How this wiki is graded and wired. Mostly notes-to-self so the tree stays consistent.

Complexity levels

Every note carries exactly one level/* tag. The intent is a learning order, not a value judgement β€” frontier notes are often the thinnest.

TagMeaning
level/0-foundationsStandard graduate background. Bloch’s theorem, second quantization, group theory. The stuff I assume elsewhere.
level/1-coreThe working knowledge of the field: the dynamical matrix, DFT+U, BdG diagonalization.
level/2-advancedWhat I actually use at the edge: exchange tensors from DFT, DC-DFT, LSWT, the imaginary-mode diagnostic.
level/3-frontierOpen problems and my own bets: flat-band SIE, density-driven error in MLIPs.

The tag pages (/tags/level/2-advanced etc.) give a cross-area view of one depth. The Atlas lays out all levels as a single tree.

Maturity

Frontmatter status: and a glyph in the Atlas. Garden metaphor, matching the site title.

  • 🌱 seedling β€” sparse, a stub, may be wrong.
  • 🌿 growing β€” has real content, still incomplete.
  • 🌳 evergreen β€” reasonably complete and trustworthy.

Seedlings open with a > [!warning] so the reader is warned on the page itself.

Frontmatter schema

---
title: Linear spin-wave theory      # display name; sets the H1 and explorer label
tags:
  - magnetism                       # topic tags, free-form
  - spin-waves
  - level/2-advanced                # exactly one level tag
level: 2                            # integer mirror of the tag, for a future custom sort
status: evergreen                  # seedling | growing | evergreen
date: 2026-01-20                   # ISO; defaultDateType is "modified" in the config
---

Linking

  • Wikilinks, full path from content root: [[physics/linear-spin-wave-theory|alias]]. Quartz resolves shortest-path too, but full paths survive folder moves better.
  • Folder MOCs: link the folder note as [[physics/index|Physics]].
  • Prerequisites are links, not prose. Put them in a callout so they create graph edges:
    > [!abstract] Prerequisites
    > [[mathematics/bogoliubov-de-gennes|BdG diagonalization]] Β· [[physics/dzyaloshinskii-moriya|DMI]]
  • Citations. A bibliography goes in a > [!cite] callout written as a > - list β€” bare > lines collapse onto one line under CommonMark; inline (Author, J. vol, page (year)) is reserved for the primary source of a single result.

Math β€” different from the main site

This is the one place the GRID and the main site diverge, so it’s worth stating loudly:

Inline $…$ is fine here.

Quartz parses Markdown with remark + Plugin.Latex, not kramdown. The $$…$$-only rule from the Minimal-Mistakes site (where kramdown reads _ as italics and single-$ math breaks) does not apply. Use $…$ for inline and $$…$$ for display freely. Bare | inside math is also safe β€” no Markdown-table ambiguity in remark math.