Mapping State

DEVELOPERS SECTION

In order to remove the use of funcs and procs it needs to be possible to map all state and control mechanisms into the observable space. For example, all agent meta data and loading ought to be possible via the control and observation of observables. The purpose of this page is to identify all the things that need to be mapped and how this might best be done.

Agents

There are two aspects of the agent system that need to be mapped to observables. 1) The browsing of all available agents. 2) The meta data and control of loaded (and unloaded??) agents. This means that observables must exist which contain lists of all available agents, along with observables for all versions of a particular agent, the title of an agent etc.

• agents_list - List of all agents available in the system???
• agents_loaded - List of all agents that are currently loaded.
• agent_[name]_versions - List of all versions for this agent
• agent_[name]_title - As found in the doxygen header.
• agent_[name]_author
• agent_[name]_script - The string of the agents script.
• agent_[name]_history - A list of history entries for the current version loaded.
• agent_[name]_active - Boolean to activate the agent (the when statements)
• agent_[name]_errors - List of all current syntax and runtime errors for this agent
• agent_[name]_hasErrors
• agent_[name]_version - Currently active version name/number
• agent_[name]_children - List of all child agents of this agent.