6 Release date: 2002-11-??
7 _________________________________________________________________
11 Major changes in this release:
14 Schemas allow users to create objects in their own namespace so
15 two people or applications can have tables with the same name.
16 There is also a public schema for shared tables. Table/index
17 creation can be restricted by removing permissions on the
21 PostgreSQL now supports the ALTER TABLE ... DROP COLUMN
25 Functions returning multiple rows and/or multiple columns are
26 now much easier to use than before. You can call such a "table
27 function" in the SELECT FROM clause, treating its output like a
28 table. Also, PL/pgSQL functions can now return sets.
31 PostgreSQL now supports prepared queries, for improved
35 PostgreSQL now records object dependencies, which allows
36 improvements in many areas. "DROP" statements now take either
37 CASCADE or RESTRICT to control whether dependent objects are
41 Functions and procedural languages now have privileges, and
42 functions can be defined to run with the privileges of their
46 Both multibyte and locale support are now always enabled.
49 A variety of logging options have been enhanced.
52 A large number of interfaces have been moved to
53 http://gborg.postgresql.org where they can be developed and
54 released independently.
57 By default, functions can now take up to 32 parameters, and
58 identifiers can be up to 63 bytes long.
59 _________________________________________________________________
61 Migration to version 7.3
63 A dump/restore using pg_dump is required for those wishing to migrate
64 data from any previous release. If your application examines the
65 system catalogs, additional changes will be required due to the
66 introduction of schemas in 7.3; for more information, see:
67 http://www.ca.postgresql.org/docs/momjian/upgrade_tips_7.3.
69 Observe the following incompatibilities:
71 * Pre-6.3 clients are no longer supported.
72 * "pg_hba.conf" now has a column for the user name and additional
73 features. Existing files need to be adjusted.
74 * Several "postgresql.conf" logging parameters have been renamed and
76 * LIMIT #,# has been disabled; use LIMIT # OFFSET #.
77 * "INSERT" statements with column lists must specify all values;
78 e.g., INSERT INTO tab (col1, col2) VALUES ('val1') is now invalid.
79 * An index is now not automatically created for serial columns.
80 * A "SET" command inside an aborted transaction is now rolled back.
81 * "COPY" no longer considers missing trailing columns to be null.
82 All columns need to be specified.
83 * The data type timestamp is now equivalent to timestamp without
84 timezone, instead of timestamp with timezone.
85 * Pre-7.3 databases loaded into 7.3 will not have the new object
86 dependencies for serial columns, unique constraints, and foreign
87 keys. See the directory "contrib/adddepend/" for a detailed
88 description and a script that will add such dependencies.
89 _________________________________________________________________
95 Add pg_locks view to show locks (Neil)
96 Security fixes for password negotiation memory allocation (Neil)
97 Remove support for version 0 FE/BE protocol (PostgreSQL 6.2 and
99 Reserve the last few backend slots for superusers, add parameter
100 superuser_reserved_connections to control this (Nigel J. Andrews)
101 _________________________________________________________________
105 Improve startup by calling localtime() only once (Tom)
106 Cache system catalog information in flat files for faster startup
108 Improve caching of index information (Tom)
109 Optimizer improvements (Tom, Fernando Nasser)
110 Catalog caches now store failed lookups (Tom)
111 Hash function improvements (Neil)
112 Improve performance of query tokenization and network handling (Peter)
113 Speed improvement for large object restore (Mario Weilguni)
114 Mark expired index entries on first lookup, saving later heap fetches
116 Eliminate NULL bitmap padding when not required (Manfred)
117 Add BSD-licensed qsort() for Solaris, for performance (Bruce)
118 Reduce per-row overhead by four bytes (Manfred Koizar)
119 Fix GEQO optimizer bug (Neil Conway)
120 Make WITHOUT OID actually save four bytes per row (Manfred Koizar)
121 Add default_statistics_target variable to specify ANALYZE buckets
123 Use local buffer cache for temporary tables so no WAL overhead (Tom)
124 Improve free space map performance on large tables (Stephen Marshall,
126 Improved WAL write concurrency (Tom)
127 _________________________________________________________________
131 Add privileges on functions and procedural languages (Peter)
132 Add OWNER to CREATE DATABASE so superusers can create databases
133 on behalf of unprivileged users. (Gavin Sherry, Tom)
134 Add new object permission bits EXECUTE and USAGE (Tom)
135 Add SET SESSION AUTHORIZATION DEFAULT and RESET SESSION AUTHORIZATION
137 Allow functions to be executed with the privilege of the function
139 _________________________________________________________________
143 Server log messages now tagged with LOG, not DEBUG (Bruce)
144 Add user column to pg_hba.conf (Bruce)
145 Have log_connections output two lines in log file (Tom)
146 Remove debug_level from postgresql.conf, now server_min_messages
148 New ALTER DATABASE/USER ... SET command for per-user/database
149 initialization (Peter)
150 New parameters server_min_messages and client_min_messages to
151 control which messages are sent to the server logs or client
153 Allow pg_hba.conf to specify lists of users/databases separated by
154 commas, group names prepended with +, and file names prepended
156 Remove secondary password file capability and pg_password utility
158 Add variable db_user_namespace for database-local user names (Bruce)
159 SSL improvements (Bear Giles)
160 Make encryption of stored passwords the default (Bruce)
161 Allow pg_statistics to be reset by calling pg_stat_reset()
163 Add log_duration parameter (Bruce)
164 Rename debug_print_query to log_statement (Bruce)
165 Rename show_query_stats to show_statement_stats (Bruce)
166 Add param log_min_error_statement to print commands to logs on error
168 _________________________________________________________________
172 Make cursors insensitive, meaning their contents do not change (Tom)
173 Disable LIMIT #,# syntax; now only LIMIT # OFFSET # supported (Bruce)
174 Increase identifier length to 63 (Neil, Bruce)
175 UNION fixes for merging >= 3 columns of different lengths (Tom)
176 Add DEFAULT keyword to INSERT, e.g., INSERT ... (..., DEFAULT, )
178 Allow views to have default values using ALTER COLUMN ... SET DEFAULT
180 Fail on INSERTs with column lists that don't supply all column
181 values, e.g., INSERT INTO tab (col1, col2) VALUES ('val1'); (Rod)
182 Fix for join aliases (Tom)
183 Fix for FULL OUTER JOINs (Tom)
184 Improve reporting of invalid identifier and location (Tom, Gavin)
185 Fix OPEN cursor(args) (Tom)
186 Allow 'ctid' to be used in a view and currtid(viewname) (Hiroshi)
187 Fix for CREATE TABLE AS with UNION (Tom)
188 SQL99 syntax improvements (Thomas)
189 Add statement_timeout variable to cancel queries (Bruce)
190 Allow prepared queries with PREPARE/EXECUTE (Neil)
191 Allow FOR UPDATE to appear after LIMIT/OFFSET (Bruce)
192 Add variable autocommit (Tom, David Van Wie)
193 _________________________________________________________________
197 Make equals signs optional in CREATE DATABASE (Gavin Sherry)
198 Make ALTER TABLE OWNER to change index ownership (Neil)
199 New ALTER TABLE tabname ALTER COLUMN colname SET STORAGE controls
200 TOAST storage, compression (John Gray)
201 Add schema support, CREATE/DROP SCHEMA (Tom)
202 Create schema for temporary tables (Tom)
203 Add variable search_path for schema search (Tom)
204 Add ALTER TABLE SET/DROP NOT NULL (Christopher)
205 New CREATE FUNCTION volatility levels (Tom)
206 Make rule names unique only per table (Tom)
207 Add 'ON tablename' clause to DROP RULE and COMMENT ON RULE (Tom)
208 Add ALTER TRIGGER RENAME (Joe)
209 New current_schema() and current_schemas() inquiry functions (Tom)
210 Allow functions to return multiple rows (table functions) (Joe)
211 Make WITH optional in CREATE DATABASE, for consistency (Bruce)
212 Add object dependency tracking (Rod, Tom)
213 Add RESTRICT/CASCADE to DROP commands (Rod)
214 Add ALTER TABLE DROP for non-CHECK CONSTRAINT (Rod)
215 Autodestroy sequence on DROP of table with SERIAL (Rod)
216 Prevent column dropping if column is used by foreign key (Rod)
217 Automatically drop constraints/functions when object is dropped (Rod)
218 Add CREATE/DROP OPERATOR CLASS (Bill Studenmund, Tom)
219 Add ALTER TABLE DROP COLUMN (Christopher, Tom, Hiroshi)
220 Prevent inherited columns from being removed or renamed (Alvaro
222 Fix foreign key constraints to not error on intermediate database
224 Propagate column or table renaming to foreign key constraints
225 Add CREATE OR REPLACE VIEW (Gavin, Neil, Tom)
226 Add CREATE OR REPLACE RULE (Gavin, Neil, Tom)
227 Have rules execute alphabetically, returning more predictable values
229 Triggers are now fired in alphabetical order (Tom)
230 Add /contrib/adddepend to handle pre-7.3 object dependencies (Rod)
231 Allow better casting when inserting/updating values (Tom)
232 _________________________________________________________________
236 Have COPY TO output embedded carriage returns and newlines as \r and
238 Allow DELIMITER in COPY FROM to be 8-bit clean (Tatsuo)
239 Make pg_dump use ALTER TABLE ADD PRIMARY KEY, for performance (Neil)
240 Disable brackets in multistatement rules (Bruce)
241 Disable VACUUM from being called inside a function (Bruce)
242 Allow dropdb and other scripts to use identifiers with spaces (Bruce)
243 Restrict database comment changes to the current database
244 Allow comments on operators, independent of the underlying function
246 Rollback SET commands in aborted transactions (Tom)
247 EXPLAIN now outputs as a query (Tom)
248 Display sort keys in EXPLAIN (Tom)
249 Add 'SET LOCAL var = value' to set configuration variables for a
250 single transaction (Tom)
251 Allow ANALYZE to run in a transaction (Bruce)
252 Improve COPY syntax using new WITH clauses, keep backward
253 compatibility (Bruce)
254 Fix pg_dump to consistently output tags in non-ASCII dumps (Bruce)
255 Make foreign key constraints clearer in dump file (Rod)
256 Add COMMENT ON CONSTRAINT (Rod)
257 Allow COPY TO/FROM to specify column names (Brent Verner)
258 Dump UNIQUE and PRIMARY KEY contraints as ALTER TABLE (Rod)
259 Have SHOW output a query result (Joe)
260 Generate failure on short COPY lines rather than pad NULLs (Neil)
261 Fix CLUSTER to preserve all table attributes (Alvaro Herrera)
262 New pg_settings table to view/modify GUC settings (Joe)
263 Add smart quoting, portability improvements to pg_dump output (Peter)
264 Dump serial columns out as SERIAL (Tom)
265 Enable large file support, >2G for pg_dump (Peter, Philip Warner,
267 Disallow TRUNCATE on tables that are involved in referential
269 Have TRUNCATE also auto-truncate the toast table of the relation (Tom)
270 Add clusterdb utility that will auto-cluster an entire database
271 based on previous CLUSTER operations (Alvaro Herrera)
272 Overhaul pg_dumpall (Peter)
273 Allow REINDEX of TOAST tables (Tom)
274 Implemented START TRANSACTION, per SQL99 (Neil)
275 Fix rare index corruption when a page split affects bulk delete (Tom)
276 Fix ALTER TABLE ... ADD COLUMN for inheritance (Alvaro Herrera)
277 _________________________________________________________________
279 Data Types and Functions
281 Fix factorial(0) to return 1 (Bruce)
282 Date/time/timezone improvements (Thomas)
283 Fix for array slice extraction (Tom)
284 Fix extract/date_part to report proper microseconds for timestamp
286 Allow text_substr() and bytea_substr() to read TOAST values more
287 efficiently (John Gray)
288 Add domain support (Rod)
289 Make WITHOUT TIME ZONE the default for TIMESTAMP and TIME data types
291 Allow alternate storage scheme of 64-bit integers for date/time types
292 using --enable-integer-datetimes in configure (Thomas)
293 Make timezone(timestamptz) return timestamp rather than a string
295 Allow fractional seconds in date/time types for dates prior to 1BC
297 Limit timestamp data types to 6 decimal places of precision (Thomas)
298 Change timezone conversion functions from timetz() to timezone()
300 Add configuration variables datestyle and timezone (Tom)
301 Add OVERLAY(), which allows substitution of a substring in a string
303 Add SIMILAR TO (Thomas, Tom)
304 Add regular expression SUBSTRING(string FROM pat FOR escape) (Thomas)
305 Add LOCALTIME and LOCALTIMESTAMP functions (Thomas)
306 Add named composite types using CREATE TYPE typename AS (column)
308 Allow composite type definition in the table alias clause (Joe)
309 Add new API to simplify creation of C language table functions (Joe)
310 Remove ODBC-compatible empty parentheses from calls to SQL99
311 functions for which these parentheses do not match the standard
313 Allow macaddr data type to accept 12 hex digits with no separators
315 Add CREATE/DROP CAST (Peter)
316 Add IS DISTINCT FROM operator (Thomas)
317 Add SQL99 TREAT() function, synonym for CAST() (Thomas)
318 Add pg_backend_pid() to output backend pid (Bruce)
319 Add IS OF / IS NOT OF type predicate (Thomas)
320 Allow bit string constants without fully-specified length (Thomas)
321 Allow conversion between 8-byte integers and bit strings (Thomas)
322 Implement hex literal conversion to bit string literal (Thomas)
323 Allow table functions to appear in the FROM clause (Joe)
324 Increase maximum number of function parameters to 32 (Bruce)
325 No longer automatically create index for SERIAL column (Tom)
326 Add current_database() (Rod)
327 Fix cash_words() to not overflow buffer (Tom)
328 Add functions replace(), split_part(), to_hex() (Joe)
329 Fix LIKE for bytea as a right-hand argument (Joe)
330 Prevent crashes caused by SELECT cash_out(2) (Tom)
331 Fix to_char(1,'FM999.99') to return a period (Karel)
332 Fix trigger/type/language functions returning OPAQUE to return
334 _________________________________________________________________
338 Add additional encodings (Korean (JOHAB), Thai (WIN874), Vietnamese
339 (TCVN), Arabic (WIN1256), Simplified Chinese (GBK), Korean (UHC)
