Release notes for 423

[martint]

No description provided.

[sopel39]

General
* Increase insert and CTAS performance. ({issue}`18212`)

#18212

[raunaqmorarka]

# Hive, Hudi, Delta, Iceberg
* Fix `OPTIMIZE` to avoid creating files smaller than `file_size_threshold` due to task writer scaling. ({issue}`18388`)

#18388

[raunaqmorarka]

# General
* Fix query planning failure due to dynamic filters in fault tolerant execution mode. ({issue}`18383`)

#18383 #18385

[findepi]

Iceberg
* Improve performance of `information_schema.columns` queries for tables managed by Trino in AWS Glue.

#18315

[raunaqmorarka]

# General
* Reduce defaults for `task.partitioned-writer-count` and `task.scale-writers.max-writer-count` to reduce memory requirements of write queries. ({issue}`18488`)

#18488

[raunaqmorarka]

# General
* Improve CBO estimates for queries with BETWEEN clause. ({issue}`18501`)

#18501

[raunaqmorarka]

# Iceberg
* Support reading parquet files with time stored in millisecond precision. ({issue}`18535`)

#18535

[sopel39]

# General
* Remove the deprecated `optimizer.use-mark-distinct` configuration property.
  The `optimizer.mark-distinct-strategy`  should be used instead. ({issue}`18540`)

#18540

[elonazoulay]

# Kafka Connector
* Renamed EmptyFieldStrategy value ADD_DUMMY to MARK. This is used by the configuration parameter kafka.empty-field-strategy and the corresponding session property empty_field_strategy. 

#18485

[sopel39]

SPI
* Remove deprecated `ConnectorMetadata#supportsReportingWrittenBytes`. ({issue}`18617`)

#18617

[raunaqmorarka]

# Hive, Hudi, Iceberg, Delta
* Improve memory usage accounting of parquet writer. ({issue}`18564`)

#18564