Yunfei He

@kazupon > the Rust side write does not receive output otpions from the JS side Nope. It does take the output options from js side, which is the options that...

A native copy plugin won't improve much the performance, since it's more of a IO-bound task. I think you might just use https://www.npmjs.com/package/rollup-plugin-copy in rolldown directly.

> [@hyf0](https://github.com/hyf0) hi, will this feature be added in the future? If so, I'd like to give it a try. It's not planned.

Thanks for reporting this. I don't think users would rely on `@rolldown/node-binding` directly. So it's probably enough for just deprecating `@rolldown/node`.

I think it's hard to support this one. Rolldown might emits multiple errors in the same time. To be compatible with rollup's API interface, rolldown will merge the errors into...

This is fine to be merged. Could you run `just test-rust` and update all changed snapshots by `cargo insta accept`.

A potential direction is using DSL introduced in https://github.com/rolldown/rolldown/pull/4346. ```js test: [ exclude(and(id(aReg), id(bReg))), include(and(code(cReg), not(code(dReg)))) ] ```

Just to clarify: The team think it's ok to have a function support.

I think this is a classic XY problem. Making your exmaple work doesn't require rolldown to rewrite `require` to static import. In your case, settting `platform: 'node'` (default is `browser`)...

Writing `require` and expecting it to behave like static `import` are improper. Bundler's first job is to concat code together with preserving the original semantic. If the need conflicts with...