341 Enable locale support by default (Peter)
342 Add locale variables (Peter)
343 Escape byes >= 0x7f for multibyte in PQescapeBytea/PQunescapeBytea
345 Add locale awareness to regular expression character classes
346 Enable multibyte support by default (Tatso)
347 Add GB18030 multibyte support (Bill Huang)
348 Add CREATE/DROP CONVERSION, allowing loadable encodings (Tatsuo,
350 Add pg_conversion table (Tatsuo)
351 Add SQL99 CONVERT() function (Tatsuo)
352 pg_dumpall, pg_controldata, and pg_resetxlog now national-language
354 New and updated translations
355 _________________________________________________________________
357 Server-side Languages
359 Allow recursive SQL function (Peter)
360 Change PL/Tcl build to use configured compiler and Makefile.shlib
362 Overhaul the PL/pgSQL FOUND variable to be more Oracle-compatible
364 Allow PL/pgSQL to handle quoted identifiers (Tom)
365 Allow set-returning PL/pgSQL functions (Neil)
366 Make PL/pgSQL schema-aware (Joe)
367 Remove some memory leaks (Nigel J. Andrews, Tom)
368 _________________________________________________________________
372 Don't lowercase psql \connect database name for 7.2.0 compatibility
374 Add psql \timing to time user queries (Greg Sabino Mullane)
375 Have psql \d show index information (Greg Sabino Mullane)
376 New psql \dD shows domains (Jonathan Eisler)
377 Allow psql to show rules on views (Paul ?)
378 Fix for psql variable substitution (Tom)
379 Allow psql \d to show temporary table structure (Tom)
380 Allow psql \d to show foreign keys (Rod)
381 Fix \? to honor \pset pager (Bruce)
382 Have psql reports its version number on startup (Tom)
383 Allow \copy to specify column names (Tom)
384 _________________________________________________________________
388 Add $HOME/.pgpass to store host/user password combinations (Alvaro
390 Add PQunescapeBytea() function to libpq (Patrick Welche)
391 Fix for sending large queries over non-blocking connections
393 Fix for libpq using timers on Win9X (David Ford)
394 Allow libpq notify to handle servers with different-length
396 Add libpq PQescapeString() and PQescapeBytea() to Win32 (Bruce)
397 Fix for SSL with non-blocking connections (Jack Bates)
398 Add libpq connection timeout parameter (Denis A Ustimenko)
399 _________________________________________________________________
403 Allow JDBC to compile with JDK 1.4 (Dave)
404 Add JDBC 3 support (Barry)
405 Allows JDBC to set loglevel by adding ?loglevel=X to the connection
407 Add Driver.info() message that prints out the version number (Barry)
408 Add updateable result sets (Raghu Nidagal, Dave)
409 Add support for callable statements (Paul Bethe)
410 Add query cancel capability
411 Add refresh row (Dave)
412 Fix MD5 encryption handling for multibyte servers (Jun Kawai)
413 Add support for prepared statements (Barry)
414 _________________________________________________________________
416 Miscellaneous Interfaces
418 Fixed ECPG bug concerning octal numbers in single quotes (Michael)
419 Move src/interfaces/libpgeasy to http://gborg.postgresql.org (Marc,
421 Improve Python interface (Elliot Lee, Andrew Johnson, Greg Copeland)
422 Add libpgtcl connection close event (Gerhard Hintermayer)
423 Move src/interfaces/libpq++ to http://gborg.postgresql.org (Marc,
425 Move src/interfaces/odbc to http://gborg.postgresql.org (Marc)
426 Move src/interfaces/libpgeasy to http://gborg.postgresql.org (Marc,
428 Move src/interfaces/perl5 to http://gborg.postgresql.org (Marc,
430 Remove src/bin/pgaccess from main tree, now at
431 http://www.pgaccess.org (Bruce)
432 Add pg_on_connection_loss command to libpgtcl (Gerhard Hintermayer,
434 _________________________________________________________________
438 Fix for parallel make (Peter)
439 AIX fixes for linking Tcl (Andreas Zeugswetter)
440 Allow PL/Perl to build under Cygwin (Jason Tishler)
441 Improve MIPS compiles (Peter, Oliver Elphick)
442 Require Autoconf version 2.53 (Peter)
443 Require readline and zlib by default in configure (Peter)
444 Allow Solaris to use Intimate Shared Memory (ISM), for performance
445 (Scott Brunza, P.J. Josh Rovero)
446 Always enable syslog in compile, remove --enable-syslog option
448 Always enable multibyte in compile, remove --enable-multibyte option
450 Always enable locale in compile, remove --enable-locale option
452 Fix for Win9x DLL creation (Magnus Naeslund)
453 Fix for link() usage by WAL code on Win32, BeOS (Jason Tishler)
454 Add sys/types.h to c.h, remove from main files (Peter, Bruce)
455 Fix AIX hang on SMP machines (Tomoyuki Niijima)
456 AIX SMP hang fix (Tomoyuki Niijima)
457 Fix pre-1970 date handling on newer glibc libraries (Tom)
458 Fix PowerPC SMP locking (Tom)
459 Prevent gcc -ffast-math from being used (Peter, Tom)
460 Bison >= 1.50 now required for developer builds
461 Kerberos 5 support now builds with Heimdal (Peter)
462 Add appendix in the User's Guide which lists SQL features (Thomas)
463 Improve loadable module linking to use RTLD_NOW (Tom)
464 New error levels WARNING, INFO, LOG, DEBUG[1-5] (Bruce)
465 New src/port directory holds replaced libc functions (Peter, Bruce)
466 New pg_namespace system catalog for schemas (Tom)
467 Add pg_class.relnamespace for schemas (Tom)
468 Add pg_type.typnamespace for schemas (Tom)
469 Add pg_proc.pronamespace for schemas (Tom)
470 Restructure aggregates to have pg_proc entries (Tom)
471 System relations now have their own namespace, pg_* test not required
473 Rename TOAST index names to be *_index rather than *_idx (Neil)
474 Add namespaces for operators, opclasses (Tom)
475 Add additional checks to server control file (Thomas)
476 New Polish FAQ (Marcin Mazurek)
477 Add Posix semaphore support (Tom)
478 Document need for reindex (Bruce)
479 Rename some internal identifiers to simplify Win32 compile (Jan,
481 Add documentation on computing disk space (Bruce)
482 Remove KSQO from GUC (Bruce)
483 Fix memory leak in rtree (Kenneth Been)
484 Modify a few error messages for consistency (Bruce)
485 Remove unused system table columns (Peter)
486 Make system columns NOT NULL where appropriate (Tom)
487 Clean up use of sprintf in favor of snprintf() (Neil, Jukka Holappa)
488 Remove OPAQUE and create specific subtypes (Tom)
489 Cleanups in array internal handling (Joe, Tom)
490 Disallow pg_atoi('') (Bruce)
491 Remove parameter wal_files because WAL files are now recycled (Bruce)
492 Add version numbers to heap pages (Tom)
493 _________________________________________________________________
497 Allow inet arrays in /contrib/array (Neil)
498 Gist fixes (Teodor Sigaev, Neil)
499 Upgrade /contrib/mysql
500 Add /contrib/dbsize which shows table sizes without vacuum (Peter)
501 Add /contrib/intagg, integer aggregator routines (mlw)
502 Improve /contrib/oid2name (Neil, Bruce)
503 Improve /contrib/tsearch (Oleg, Teodor Sigaev)
504 Cleanups of /contrib/rserver (Alexey V. Borzov)
505 Update /contrib/oracle conversion utility (Gilles Darold)
506 Update /contrib/dblink (Joe)
507 Improve options supported by /contrib/vacuumlo (Mario Weilguni)
508 Improvements to /contrib/intarray (Oleg, Teodor Sigaev, Andrey
510 Add /contrib/reindexdb utility (Shaun Thomas)
511 Add indexing to /contrib/isbn_issn (Dan Weston)
512 Add /contrib/dbmirror (Steven Singer)
513 Improve /contrib/pgbench (Neil)
514 Add /contrib/tablefunc table function examples (Joe)
515 Add /contrib/ltree data type for tree structures (Teodor Sigaev,
517 Move /contrib/pg_controldata into main tree (Bruce)
518 Fixes to /contrib/cube (Bruno Wolff)
519 Improve /contrib/fulltextindex (Christopher)
520 _________________________________________________________________
524 Release date: 2002-10-01
526 This has a variety of fixes from 7.2.2, including fixes to prevent
528 _________________________________________________________________
530 Migration to version 7.2.3
532 A dump/restore is *not* required for those running 7.2.X.
533 _________________________________________________________________
537 Prevent possible compressed transaction log loss (Tom)
538 Prevent non-superuser from increasing most recent vacuum info (Tom)
539 Handle pre-1970 date values in newer versions of glibc (Tom)
540 Fix possible hang during server shutdown
541 Prevent spinlock hangs on SMP PPC machines (Tomoyuki Niijima)
542 Fix pg_dump to properly dump FULL JOIN USING (Tom)
543 _________________________________________________________________
547 Release date: 2002-08-23
549 This has a variety of fixes from 7.2.1.
550 _________________________________________________________________
552 Migration to version 7.2.2
554 A dump/restore is *not* required for those running 7.2.X.
555 _________________________________________________________________
559 Allow EXECUTE of "CREATE TABLE AS ... SELECT" in PL/pgSQL (Tom)
560 Fix for compressed transaction log id wraparound (Tom)
561 Fix PQescapeBytea/PQunescapeBytea so that they handle bytes > 0x7f (Ta
563 Fix for psql and pg_dump crashing when invoked with non-existand long
565 Fix crash when invoking geometric operators (Tom)
566 Allow OPEN cursor(args) (Tom)
567 Fix for rtree_gist index build (Teodor)
568 Fix for dumping user-defined aggregates (Tom)
569 Contrib/intarray fixes (Oleg)
570 Fix for complex UNION/EXCEPT/INTERSECT queries using parens (Tom)
571 Fix to pg_convert (Tatsuo)
572 Fix for crash with long DATA strings (Thomes, Neil)
573 Fix for repeat(), lpad(), rpad() and long strings (Neil)
574 _________________________________________________________________
578 Release date: 2002-03-21
580 This has a variety of fixes from 7.2.
581 _________________________________________________________________
583 Migration to version 7.2.1
585 A dump/restore is *not* required for those running 7.2.
586 _________________________________________________________________
590 Ensure that sequence counters do not go backwards after a crash (Tom)
591 Fix pgaccess kanji-coversion key binding (Tatsuo)
592 Optimizer improvements (Tom)
593 cash I/O improvements (Tom)
595 Compile fix for missing AuthBlockSig (Heiko)
596 Additional time zones and time zone fixes (Thomas)
597 Allow psql \connect to handle mixed case database and user names (Tom)
598 Return proper OID on command completion even with ON INSERT rules (Tom
600 Allow COPY FROM to use 8-bit DELIMITERS (Tatsuo)
601 Fix bug in extract/date_part for milliseconds/microseconds (Tatsuo)
602 Improve handling of multiple UNIONs with different lengths (Tom)
603 contrib/btree_gist improvements (Teodor Sigaev)
604 contrib/tsearch dictionary improvements, see README.tsearch for
605 an additional installation step (Thomas T. Thai, Teodor Sigaev)
606 Fix for array subscripts handling (Tom)
607 Allow EXECUTE of "CREATE TABLE AS ... SELECT" in PL/pgSQL (Tom)
608 _________________________________________________________________
612 Release date: 2002-02-04
613 _________________________________________________________________
617 This release improves PostgreSQL for use in high-volume applications.
619 Major changes in this release:
622 Vacuuming no longer locks tables, thus allowing normal user
623 access during the vacuum. A new "VACUUM FULL" command does
624 old-style vacuum by locking the table and shrinking the on-disk
628 There is no longer a problem with installations that exceed
629 four billion transactions.
632 OIDs are now optional. Users can now create tables without OIDs
633 for cases where OID usage is excessive.
636 The system now computes histogram column statistics during
637 "ANALYZE", allowing much better optimizer choices.
640 A new MD5 encryption option allows more secure storage and
641 transfer of passwords. A new Unix-domain socket authentication
642 option is available on Linux and BSD systems.
645 Administrators can use the new table access statistics module
646 to get fine-grained information about table and index usage.
649 Program and library messages can now be displayed in several
651 _________________________________________________________________
653 Migration to version 7.2
655 A dump/restore using "pg_dump" is required for those wishing to
656 migrate data from any previous release.
658 Observe the following incompatibilities:
660 * The semantics of the "VACUUM" command have changed in this
661 release. You may wish to update your maintenance procedures
663 * In this release, comparisons using = NULL will always return false
664 (or NULL, more precisely). Previous releases automatically
665 transformed this syntax to IS NULL. The old behavior can be
666 re-enabled using a "postgresql.conf" parameter.
667 * The "pg_hba.conf" and "pg_ident.conf" configuration is now only
668 reloaded after receiving a SIGHUP signal, not with each
670 * The function "octet_length()" now returns the uncompressed data
672 * The date/time value 'current' is no longer available. You will
673 need to rewrite your applications.
674 * The timestamp(), time(), and interval() functions are no longer
675 available. Instead of timestamp(), use timestamp 'string' or CAST.
677 The SELECT ... LIMIT #,# syntax will be removed in the next release.
678 You should change your queries to use separate LIMIT and OFFSET
679 clauses, e.g. LIMIT 10 OFFSET 20.
