YAML indentation bugs cost me a Friday night once. Now I use TOML for most configs. Here's my practical guide to picking the right format.
I used to think YAML was fine. Then one Friday night, a misaligned space in a CI config took down our staging environment. That's when I started looking at TOML seriously.
TOML (Tom's Obvious Minimal Language) was created by GitHub's co-founder specifically to avoid this pain. No whitespace sensitivity, explicit syntax, and it's now the standard for Rust and Python projects. Here's when I use each format.
Each configuration format has distinct strengths and weaknesses. Understanding these helps you choose the right format for your project.
TOML excels for application configuration where readability and explicit syntax matter. Its lack of whitespace sensitivity makes it less error-prone than YAML.
YAML is ideal for complex nested configurations and data serialization where its features like anchors and aliases provide value.
JSON is the standard for data interchange between systems. Use it when your configuration needs to be consumed by APIs or generated programmatically.
Converting between TOML, YAML, and JSON is straightforward since they all represent similar data structures. However, some features are format-specific.
TOML uses a simple key-value syntax with tables for grouping related settings. Square brackets define table headers, and double brackets create arrays of tables.
For configs humans edit? Yes. TOML has no whitespace sensitivity, so you can't break things with a misplaced space. YAML is better for deeply nested stuff like Kubernetes manifests.
Stick with YAML for Kubernetes. The entire ecosystem uses it. Fighting that convention isn't worth it.
Unambiguous syntax. No whitespace gotchas. Supports comments. Rust wanted a config format that wouldn't cause debugging nightmares.
Paste it into our TOML formatter. It validates syntax and highlights errors instantly.
Founder of CodeUtil. Web developer building tools I actually use. When I'm not coding, I experiment with productivity techniques (with mixed success).
I convert CSV to JSON probably weekly - data imports, API migrations, analytics exports. Here is everything I have learned about doing it right, including the edge cases that used to break my scripts.
After pasting sensitive API responses into random online tools for years, I finally did a proper comparison. Here's what I actually use now and why privacy matters more than features.
After years of config file debates at Šikulovi s.r.o., here's my honest take on when to use what. Spoiler: I use all three, but for very different reasons.