Release 0.171#

General#

  • Fix planning regression for queries that compute a mix of distinct and non-distinct aggregations.

  • Fix casting from certain complex types to JSON when source type contains JSON or DECIMAL.

  • Fix issue for data definition queries that prevented firing completion events or purging them from the coordinator’s memory.

  • Add support for capture in lambda expressions.

  • Add support for ARRAY and ROW type as the compared value in min_by() and max_by().

  • Add support for CHAR(n) data type to common string functions.

  • Add codepoint(), skewness() and kurtosis() functions.

  • Improve validation of resource group configuration.

  • Fail queries when casting unsupported types to JSON; see JSON functions and operators for supported types.

Web UI#

  • Fix the threads UI (/ui/thread).

Hive#

  • Fix issue where some files are not deleted on cancellation of INSERT or CREATE queries.

  • Allow writing to non-managed (external) Hive tables. This is disabled by default but can be enabled via the hive.non-managed-table-writes-enabled configuration option.