Knowledge Node

Finite Automaton of Knowledge States

An RWX knowledge node is simply some amount of knowledge at its most granular level. If you imagine your brain as a cognitive, finite state machine. Then a knowledge node is one finite state, which links to may others as well. Each time you follow a link you change the state of your cognitive circuitry. A knowledge base is therefore cognitive programming you can load and run. Your brain processes the same nodes following different paths and KNode states.

The practical implementation of a knowledge node manifests as a simple subdirectory containing a single README.md within the subdirectory. Knowledge nodes may contain any other assets — including other subnodes — but should not contain other Markdown files unless those files are examples or other reference material.

Only the main README.md Pandoc Markdown file will be rendered as HTML.

Limited Size

Like a good program, a knowledge node should not be longer than a 2-3 screens of text (about 300 lines of Markdown). This is to limit the required computing resources to display a single node and afford the best opportunity for node aggregation through hyperlinking.

Aggregation, Not Composition

A single knowledge node can be logically aggregated from many others. Aggregation takes the simple form of hyperlinking to other nodes. Even though the linked content does not actually become a physical part of the linking node the reference to other nodes is a critical dependency since the linking node cannot exist without the linked node.

Aggregation is a critical dependency from one thing to others whereby if the thing depended on goes away it breaks. Composition is when the that dependency goes both ways and if either thing goes away both break.