Since the previous update 30 additional patches have been committed
from the CF. This leaves us with 120 Needs Review and 20 Ready for
Committer. There's only a few days left until the end of the month.
* Add comment about startup process getting a free procState array slot always
* Consistent use of SSL/TLS in docs
* Allow COPY "text" to output a header and add header matching mode to COPY FROM
* Enable SSL library detection via PQsslAttribute
* Fully WAL logged CREATE DATABASE - No Checkpoints
* Skipping logical replication transactions on subscriber side
* Add system view pg_ident_file_mappings
* document the need to analyze partitioned tables
* Expose get_query_def()
* JSON path numeric literal syntax
* use has_privs_for_role for predefined roles
* Support for MERGE
* Column filtering in logical replication
* Corruption during WAL replay
* Doc patch for retryable xacts
* pg_statio_all_tables: several rows per table due to invalid TOAST index
* Parameter for planner estimates of recursive queries
* logical decoding and replication of sequences
* Add relation and block-level filtering to pg_waldump
* pgbench - allow retries on some errors
* pgcrypto: Remove internal padding implementation
* Preserving db/ts/relfilenode OIDs across pg_upgrade
* Add new reloption to views for enabling row level security
* Fix handling of outer GroupingFunc within subqueries
* Fix concurrent deadlock scenario with DROP INDEX on partitioned index
* Fix bogus dependency management for GENERATED expressions
* Fix firing of RI triggers during cross-partition updates of partitioned tables
* Add fix to table_to_xmlschema regex when timestamp has time zone
* ltree_gist indexes broken after pg_upgrade from 12
* ExecTypeSetColNames is fundamentally broken
I'm going to start moving any patches that are Waiting on Author to
the next CF if they made any progress recently.
Patches that are Waiting on Author and haven't had activity in months
-- traditionally they were set to Returned with Feedback. It seems the
feeling these days is to not lose state on them and just move them to
the next CF. I'm not sure that's wise, it ends up just filling up the
list with patches nobody's working on.
--
greg
