0.15.3 - Paradedb — ContextQMD

New Features 🎉

Updated the index tuning settings for memory and parallelism to no longer be restricted to superuser privileges.

Fixed an issue where paradedb.score() returned empty values when using parametrized queries/plans.
Fixed an issue where text key_field values would get truncated to 255 characters. If you are using a text key field, you must REINDEX.
Fixed an issue with paradedb.term() could not find docs matching a full UUID value.
Fixed an issue where the ICU tokenizer would block inserts if libicu72 was installed.
Fixed an issue where, if ParadeDB was running as a logical replica, bulk insert/update statements would take too long and cause the replication stream to freeze.

Improved our segment merge policy such that a more optimal segment layout is created when the index experiences frequent inserts/updates

The full changelog is available on the GitHub Release.