The core part of YaST providing the components system, the agents infrastructure and the implementation of the YCP language for backward compatibility.
The YaST agents are used to modify the underlying system. Their most important feature is that they can be switched to a different target system, which is heavily used during installation.
The agents for the target system are managed by a SCR instance. Each agent is attached to its unique YaST path. Read more about SCR, YaST paths and their relationship with the agents in the "YaST architecture" document at the YaST documentation page.
Each agent is defined by its scrconf file which can contain either a base with some parameters either a path to a binary if the agent is implemented by a script.
Agents that are implemented in scripts communicate via the YCP protocol and have predefined methods. Agents using a base just configure such base. The bases are libraries that are registered to SCR without a path and that usually provide parsing funcionalities for different file types.
- The agents can be registered during runtime. See SCR#RegisterAgent
- Ruby-bindings provide shortcut for constructing paths. See Yast.path
The core provides a components system that allows language agnostic communication between various parts of YaST. Code written in ruby and perl uses the respective ruby and perl bindings to attach to such system. There are also components written directly in C++ like libyui and the package bindings.
Beside providing backward compatibility, YCP is still needed as a communication protocol within the components system and with the agents. It is also used in agents scrconf file to define agent and its path.
More information about YaST can be found on its homepage.