This is a sort of qualitative question regarding AI and time investments... How big of a time sink is it to adjust let's say, the Campaign AI to appropriately handle custom units? I suppose this is a difficult question to answer, I'm sure it varies depending on how far that unit is removed from any sort of existing unit. So I suppose I can break it down into something like...
If I essentially were to duplicate a unit, would some sort of AI marker go with it? For instance, if I decided to simply duplicate all terran campaign units, and use these duplicated units (with whatever small stat modifications, or none at all) would the AI behave properly?
What if I changed the tech tree, such as changing a structure required to build a unit?
Added an existing ability to a unit that previously did not have access to that ability?
Created a new ability with a unique effect?
From what I can gather the Campaign AI more or less is a bare-bones framework, and just handles rebuilding lost units and structures, and is told via scripts to attack and expand etc, so I imagine it would be easier to tackle than a melee AI. I'm basically trying to make a judgment call on how much unique data I want to create for the project I'm working on, and have yet to dip my hands into AI at all. It's a front loaded cost, so it may be worth it in the long run, and I just wanted to get feedback from people who may have already worked with AI and custom units in depth. I don't expect exact numbers, more like estimates, educated guesses and suggestions.
If anyone has any tutorials they found tremendously useful for this topic, I'd gladly take suggestions. This is one place where messing around with existing maps is probably not going to cut it, and I'd rather not learn wholly by trial and error.
PS. I understand most AI editing would be done in Galaxy which is why I posted this here, though it's not a specific technical question, so if any mod sees this and feels it would best be placed elsewhere, please do so!
you would have to add alot to the AI to get it to use the new units
the only thing you listed that wouldn't require is the tech tree.
when you aisetstock the ai works out what it needs to build that unit and starts making the structures.
Rollback Post to RevisionRollBack
To post a comment, please login or register a new account.
This is a sort of qualitative question regarding AI and time investments... How big of a time sink is it to adjust let's say, the Campaign AI to appropriately handle custom units? I suppose this is a difficult question to answer, I'm sure it varies depending on how far that unit is removed from any sort of existing unit. So I suppose I can break it down into something like...
From what I can gather the Campaign AI more or less is a bare-bones framework, and just handles rebuilding lost units and structures, and is told via scripts to attack and expand etc, so I imagine it would be easier to tackle than a melee AI. I'm basically trying to make a judgment call on how much unique data I want to create for the project I'm working on, and have yet to dip my hands into AI at all. It's a front loaded cost, so it may be worth it in the long run, and I just wanted to get feedback from people who may have already worked with AI and custom units in depth. I don't expect exact numbers, more like estimates, educated guesses and suggestions.
If anyone has any tutorials they found tremendously useful for this topic, I'd gladly take suggestions. This is one place where messing around with existing maps is probably not going to cut it, and I'd rather not learn wholly by trial and error.
PS. I understand most AI editing would be done in Galaxy which is why I posted this here, though it's not a specific technical question, so if any mod sees this and feels it would best be placed elsewhere, please do so!