This contains implementation details for behavioral trees.
This includes database structure, data retrieval,
node prototypes vs node instances, and how a generic compact solution was achieved
Note: This is one of the more difficult and complex topics, when we talk about
the structure of generic behavioral tree redaction system...
So first off, start with what we require from the system?
We want to be able to use behavioral trees in a game server
We want to be able to edit behavioral trees (have a WYSIHYG editor)
Make adding a new type of node (or decorator) as easy and as simple as possible
In order to achieve this, we will use a SQL database to store the behavioral trees.
We will also store the node prototypes into the database,
and allow to add / edit / delete node types from this website.
The same CRUD operations will be available for node attributes.
2 Database structure
Let's start with database structure. The following diagram will explain more than thousand words: