Knowledge management¶

How knowledge gets organised, independently of any tool that stores it.

A domain directory rather than a nature one, and a deliberate exception to the folder/tag split: these pages are read as a group rather than found individually.

Layer build-knowledge as a values-to-blueprints derivation pipeline - the five layers build-knowledge stratifies into, and why each is its own layer
Separate audiences with separate bundles, not with folders or tags - why visibility is decided one level up from the content, and what that frees the folders to do
Split orthogonal classification axes across folders and tags - what to do when one hierarchy cannot express two independent questions
Categorize by what content is, not why you made it - the drift that files a security review under whatever you were doing when you wrote it
Record an abandoned exploration as an exploration, not a superseded decision - what to do with the record of a commitment you have since withdrawn
Give a knowledge base a tier that is not asked to earn its place - why the cheapest pages are the ones that fill, and what promotion does instead of ceremony
Date a page whose claim is about a version - when a date belongs in a filename, and why putting one on a principle reverses its meaning
Automatic session capture is not an inbox - why a queue that fills itself never reaches zero, and what capture must refuse to ask
Keep the GTD inbox inside the private bundle, not beside it - where a queue of pages meant to be deleted can live in a knowledge base without breaking it

Backlinks¶

The following pages link to this page: