Abhishek Agarwal

btw have you looked at https://github.com/apache/druid/pull/15521 (flatting arrays of objects using UNNEST and JSON_QUERY_ARRAY) ?

You are right. This function has broader usage. I think when you say "arrays of objects", you don't mean arrays in a single record but rather multiple records. Since `fold`...

@lkm - are you going to revive this PR?

You might have to open a separate PR since I can't figure out a way to re-open it myself.

FWIW the PR looks good to me in the current shape so we should be able to merge your PR pretty quick.

Extensions don't access jars that are in the other extension folders. Have you considered that?

> > Extensions don't access jars that are in the other extension folders. Have you considered that? > > No, I did not know about this. Will they have access...

You may want to set these job properties in your ingestion spec ``` "jobProperties": { "mapreduce.job.classloader": "true", "mapreduce.job.classloader.system.classes": "-javax.el.,org.apache.commons.logging.,org.apache.log4j.,org.apache.hadoop.,com.google.cloud.hadoop.fs.gcs.", "mapreduce.framework.name": "yarn" }, ```

There has to be some docs changes. How are you going to infer the time in publishing tasks (btw what does supervisor publishing a task mean exactly)? And how do...

I have seen many similar PRs recently. Is there a reason not to do all of these in one single PR?