| AI Rewrite - How Reasonable[message #303936]
|
Thu, 26 April 2012 13:26 
|
|
 storytime |
 |
Messages:19
Registered:March 2004
Location: Ontario |
|
|
I have an idea and I'm looking for advice on this from people familiar with the source. I started learning C++ recently and I'm mainly interested in deep logic systems. I don't have any "substance" of my own to work with at this time. So, I'm thinking that because the whole source is available for this game it would be a good platform to test things, whereas generally not much can be done with other games due to lack of accessible variables to base decisions on. Basically, the idea in this context would be to just make the individual AI soldier player as tricky as possible. What I would not do is have a god perspective decision making like the player does. There would probably be some interaction based on soldier rank. It's not primarily for fun play: I would optimize, but not limit it overall for the sake of quick turns on slow machines.
I haven't looked at the source yet. The trillion dollar question: does this idea seem reasonable based on how the code is organized and commented and how much engine stuff there is to use for logic? Would it take all day to find the variable for something that I want to affect a decision? I've read things in other threads that scared me a bit. Another thing I'm wondering about is how to keep updated with frequent changes to the SVN, because I would like it to be a usable mod. Let's say I'm working with a version of a file and it gets changed. Is it a hair puller merging the changes every time?
Report message to a moderator
|
Private
|
|
|
|
|
|
Blog
Search
Help
Members
Register
Login
Home