680 _________________________________________________________________
686 Create temporary files in a separate directory (Bruce)
687 Delete orphaned temporary files on postmaster startup (Bruce)
688 Added unique indexes to some system tables (Tom)
689 System table operator reorganization (Oleg Bartunov, Teodor Sigaev, To
691 Renamed pg_log to pg_clog (Tom)
692 Enable SIGTERM, SIGQUIT to kill backends (Jan)
693 Removed compile-time limit on number of backends (Tom)
694 Better cleanup for semaphore resource failure (Tatsuo, Tom)
695 Allow safe transaction ID wraparound (Tom)
696 Removed OIDs from some system tables (Tom)
697 Removed "triggered data change violation" error check (Tom)
698 SPI portal creation of prepared/saved plans (Jan)
699 Allow SPI column functions to work for system columns (Tom)
700 Long value compression improvement (Tom)
701 Statistics collector for table, index access (Jan)
702 Truncate extra-long sequence names to a reasonable value (Tom)
703 Measure transaction times in milliseconds (Thomas)
704 Fix TID sequential scans (Hiroshi)
705 Superuser ID now fixed at 1 (Peter E)
706 New pg_ctl "reload" option (Tom)
707 _________________________________________________________________
711 Optimizer improvements (Tom)
712 New histogram column statistics for optimizer (Tom)
713 Reuse write-ahead log files rather than discarding them (Tom)
714 Cache improvements (Tom)
715 IS NULL, IS NOT NULL optimizer improvement (Tom)
716 Improve lock manager to reduce lock contention (Tom)
717 Keep relcache entries for index access support functions (Tom)
718 Allow better selectivity with NaN and infinities in NUMERIC (Tom)
719 R-tree performance improvements (Kenneth Been)
720 B-tree splits more efficient (Tom)
721 _________________________________________________________________
725 Change UPDATE, DELETE permissions to be distinct (Peter E)
726 New REFERENCES, TRIGGER privileges (Peter E)
727 Allow GRANT/REVOKE to/from more than one user at a time (Peter E)
728 New has_table_privilege() function (Joe Conway)
729 Allow non-superuser to vacuum database (Tom)
730 New SET SESSION AUTHORIZATION command (Peter E)
731 Fix bug in privilege modifications on newly created tables (Tom)
732 Disallow access to pg_statistic for non-superuser, add user-accessible
734 _________________________________________________________________
736 Client Authentication
738 Fork postmaster before doing authentication to prevent hangs (Peter E)
739 Add ident authentication over Unix domain sockets on Linux, *BSD (Helg
740 e Bahmann, Oliver Elphick, Teodor Sigaev, Bruce)
741 Add a password authentication method that uses MD5 encryption (Bruce)
742 Allow encryption of stored passwords using MD5 (Bruce)
743 PAM authentication (Dominic J. Eidson)
744 Load pg_hba.conf and pg_ident.conf only on startup and SIGHUP (Bruce)
745 _________________________________________________________________
749 Interpretation of some time zone abbreviations as Australian rather th
750 an North American now settable at run time (Bruce)
751 New parameter to set default transaction isolation level (Peter E)
752 New parameter to enable conversion of "expr = NULL" into "expr IS NULL
753 ", off by default (Peter E)
754 New parameter to control memory usage by VACUUM (Tom)
755 New parameter to set client authentication timeout (Tom)
756 New parameter to set maximum number of open files (Tom)
757 _________________________________________________________________
761 Statements added by INSERT rules now execute after the INSERT (Jan)
762 Prevent unadorned relation names in target list (Bruce)
763 NULLs now sort after all normal values in ORDER BY (Tom)
764 New IS UNKNOWN, IS NOT UNKNOWN Boolean tests (Tom)
765 New SHARE UPDATE EXCLUSIVE lock mode (Tom)
766 New EXPLAIN ANALYZE command that shows run times and row counts (Marti
768 Fix problem with LIMIT and subqueries (Tom)
769 Fix for LIMIT, DISTINCT ON pushed into subqueries (Tom)
770 Fix nested EXCEPT/INTERSECT (Tom)
771 _________________________________________________________________
775 Fix SERIAL in temporary tables (Bruce)
776 Allow temporary sequences (Bruce)
777 Sequences now use int8 internally (Tom)
778 New SERIAL8 creates int8 columns with sequences, default still SERIAL4
780 Make OIDs optional using WITHOUT OIDS (Tom)
781 Add %TYPE syntax to CREATE TYPE (Ian Lance Taylor)
782 Add ALTER TABLE / DROP CONSTRAINT for CHECK constraints (Christopher K
784 New CREATE OR REPLACE FUNCTION to alter existing function (preserving
785 the function OID) (Gavin Sherry)
786 Add ALTER TABLE / ADD [ UNIQUE | PRIMARY ] (Christopher Kings-Lynne)
787 Allow column renaming in views
788 Make ALTER TABLE / RENAME COLUMN update column names of indexes (Brent
790 Fix for ALTER TABLE / ADD CONSTRAINT ... CHECK with inherited tables (
792 ALTER TABLE RENAME update foreign-key trigger arguments correctly (Bre
794 DROP AGGREGATE and COMMENT ON AGGREGATE now accept an aggtype (Tom)
795 Add automatic return type data casting for SQL functions (Tom)
796 Allow GiST indexes to handle NULLs and multikey indexes (Oleg Bartunov
797 , Teodor Sigaev, Tom)
798 Enable partial indexes (Martijn van Oosterhout)
799 _________________________________________________________________
803 Add RESET ALL, SHOW ALL (Marko Kreen)
804 CREATE/ALTER USER/GROUP now allow options in any order (Vince)
805 Add LOCK A, B, C functionality (Neil Padgett)
806 New ENCRYPTED/UNENCRYPTED option to CREATE/ALTER USER (Bruce)
807 New light-weight VACUUM does not lock table; old semantics are availab
808 le as VACUUM FULL (Tom)
809 Disable COPY TO/FROM on views (Bruce)
810 COPY DELIMITERS string must be exactly one character (Tom)
811 VACUUM warning about index tuples fewer than heap now only appears whe
812 n appropriate (Martijn van Oosterhout)
813 Fix permission checks for CREATE INDEX (Tom)
814 Disallow inappropriate use of CREATE/DROP INDEX/TRIGGER/VIEW (Tom)
815 _________________________________________________________________
817 Data Types and Functions
819 SUM(), AVG(), COUNT() now uses int8 internally for speed (Tom)
820 Add convert(), convert2() (Tatsuo)
821 New function bit_length() (Peter E)
822 Make the "n" in CHAR(n)/VARCHAR(n) represents letters, not bytes (Tats
824 CHAR(), VARCHAR() now reject strings that are too long (Peter E)
825 BIT VARYING now rejects bit strings that are too long (Peter E)
826 BIT now rejects bit strings that do not match declared size (Peter E)
827 INET, CIDR text conversion functions (Alex Pilosov)
828 INET, CIDR operators << and <<= indexable (Alex Pilosov)
829 Bytea \### now requires valid three digit octal number
830 Bytea comparison improvements, now supports =, <>, >, >=, <, and <=
831 Bytea now supports B-tree indexes
832 Bytea now supports LIKE, LIKE...ESCAPE, NOT LIKE, NOT LIKE...ESCAPE
833 Bytea now supports concatenation
834 New bytea functions: position, substring, trim, btrim, and length
835 New encode() function mode, "escaped", converts minimally escaped byte
837 Add pg_database_encoding_max_length() (Tatsuo)
838 Add pg_client_encoding() function (Tatsuo)
839 now() returns time with millisecond precision (Thomas)
840 New TIMESTAMP WITHOUT TIMEZONE data type (Thomas)
841 Add ISO date/time specification with "T", yyyy-mm-ddThh:mm:ss (Thomas)
842 New xid/int comparison functions (Hiroshi)
843 Add precision to TIME, TIMESTAMP, and INVERVAL data types (Thomas)
844 Modify type coercion logic to attempt binary-compatible functions firs
846 New encode() function installed by default (Marko Kreen)
847 Improved to_*() conversion functions (Karel Zak)
848 Optimize LIKE/ILIKE when using single-byte encodings (Tatsuo)
849 New functions in contrib/pgcrypto: crypt(), hmac(), encrypt(), gen_sal
851 Correct description of translate() function (Bruce)
852 Add INTERVAL argument for SET TIME ZONE (Thomas)
853 Add INTERVAL YEAR TO MONTH (etc.) syntax (Thomas)
854 Optimize length functions when using single-byte encodings (Tatsuo)
855 Fix path_inter, path_distance, path_length, dist_ppath to handle close
856 d paths (Curtis Barrett, Tom)
857 octet_length(text) now returns non-compressed length (Tatsuo, Bruce)
858 Handle "July" full name in date/time literals (Greg Sabino Mullane)
859 Some datatype() function calls now evaluated differently
860 Add support for Julian and ISO time specifications (Thomas)
861 _________________________________________________________________
865 National language support in psql, pg_dump, libpq, and server (Peter E
867 Message translations in Chinese (simplified, traditional), Czech, Fren
868 ch, German, Hungarian, Russian, Swedish (Peter E, Serguei A. Mokhov, K
869 arel Zak, Weiping He, Zhenbang Wei, Kovacs Zoltan)
870 Make trim, ltrim, rtrim, btrim, lpad, rpad, translate multibyte aware
872 Add LATIN5,6,7,8,9,10 support (Tatsuo)
873 Add ISO 8859-5,6,7,8 support (Tatsuo)
874 Correct LATIN5 to mean ISO-8859-9, not ISO-8859-5 (Tatsuo)
875 Make mic2ascii() non-ASCII aware (Tatsuo)
876 Reject invalid multibyte character sequences (Tatsuo)
877 _________________________________________________________________
881 Now uses portals for SELECT loops, allowing huge result sets (Jan)
882 CURSOR and REFCURSOR support (Jan)
883 Can now return open cursors (Jan)
884 Add ELSEIF (Klaus Reger)
885 Improve PL/pgSQL error reporting, including location of error (Tom)
886 Allow IS or FOR key words in cursor declaration, for compatibility (Br
888 Fix for SELECT ... FOR UPDATE (Tom)
889 Fix for PERFORM returning multiple rows (Tom)
890 Make PL/pgSQL use the server's type coercion code (Tom)
891 Memory leak fix (Jan, Tom)
892 Make trailing semicolon optional (Tom)
893 _________________________________________________________________
897 New untrusted PL/Perl (Alex Pilosov)
898 PL/Perl is now built on some platforms even if libperl is not shared (
900 _________________________________________________________________
904 Now reports errorInfo (Vsevolod Lobko)
905 Add spi_lastoid function (bob@redivi.com)
906 _________________________________________________________________
910 ...is new (Andrew Bosma)
911 _________________________________________________________________
915 \d displays indexes in unique, primary groupings (Christopher Kings-Ly
917 Allow trailing semicolons in backslash commands (Greg Sabino Mullane)
918 Read password from /dev/tty if possible
919 Force new password prompt when changing user and database (Tatsuo, Tom
921 Format the correct number of columns for Unicode (Patrice)
922 _________________________________________________________________
926 New function PQescapeString() to escape quotes in command strings (Flo
928 New function PQescapeBytea() escapes binary strings for use as SQL str
930 _________________________________________________________________
934 Return OID of INSERT (Ken K)
935 Handle more data types (Ken K)
936 Handle single quotes and newlines in strings (Ken K)
937 Handle NULL variables (Ken K)
938 Fix for time zone handling (Barry Lind)
939 Improved Druid support
940 Allow eight-bit characters with non-multibyte server (Barry Lind)
941 Support BIT, BINARY types (Ned Wolpert)
942 Reduce memory usage (Michael Stephens, Dave Cramer)
943 Update DatabaseMetaData (Peter E)
944 Add DatabaseMetaData.getCatalogs() (Peter E)
945 Encoding fixes (Anders Bengtsson)
946 Get/setCatalog methods (Jason Davies)
947 DatabaseMetaData.getColumns() now returns column defaults (Jason Davie
949 DatabaseMetaData.getColumns() performance improvement (Jeroen van Vian
951 Some JDBC1 and JDBC2 merging (Anders Bengtsson)
952 Transaction performance improvements (Barry Lind)
953 Array fixes (Greg Zoller)
955 Fix batch processing (Rene Pijlman)
956 ExecSQL method reorganization (Anders Bengtsson)
957 GetColumn() fixes (Jeroen van Vianen)
958 Fix isWriteable() function (Rene Pijlman)
959 Improved passage of JDBC2 conformance tests (Rene Pijlman)
960 Add bytea type capability (Barry Lind)
961 Add isNullable() (Rene Pijlman)
962 JDBC date/time test suite fixes (Liam Stewart)
963 Fix for SELECT 'id' AS xxx FROM table (Dave Cramer)
964 Fix DatabaseMetaData to show precision properly (Mark Lillywhite)
965 New getImported/getExported keys (Jason Davies)
966 MD5 password encryption support (Jeremy Wohl)
967 Fix to actually use type cache (Ned Wolpert)
968 _________________________________________________________________
972 Remove query size limit (Hiroshi)
973 Remove text field size limit (Hiroshi)
974 Fix for SQLPrimaryKeys in multibyte mode (Hiroshi)
975 Allow ODBC procedure calls (Hiroshi)
976 Improve boolean handing (Aidan Mountford)
977 Most configuration options on setable via DSN (Hiroshi)
978 Multibyte, performance fixes (Hiroshi)
979 Allow driver to be used with iODBC or unixODBC (Peter E)
980 MD5 password encryption support (Bruce)
981 Add more compatibility functions to odbc.sql (Peter E)
982 _________________________________________________________________
986 EXECUTE ... INTO implemented (Christof Petig)
987 Multiple row descriptor support (e.g. CARDINALITY) (Christof Petig)
988 Fix for GRANT parameters (Lee Kindness)
989 Fix INITIALLY DEFERRED bug
990 Various bug fixes (Michael, Christof Petig)
991 Auto allocation for indicator variable arrays (int *ind_p=NULL)
992 Auto allocation for string arrays (char **foo_pp=NULL)
993 ECPGfree_auto_mem fixed
994 All function names with external linkage are now prefixed by ECPG
995 Fixes for arrays of structures (Michael)
996 _________________________________________________________________
1000 Python fix fetchone() (Gerhard Haring)
1001 Use UTF, Unicode in Tcl where appropriate (Vsevolod Lobko, Reinhard Ma
1003 Add Tcl COPY TO/FROM (ljb)
1004 Prevent output of default index op class in pg_dump (Tom)
1005 Fix libpgeasy memory leak (Bruce)
1006 _________________________________________________________________
1010 Configure, dynamic loader, and shared library fixes (Peter E)
1011 Fixes in QNX 4 port (Bernd Tegge)
1012 Fixes in Cygwin and Win32 ports (Jason Tishler, Gerhard Haring, Dmitry
1013 Yurtaev, Darko Prenosil, Mikhail Terekhov)
1014 Fix for Win32 socket communication failures (Magnus, Mikhail Terekhov)
1015 Hurd compile fix (Oliver Elphick)
1016 BeOS fixes (Cyril Velter)
1017 Remove configure --enable-unicode-conversion, now enabled by multibyte
1019 AIX fixes (Tatsuo, Andreas)
1020 Fix parallel make (Peter E)
1021 Install SQL language manual pages into OS-specific directories (Peter
1023 Rename config.h to pg_config.h (Peter E)
1024 Reorganize installation layout of header files (Peter E)
1025 _________________________________________________________________
1029 Remove SEP_CHAR (Bruce)
1031 Merge GUC and command line handling (Marko Kreen)
1032 Remove EXTEND INDEX (Martijn van Oosterhout, Tom)
1033 New pgjindent utility to indent java code (Bruce)
1034 Remove define of true/false when compiling under C++ (Leandro Fanzone,
1036 pgindent fixes (Bruce, Tom)
1037 Replace strcasecmp() with strcmp() where appropriate (Peter E)
1038 Dynahash portability improvements (Tom)
1039 Add 'volatile' usage in spinlock structures
1040 Improve signal handling logic (Tom)
1041 _________________________________________________________________
1045 New contrib/rtree_gist (Oleg Bartunov, Teodor Sigaev)
1046 New contrib/tsearch full-text indexing (Oleg, Teodor Sigaev)
1047 Add contrib/dblink for remote database access (Joe Conway)
1048 contrib/ora2pg Oracle conversion utility (Gilles Darold)
1049 contrib/xml XML conversion utility (John Gray)
1050 contrib/fulltextindex fixes (Christopher Kings-Lynne)
1051 New contrib/fuzzystrmatch with levenshtein and metaphone, soundex merg
1053 Add contrib/intarray boolean queries, binary search, fixes (Oleg Bartu
1055 New pg_upgrade utility (Bruce)
1056 Add new pg_resetxlog options (Bruce, Tom)
1057 _________________________________________________________________
1061 Release date: 2001-08-15
1062 _________________________________________________________________
1064 Migration to version 7.1.3
1066 A dump/restore is *not* required for those running 7.1.X.
1067 _________________________________________________________________
1071 Remove unused WAL segements of large transactions (Tom)
1072 Multiaction rule fix (Tom)
1073 PL/pgSQL memory allocation fix (Jan)
1074 VACUUM buffer fix (Tom)
1075 Regression test fixes (Tom)
1076 pg_dump fixes for GRANT/REVOKE/comments on views, user-defined types (Tom)
1077 Fix subselects with DISTINCT ON or LIMIT (Tom)
1079 Disable COPY TO/FROM a view (Tom)
1080 Cygwin build (Jason Tishler)
1082 _________________________________________________________________
1086 Release date: 2001-05-11
1088 This has one fix from 7.1.1.
1089 _________________________________________________________________
1091 Migration to version 7.1.2
1093 A dump/restore is *not* required for those running 7.1.X.
1094 _________________________________________________________________
1098 Fix PL/pgSQL SELECTs when returning no rows
1099 Fix for psql backslash core dump
1100 Referential integrity permission fix
1104 _________________________________________________________________
1108 Release date: 2001-05-05
1110 This has a variety of fixes from 7.1.
1111 _________________________________________________________________
1113 Migration to version 7.1.1
1115 A dump/restore is *not* required for those running 7.1.
1116 _________________________________________________________________
1120 Fix for numeric MODULO operator (Tom)
1121 pg_dump fixes (Philip)
1122 pg_dump can dump 7.0 databases (Philip)
1123 readline 4.2 fixes (Peter E)
1125 AIX, MSWIN, VAX, N32K fixes (Tom)
1126 Multibytes fixes (Tom)
1127 Unicode fixes (Tatsuo)
1128 Optimizer improvements (Tom)
1129 Fix for whole tuples in functions (Tom)
1130 Fix for pg_ctl and option strings with spaces (Peter E)
1131 ODBC fixes (Hiroshi)
1132 EXTRACT can now take string argument (Thomas)
1133 Python fixes (Darcy)
1135 _________________________________________________________________
1139 Release date: 2001-04-13
1141 This release focuses on removing limitations that have existed in the
1142 PostgreSQL code for many years.
1144 Major changes in this release:
1146 Write-ahead Log (WAL)
1147 To maintain database consistency in case of an operating system
1148 crash, previous releases of PostgreSQL have forced all data
1149 modifications to disk before each transaction commit. With WAL,
1150 only one log file must be flushed to disk, greatly improving
1151 performance. If you have been using -F in previous releases to
1152 disable disk flushes, you may want to consider discontinuing
1156 TOAST - Previous releases had a compiled-in row length limit,
1157 typically 8k - 32k. This limit made storage of long text fields
1158 difficult. With TOAST, long rows of any length can be stored
1159 with good performance.
