After reading @sindresorhus tweet I am a
little bit torn to move @bifravst to pure ESM.
The benefits are a little hard to justify, especially because it means that some
tooling needs to be replaced (e.g. jest) and other need not so clean workaround
(e.g. need .cjs config files).
/status/1349776733986312195
2 replies
Replying to @coderbyheart
OTOH it might be a good time to do it now, before a "1.0" release.
The upside is that it moves the code base further in the right direction, but
since it's written in TypeScript I already have the ability to author pure ESM
JS code, so I can do the switch in the future as well.
Replying to @coderbyheart
The needed changes to the codebase can be automated: adding a .js extension to
imports, might even not be necessary in future TypeScript versions. And from a
user point of view it also has no immediate benefit
(@bifravst code mostly targets Node.js, with a
few browser libraries).