Jump to content

Artificial Opponents, Overview of Methods of 0 AD AI Programmers


Recommended Posts

I don't mean the thread title to be misleading, you just have limited space to describe the topic.

As I work on my own game engine I have slowly been working on the concept for my AI system. I want it to be able to deal with new maps and units without my specific input. Now realistically I don't expect to have many modders making game factions, but I am sort of compulsive and perfectionist so I want it to work theoretically even if I am just making games to play by myself.

On that topic, what sort of AI are people making for 0 AD? Any of the standard approaches? Anything original? How much simulation data is exposed to the AI? How does your AI read the map? How does it decide what units are for? Is it a tag system where the faction designer specifies a type tag? Can the AI look at the units stats and figure out what they are for?

How does it work? I had been considering trying out some of my AI theories using 0AD since my engine does the AI in source and I don't have multiple well designed and completed factions. How much power does the API have to work with?

Also I just like discussing AI and only one or two sites on the web are dedicated to AI for video games.

Link to comment
Share on other sites

Ok, I think we really need a centralized wiki-article about everything related to AI. If you are interested in AI and able to understand JavaScript you can have a look at the sourcecode in the directory /binaries/data/mods/public/simulation/ai/

or in here: http://trac.wildfiregames.com/browser/ps/trunk/binaries/data/mods/public/simulation/ai.

I also started to get interested in AI and sadly there's not a little bit of documentation that I've seen so far, well except of one wiki-article with two links which are outdated: http://trac.wildfiregames.com/wiki/TDD_AI

The best documentation you get is what you find out by reading the source, sorry! ;)

I think I'm starting to document a little bit in the above wiki-article, but this month I've got to succeed in my A-levels-tests, then I've got a few months without school and studying were I've got enough time. But for now you may ask programmers of the team who have worked the AI out and know much better than me, e.g. quantumstate. He's a nice guy and will also answer your questions if you enter the IRC-Chat. ;)

Link to comment
Share on other sites

Ok, I think we really need a centralized wiki-article about everything related to AI. If you are interested in AI and able to understand JavaScript you can have a look at the sourcecode in the directory /binaries/data/mods/public/simulation/ai/

or in here: http://trac.wildfire.../simulation/ai.

I also started to get interested in AI and sadly there's not a little bit of documentation that I've seen so far, well except of one wiki-article with two links which are outdated: http://trac.wildfire...com/wiki/TDD_AI

The AI/Engine low-level API page which you linked to is up to date, it is just lacking information.

Link to comment
Share on other sites

So it looks like a pretty standard planner type AI. How well is it holding up against medium level players? And does it do anything that is commonly referred to as cheating like seeing the whole map and so forth?

Link to comment
Share on other sites

The qBot Structure Overview should give some overview of qbot. (Most of it should still be correct)

Ahh, now this is nice!

The AI/Engine low-level API page which you linked to is up to date, it is just lacking information.

Ooops, sorry, I didn't know that, it seemed a little bit too short for me. I didn't want to offend you or anything!

Edited by Almin
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...