1162 We now support outer joins. The UNION/NOT IN workaround for
1163 outer joins is no longer required. We use the SQL92 outer join
1167 The previous C function manager did not handle null values
1168 properly, nor did it support 64-bit CPU's (Alpha). The new
1169 function manager does. You can continue using your old custom
1170 functions, but you may want to rewrite them in the future to
1171 use the new function manager call interface.
1174 A large number of complex queries that were unsupported in
1175 previous releases now work. Many combinations of views,
1176 aggregates, UNION, LIMIT, cursors, subqueries, and inherited
1177 tables now work properly. Inherited tables are now accessed by
1178 default. Subqueries in FROM are now supported.
1179 _________________________________________________________________
1181 Migration to version 7.1
1183 A dump/restore using pg_dump is required for those wishing to migrate
1184 data from any previous release.
1185 _________________________________________________________________
1191 Many multibyte/Unicode/locale fixes (Tatsuo and others)
1192 More reliable ALTER TABLE RENAME (Tom)
1193 Kerberos V fixes (David Wragg)
1194 Fix for INSERT INTO...SELECT where targetlist has subqueries (Tom)
1195 Prompt username/password on standard error (Bruce)
1196 Large objects inv_read/inv_write fixes (Tom)
1197 Fixes for to_char(), to_date(), to_ascii(), and to_timestamp() (Karel,
1199 Prevent query expressions from leaking memory (Tom)
1200 Allow UPDATE of arrays elements (Tom)
1201 Wake up lock waiters during cancel (Hiroshi)
1202 Fix rare cursor crash when using hash join (Tom)
1203 Fix for DROP TABLE/INDEX in rolled-back transaction (Hiroshi)
1204 Fix psql crash from \l+ if MULTIBYTE enabled (Peter E)
1205 Fix truncation of rule names during CREATE VIEW (Ross Reedstrom)
1206 Fix PL/perl (Alex Kapranoff)
1207 Disallow LOCK on views (Mark Hollomon)
1208 Disallow INSERT/UPDATE/DELETE on views (Mark Hollomon)
1209 Disallow DROP RULE, CREATE INDEX, TRUNCATE on views (Mark Hollomon)
1210 Allow PL/pgSQL accept non-ASCII identifiers (Tatsuo)
1211 Allow views to proper handle GROUP BY, aggregates, DISTINCT (Tom)
1212 Fix rare failure with TRUNCATE command (Tom)
1213 Allow UNION/INTERSECT/EXCEPT to be used with ALL, subqueries, views,
1214 DISTINCT, ORDER BY, SELECT...INTO (Tom)
1215 Fix parser failures during aborted transactions (Tom)
1216 Allow temporary relations to properly clean up indexes (Bruce)
1217 Fix VACUUM problem with moving rows in same page (Tom)
1218 Modify pg_dump to better handle user-defined items in template1 (Philip)
1219 Allow LIMIT in VIEW (Tom)
1220 Require cursor FETCH to honor LIMIT (Tom)
1221 Allow PRIMARY/FOREIGN Key definitions on inherited columns (Stephan)
1222 Allow ORDER BY, LIMIT in sub-selects (Tom)
1223 Allow UNION in CREATE RULE (Tom)
1224 Make ALTER/DROP TABLE rollback-able (Vadim, Tom)
1225 Store initdb collation in pg_control so collation cannot be changed (Tom)
1226 Fix INSERT...SELECT with rules (Tom)
1227 Fix FOR UPDATE inside views and subselects (Tom)
1228 Fix OVERLAPS operators conform to SQL92 spec regarding NULLs (Tom)
1229 Fix lpad() and rpad() to handle length less than input string (Tom)
1230 Fix use of NOTIFY in some rules (Tom)
1231 Overhaul btree code (Tom)
1232 Fix NOT NULL use in Pl/pgSQL variables (Tom)
1233 Overhaul GIST code (Oleg)
1234 Fix CLUSTER to preserve constraints and column default (Tom)
1235 Improved deadlock detection handling (Tom)
1236 Allow multiple SERIAL columns in a table (Tom)
1237 Prevent occasional index corruption (Vadim)
1241 Add OUTER JOINs (Tom)
1242 Function manager overhaul (Tom)
1243 Allow ALTER TABLE RENAME on indexes (Tom)
1244 Improve CLUSTER (Tom)
1245 Improve ps status display for more platforms (Peter E, Marc)
1246 Improve CREATE FUNCTION failure message (Ross)
1247 JDBC improvements (Peter, Travis Bauer, Christopher Cain, William Webber,
1249 Grand Unified Configuration scheme/GUC. Many options can now be set in
1250 data/postgresql.conf, postmaster/postgres flags, or SET commands (Peter E)
1251 Improved handling of file descriptor cache (Tom)
1252 New warning code about auto-created table alias entries (Bruce)
1253 Overhaul initdb process (Tom, Peter E)
1254 Overhaul of inherited tables; inherited tables now accessed by default;
1255 new ONLY keyword prevents it (Chris Bitmead, Tom)
1256 ODBC cleanups/improvements (Nick Gorham, Stephan Szabo, Zoltan Kovacs,
1258 Allow renaming of temp tables (Tom)
1259 Overhaul memory manager contexts (Tom)
1260 pg_dumpall uses CREATE USER or CREATE GROUP rather using COPY (Peter E)
1261 Overhaul pg_dump (Philip Warner)
1262 Allow pg_hba.conf secondary password file to specify only username (Peter E)
1263 Allow TEMPORARY or TEMP keyword when creating temporary tables (Bruce)
1264 New memory leak checker (Karel)
1265 New SET SESSION CHARACTERISTICS (Thomas)
1266 Allow nested block comments (Thomas)
1267 Add WITHOUT TIME ZONE type qualifier (Thomas)
1268 New ALTER TABLE ADD CONSTRAINT (Stephan)
1269 Use NUMERIC accumulators for INTEGER aggregates (Tom)
1270 Overhaul aggregate code (Tom)
1271 New VARIANCE and STDDEV() aggregates
1272 Improve dependency ordering of pg_dump (Philip)
1273 New pg_restore command (Philip)
1274 New pg_dump tar output option (Philip)
1275 New pg_dump of large objects (Philip)
1276 New ESCAPE option to LIKE (Thomas)
1277 New case-insensitive LIKE - ILIKE (Thomas)
1278 Allow functional indexes to use binary-compatible type (Tom)
1279 Allow SQL functions to be used in more contexts (Tom)
1280 New pg_config utility (Peter E)
1281 New PL/pgSQL EXECUTE command which allows dynamic SQL and utility statements
1283 New PL/pgSQL GET DIAGNOSTICS statement for SPI value access (Jan)
1284 New quote_identifiers() and quote_literal() functions (Jan)
1285 New ALTER TABLE table OWNER TO user command (Mark Hollomon)
1286 Allow subselects in FROM, i.e. FROM (SELECT ...) [AS] alias (Tom)
1287 Update PyGreSQL to version 3.1 (D'Arcy)
1288 Store tables as files named by OID (Vadim)
1289 New SQL function setval(seq,val,bool) for use in pg_dump (Philip)
1290 Require DROP VIEW to remove views, no DROP TABLE (Mark)
1291 Allow DROP VIEW view1, view2 (Mark)
1292 Allow multiple objects in DROP INDEX, DROP RULE, and DROP TYPE (Tom)
1293 Allow automatic conversion to/from Unicode (Tatsuo, Eiji)
1294 New /contrib/pgcrypto hashing functions (Marko Kreen)
1295 New pg_dumpall --globals-only option (Peter E)
1296 New CHECKPOINT command for WAL which creates new WAL log file (Vadim)
1297 New AT TIME ZONE syntax (Thomas)
1298 Allow location of Unix domain socket to be configurable (David J. MacKenzie)
1299 Allow postmaster to listen on a specific IP address (David J. MacKenzie)
1300 Allow socket path name to be specified in hostname by using leading slash
1301 (David J. MacKenzie)
1302 Allow CREATE DATABASE to specify template database (Tom)
1303 New utility to convert MySQL schema dumps to SQL92 and PostgreSQL (Thomas)
1304 New /contrib/rserv replication toolkit (Vadim)
1305 New file format for COPY BINARY (Tom)
1306 New /contrib/oid2name to map numeric files to table names (B Palmer)
1307 New "idle in transaction" ps status message (Marc)
1308 Update to pgaccess 0.98.7 (Constantin Teodorescu)
1309 pg_ctl now defaults to -w (wait) on shutdown, new -l (log) option
1310 Add rudimentary dependency checking to pg_dump (Philip)
1314 Fix INET/CIDR type ordering and add new functions (Tom)
1315 Make OID behave as an unsigned type (Tom)
1316 Allow BIGINT as synonym for INT8 (Peter E)
1317 New int2 and int8 comparison operators (Tom)
1318 New BIT and BIT VARYING types (Adriaan Joubert, Tom, Peter E)
1319 CHAR() no longer faster than VARCHAR() because of TOAST (Tom)
1320 New GIST seg/cube examples (Gene Selkov)
1321 Improved round(numeric) handling (Tom)
1322 Fix CIDR output formatting (Tom)
1323 New CIDR abbrev() function (Tom)
1327 Write-Ahead Log (WAL) to provide crash recovery with less performance
1329 ANALYZE stage of VACUUM no longer exclusively locks table (Bruce)
1330 Reduced file seeks (Denis Perchine)
1331 Improve BTREE code for duplicate keys (Tom)
1332 Store all large objects in a single table (Denis Perchine, Tom)
1333 Improve memory allocation performance (Karel, Tom)
1337 New function manager call conventions (Tom)
1338 SGI portability fixes (David Kaelbling)
1339 New configure --enable-syslog option (Peter E)
1340 New BSDI README (Bruce)
1341 configure script moved to top level, not /src (Peter E)
1342 Makefile/configuration/compilation overhaul (Peter E)
1343 New configure --with-python option (Peter E)
1344 Solaris cleanups (Peter E)
1345 Overhaul /contrib Makefiles (Karel)
1346 New OpenSSL configuration option (Magnus, Peter E)
1348 QNX fixes (Maurizio)
1349 New heap_open(), heap_openr() API (Tom)
1350 Remove colon and semi-colon operators (Thomas)
1351 New pg_class.relkind value for views (Mark Hollomon)
1352 Rename ichar() to chr() (Karel)
1353 New documentation for btrim(), ascii(), chr(), repeat() (Karel)
1354 Fixes for NT/Cygwin (Pete Forman)
1355 AIX port fixes (Andreas)
1356 New BeOS port (David Reid, Cyril Velter)
1357 Add proofreader's changes to docs (Addison-Wesley, Bruce)
1358 New Alpha spinlock code (Adriaan Joubert, Compaq)
1359 UnixWare port overhaul (Peter E)
1360 New Darwin/MacOS X port (Peter Bierman, Bruce Hartzler)
1361 New FreeBSD Alpha port (Alfred)
1362 Overhaul shared memory segments (Tom)
1363 Add IBM S/390 support (Neale Ferguson)
1364 Moved macmanuf to /contrib (Larry Rosenman)
1365 Syslog improvements (Larry Rosenman)
1366 New template0 database that contains no user additions (Tom)
1367 New /contrib/cube and /contrib/seg GIST sample code (Gene Selkov)
1368 Allow NetBSD's libedit instead of readline (Peter)
1369 Improved assembly language source code format (Bruce)
1370 New contrib/pg_logger
1371 New --template option to createdb
1372 New contrib/pg_control utility (Oliver)
1373 New FreeBSD tools ipc_check, start-scripts/freebsd
1375 _________________________________________________________________
1379 Release date: 2000-11-11
1381 This has a variety of fixes from 7.0.2.
1382 _________________________________________________________________
1384 Migration to version 7.0.3
1386 A dump/restore is *not* required for those running 7.0.*.
1387 _________________________________________________________________
1392 Large object fix (Tom)
1393 Fix lean in COPY WITH OIDS leak (Tom)
1394 Fix backwards-index-scan (Tom)
1395 Fix SELECT ... FOR UPDATE so it checks for duplicate keys (Hiroshi)
1396 Add --enable-syslog to configure (Marc)
1397 Fix abort transaction at backend exit in rare cases (Tom)
1398 Fix for psql \l+ when multibyte enabled (Tatsuo)
1399 Allow PL/pgSQL to accept non ascii identifiers (Tatsuo)
1400 Make vacuum always flush buffers (Tom)
1401 Fix to allow cancel while waiting for a lock (Hiroshi)
1402 Fix for memory aloocation problem in user authentication code (Tom)
1403 Remove bogus use of int4out() (Tom)
1404 Fixes for multiple subqueries in COALESCE or BETWEEN (Tom)
1405 Fix for failure of triggers on heap open in certain cases (Jeroen van
1407 Fix for erroneous selectivity of not-equals (Tom)
1408 Fix for erroneous use of strcmp() (Tom)
1409 Fix for bug where storage manager accesses items beyond end of file
1411 Fix to include kernel errno message in all smgr elog messages (Tom)
1412 Fix for '.' not in PATH at build time (SL Baur)
1413 Fix for out-of-file-descriptors error (Tom)
1414 Fix to make pg_dump dump 'iscachable' flag for functions (Tom)
1415 Fix for subselect in targetlist of Append node (Tom)
1416 Fix for mergejoin plans (Tom)
1417 Fix TRUNCATE failure on relations with indexes (Tom)
1418 Avoid database-wide restart on write error (Hiroshi)
1419 Fix nodeMaterial to honor chgParam by recomputing its output (Tom)
1420 Fix VACUUM problem with moving chain of update tuples when source and
1421 destination of a tuple lie on the same page (Tom)
1422 Fix user.c CommandCounterIncrement (Tom)
1423 Fix for AM/PM boundary problem in to_char() (Karel Zak)
1424 Fix TIME aggregate handling (Tom)
1425 Fix to_char() to avoid coredump on NULL input (Tom)
1427 Fix for inserting/copying longer multibyte strings into char() data
1429 Fix for crash of backend, on abort (Tom)
1431 _________________________________________________________________
1435 Release date: 2000-06-05
1437 This is a repackaging of 7.0.1 with added documentation.
1438 _________________________________________________________________
1440 Migration to version 7.0.2
1442 A dump/restore is *not* required for those running 7.*.
1443 _________________________________________________________________
1447 Added documentation to tarball.
1449 _________________________________________________________________
1453 Release date: 2000-06-01
1455 This is a cleanup release for 7.0.
1456 _________________________________________________________________
1458 Migration to version 7.0.1
1460 A dump/restore is *not* required for those running 7.0.
1461 _________________________________________________________________
1465 Fix many CLUSTER failures (Tom)
1466 Allow ALTER TABLE RENAME works on indexes (Tom)
1467 Fix plpgsql to handle datetime->timestamp and timespan->interval (Bruce)
1468 New configure --with-setproctitle switch to use setproctitle() (Marc, Bruce)
1469 Fix the off by one errors in ResultSet from 6.5.3, and more.
1470 jdbc ResultSet fixes (Joseph Shraibman)
1471 optimizer tunings (Tom)
1472 Fix create user for pgaccess
1473 Fix for UNLISTEN failure
1474 IRIX fixes (David Kaelbling)
1475 QNX fixes (Andreas Kardos)
1476 Reduce COPY IN lock level (Tom)
1477 Change libpqeasy to use PQconnectdb() style parameters (Bruce)
1478 Fix pg_dump to handle OID indexes (Tom)
1479 Fix small memory leak (Tom)
1480 Solaris fix for createdb/dropdb (Tatsuo)
1481 Fix for non-blocking connections (Alfred Perlstein)
1482 Fix improper recovery after RENAME TABLE failures (Tom)
1483 Copy pg_ident.conf.sample into /lib directory in install (Bruce)
1484 Add SJIS UDC (NEC selection IBM kanji) support (Eiji Tokuya)
1485 Fix too long syslog message (Tatsuo)
1486 Fix problem with quoted indexes that are too long (Tom)
1487 JDBC ResultSet.getTimestamp() fix (Gregory Krasnow & Floyd Marinescu)
1488 ecpg changes (Michael)
1490 _________________________________________________________________
1494 Release date: 2000-05-08
1496 This release contains improvements in many areas, demonstrating the
1497 continued growth of PostgreSQL. There are more improvements and fixes
1498 in 7.0 than in any previous release. The developers have confidence
1499 that this is the best release yet; we do our best to put out only
1500 solid releases, and this one is no exception.
