Release 0.5 [2016-06-23]

• Added an option to Rebuild Indexes after data load
• Added an option to partition table by date column for Postgres (using a trigger and a function)
• Removed schema name from index name, shortened long staging tables names
• Added auto-decode from timezone 'UTC' for Unix timestamps

Release 0.4 [2015-11-25]

• Changed Distkey/Sortkey from position to flags 1/0
• Use Distkey/Sortkey in Redshift
• Add indexes on "Distkey/Sortkey Pos != 0" columns to Postgres
• By default comment out pk_lookup DDL section for non-main entities
• Moved the FK BK expressions to a lower level in source table select, where BK and Hash are
• Fixed more than one date_updated columns
• Added "nulls last" to date_updated columns
• Added timestamp with/without timezone to Hash definitions to use formatting
• Added is_deleted flag in batch_info (is_track_deleted for snapshot load tables) for snapshot loads (move old row to history, keep pk_lookup, remove the row in main)
• Added session id to inferred stage table names, to avoid overwriting with parallel scripts
• Disabled dropping of temporary staging batch table, to avoid unnoticed collisions between running scripts
• Added begin/lock table/commit to avoid conlicts when loading inferred tables with no direct depencendies.
• Showing dropdown tables for FKs even from other stage DBs within one project
• Fixed that "Import Column" didn't populate target attribute name/type
• Disabled Analyze for inferred entities, because it takes too much time
• Showing stage table name in entities tree
• Allowing now to check Sortkey or Distkey for BK columns, because separate indexes might be required for composite (multi-column) indexes
• Removed intendation for "create index" statements

Release 0.3 [2015-06-29]

• Add "Track changes", "Keep history" flags
• When a column is ignored, still use it for last_updated
• Add optional tablespace to tables
• Add target schema to generated file name
• Add ".sql" to save file name in Linux
• Add "Is Truncate Stage" flag
• Add "Is Ignore Changes" flag for columns
• Use tablespace for stage tables and indexes
• Add indexes to stage tables (all entity_bks, row_number, batch_number_old ?)
• Add indexes on target tables DDL (entity_bk for pk_lookup, all entity_keys)
• Remove primary key from history tables
• Setting (use "Is Fk Mandatory") to ignore NULL for entity/inferred BK

Release 0.2 [2015-06-19]

• For inferred: subname of another entity (to populate a column in)
• is_delete_temp_tables flag
• Add DDL generation
• Fix bpchar and int4 -> timestamp for unix timestamps
• Add settings for temporary/permanent stage tables, drop after use or not.
• Fix numeric type without precision

Release 0.1 [2015-05-30]

• add entity bk
• build hash from all columns as char
• add order by updated date and hash
• build inferred section from each inferred row
• build middle lookup sections with entity name only
• build stage2 - entity key with required joins, then all other columns mixed with fk column keys plus left joins (alias by column name).
• in history, first add system columns, then all table columns.
• Add custom table expression field
• Add custom column expression field
• Add stage schema to stage_db table, use it in all scripts for staging tables