Shixiong Zhu

> add jars to the HIVE_AUX_JARS_PATH, and now i'm getting error: What are the jars you added to `HIVE_AUX_JARS_PATH`? You would need Scala jars and the delta-hive-assembly jar.

Have you seen any exception in the hive server logs?

> when i was getting error `java.lang.NoSuchMethodError: scala.Some.value()Ljava/lang/Object;` i found this error: I feel there is an old Scala version on Hive classpath. Would you be able to search jars...

@wedesoft Spark doesn't support it yet. The sql syntax supported for GENERATED COLUMN is tracked by #1100

Although vacuum is a best-effort operation, dropping exceptions such as permission errors. Log all of them is also not ideal. One thing we can improve is making `vacuum` still log...

> The reason I ran `VACUUM` was because querying the table changes was taking a long time. Just to set the right expectation. `VACUUM` will not speed up any queries....

> uses Observable API which is based on query listener metrics ~~Do you mean reading the metrics from `StreamingQueryListener`? Here we are not doing that. This is a batch query...

> add a new UT with the current behavior (no nested pruning), with a comment that it should fail when upgrading to Spark 3.4. This sounds better. It's worth to...

could you resolve the conflicts?

Closing this. It has been merged: https://github.com/delta-io/delta/commit/acd0bd4fa6f20da6af8f149b37b5d414d3174023