YAML — YAML Ain't Markup Language

Reference & Beautifier

# Strings — bare or quoted
name:        Ada Lovelace
title:       "Countess of Computing"
bio: |
  First programmer.
  Wrote the algorithm.

# Numbers, booleans, null
age:         36
pi:          3.14159
active:      true
nickname:    null

# Sequence (list)
scores:
  - 98
  - 100
  - 87

# Nested mapping (object)
address:
  city:      London
  postcode:  W1A 1AA

# Anchors & aliases
defaults: &defaults
  timeout: 30
production:
  <<: *defaults
  timeout: 60

key: — mapping key string — text value 42 — number true/false — boolean null — nothing &anchor *alias — reuse # comment — ignored

string

bare · "quoted" · 'single'

Bare words are strings. Quote when special chars needed.

number

42 · -7 · 3.14 · 1e6

Integer or float. Written without quotes.

boolean

true · false · yes · no

true/false (also yes/no, on/off in YAML 1.1).

null

null · ~ · (empty)

Explicit absence of value. Tilde ~ is shorthand.

mapping

key: value

Key–value pairs. Keys end with a colon and space.

sequence

- item1 - item2

Ordered list. Each item prefixed with "- ".

block scalar

key: | line one line two

| preserves newlines. > folds them into spaces.

anchor/alias

&name · *name · <<:

Define once with &, reuse with *. Merge maps with <<:.

1996

Origins: Inspiration from Perl & Python

YAML draws early inspiration from Perl's data serialization module (Data::Dumpster) and Python's indentation-based block structure. The goal: a format more human-friendly than XML.

2001

YAML 0.1 — First Draft

Clark Evans, Ingy döt Net, and Oren Ben-Kiki collaborate on the first YAML draft, presenting it at a Python developer conference. The name originally stood for "Yet Another Markup Language" before being backronymed to "YAML Ain't Markup Language."

2004

YAML 1.0 Released

The first stable specification is published. It introduces the core concepts: mappings, sequences, scalars, anchors, aliases, and document markers (--- / ...). The format is already widely used in Perl and Python communities.

2005

YAML 1.1 — Boolean Expansion

YAML 1.1 adds alternate boolean values: yes, no, on, off. While convenient, this later becomes a source of the infamous "Norway problem" — the country code NO is parsed as false.

2006

Ruby on Rails Adopts YAML

Rails uses YAML for database.yml and i18n locale files, dramatically boosting YAML's mainstream adoption. YAML becomes the de facto config format for Ruby projects.

2009

YAML 1.2 — JSON Compatibility

YAML 1.2 is published, making every valid JSON document a valid YAML document. The alternate boolean values (yes/no/on/off) are removed from the core schema, addressing the Norway problem. This version is still the current specification.

2012

Ansible & DevOps Rise

Ansible (2012) chooses YAML for its playbook format, cementing YAML as the language of infrastructure-as-code. Soon after, Docker Compose, Kubernetes manifests, and GitHub Actions all follow suit.

2014

Kubernetes Uses YAML

Kubernetes is open-sourced by Google using YAML for all resource manifests (Deployments, Services, ConfigMaps). This becomes the single largest driver of YAML usage globally, with millions of engineers writing it daily.

2019

GitHub Actions Launches

GitHub Actions uses YAML workflow files (.github/workflows/*.yml), adding hundreds of millions of YAML files to public repositories. CI/CD pipelines across the industry consolidate on YAML.

2021

YAML 1.2.2 — Formal Errata

YAML 1.2.2 is released, fixing ambiguities and errata in the 2009 spec. It introduces the "Core Schema" as the recommended default, dropping legacy 1.1 boolean aliases while preserving compatibility with 1.2 compliant parsers.

2020s

YAML Everywhere — and Alternatives Emerge

YAML dominates DevOps, cloud, and config tooling. Meanwhile, TOML (simpler configs), CUE (typed configs), and Jsonnet (templating) emerge to address YAML's complexity at scale. YAML remains the most widely used human-writable data format.