1502 Major changes in this release:
1505 Foreign keys are now implemented, with the exception of PARTIAL
1506 MATCH foreign keys. Many users have been asking for this
1507 feature, and we are pleased to offer it.
1510 Continuing on work started a year ago, the optimizer has been
1511 improved, allowing better query plan selection and faster
1512 performance with less memory usage.
1515 psql, our interactive terminal monitor, has been updated with a
1516 variety of new features. See the psql manual page for details.
1519 SQL92 join syntax is now supported, though only as INNER JOIN
1520 for this release. JOIN, NATURAL JOIN, JOIN/USING, and JOIN/ON
1521 are available, as are column correlation names.
1522 _________________________________________________________________
1524 Migration to version 7.0
1526 A dump/restore using pg_dump is required for those wishing to migrate
1527 data from any previous release of PostgreSQL. For those upgrading from
1528 6.5.*, you may instead use pg_upgrade to upgrade to this release;
1529 however, a full dump/reload installation is always the most robust
1530 method for upgrades.
1532 Interface and compatibility issues to consider for the new release
1535 * The date/time types datetime and timespan have been superseded by
1536 the SQL92-defined types timestamp and interval. Although there has
1537 been some effort to ease the transition by allowing PostgreSQL to
1538 recognize the deprecated type names and translate them to the new
1539 type names, this mechanism may not be completely transparent to
1540 your existing application.
1541 * The optimizer has been substantially improved in the area of query
1542 cost estimation. In some cases, this will result in decreased
1543 query times as the optimizer makes a better choice for the
1544 preferred plan. However, in a small number of cases, usually
1545 involving pathological distributions of data, your query times may
1546 go up. If you are dealing with large amounts of data, you may want
1547 to check your queries to verify performance.
1548 * The JDBC and ODBC interfaces have been upgraded and extended.
1549 * The string function CHAR_LENGTH is now a native function. Previous
1550 versions translated this into a call to LENGTH, which could result
1551 in ambiguity with other types implementing LENGTH such as the
1553 _________________________________________________________________
1559 Prevent function calls exceeding maximum number of arguments (Tom)
1560 Improve CASE construct (Tom)
1561 Fix SELECT coalesce(f1,0) FROM int4_tbl GROUP BY f1 (Tom)
1562 Fix SELECT sentence.words[0] FROM sentence GROUP BY sentence.words[0] (Tom)
1563 Fix GROUP BY scan bug (Tom)
1564 Improvements in SQL grammar processing (Tom)
1565 Fix for views involved in INSERT ... SELECT ... (Tom)
1566 Fix for SELECT a/2, a/2 FROM test_missing_target GROUP BY a/2 (Tom)
1567 Fix for subselects in INSERT ... SELECT (Tom)
1568 Prevent INSERT ... SELECT ... ORDER BY (Tom)
1569 Fixes for relations greater than 2GB, including vacuum
1570 Improve propagating system table changes to other backends (Tom)
1571 Improve propagating user table changes to other backends (Tom)
1572 Fix handling of temp tables in complex situations (Bruce, Tom)
1573 Allow table locking at table open, improving concurrent reliability (Tom)
1574 Properly quote sequence names in pg_dump (Ross J. Reedstrom)
1575 Prevent DROP DATABASE while others accessing
1576 Prevent any rows from being returned by GROUP BY if no rows processed (Tom)
1577 Fix SELECT COUNT(1) FROM table WHERE ...' if no rows matching WHERE (Tom)
1578 Fix pg_upgrade so it works for MVCC (Tom)
1579 Fix for SELECT ... WHERE x IN (SELECT ... HAVING SUM(x) > 1) (Tom)
1580 Fix for "f1 datetime DEFAULT 'now'" (Tom)
1581 Fix problems with CURRENT_DATE used in DEFAULT (Tom)
1582 Allow comment-only lines, and ;;; lines too. (Tom)
1583 Improve recovery after failed disk writes, disk full (Hiroshi)
1584 Fix cases where table is mentioned in FROM but not joined (Tom)
1585 Allow HAVING clause without aggregate functions (Tom)
1586 Fix for "--" comment and no trailing newline, as seen in perl interface
1587 Improve pg_dump failure error reports (Bruce)
1588 Allow sorts and hashes to exceed 2GB file sizes (Tom)
1589 Fix for pg_dump dumping of inherited rules (Tom)
1590 Fix for NULL handling comparisons (Tom)
1591 Fix inconsistent state caused by failed CREATE/DROP commands (Hiroshi)
1592 Fix for dbname with dash
1593 Prevent DROP INDEX from interfering with other backends (Tom)
1594 Fix file descriptor leak in verify_password()
1595 Fix for "Unable to identify an operator =$" problem
1596 Fix ODBC so no segfault if CommLog and Debug enabled (Dirk Niggemann)
1597 Fix for recursive exit call (Massimo)
1598 Fix for extra-long timezones (Jeroen van Vianen)
1599 Make pg_dump preserve primary key information (Peter E)
1600 Prevent databases with single quotes (Peter E)
1601 Prevent DROP DATABASE inside transaction (Peter E)
1602 ecpg memory leak fixes (Stephen Birch)
1603 Fix for SELECT null::text, SELECT int4fac(null) and SELECT 2 + (null) (Tom)
1604 Y2K timestamp fix (Massimo)
1605 Fix for VACUUM 'HEAP_MOVED_IN was not expected' errors (Tom)
1606 Fix for views with tables/columns containing spaces (Tom)
1607 Prevent permissions on indexes (Peter E)
1608 Fix for spinlock stuck problem when error is generated (Hiroshi)
1609 Fix ipcclean on Linux
1610 Fix handling of NULL constraint conditions (Tom)
1611 Fix memory leak in odbc driver (Nick Gorham)
1612 Fix for permission check on UNION tables (Tom)
1613 Fix to allow SELECT 'a' LIKE 'a' (Tom)
1614 Fix for SELECT 1 + NULL (Tom)
1616 Fix log() on numeric type (Tom)
1617 Deprecate ':' and ';' operators
1618 Allow vacuum of temporary tables
1619 Disallow inherited columns with the same name as new columns
1620 Recover or force failure when disk space is exhausted (Hiroshi)
1621 Fix INSERT INTO ... SELECT with AS columns matching result columns
1622 Fix INSERT ... SELECT ... GROUP BY groups by target columns not source columns
1624 Fix CREATE TABLE test (a char(5) DEFAULT text '', b int4) with INSERT (Tom)
1625 Fix UNION with LIMIT
1626 Fix CREATE TABLE x AS SELECT 1 UNION SELECT 2
1627 Fix CREATE TABLE test(col char(2) DEFAULT user)
1628 Fix mismatched types in CREATE TABLE ... DEFAULT
1629 Fix SELECT * FROM pg_class where oid in (0,-1)
1630 Fix SELECT COUNT('asdf') FROM pg_class WHERE oid=12
1631 Prevent user who can create databases can modifying pg_database table (Peter E)
1632 Fix btree to give a useful elog when key > 1/2 (page - overhead) (Tom)
1633 Fix INSERT of 0.0 into DECIMAL(4,4) field (Tom)
1637 New CLI interface include file sqlcli.h, based on SQL3/SQL98
1638 Remove all limits on query length, row length limit still exists (Tom)
1639 Update jdbc protocol to 2.0 (Jens Glaser <jens@jens.de>)
1640 Add TRUNCATE command to quickly truncate relation (Mike Mascari)
1641 Fix to give super user and createdb user proper update catalog rights (Peter E)
1642 Allow ecpg bool variables to have NULL values (Christof)
1643 Issue ecpg error if NULL value for variable with no NULL indicator (Christof)
1644 Allow ^C to cancel COPY command (Massimo)
1645 Add SET FSYNC and SHOW PG_OPTIONS commands(Massimo)
1646 Function name overloading for dynamically-loaded C functions (Frankpitt)
1647 Add CmdTuples() to libpq++(Vince)
1648 New CREATE CONSTRAINT TRIGGER and SET CONSTRAINTS commands(Jan)
1649 Allow CREATE FUNCTION/WITH clause to be used for all language types
1650 configure --enable-debug adds -g (Peter E)
1651 configure --disable-debug removes -g (Peter E)
1652 Allow more complex default expressions (Tom)
1653 First real FOREIGN KEY constraint trigger functionality (Jan)
1654 Add FOREIGN KEY ... MATCH FULL ... ON DELETE CASCADE (Jan)
1655 Add FOREIGN KEY ... MATCH <unspecified> referential actions (Don Baccus)
1656 Allow WHERE restriction on ctid (physical heap location) (Hiroshi)
1657 Move pginterface from contrib to interface directory, rename to pgeasy (Bruce)
1658 Change pgeasy connectdb() parameter ordering (Bruce)
1659 Require SELECT DISTINCT target list to have all ORDER BY columns (Tom)
1660 Add Oracle's COMMENT ON command (Mike Mascari <mascarim@yahoo.com>)
1661 libpq's PQsetNoticeProcessor function now returns previous hook(Peter E)
1662 Prevent PQsetNoticeProcessor from being set to NULL (Peter E)
1663 Make USING in COPY optional (Bruce)
1664 Allow subselects in the target list (Tom)
1665 Allow subselects on the left side of comparison operators (Tom)
1666 New parallel regression test (Jan)
1667 Change backend-side COPY to write files with permissions 644 not 666 (Tom)
1668 Force permissions on PGDATA directory to be secure, even if it exists (Tom)
1669 Added psql LASTOID variable to return last inserted oid (Peter E)
1670 Allow concurrent vacuum and remove pg_vlock vacuum lock file (Tom)
1671 Add permissions check for vacuum (Peter E)
1672 New libpq functions to allow asynchronous connections: PQconnectStart(),
1673 PQconnectPoll(), PQresetStart(), PQresetPoll(), PQsetenvStart(),
1674 PQsetenvPoll(), PQsetenvAbort (Ewan Mellor)
1675 New libpq PQsetenv() function (Ewan Mellor)
1676 create/alter user extension (Peter E)
1677 New postmaster.pid and postmaster.opts under $PGDATA (Tatsuo)
1678 New scripts for create/drop user/db (Peter E)
1679 Major psql overhaul (Peter E)
1680 Add const to libpq interface (Peter E)
1681 New libpq function PQoidValue (Peter E)
1682 Show specific non-aggregate causing problem with GROUP BY (Tom)
1683 Make changes to pg_shadow recreate pg_pwd file (Peter E)
1684 Add aggregate(DISTINCT ...) (Tom)
1685 Allow flag to control COPY input/output of NULLs (Peter E)
1686 Make postgres user have a password by default (Peter E)
1687 Add CREATE/ALTER/DROP GROUP (Peter E)
1688 All administration scripts now support --long options (Peter E, Karel)
1689 Vacuumdb script now supports --all option (Peter E)
1690 ecpg new portable FETCH syntax
1691 Add ecpg EXEC SQL IFDEF, EXEC SQL IFNDEF, EXEC SQL ELSE, EXEC SQL ELIF
1692 and EXEC SQL ENDIF directives
1693 Add pg_ctl script to control backend start-up (Tatsuo)
1694 Add postmaster.opts.default file to store start-up flags (Tatsuo)
1695 Allow --with-mb=SQL_ASCII
1696 Increase maximum number of index keys to 16 (Bruce)
1697 Increase maximum number of function arguments to 16 (Bruce)
1698 Allow configuration of maximum number of index keys and arguments (Bruce)
1699 Allow unprivileged users to change their passwords (Peter E)
1700 Password authentication enabled; required for new users (Peter E)
1701 Disallow dropping a user who owns a database (Peter E)
1702 Change initdb option --with-mb to --enable-multibyte
1703 Add option for initdb to prompts for superuser password (Peter E)
1704 Allow complex type casts like col::numeric(9,2) and col::int2::float8 (Tom)
1705 Updated user interfaces on initdb, initlocation, pg_dump, ipcclean (Peter E)
1706 New pg_char_to_encoding() and pg_encoding_to_char() functions (Tatsuo)
1707 Libpq non-blocking mode (Alfred Perlstein)
1708 Improve conversion of types in casts that don't specify a length
1709 New plperl internal programming language (Mark Hollomon)
1710 Allow COPY IN to read file that do not end with a newline (Tom)
1711 Indicate when long identifiers are truncated (Tom)
1712 Allow aggregates to use type equivalency (Peter E)
1713 Add Oracle's to_char(), to_date(), to_datetime(), to_timestamp(), to_number()
1714 conversion functions (Karel Zak <zakkr@zf.jcu.cz>)
1715 Add SELECT DISTINCT ON (expr [, expr ...]) targetlist ... (Tom)
1716 Check to be sure ORDER BY is compatible with the DISTINCT operation (Tom)
1717 Add NUMERIC and int8 types to ODBC
1718 Improve EXPLAIN results for Append, Group, Agg, Unique (Tom)
1719 Add ALTER TABLE ... ADD FOREIGN KEY (Stephan Szabo)
1720 Allow SELECT .. FOR UPDATE in PL/pgSQL (Hiroshi)
1721 Enable backward sequential scan even after reaching EOF (Hiroshi)
1722 Add btree indexing of boolean values, >= and <= (Don Baccus)
1723 Print current line number when COPY FROM fails (Massimo)
1724 Recognize POSIX time zone e.g. "PST+8" and "GMT-8" (Thomas)
1725 Add DEC as synonym for DECIMAL (Thomas)
1726 Add SESSION_USER as SQL92 keyword, same as CURRENT_USER (Thomas)
1727 Implement SQL92 column aliases (aka correlation names) (Thomas)
1728 Implement SQL92 join syntax (Thomas)
1729 Make INTERVAL reserved word allowed as a column identifier (Thomas)
1730 Implement REINDEX command (Hiroshi)
1731 Accept ALL in aggregate function SUM(ALL col) (Tom)
1732 Prevent GROUP BY from using column aliases (Tom)
1733 New psql \encoding option (Tatsuo)
1734 Allow PQrequestCancel() to terminate when in waiting-for-lock state (Hiroshi)
1735 Allow negation of a negative number in all cases
1736 Add ecpg descriptors (Christof, Michael)
1737 Allow CREATE VIEW v AS SELECT f1::char(8) FROM tbl
1738 Allow casts with length, like foo::char(8)
1739 New libpq functions PQsetClientEncoding(), PQclientEncoding() (Tatsuo)
1740 Add support for SJIS user defined characters (Tatsuo)
1741 Larger views/rules supported
1742 Make libpq's PQconndefaults() thread-safe (Tom)
1743 Disable // as comment to be ANSI conforming, should use -- (Tom)
1744 Allow column aliases on views CREATE VIEW name (collist)
1745 Fixes for views with subqueries (Tom)
1746 Allow UPDATE table SET fld = (SELECT ...) (Tom)
1747 SET command options no longer require quotes
1748 Update pgaccess to 0.98.6
