Release 0.180¶

General Changes¶

Fix a rare bug where rows containing only null values are not returned to the client. This only occurs when an entire result page contains only null values. The only known case is a query over an ORC encoded Hive table that does not perform any transformation of the data.
Fix incorrect results when performing comparisons between values of approximate data types (REAL, DOUBLE) and columns of certain exact numeric types (INTEGER, BIGINT, DECIMAL).
Fix memory accounting for min_by() and max_by() on complex types.
Fix query failure due to NoClassDefFoundError when scalar functions declared in plugins are implemented with instance methods.
Improve performance of map subscript from O(n) to O(1) in all cases. Previously, only maps produced by certain functions and readers could take advantage of this improvement.
Skip unknown costs in EXPLAIN output.
Support Secure Internal Communication between Presto nodes.
Add initial support for CROSS JOIN against LATERAL derived tables.
Add support for VARBINARY concatenation.
Add Thrift Connector that makes it possible to use Presto with external systems without the need to implement a custom connector.
Add experimental /v1/resourceGroupState REST endpoint on coordinator.

Fix skipping short decimal values in the optimized Parquet reader when they are backed by the int32 or int64 types.
Ignore partition bucketing if table is not bucketed. This allows dropping the bucketing from table metadata but leaving it for old partitions.
Improve error message for Hive partitions dropped during execution.
The optimized RCFile writer is enabled by default, but can be disabled with the hive.rcfile-optimized-writer.enabled config option. The writer supports validation which reads back the entire file after writing. Validation is disabled by default, but can be enabled with the hive.rcfile.writer.validate config option.