Say no - to lodash

[uboness]

Get rid of lodash dependency!

[myasonik]

Is the impetus to remove lodash just bundle size or is there more to it than that?

[chandlerprall]

Bundle size and limiting number & scope of external dependencies.

[smhutch]

I started looking into this today, and would like to help out. As a starting point, would it be suitable to open a PR which changes a couple of lodash imports as shown below?

# current
import _ from 'lodash'
import { utilA, utilB } from 'lodash'

# proposed
import utilA from 'lodash/utilA'
import utilB from 'lodash/utilB'

Although the approach I'm proposing here is a bit repetitive, my understanding is that it enables Webpack to bundle-split lodash. When I perform bundle analysis on my current application (which consumes eui), it appears the entire lodash library is included. This guide, and this summary post led me to make this suggestion, though I'm unsure if other webpack configuration changes would be required.

It appears that these two modules would need to change:

eui/src-docs/src/views/collapsible_nav/collapsible_nav_all.tsx

Line 2 in 52c326b

import _ from 'lodash';

eui/src/components/token/token.tsx

Line 21 in 52c326b

import { defaults } from 'lodash';

Getting rid of lodash completely would also be good. My idea is to use this as a low-risk stepping stone towards that. 😄

[chandlerprall]

Removing lodash completely is the goal. Replacing each instance of its use in small batches - either by file by lodash method - will likely be the easiest way for us to review.

[smhutch]

Awesome, I'll likely make that contribution at the start of next week 😄

[github-actions]

👋 Hey there. This issue hasn't had any activity for 180 days. We'll automatically close it if that trend continues for another week. If you feel this issue is still valid and needs attention please let us know with a comment.

[smith]

Note that if you remove lodash from EUI (a noble goal!) when you use it in Kibana you will still have lodash.

[cee-chen]

Agreed that this is a noble goal, but given the size of EUI's backlog and many other higher priority tech debt items to work on, it's unlikely we'll be working on this as a team anytime soon. We're going to go ahead and close this as a result, but if anyone else is interested in contributing to EUI or taking lead on this, we will gladly welcome your contributions.