1749 New SET SEED command
1750 New pg_options.sample file
1751 New SET FSYNC command (Massimo)
1752 Allow pg_descriptions when creating tables
1753 Allow pg_descriptions when creating types, columns, and functions
1754 Allow psql \copy to allow delimiters (Peter E)
1755 Allow psql to print nulls as distinct from "" [null] (Peter E)
1759 Many array fixes (Tom)
1760 Allow bare column names to be subscripted as arrays (Tom)
1761 Improve type casting of int and float constants (Tom)
1762 Cleanups for int8 inputs, range checking, and type conversion (Tom)
1763 Fix for SELECT timespan('21:11:26'::time) (Tom)
1764 netmask('x.x.x.x/0') is 255.255.255.255 instead of 0.0.0.0 (Oleg Sharoiko)
1765 Add btree index on NUMERIC (Jan)
1766 Perl fix for large objects containing NUL characters (Douglas Thomson)
1767 ODBC fix for for large objects (free)
1768 Fix indexing of cidr data type
1769 Fix for Ethernet MAC addresses (macaddr type) comparisons
1770 Fix for date/time types when overflows happened in computations (Tom)
1771 Allow array on int8 (Peter E)
1772 Fix for rounding/overflow of NUMERIC type, like NUMERIC(4,4) (Tom)
1773 Allow NUMERIC arrays
1774 Fix bugs in NUMERIC ceil() and floor() functions (Tom)
1775 Make char_length()/octet_length including trailing blanks (Tom)
1776 Made abstime/reltime use int4 instead of time_t (Peter E)
1777 New lztext data type for compressed text fields
1778 Revise code to handle coercion of int and float constants (Tom)
1779 Start at new code to implement a BIT and BIT VARYING type (Adriaan Joubert)
1780 NUMERIC now accepts scientific notation (Tom)
1781 NUMERIC to int4 rounds (Tom)
1782 Convert float4/8 to NUMERIC properly (Tom)
1783 Allow type conversion with NUMERIC (Thomas)
1784 Make ISO date style (2000-02-16 09:33) the default (Thomas)
1785 Add NATIONAL CHAR [ VARYING ] (Thomas)
1786 Allow NUMERIC round and trunc to accept negative scales (Tom)
1787 New TIME WITH TIME ZONE type (Thomas)
1788 Add MAX()/MIN() on time type (Thomas)
1789 Add abs(), mod(), fac() for int8 (Thomas)
1790 Rename functions to round(), sqrt(), cbrt(), pow() for float8 (Thomas)
1791 Add transcendental math functions (e.g. sin(), acos()) for float8 (Thomas)
1792 Add exp() and ln() for NUMERIC type
1793 Rename NUMERIC power() to pow() (Thomas)
1794 Improved TRANSLATE() function (Edwin Ramirez, Tom)
1795 Allow X=-Y operators (Tom)
1796 Allow SELECT float8(COUNT(*))/(SELECT COUNT(*) FROM t) FROM t GROUP BY f1; (Tom
1798 Allow LOCALE to use indexes in regular expression searches (Tom)
1799 Allow creation of functional indexes to use default types
1803 Prevent exponential space consumption with many AND's and OR's (Tom)
1804 Collect attribute selectivity values for system columns (Tom)
1805 Reduce memory usage of aggregates (Tom)
1806 Fix for LIKE optimization to use indexes with multibyte encodings (Tom)
1807 Fix r-tree index optimizer selectivity (Thomas)
1808 Improve optimizer selectivity computations and functions (Tom)
1809 Optimize btree searching for cases where many equal keys exist (Tom)
1810 Enable fast LIKE index processing only if index present (Tom)
1811 Re-use free space on index pages with duplicates (Tom)
1812 Improve hash join processing (Tom)
1813 Prevent descending sort if result is already sorted(Hiroshi)
1814 Allow commuting of index scan query qualifications (Tom)
1815 Prefer index scans in cases where ORDER BY/GROUP BY is required (Tom)
1816 Allocate large memory requests in fix-sized chunks for performance (Tom)
1817 Fix vacuum's performance by reducing memory allocation requests (Tom)
1818 Implement constant-expression simplification (Bernard Frankpitt, Tom)
1819 Use secondary columns to be used to determine start of index scan (Hiroshi)
1820 Prevent quadruple use of disk space when doing internal sorting (Tom)
1821 Faster sorting by calling fewer functions (Tom)
1822 Create system indexes to match all system caches (Bruce, Hiroshi)
1823 Make system caches use system indexes (Bruce)
1824 Make all system indexes unique (Bruce)
1825 Improve pg_statistics management for VACUUM speed improvement (Tom)
1826 Flush backend cache less frequently (Tom, Hiroshi)
1827 COPY now reuses previous memory allocation, improving performance (Tom)
1828 Improve optimization cost estimation (Tom)
1829 Improve optimizer estimate of range queries x > lowbound AND x < highbound (Tom
1831 Use DNF instead of CNF where appropriate (Tom, Taral)
1832 Further cleanup for OR-of-AND WHERE-clauses (Tom)
1833 Make use of index in OR clauses (x = 1 AND y = 2) OR (x = 2 AND y = 4) (Tom)
1834 Smarter optimizer computations for random index page access (Tom)
1835 New SET variable to control optimizer costs (Tom)
1836 Optimizer queries based on LIMIT, OFFSET, and EXISTS qualifications (Tom)
1837 Reduce optimizer internal housekeeping of join paths for speedup (Tom)
1838 Major subquery speedup (Tom)
1839 Fewer fsync writes when fsync is not disabled (Tom)
1840 Improved LIKE optimizer estimates (Tom)
1841 Prevent fsync in SELECT-only queries (Vadim)
1842 Make index creation use psort code, because it is now faster (Tom)
1843 Allow creation of sort temp tables > 1 Gig
1847 Fix for linux PPC compile
1848 New generic expression-tree-walker subroutine (Tom)
1849 Change form() to varargform() to prevent portability problems
1850 Improved range checking for large integers on Alphas
1851 Clean up #include in /include directory (Bruce)
1852 Add scripts for checking includes (Bruce)
1853 Remove un-needed #include's from *.c files (Bruce)
1854 Change #include's to use <> and "" as appropriate (Bruce)
1855 Enable WIN32 compilation of libpq
1856 Alpha spinlock fix from Uncle George <gatgul@voicenet.com>
1857 Overhaul of optimizer data structures (Tom)
1858 Fix to cygipc library (Yutaka Tanida)
1859 Allow pgsql to work on newer Cygwin snapshots (Dan)
1860 New catalog version number (Tom)
1862 Rename heap_replace to heap_update
1863 Update for QNX (Dr. Andreas Kardos)
1864 New platform-specific regression handling (Tom)
1865 Rename oid8 -> oidvector and int28 -> int2vector (Bruce)
1866 Included all yacc and lex files into the distribution (Peter E.)
1867 Remove lextest, no longer needed (Peter E)
1868 Fix for libpq and psql on Win32 (Magnus)
1869 Internally change datetime and timespan into timestamp and interval (Thomas)
1870 Fix for plpgsql on BSD/OS
1871 Add SQL_ASCII test case to the regression test (Tatsuo)
1872 configure --with-mb now deprecated (Tatsuo)
1874 NetBSD fixes (Johnny C. Lam <lamj@stat.cmu.edu>)
1875 Fixes for Alpha compiles
1876 New multibyte encodings
1878 _________________________________________________________________
1882 Release date: 1999-10-13
1884 This is basically a cleanup release for 6.5.2. We have added a new
1885 PgAccess that was missing in 6.5.2, and installed an NT-specific fix.
1886 _________________________________________________________________
1888 Migration to version 6.5.3
1890 A dump/restore is *not* required for those running 6.5.*.
1891 _________________________________________________________________
1895 Updated version of pgaccess 0.98
1897 Fix dumping rules on inherited tables
1899 _________________________________________________________________
1903 Release date: 1999-09-15
1905 This is basically a cleanup release for 6.5.1. We have fixed a variety
1906 of problems reported by 6.5.1 users.
1907 _________________________________________________________________
1909 Migration to version 6.5.2
1911 A dump/restore is *not* required for those running 6.5.*.
1912 _________________________________________________________________
1916 subselect+CASE fixes(Tom)
1917 Add SHLIB_LINK setting for solaris_i386 and solaris_sparc ports(Daren Sefcik)
1918 Fixes for CASE in WHERE join clauses(Tom)
1919 Fix BTScan abort(Tom)
1920 Repair the check for redundant UNIQUE and PRIMARY KEY indexes(Thomas)
1921 Improve it so that it checks for multicolumn constraints(Thomas)
1922 Fix for Win32 making problem with MB enabled(Hiroki Kataoka)
1923 Allow BSD yacc and bison to compile pl code(Bruce)
1924 Fix SET NAMES working
1926 Fix vacuum's memory consumption(Hiroshi,Tatsuo)
1927 Reduce the total memory consumption of vacuum(Tom)
1928 Fix for timestamp(datetime)
1929 Rule deparsing bugfixes(Tom)
1930 Fix quoting problems in mkMakefile.tcldefs.sh.in and mkMakefile.tkdefs.sh.in(To
1932 This is to re-use space on index pages freed by vacuum(Vadim)
1933 document -x for pg_dump(Bruce)
1934 Fix for unary operators in rule deparser(Tom)
1935 Comment out FileUnlink of excess segments during mdtruncate()(Tom)
1936 IRIX linking fix from Yu Cao >yucao@falcon.kla-tencor.com<
1937 Repair logic error in LIKE: should not return LIKE_ABORT
1938 when reach end of pattern before end of text(Tom)
1939 Repair incorrect cleanup of heap memory allocation during transaction abort(Tom
1941 Updated version of pgaccess 0.98
1943 _________________________________________________________________
1947 Release date: 1999-07-15
1949 This is basically a cleanup release for 6.5. We have fixed a variety
1950 of problems reported by 6.5 users.
1951 _________________________________________________________________
1953 Migration to version 6.5.1
1955 A dump/restore is *not* required for those running 6.5.
1956 _________________________________________________________________
1961 Portability fixes for linux_ppc, IRIX, linux_alpha, OpenBSD, alpha
1962 Remove QUERY_LIMIT, use SELECT...LIMIT
1963 Fix for EXPLAIN on inheritance(Tom)
1964 Patch to allow vacuum on multisegment tables(Hiroshi)
1965 R-Tree optimizer selectivity fix(Tom)
1966 ACL file descriptor leak fix(Atsushi Ogawa)
1967 New expresssion subtree code(Tom)
1968 Avoid disk writes for read-only transactions(Vadim)
1969 Fix for removal of temp tables if last transaction was aborted(Bruce)
1970 Fix to prevent too large tuple from being created(Bruce)
1972 Allow port numbers 32k - 64k(Bruce)
1973 Add ^ precidence(Bruce)
1974 Rename sort files called pg_temp to pg_sorttemp(Bruce)
1975 Fix for microseconds in time values(Tom)
1976 Tutorial source cleanup
1978 Fix for sorting of NULL's in some cases(Tom)
1979 Shared library dependencies fixed (Tom)
1980 Fixed glitches affecting GROUP BY in subselects(Tom)
1981 Fix some compiler warnings (Tomoaki Nishiyama)
1982 Add Win1250 (Czech) support (Pavel Behal)
1984 _________________________________________________________________
1988 Release date: 1999-06-09
1990 This release marks a major step in the development team's mastery of
1991 the source code we inherited from Berkeley. You will see we are now
1992 easily adding major features, thanks to the increasing size and
1993 experience of our world-wide development team.
1995 Here is a brief summary of the more notable changes:
1997 Multiversion concurrency control(MVCC)
1998 This removes our old table-level locking, and replaces it with
1999 a locking system that is superior to most commercial database
2000 systems. In a traditional system, each row that is modified is
2001 locked until committed, preventing reads by other users. MVCC
2002 uses the natural multiversion nature of PostgreSQL to allow
2003 readers to continue reading consistent data during writer
2004 activity. Writers continue to use the compact pg_log
2005 transaction system. This is all performed without having to
2006 allocate a lock for every row like traditional database
2007 systems. So, basically, we no longer are restricted by simple
2008 table-level locking; we have something better than row-level
2011 Hot backups from pg_dump
2012 pg_dump takes advantage of the new MVCC features to give a
2013 consistent database dump/backup while the database stays online
2014 and available for queries.
2017 We now have a true numeric data type, with user-specified
2021 Temporary tables are guaranteed to have unique names within a
2022 database session, and are destroyed on session exit.
2025 We now have CASE, INTERSECT, and EXCEPT statement support. We
2026 have new LIMIT/OFFSET, SET TRANSACTION ISOLATION LEVEL, SELECT
2027 ... FOR UPDATE, and an improved LOCK TABLE command.
2030 We continue to speed up PostgreSQL, thanks to the variety of
2031 talents within our team. We have sped up memory allocation,
2032 optimization, table joins, and row transfer routines.
2035 We continue to expand our port list, this time including
2036 Windows NT/ix86 and NetBSD/arm32.
2039 Most interfaces have new versions, and existing functionality
2043 New and updated material is present throughout the
2044 documentation. New FAQs have been contributed for SGI and AIX
2045 platforms. The Tutorial has introductory information on SQL
2046 from Stefan Simkovics. For the User's Guide, there are
2047 reference pages covering the postmaster and more utility
2048 programs, and a new appendix contains details on date/time
2049 behavior. The Administrator's Guide has a new chapter on
2050 troubleshooting from Tom Lane. And the Programmer's Guide has a
2051 description of query processing, also from Stefan, and details
2052 on obtaining the PostgreSQL source tree via anonymous CVS and
2054 _________________________________________________________________
2056 Migration to version 6.5
2058 A dump/restore using pg_dump is required for those wishing to migrate
2059 data from any previous release of PostgreSQL. pg_upgrade can *not* be
2060 used to upgrade to this release because the on-disk structure of the
2061 tables has changed compared to previous releases.
