Throw as instance of JS new Error()

[aspeddro]

Throws an instance of JavaScript's new Error() and adds the extension payload for cause option

Breaking Changes

• cause has a good support on browsers. See caniuse.com
• Supported only in Node >= 16.9.0 MDN Browser Compatibility
  

Adapted from melange-re/melange#1036

Thanks @anmonteiro

[DZakh]

Should it be somehow marked as a ReScript exception?

[DZakh]

Ah, I see there's a discussion exactly about this in the Melange PR melange-re/melange#1036 (comment)

[fhammerschmidt]

Related issue: #4898

[aspeddro]

Ready for review. I don't know if it should go to v12 or v11

[aspeddro]

Since cause is only supported above Node 16.9 we can emit RescriptError and use instanceof

instanceof support

[tsnobip]

Can't we just drop support for nodejs 16 and only support node > 18? They dropped support for node 16 last August already.

[cometkim]

Can't we just drop support for nodejs 16 and only support node > 18? They dropped support for node 16 last August already.

related: #6429

[zth]

Sorry, missed this. Looks like a good change! @cristianoc any thoughts?

[cristianoc]

Would be useful to have an assessment of the consequences of this change, as well as a description.
I guess this changes the internal representation of exceptions raised from ReScript.
First things that come to mind is if this affects the ability of recognising what exceptions are from ReScript, and e.g. pattern match on them.
Are there any observable changes after this in user code, any breaks, etc?

[zth]

@aspeddro and anyone else interested - some more research needed here, let us know if you're interested in tackling it.

[cknitt]

@aspeddro Could you rebase?

[aspeddro]

Rebase done!! Are we going to merge this?

I need to update the changelog

[cknitt]

LGTM and could be merged after after updating the CHANGELOG.
Agreed @zth?

We can investigate whether to port melange-re/melange#1043 separately later on.

[aspeddro]

CHANGELOG updated 15ea253

[zth]

Good to go from my end.

[cknitt]

Was wondering why CI wouldn't run, but then noticed the [skip ci] on the CHANGELOG commit. 😄
Will merge now as CI was successful on the previous commit.