James Talmage

I disagree with the rule as proposed. I do agree with the sentiments from [@timoxley's blogpost](http://blog.timoxley.com/post/47041269194/avoid-else-return-early). But those are different from the code being discussed in https://github.com/sindresorhus/execa/pull/15#discussion-diff-60815913. In Tim's blog...

> modifying function parameters will also mutate the arguments object Wow. Did not know that. So yeah, :+1:. For non-`esnext`, could we still enable it, but relax if `arguments` is...

> The important part for me is to protect about modifying user supplied objects, like options. OK, is there a way to have it just be about modifying a properties...

> I guess we could do a quick fork of it and modify it to only be about props (If so, should be a separate plugin). That would be my...

Isn't this exactly what `eslint-config-xo` is?

I think this idea is sort of interesting. He is right that there is no XO config for WebStorm/IDEA, and the existing eslint plugin can't be forced to work with...

> A flat node_modules dir is not guaranteed with npm3 I'm aware, but if you're installing eslint plugins directly into devDependencies, they should probably take precedence anyways. > I think...

Arrow functions would be excluded (that's an option even in ESLints plugin).

The eslint rule (in `declaration` mode, with arrow functions excluded), only flags a `FunctionExpression` if it's immediate parent is a `VariableDeclarator`, so you only run afoul doing `var fn =...

A bit dated, but: https://kangax.github.io/nfe/