Yes, this is exactly right. YAML started way before JSON and isn't a derivative work; JSON started later and isn't a derivative of YAML, either. It just happened that JSON was oh-so-close to being a proper subset of YAML at the time, so people were complaining about the few remaining edge cases. We therefore went over the YAML spec and tweaked a few things to ensure YAML was a proper super-set of JSON. We couldn't do that without referring to JSON in the spec, of course.

Given JSON's popularity, this was a good move IMO. We now have a very simple machine-readable format (JSON) and a clear growth path for increased human readability and additional features (tagging, references, etc.). It also allows using YAML libraries to directly load JSON data.

Have fun,

    Oren Ben-Kiki

On Tue, Jun 28, 2011 at 8:15 PM, William Spitzak <> wrote:
Isn't this due to the desire to have all JSON files be valid YAML files?
If that is desired then I see no way to avoid referring to JSON.

Trans wrote:
> I was reading through the Specification today (
> spec/1.2/spec.html), something I hadn't done in some time, and found
> it rather curious that JSON has taken such a prominent role in the
> specs. I understand having a section to discuss the relationship
> between YAML and JSON, but not some of the references that appear to
> actually co-op aspects of the JSON spec for YAML.

All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
Yaml-core mailing list