2063 The new Multiversion Concurrency Control (MVCC) features can give
2064 somewhat different behaviors in multiuser environments. *Read and
2065 understand the following section to ensure that your existing
2066 applications will give you the behavior you need.*
2067 _________________________________________________________________
2069 Multiversion Concurrency Control
2071 Because readers in 6.5 don't lock data, regardless of transaction
2072 isolation level, data read by one transaction can be overwritten by
2073 another. In other words, if a row is returned by "SELECT" it doesn't
2074 mean that this row really exists at the time it is returned (i.e.
2075 sometime after the statement or transaction began) nor that the row is
2076 protected from being deleted or updated by concurrent transactions
2077 before the current transaction does a commit or rollback.
2079 To ensure the actual existence of a row and protect it against
2080 concurrent updates one must use "SELECT FOR UPDATE" or an appropriate
2081 "LOCK TABLE" statement. This should be taken into account when porting
2082 applications from previous releases of PostgreSQL and other
2085 Keep the above in mind if you are using "contrib/refint.*" triggers
2086 for referential integrity. Additional techniques are required now. One
2087 way is to use "LOCK parent_table IN SHARE ROW EXCLUSIVE MODE" command
2088 if a transaction is going to update/delete a primary key and use "LOCK
2089 parent_table IN SHARE MODE" command if a transaction is going to
2090 update/insert a foreign key.
2092 Note: Note that if you run a transaction in SERIALIZABLE mode then
2093 you must execute the "LOCK" commands above before execution of any
2094 DML statement ("SELECT/INSERT/DELETE/UPDATE/FETCH/COPY_TO") in the
2097 These inconveniences will disappear in the future when the ability to
2098 read dirty (uncommitted) data (regardless of isolation level) and true
2099 referential integrity will be implemented.
2100 _________________________________________________________________
2106 Fix text<->float8 and text<->float4 conversion functions(Thomas)
2107 Fix for creating tables with mixed-case constraints(Billy)
2108 Change exp()/pow() behavior to generate error on underflow/overflow(Jan)
2109 Fix bug in pg_dump -z
2110 Memory overrun cleanups(Tatsuo)
2111 Fix for lo_import crash(Tatsuo)
2112 Adjust handling of data type names to suppress double quotes(Thomas)
2113 Use type coercion for matching columns and DEFAULT(Thomas)
2114 Fix deadlock so it only checks once after one second of sleep(Bruce)
2115 Fixes for aggregates and PL/pgsql(Hiroshi)
2116 Fix for subquery crash(Vadim)
2117 Fix for libpq function PQfnumber and case-insensitive names(Bahman Rafatjoo)
2118 Fix for large object write-in-middle, no extra block, memory consumption(Tatsuo
2120 Fix for pg_dump -d or -D and quote special characters in INSERT
2121 Repair serious problems with dynahash(Tom)
2122 Fix INET/CIDR portability problems
2123 Fix problem with selectivity error in ALTER TABLE ADD COLUMN(Bruce)
2124 Fix executor so mergejoin of different column types works(Tom)
2125 Fix for Alpha OR selectivity bug
2126 Fix OR index selectivity problem(Bruce)
2127 Fix so \d shows proper length for char()/varchar()(Ryan)
2128 Fix tutorial code(Clark)
2129 Improve destroyuser checking(Oliver)
2130 Fix for Kerberos(Rodney McDuff)
2131 Fix for dropping database while dirty buffers(Bruce)
2132 Fix so sequence nextval() can be case-sensitive(Bruce)
2134 Drop buffers before destroying database files(Bruce)
2135 Fix case where executor evaluates functions twice(Tatsuo)
2136 Allow sequence nextval actions to be case-sensitive(Bruce)
2137 Fix optimizer indexing not working for negative numbers(Bruce)
2138 Fix for memory leak in executor with fjIsNull
2139 Fix for aggregate memory leaks(Erik Riedel)
2140 Allow username containing a dash GRANT permissions
2141 Cleanup of NULL in inet types
2142 Clean up system table bugs(Tom)
2143 Fix problems of PAGER and \? command(Masaaki Sakaida)
2144 Reduce default multisegment file size limit to 1GB(Peter)
2145 Fix for dumping of CREATE OPERATOR(Tom)
2146 Fix for backward scanning of cursors(Hiroshi Inoue)
2147 Fix for COPY FROM STDIN when using \i(Tom)
2148 Fix for subselect is compared inside an expression(Jan)
2149 Fix handling of error reporting while returning rows(Tom)
2150 Fix problems with reference to array types(Tom,Jan)
2151 Prevent UPDATE SET oid(Jan)
2152 Fix pg_dump so -t option can handle case-sensitive tablenames
2153 Fixes for GROUP BY in special cases(Tom, Jan)
2154 Fix for memory leak in failed queries(Tom)
2155 DEFAULT now supports mixed-case identifiers(Tom)
2156 Fix for multisegment uses of DROP/RENAME table, indexes(Ole Gjerde)
2157 Disable use of pg_dump with both -o and -d options(Bruce)
2158 Allow pg_dump to properly dump GROUP permissions(Bruce)
2159 Fix GROUP BY in INSERT INTO table SELECT * FROM table2(Jan)
2160 Fix for computations in views(Jan)
2161 Fix for aggregates on array indexes(Tom)
2162 Fix for DEFAULT handles single quotes in value requiring too many quotes
2163 Fix security problem with non-super users importing/exporting large objects(Tom
2165 Rollback of transaction that creates table cleaned up properly(Tom)
2166 Fix to allow long table and column names to generate proper serial names(Tom)
2170 Add "vacuumdb" utility
2171 Speed up libpq by allocating memory better(Tom)
2172 EXPLAIN all indexes used(Tom)
2173 Implement CASE, COALESCE, NULLIF expression(Thomas)
2174 New pg_dump table output format(Constantin)
2175 Add string min()/max() functions(Thomas)
2176 Extend new type coercion techniques to aggregates(Thomas)
2177 New moddatetime contrib(Terry)
2178 Update to pgaccess 0.96(Constantin)
2179 Add routines for single-byte "char" type(Thomas)
2180 Improved substr() function(Thomas)
2181 Improved multibyte handling(Tatsuo)
2182 Multiversion concurrency control/MVCC(Vadim)
2183 New Serialized mode(Vadim)
2184 Fix for tables over 2gigs(Peter)
2185 New SET TRANSACTION ISOLATION LEVEL(Vadim)
2186 New LOCK TABLE IN ... MODE(Vadim)
2187 Update ODBC driver(Byron)
2188 New NUMERIC data type(Jan)
2189 New SELECT FOR UPDATE(Vadim)
2190 Handle "NaN" and "Infinity" for input values(Jan)
2191 Improved date/year handling(Thomas)
2192 Improved handling of backend connections(Magnus)
2193 New options ELOG_TIMESTAMPS and USE_SYSLOG options for log files(Massimo)
2194 New TCL_ARRAYS option(Massimo)
2195 New INTERSECT and EXCEPT(Stefan)
2196 New pg_index.indisprimary for primary key tracking(D'Arcy)
2197 New pg_dump option to allow dropping of tables before creation(Brook)
2198 Speedup of row output routines(Tom)
2199 New READ COMMITTED isolation level(Vadim)
2200 New TEMP tables/indexes(Bruce)
2201 Prevent sorting if result is already sorted(Jan)
2202 New memory allocation optimization(Jan)
2203 Allow psql to do \p\g(Bruce)
2204 Allow multiple rule actions(Jan)
2205 Added LIMIT/OFFSET functionality(Jan)
2206 Improve optimizer when joining a large number of tables(Bruce)
2207 New intro to SQL from S. Simkovics' Master's Thesis (Stefan, Thomas)
2208 New intro to backend processing from S. Simkovics' Master's Thesis (Stefan)
2209 Improved int8 support(Ryan Bradetich, Thomas, Tom)
2210 New routines to convert between int8 and text/varchar types(Thomas)
2211 New bushy plans, where meta-tables are joined(Bruce)
2212 Enable right-hand queries by default(Bruce)
2213 Allow reliable maximum number of backends to be set at configure time
2214 (--with-maxbackends and postmaster switch (-N backends))(Tom)
2215 GEQO default now 10 tables because of optimizer speedups(Tom)
2216 Allow NULL=Var for MS-SQL portability(Michael, Bruce)
2217 Modify contrib check_primary_key() so either "automatic" or "dependent"(Anand)
2218 Allow psql \d on a view show query(Ryan)
2219 Speedup for LIKE(Bruce)
2220 Ecpg fixes/features, see src/interfaces/ecpg/ChangeLog file(Michael)
2221 JDBC fixes/features, see src/interfaces/jdbc/CHANGELOG(Peter)
2222 Make % operator have precedence like /(Bruce)
2223 Add new postgres -O option to allow system table structure changes(Bruce)
2224 Update contrib/pginterface/findoidjoins script(Tom)
2225 Major speedup in vacuum of deleted rows with indexes(Vadim)
2226 Allow non-SQL functions to run different versions based on arguments(Tom)
2227 Add -E option that shows actual queries sent by \dt and friends(Masaaki Sakaida
2229 Add version number in start-up banners for psql(Masaaki Sakaida)
2230 New contrib/vacuumlo removes large objects not referenced(Peter)
2231 New initialization for table sizes so non-vacuumed tables perform better(Tom)
2232 Improve error messages when a connection is rejected(Tom)
2233 Support for arrays of char() and varchar() fields(Massimo)
2234 Overhaul of hash code to increase reliability and performance(Tom)
2235 Update to PyGreSQL 2.4(D'Arcy)
2236 Changed debug options so -d4 and -d5 produce different node displays(Jan)
2237 New pg_options: pretty_plan, pretty_parse, pretty_rewritten(Jan)
2238 Better optimization statistics for system table access(Tom)
2239 Better handling of non-default block sizes(Massimo)
2240 Improve GEQO optimizer memory consumption(Tom)
2241 UNION now suppports ORDER BY of columns not in target list(Jan)
2242 Major libpq++ improvements(Vince Vielhaber)
2243 pg_dump now uses -z(ACL's) as default(Bruce)
2244 backend cache, memory speedups(Tom)
2245 have pg_dump do everything in one snapshot transaction(Vadim)
2246 fix for large object memory leakage, fix for pg_dumping(Tom)
2247 INET type now respects netmask for comparisons
2248 Make VACUUM ANALYZE only use a readlock(Vadim)
2249 Allow VIEWs on UNIONS(Jan)
2250 pg_dump now can generate consistent snapshots on active databases(Vadim)
2254 Improve port matching(Tom)
2255 Portability fixes for SunOS
2256 Add NT/Win32 backend port and enable dynamic loading(Magnus and Daniel Horak)
2257 New port to Cobalt Qube(Mips) running Linux(Tatsuo)
2258 Port to NetBSD/m68k(Mr. Mutsuki Nakajima)
2259 Port to NetBSD/sun3(Mr. Mutsuki Nakajima)
2260 Port to NetBSD/macppc(Toshimi Aoki)
2261 Fix for tcl/tk configuration(Vince)
2262 Removed CURRENT keyword for rule queries(Jan)
2263 NT dynamic loading now works(Daniel Horak)
2264 Add ARM32 support(Andrew McMurry)
2265 Better support for HP-UX 11 and UnixWare
2266 Improve file handling to be more uniform, prevent file descriptor leak(Tom)
2267 New install commands for plpgsql(Jan)
2269 _________________________________________________________________
2273 Release date: 1998-12-20
2275 The 6.4.1 release was improperly packaged. This also has one
2277 _________________________________________________________________
2279 Migration to version 6.4.2
2281 A dump/restore is *not* required for those running 6.4.*.
2282 _________________________________________________________________
2286 Fix for datetime constant problem on some platforms(Thomas)
2287 _________________________________________________________________
2291 Release date: 1998-12-18
2293 This is basically a cleanup release for 6.4. We have fixed a variety
2294 of problems reported by 6.4 users.
2295 _________________________________________________________________
2297 Migration to version 6.4.1
2299 A dump/restore is *not* required for those running 6.4.
2300 _________________________________________________________________
2304 Add pg_dump -N flag to force double quotes around identifiers. This is
2306 Fix for NOT in where clause causing crash(Bruce)
2307 EXPLAIN VERBOSE coredump fix(Vadim)
2308 Fix shared-library problems on Linux
2309 Fix test for table existance to allow mixed-case and whitespace in
2310 the table name(Thomas)
2311 Fix a couple of pg_dump bugs
2312 Configure matches template/.similar entries better(Tom)
2313 Change builtin function names from SPI_* to spi_*
2314 OR WHERE clause fix(Vadim)
2315 Fixes for mixed-case table names(Billy)
2316 contrib/linux/postgres.init.csh/sh fix(Thomas)
2317 libpq memory overrun fix
2319 Change exp() behavior to generate error on underflow(Thomas)
2320 pg_dump fixes for memory leak, inheritance constraints, layout change
2321 update pgaccess to 0.93
2322 Fix prototype for 64-bit platforms
2323 Multibyte fixes(Tatsuo)
2325 Fix memory overruns(Tatsuo)
2326 Fix for lo_import() crash(Bruce)
2327 Better search for install program(Tom)
2330 Use implicit type coercion for matching DEFAULT values(Thomas)
2331 Add routines to help with single-byte (internal) character type(Thomas)
2332 Compilation of libpq for Win32 fixes(Magnus)
2333 Upgrade to PyGreSQL 2.2(D'Arcy)
2334 _________________________________________________________________
2338 Release date: 1998-10-30
2340 There are *many* new features and improvements in this release. Thanks
2341 to our developers and maintainers, nearly every aspect of the system
2342 has received some attention since the previous release. Here is a
2343 brief, incomplete summary:
2345 * Views and rules are now functional thanks to extensive new code in
2346 the rewrite rules system from Jan Wieck. He also wrote a chapter
2347 on it for the Programmer's Guide.
2348 * Jan also contributed a second procedural language, PL/pgSQL, to go
2349 with the original PL/pgTCL procedural language he contributed last
2351 * We have optional multiple-byte character set support from Tatsuo
2352 Ishii to complement our existing locale support.
2353 * Client/server communications has been cleaned up, with better
2354 support for asynchronous messages and interrupts thanks to Tom
2356 * The parser will now perform automatic type coercion to match
2357 arguments to available operators and functions, and to match
2358 columns and expressions with target columns. This uses a generic
2359 mechanism which supports the type extensibility features of
2360 PostgreSQL. There is a new chapter in the User's Guide which
2362 * Three new data types have been added. Two types, inet and cidr,
2363 support various forms of IP network, subnet, and machine
2364 addressing. There is now an 8-byte integer type available on some
2365 platforms. See the chapter on data types in the User's Guide for
2366 details. A fourth type, serial, is now supported by the parser as
2367 an amalgam of the int4 type, a sequence, and a unique index.
2368 * Several more SQL92-compatible syntax features have been added,
2369 including "INSERT DEFAULT VALUES"
2370 * The automatic configuration and installation system has received
2371 some attention, and should be more robust for more platforms than
2373 _________________________________________________________________
2375 Migration to version 6.4
2377 A dump/restore using pg_dump or pg_dumpall is required for those
2378 wishing to migrate data from any previous release of PostgreSQL.
