Why lodash instead of underscore?

[adammlr]

Seems like you have a strong preference towards lodash in your projects, I've read a couple comparison articles, but was curious as to your reasoning.

[kentcdodds]

Underscore is super cool. But yes, I do prefer lodash. Here are a few of my reasons:

• lodash follows semver, underscore does not. I cannot trust it.
• lodash is faster (sorry, looks like benchmarks are down currently).
• lodash has more functionality
• @jdalton is one of my favorite devs ever (maybe he can chime in with some additional reasons).

It is notable that lodash and underscore are in the process of merging (🎉) into underdash (for the record, I'm not a fan of the name).

Hope that helps you understand the reasoning behind my preference.

[jdalton]

Thanks for the ping Kent!

Other reasons are:

• Modular
  
  • lodash supports cherry-picking require('lodash/array/chunk') or require('lodash.chunk')
• Embraces ES6
  
  • Support for es6 template delimiters
  • Support for Map or WeakMap backed _.memoize caches
  • Uses Set and WeakMap internally for performance gains
  • Has methods like _.includes, _.find, _.findIndex, _.spread, & _.restParam
  • Matches values based off SameValueZero
  • Detects array-like values based on the reverse of ToLength
  • Provides an es6 build.
• Gives devs features by popular demand. So you get deep path looks ups, callback shorthands, deep cloning, and other features not available in Underscore.
• ~100% code coverage
• Customizable with lodash-cli
• Supports a more FP style with lodash-fp

Lodash's v4 is a proof of concept of the Lodash+Underscore merge and should be out soon.