eslint: enable recommended ruleset

[alxndrsn]

No description provided.

[alxndrsn]

@brianc is there anything I can do to help get this PR considered for review?

[charmander]

I’d merge this if it were just the ESLint rule. Don’t want to combine two unrelated changes.

[alxndrsn]

I’d merge this if it were just the ESLint rule. Don’t want to combine two unrelated changes.

Thanks for the feedback. I've reverted the unrelated changes.

[charmander]

Oh, I see there are no code changes associated with this. How about applying the entire ESLint recommended configuration, then?

[alxndrsn]

How about applying the entire ESLint recommended configuration, then?

I've done this, and updated the PR title to reflect the change.

Eslint's ECMAScript version is bumped to 2020 as this was required to support globalThis (already in use before this PR), and 2020 seems to be fully supported by node 16, which is the oldest version supported in this repo's CI tests:

node-postgres/.github/workflows/ci.yml

Lines 40 to 45 in a9fd34f

	node:
	- '16'
	- '18'
	- '20'
	- '22'
	- '23'

It's picked up some obvious errors, and some stuff which is more debatable.

[charmander]

Thank you! It’s nice to see this change catching some important issues.

[charmander]

(note for other readers: the difference is a zero-width space after “SHA-256”)

[charmander]

Eslint's ECMAScript version is bumped to 2020 as this was required to support globalThis (already in use before this PR), and 2020 seems to be fully supported by node 16, which is the oldest version supported in this repo's CI tests:

You can define navigator as a global instead. Despite the lack of CI, pg 8 still supports Node 8(!).

[alxndrsn]

You can define navigator as a global instead. Despite the lack of CI, pg 8 still supports Node 8(!).

Any thoughts on what to do with the existing globalThis in packages/pg/lib/crypto/utils-webcrypto.js? From the filename, maybe it doesn't get used in node, so a specific exception would be ok?