Matteo Collina

> If it’s so important that async hooks are registered before any async activity, doesn’t this pending promise already violate that requirement? If so, what specific consequences entail as a...

Could you add a test to validate we don't regress this?

I have no clue on how to test it :(.

They are tested but the coverage report system cannot detect them.

The reason why a test is needed it's because I don't want to claim this use case is supported without actual proof this is supported.

This looks ok to me, I'll land and release.

You can do it with https://getpino.io/#/docs/api?id=pino-multistream and split2. Take a look at: - https://github.com/pinojs/pino/blob/37337510704bf81cfb1015d7729439cd599ef5b2/test/helper.js#L31-L42 - https://github.com/pinojs/pino/blob/0a56154fa0a24ac2b4015b597e6ae06468821699/test/timestamp.test.js#L21 I think a less convoluted API to do that would be nice indeed.

I agree with @kibertoad, I think there is a lot that can be improved here. I think we can have 1. A guide in the docs "how to test logging"...

Would you mind trying passing process.stdout instead? I don't expect much difference, but let me know. My running theory is that systemd is playing havoc. Systemd journal has a rate...

No, I meant something like this: ```js const logger = pino({ name: 'bot-sales', timestamp: pino.stdTimeFunctions.isoTime, }, process.stddout) ```