LÖVE AI

[ajinkya]

I want to know whether anyone has written AI module for LÖVE.
If not then i am interesting in writing that.
So should i start coding?

[Robin]

So... that's pretty general.

I've written plenty of AIs for LÖVE games. But a general "AI module" is useless.

[trubblegum]

I would tend to agree, unless it is able to efficiently make fuzzy judgement calls based on a range of parameters and simplify abstraction considerably.
Even a simplified-for-efficiency and pre-programmable condition-based neural net could be a nifty and time-saving addition to many (but not necessarily most) game engines.

I think it could certainly be an interesting project, and worth watching, but more details re just what you mean would be helpful.

[teh8bits]

viewtopic.php?f=5&t=8177
That one is pretty cool imo

All you need though (unless you're making conversation) is a few lines of code that make the enemy draw closer to the player really, maybe throwing something from time to time

[Inny]

Game AI typically comes in two forms: pathfinding and decision making. The pathfinding part is already well covered by A* libraries (here's the one on the wiki: https://love2d.org/wiki/LoveAStar). As for decision making there's not much that would be useful to you, but the best tools to use there are either state machines, or lua's coroutines which I like a lot.