ontological approach, some notes on

tl;dr capture widely → select aggressively → structure deliberately → explain clearly

The State of Play

  1. Input is garbage. Everything starts as noise. If you don’t filter it aggressively, your “Information Architecture” is just a digital landfill or hoarders paradise.
  2. Context is the only thing that matters. A note without a link is a dead end. Meaning only exists in the relationship between the parts.
  3. If it isn’t structured, it doesn’t exist. If a machine can’t parse it123 and a human can’t find it, you didn’t “capture” it; you just lost it in a different way.
  4. The pipe is the process. The “Matryoshka” isn’t just for the code; it’s for the thoughts. Micro → Meso → Macro is just a fancy way of saying “Stop, Think, and Edit.”
  5. One is enough. You are the only audience that matters at the start. If it doesn’t work for One, it won’t work for anyone.

This is nothing novel. Spelling it out and putting it together in a deliberate configuration is intentional and deliberate.w

Hierarchy of Needs

knowledge management := data → information → knowledge

systems theory := parts → subsystems → whole

If we apply these concepts to build an information theory framework:

scale := micro → meso → macro

meaning := raw → collected → interpreted

DIKW set := (Data → Information → Knowledge → Wisdom)

Movement

There is a fundamental simultaneous hierarchy (micro → meso → macro) and progression of information and complexity (simple → gathered → explained) and content (atomic → compound → amalgam).

micro (atomic units)

note
random snippets of information
card
the infamous data point written on a virtual 3 × 5 index card
social
tweet / toot / whatever

meso (aggregation and filtering)

summary
list of many micro elements
briefing
many micro elements
social
note / post

macro (synthesis and meaning)

essay
from summary and briefing
social
post

$LastModified: 2026-05-12 08:55:10Z (Tue, 12 May 2026) $

  1. The solution to having something machine readable is not JSON-LD. On the one-hand, it is a wonderfully extensible graph that appears to bring order to the chaos and makes many ontologists smile when they see it. On the other-hand, it is an extensible graph, and the trap is TIMTOWTDI; flexibility allows ambiguity, reduces rigor, and causes tears when somebody has to consume this mess. ↩︎

  2. The solution is not microformats, because that is mixing presentation (HTML) with data (MF/MF2) in the same layer. The indieweb.org people are obsessed with not violating DRY, but they are idealists who are stuck with self-imposed constraints and have failed to realize and materialize any gains in two decades (which in internet years is a lot). Let us not forget that while the name “microformats” was coined in 2004, the underlying concepts had been around previously (and, ahem, were a bit of a cul-de-sac); they took something and “productized” it. ↩︎

  3. The pragmatic solution is to have something strictly-loose, or loosely-but-strict, depending on your perspective. Have it a simple key=value mapping (which is the loose part) but do not allow it to nest (which is the strict part). Essentially, a graph is presented as a flat structure. Use the graph for processing, but the flat-structure for information ingress. This is not new, PDF did it, and it was an elegant solution that few people ever saw and fewer recognized for its beauty (but I did, hehe). ↩︎