This is an experimental Platformer AI (V0.0.3) Should I release a B.E.T.A. tag game with this AI? In this environment, you can edit the 16*16 grid by clicking on the tiles and setting the goal for the AI who will then try to reach it if he can. Any feedback would be very useful, thank you in advance for your help. inputs : - Click the grid to edit the block hovered - Press e to set the goal of the AI (put it somewhere accessible for the AI or it won't move, meaning it needs to be touching red lines and have them connecting the AI and the tile you are selecting) Information / warning: - uses a graphing algorithm - uses an A* algorithm - The AI core ( engine + movement ) is 700 blocks - The movement can clog sometimes I'm working on it. You can simply move the goal somewhere else or edit the block clogging the bot to fix this. - The graphing isn't up to date making it so the AI can't fall down from more than 2 blocks. I'm currently thinking about how to implement bigger falls. How does it work : Each time the level is edited a new graph will be made. The graph is what the lines represent, showing which blocks can access which. Then an A* algorithm is applied to the graph when the goal is set to figure out the fastest path. After that, the AI will use its platformer movement mechanics, a movement that is exactly similar to the ones people use to control their player in platformer games but with the inputs controlled by the AI so he can move left and right as well as jump. Aim of this experiment : My goal is to add reliable enemies that can challenge the player in all your grid platformers. When this experiment will reach its end scratch will be provided with its very own platformer bot sprite that would be very easy to implement and use in your own projects with simple functions and utility. Log : -v0.0.1 Graphing algorithm (not polished) A* algorithm AI movement (not polished) -v0.0.2 Added gap jumping as well as different jump powers -v0.0.3 Added smart jumping What else should I add/fix?
