I’m not sure I agree with all the points made in the article. They are written and presented well enough, but it’s woefully incomplete to declare a single choice for all instances.
They chose only a handful of languages against which to compare, for one. Picking your targets based on the arguments you want to make is kind of disingenuous. Those are pretty popular but Cue, KCL, XML, XSLT, RCL, Dhall, ELCL, LTSV, Hocon, CCL,HCL, and UCL among others that all have their own communities and their own sites talking about their benefits for configuration and sometimes also for data interchange.
There’s no discussion about how much configuration your project needs and whether objects in your configuration are a benefit.
There’s nothing about what the other tools in your stack already use. Often it’s simpler to stick to as few formats as possible for a single stack.
I think “the only choice” is a pretty big claim given these drawbacks, but it’s a pretty informative article for what it covers.
I’m not sure I agree with all the points made in the article. They are written and presented well enough, but it’s woefully incomplete to declare a single choice for all instances.
They chose only a handful of languages against which to compare, for one. Picking your targets based on the arguments you want to make is kind of disingenuous. Those are pretty popular but Cue, KCL, XML, XSLT, RCL, Dhall, ELCL, LTSV, Hocon, CCL,HCL, and UCL among others that all have their own communities and their own sites talking about their benefits for configuration and sometimes also for data interchange.
There’s no discussion about how much configuration your project needs and whether objects in your configuration are a benefit.
There’s nothing about what the other tools in your stack already use. Often it’s simpler to stick to as few formats as possible for a single stack.
I think “the only choice” is a pretty big claim given these drawbacks, but it’s a pretty informative article for what it covers.
TOML - Toms Obvious Minimal Language