Skip to content

📚 Study Method#

"Reading isn't understanding. Understanding is being able to break it and put it back together."

This path is not a reading list — it's a curriculum. The method below exists to turn what you read into competence. Read little, do much, prove it.

⚖️ Read/do ratio#

Rough target: 1 unit reading → 2 units doing. Don't stop once you finish a module's "Read first" table — the real learning happens in the lab and in the broken lab. Reading a concept makes you think you understand it; applying it to a system and breaking it is when you actually learn it.

Stage What you do Goal
Read Go through the sources in the "Read first" table Concept + the "why"
Build Finish the labs/build/ task Being able to do it by hand
Break & fix Solve the labs/broken/ broken lab Diagnostic intuition
Prove Pass the acceptance criteria with a command Objective verification
Explain Write a concept in your own words Retention (active recall)

🔁 Active recall and spaced repetition#

  • At the end of each module, answer the 🧪 Test yourself questions closed-book.
  • After taking a week's break, re-run one of the previous block's acceptance criteria.
  • If you can't explain the concept to someone else (or to a blank page), you haven't learned it.

This repo is a curator, not a copy of the entire world. The goal isn't "no need for any other source" — it's "no moment left without direction." That's why every external link in this path fills these four fields:

Source Why you're going Exactly what you'll do there Duration How it's verified when you return
(example) official Terraform tutorial See the state concept from the official source terraform apply → inspect the state file ~30 min Compare terraform state list output against the module's lab

An external link without all four fields is a link dump, and it's forbidden in this path. If you see a link in a module that just says "check this address," that's a bug — it gets fixed in the Phase 9 audit.

Two exceptions — these are not link dumps: 1. Just-in-time single-fact reference. A source you check to grab a single piece of information instantly — a man page, a tool's wiki, or the current version number on a GitHub release — rather than "going to learn something," doesn't need all four fields. It's a dictionary lookup, not directed reading. 2. In-repo "Read first" links. The deep-dive links in a module's 📖 Read first table are not external sources; source + purpose + duration (three fields) is enough. The module's acceptance criteria already handle the "verify on return" job.

🧱 When you get stuck#

  1. Check the module's 🆘 If you're stuck table (symptom → likely cause → what to do).
  2. General errors: TROUBLESHOOTING.md.
  3. In the broken lab, the hints/ folder is staged: hint-1 (direction) → hint-2 (narrow down) → hint-3 (near-answer). Open them in order; don't open solution.md on your first move.

Getting stuck isn't a malfunction — it's the process itself. Narrowing down an error without help is the actual skill this path teaches.

🚫 Anti-patterns#

Anti-pattern Why it's bad Correct approach
Reading an entire block without doing any labs Reading isn't competence; you'll collapse on the first real task Finish at least one lab per module
Passing an acceptance criterion by saying "I get it" Subjective; unprovable Prove it with a command / file existence / written output
Opening solution.md at the first difficulty You never exercise your diagnostic muscle Start with hint-1, form your own hypothesis first
Skipping a block ("I already know this") The gap shows up later as a wall Pass the check test, then skip
Diving into an external link and never coming back You end up directionless, hours lost Follow the four-field contract: duration + return verification

📋 Checklist — before counting a module as "done"#

[ ] Went through the "Read first" sources
[ ] Finished the lab, verify.sh passed with zero errors
[ ] (If present) solved the broken lab without help (at most with hint-1/2)
[ ] Proved every acceptance criterion with a command/output
[ ] Answered the test-yourself questions closed-book
[ ] Saved the portfolio artifact (if any)

"Speed isn't the goal. Actually passing one block is worth more than half-knowing two."