKNY2IKXo1AY
If you need help figuring out how anything worked in the original, the excellent Pac-Man Dossier is essential reading and very interesting, I might add.
I've already got some experience with coding Pac-Man AI, so when you get to that point, check out my blog post here on the BBS: Animation, collision and AI, oh my!, where I explain in good detail how the AI works, and how to implement the pathfinding algorithm, with working examples throughout.
Basically the heart of it is this:
function dst(fx,tx,fy,ty) return sqrt((fx-tx)^2+(fy-ty)^2) end |
What it does is it draws an imaginary line from the coordinates of each tile that it can choose to move to, to the coordinates of a target that it's trying to reach. Whichever one has the shortest line, that's the direction it moves. Very much like what is illustrated here from the aforementioned Pac-Man Dossier:
As you can see in this example, the red enemy can choose to move up, left, or right (in most cases in Pac-Man, the AI can not turn around and move in the opposite direction of what it was just moving, so down is not an option in this case). It measures the distance, using the function I posted above, from each option to the green target, and we can clearly see that it will choose to go right here because that's the shortest option to reach the target :)
So it's very simple really, not like a complex algorithm like A*, but the simplicity of it does introduce flaws. Since it's only planning ahead 1 tile instead of actually planning an entire path and ensuring that there are no obstructions anywhere between that path and the target, it is possible that it can be tricked or follow a path that will not actually end up leading to the target. As you can see, it's possible to "hide" from the ghosts in Pac-Man: https://www.youtube.com/watch?v=aFGoiQAzG-I
Scathe! this material is awesome!! I'll read it later :D
I, too, have wondered about the wandering mechanics of the ghosts in Pac-Man. I did notice that there was always one ghost who was out for me and only me and the others seemed to just follow a trail.
Curious to see some research on the subject !
I think there a gamecenter cx episode who they talk with the Toru Iwatani (the pacman father), and they said how the ghost works, a very nice thing, he show a spot where no ghost can go! I'll try to find and post the link here :)
Another thing, no new lives, why?? I'm working in a new project, a new original arcade machine (not mame, a real thing!), so, I'll do the live just on friday :D I know, I can keep the development offline, but it isn't the same thing as I'd like to document the entire development :)
here my channel, I'll post some videos about gamedev and tutorials very soon ;D
https://www.youtube.com/channel/UCV6lLuuVfN0Agq2xfaWIPjg
[Please log in to post a comment]
