Restore previous FORCE_COLOR=0 behavior

[nicolo-ribaudo]

Q	A
Fixed Issues?	Fixes #1, Fixes #2
Patch: Bug Fix?
Major: Breaking Change?
Minor: New Feature?
Tests Added + Pass?	Yes
Documentation PR Link
Any Dependency Changes?
License	MIT

See alexeyraspopov/picocolors#62 and jestjs/jest#14976 (comment). Until the last released we interpreted FORCE_COLOR=0 as disabling color. There is no consensus among tools about what FORCE_COLOR=0 does, but we should not change this behavior in a patch release.

[babel-bot]

Build successful! You can test your changes in the REPL here: https://babeljs.io/repl/build/56549

[liuxingbaoyu]

https://github.com/chalk/supports-color/
I'm not sure if it's worth using this package with Babel 7, there may be other edge cases where the differences exist.

[nicolo-ribaudo]

Most of that complexity of that package comes from supporting multiple CLI flags (that we do not support) and detecting the level of color support.

I would just do the minimal changes done by this PR, and if we want to move to something more reliable we can switch to process.stdout.hasColors() (Node.js 10+). That is, unless somebody opens an issue.

[nicolo-ribaudo]

Actually, it seems like the only difference is FORCE_COLOR=false. I'll add it to this PR.

[liuxingbaoyu]

https://github.com/chalk/supports-color/blob/c214314a14bcb174b12b3014b2b0a8de375029ae/index.js#L126-L136
I'm guessing there will be differences in those too, but I think it's okay to wait for anyone to report it.

[nicolo-ribaudo]

Releasing