2379 _________________________________________________________________
2385 Fix for a tiny memory leak in PQsetdb/PQfinish(Bryan)
2386 Remove char2-16 data types, use char/varchar(Darren)
2387 Pqfn not handles a NOTICE message(Anders)
2388 Reduced busywaiting overhead for spinlocks with many backends (dg)
2389 Stuck spinlock detection (dg)
2390 Fix up "ISO-style" timespan decoding and encoding(Thomas)
2391 Fix problem with table drop after rollback of transaction(Vadim)
2392 Change error message and remove non-functional update message(Vadim)
2393 Fix for COPY array checking
2394 Fix for SELECT 1 UNION SELECT NULL
2395 Fix for buffer leaks in large object calls(Pascal)
2396 Change owner from oid to int4 type(Bruce)
2397 Fix a bug in the oracle compatibility functions btrim() ltrim() and rtrim()
2398 Fix for shared invalidation cache overflow(Massimo)
2399 Prevent file descriptor leaks in failed COPY's(Bruce)
2400 Fix memory leak in libpgtcl's pg_select(Constantin)
2401 Fix problems with username/passwords over 8 characters(Tom)
2402 Fix problems with handling of asynchronous NOTIFY in backend(Tom)
2403 Fix of many bad system table entries(Tom)
2407 Upgrade ecpg and ecpglib,see src/interfaces/ecpc/ChangeLog(Michael)
2408 Show the index used in an EXPLAIN(Zeugswetter)
2409 EXPLAIN invokes rule system and shows plan(s) for rewritten queries(Jan)
2410 Multibyte awareness of many data types and functions, via configure(Tatsuo)
2411 New configure --with-mb option(Tatsuo)
2412 New initdb --pgencoding option(Tatsuo)
2413 New createdb -E multibyte option(Tatsuo)
2414 Select version(); now returns PostgreSQL version(Jeroen)
2415 Libpq now allows asynchronous clients(Tom)
2416 Allow cancel from client of backend query(Tom)
2417 Psql now cancels query with Control-C(Tom)
2418 Libpq users need not issue dummy queries to get NOTIFY messages(Tom)
2419 NOTIFY now sends sender's PID, so you can tell whether it was your own(Tom)
2420 PGresult struct now includes associated error message, if any(Tom)
2421 Define "tz_hour" and "tz_minute" arguments to date_part()(Thomas)
2422 Add routines to convert between varchar and bpchar(Thomas)
2423 Add routines to allow sizing of varchar and bpchar into target columns(Thomas)
2424 Add bit flags to support timezonehour and minute in data retrieval(Thomas)
2425 Allow more variations on valid floating point numbers (e.g. ".1", "1e6")(Thomas
2427 Fixes for unary minus parsing with leading spaces(Thomas)
2428 Implement TIMEZONE_HOUR, TIMEZONE_MINUTE per SQL92 specs(Thomas)
2429 Check for and properly ignore FOREIGN KEY column constraints(Thomas)
2430 Define USER as synonym for CURRENT_USER per SQL92 specs(Thomas)
2431 Enable HAVING clause but no fixes elsewhere yet.
2432 Make "char" type a synonym for "char(1)" (actually implemented as bpchar)(Thoma
2434 Save string type if specified for DEFAULT clause handling(Thomas)
2435 Coerce operations involving different data types(Thomas)
2436 Allow some index use for columns of different types(Thomas)
2437 Add capabilities for automatic type conversion(Thomas)
2438 Cleanups for large objects, so file is truncated on open(Peter)
2439 Readline cleanups(Tom)
2440 Allow psql \f \ to make spaces as delimiter(Bruce)
2441 Pass pg_attribute.atttypmod to the frontend for column field lengths(Tom,Bruce)
2442 Msql compatibility library in /contrib(Aldrin)
2443 Remove the requirement that ORDER/GROUP BY clause identifiers be
2444 included in the target list(David)
2445 Convert columns to match columns in UNION clauses(Thomas)
2446 Remove fork()/exec() and only do fork()(Bruce)
2447 Jdbc cleanups(Peter)
2448 Show backend status on ps command line(only works on some platforms)(Bruce)
2449 Pg_hba.conf now has a sameuser option in the database field
2450 Make lo_unlink take oid param, not int4
2451 New DISABLE_COMPLEX_MACRO for compilers that can't handle our macros(Bruce)
2452 Libpgtcl now handles NOTIFY as a Tcl event, need not send dummy queries(Tom)
2453 libpgtcl cleanups(Tom)
2454 Add -error option to libpgtcl's pg_result command(Tom)
2455 New locale patch, see docs/README/locale(Oleg)
2456 Fix for pg_dump so CONSTRAINT and CHECK syntax is correct(ccb)
2457 New contrib/lo code for large object orphan removal(Peter)
2458 New psql command "SET CLIENT_ENCODING TO 'encoding'" for multibytes
2459 feature, see /doc/README.mb(Tatsuo)
2460 /contrib/noupdate code to revoke update permission on a column
2461 Libpq can now be compiled on win32(Magnus)
2462 Add PQsetdbLogin() in libpq
2463 New 8-byte integer type, checked by configure for OS support(Thomas)
2464 Better support for quoted table/column names(Thomas)
2465 Surround table and column names with double-quotes in pg_dump(Thomas)
2466 PQreset() now works with passwords(Tom)
2467 Handle case of GROUP BY target list column number out of range(David)
2468 Allow UNION in subselects
2469 Add auto-size to screen to \d? commands(Bruce)
2470 Use UNION to show all \d? results in one query(Bruce)
2471 Add \d? field search feature(Bruce)
2472 Pg_dump issues fewer \connect requests(Tom)
2473 Make pg_dump -z flag work better, document it in manual page(Tom)
2474 Add HAVING clause with full support for subselects and unions(Stephan)
2475 Full text indexing routines in contrib/fulltextindex(Maarten)
2476 Transaction ids now stored in shared memory(Vadim)
2477 New PGCLIENTENCODING when issuing COPY command(Tatsuo)
2478 Support for SQL92 syntax "SET NAMES"(Tatsuo)
2479 Support for LATIN2-5(Tatsuo)
2480 Add UNICODE regression test case(Tatsuo)
2481 Lock manager cleanup, new locking modes for LLL(Vadim)
2482 Allow index use with OR clauses(Bruce)
2483 Allows "SELECT NULL ORDER BY 1;"
2484 Explain VERBOSE prints the plan, and now pretty-prints the plan to
2485 the postmaster log file(Bruce)
2486 Add indexes display to \d command(Bruce)
2487 Allow GROUP BY on functions(David)
2488 New pg_class.relkind for large objects(Bruce)
2489 New way to send libpq NOTICE messages to a different location(Tom)
2490 New \w write command to psql(Bruce)
2491 New /contrib/findoidjoins scans oid columns to find join relationships(Bruce)
2492 Allow binary-compatible indexes to be considered when checking for valid
2493 Indexes for restriction clauses containing a constant(Thomas)
2494 New ISBN/ISSN code in /contrib/isbn_issn
2495 Allow NOT LIKE, IN, NOT IN, BETWEEN, and NOT BETWEEN constraint(Thomas)
2496 New rewrite system fixes many problems with rules and views(Jan)
2497 * Rules on relations work
2498 * Event qualifications on insert/update/delete work
2499 * New OLD variable to reference CURRENT, CURRENT will be remove in futu
2501 * Update rules can reference NEW and OLD in rule qualifications/actions
2502 * Insert/update/delete rules on views work
2503 * Multiple rule actions are now supported, surrounded by parentheses
2504 * Regular users can create views/rules on tables they have RULE permits
2505 * Rules and views inherit the permissions on the creator
2506 * No rules at the column level
2507 * No UPDATE NEW/OLD rules
2508 * New pg_tables, pg_indexes, pg_rules and pg_views system views
2509 * Only a single action on SELECT rules
2510 * Total rewrite overhaul, perhaps for 6.5
2512 * handle aggregates on views
2513 * handle insert into select from view works
2514 System indexes are now multikey(Bruce)
2515 Oidint2, oidint4, and oidname types are removed(Bruce)
2516 Use system cache for more system table lookups(Bruce)
2517 New backend programming language PL/pgSQL in backend/pl(Jan)
2518 New SERIAL data type, auto-creates sequence/index(Thomas)
2519 Enable assert checking without a recompile(Massimo)
2520 User lock enhancements(Massimo)
2521 New setval() command to set sequence value(Massimo)
2522 Auto-remove unix socket file on start-up if no postmaster running(Massimo)
2523 Conditional trace package(Massimo)
2524 New UNLISTEN command(Massimo)
2525 Psql and libpq now compile under win32 using win32.mak(Magnus)
2526 Lo_read no longer stores trailing NULL(Bruce)
2527 Identifiers are now truncated to 31 characters internally(Bruce)
2528 Createuser options now availble on the command line
2529 Code for 64-bit integer supported added, configure tested, int8 type(Thomas)
2530 Prevent file descriptor leaf from failed COPY(Bruce)
2531 New pg_upgrade command(Bruce)
2532 Updated /contrib directories(Massimo)
2533 New CREATE TABLE DEFAULT VALUES statement available(Thomas)
2534 New INSERT INTO TABLE DEFAULT VALUES statement available(Thomas)
2535 New DECLARE and FETCH feature(Thomas)
2536 libpq's internal structures now not exported(Tom)
2537 Allow up to 8 key indexes(Bruce)
2538 Remove ARCHIVE keyword, that is no longer used(Thomas)
2539 pg_dump -n flag to supress quotes around indentifiers
2540 disable system columns for views(Jan)
2541 new INET and CIDR types for network addresses(TomH, Paul)
2542 no more double quotes in psql output
2543 pg_dump now dumps views(Terry)
2544 new SET QUERY_LIMIT(Tatsuo,Jan)
2548 /contrib cleanup(Jun)
2549 Inline some small functions called for every row(Bruce)
2552 Multibyte regression tests(Soonmyung.)
2553 Remove --disabled options from configure
2554 Define PGDOC to use POSTGRESDIR by default
2555 Make regression optional
2556 Remove extra braces code to pgindent(Bruce)
2557 Add bsdi shared library support(Bruce)
2558 New --without-CXX support configure option(Brook)
2560 Update backend flowchart in tools/backend(Bruce)
2561 Change atttypmod from int16 to int32(Bruce, Tom)
2562 Getrusage() fix for platforms that do not have it(Tom)
2563 Add PQconnectdb, PGUSER, PGPASSWORD to libpq man page
2564 NS32K platform fixes(Phil Nelson, John Buller)
2565 SCO 7/UnixWare 2.x fixes(Billy,others)
2566 Sparc/Solaris 2.5 fixes(Ryan)
2567 Pgbuiltin.3 is obsolete, move to doc files(Thomas)
2568 Even more documention(Thomas)
2569 Nextstep support(Jacek)
2571 pginterface manual page(Bruce)
2572 shared libraries all have version numbers
2573 merged all OS-specific shared library defines into one file
2574 smarter TCL/TK configuration checking(Billy)
2575 smarter perl configuration(Brook)
2576 configure uses supplied install-sh if no install script found(Tom)
2577 new Makefile.shlib for shared library configuration(Tom)
2578 _________________________________________________________________
2582 Release date: 1998-04-07
2584 This is a bug-fix release for 6.3.x. Refer to the release notes for
2585 version 6.3 for a more complete summary of new features.
2589 * Repairs automatic configuration support for some platforms,
2590 including Linux, from breakage inadvertently introduced in version
2592 * Correctly handles function calls on the left side of BETWEEN and
2595 A dump/restore is NOT required for those running 6.3 or 6.3.1. A make
2596 distclean, make, and make install is all that is required. This last
2597 step should be performed while the postmaster is not running. You
2598 should re-link any custom applications that use PostgreSQL libraries.
2600 For upgrades from pre-6.3 installations, refer to the installation and
2601 migration instructions for version 6.3.
2602 _________________________________________________________________
2606 Configure detection improvements for tcl/tk(Brook Milligan, Alvin)
2607 Manual page improvements(Bruce)
2608 BETWEEN and LIKE fix(Thomas)
2609 fix for psql \connect used by pg_dump(Oliver Elphick)
2611 pgaccess, version 0.86
2612 qsort removed, now uses libc version, cleanups(Jeroen)
2613 fix for buffer over-runs detected(Maurice Gittens)
2614 fix for buffer overrun in libpgtcl(Randy Kunkee)
2615 fix for UNION with DISTINCT or ORDER BY(Bruce)
2616 gettimeofday configure check(Doug Winterburn)
2617 Fix "indexes not used" bug(Vadim)
2618 docs additions(Thomas)
2619 Fix for backend memory leak(Bruce)
2620 libreadline cleanup(Erwan MAS)
2621 Remove DISTDIR(Bruce)
2622 Makefile dependency cleanup(Jeroen van Vianen)
2625 _________________________________________________________________
2629 Release date: 1998-03-23
2633 * Additional support for multibyte character sets.
2634 * Repair byte ordering for mixed-endian clients and servers.
2635 * Minor updates to allowed SQL syntax.
2636 * Improvements to the configuration autodetection for installation.
2638 A dump/restore is NOT required for those running 6.3. A make
2639 distclean, make, and make install is all that is required. This last
2640 step should be performed while the postmaster is not running. You
2641 should re-link any custom applications that use PostgreSQL libraries.
2643 For upgrades from pre-6.3 installations, refer to the installation and
2644 migration instructions for version 6.3.
2645 _________________________________________________________________
2649 ecpg cleanup/fixes, now version 1.1(Michael Meskes)
2650 pg_user cleanup(Bruce)
2651 large object fix for pg_dump and tclsh (alvin)
2652 LIKE fix for multiple adjacent underscores
2653 fix for redefining builtin functions(Thomas)
2655 upgrade to pg_access 0.83
2656 updated CLUSTER manual page
2657 multibyte character set support, see doc/README.mb(Tatsuo)
2658 configure --with-pgport fix
2660 big-endian fix for backend communications(Kataoka)
2661 SUBSTR() and substring() fix(Jan)
2662 several jdbc fixes(Peter)
2663 libpgtcl improvements, see libptcl/README(Randy Kunkee)
2664 Fix for "Datasize = 0" error(Vadim)
2665 Prevent \do from wrapping(Bruce)
2666 Remove duplicate Russian character set entries
2668 Allow optional TABLE keyword in LOCK and SELECT INTO(Thomas)
2669 CREATE SEQUENCE options to allow a negative integer(Thomas)
2670 Add "PASSWORD" as an allowed column identifier(Thomas)
2671 Add checks for UNION target fields(Bruce)
2672 Fix Alpha port(Dwayne Bailey)
2673 Fix for text arrays containing quotes(Doug Gibson)
2674 Solaris compile fix(Albert Chin-A-Young)
2675 Better identify tcl and tk libs and includes(Bruce)
2677 _________________________________________________________________
2681 Release date: 1998-03-01
2683 There are *many* new features and improvements in this release. Here
2684 is a brief, incomplete summary:
2686 * Many new SQL features, including full SQL92 subselect capability
2687 (everything is here but target-list subselects).
2688 * Support for client-side environment variables to specify time zone
2690 * Socket interface for client/server connection. This is the default
2691 now so you may need to start postmaster with the "-i" flag.
2692 * Better password authorization mechanisms. Default table
2693 permissions have changed.
2694 * Old-style time travel has been removed. Performance has been
2697 Note: Bruce Momjian wrote the following notes to introduce the new
2700 There are some general 6.3 issues that I want to mention. These are
2701 only the big items that can not be described in one sentence. A review
2702 of the detailed changes list is still needed.
2704 First, we now have subselects. Now that we have them, I would like to
2705 mention that without subselects, SQL is a very limited language.
2706 Subselects are a major feature, and you should review your code for
2707 places where subselects provide a better solution for your queries. I
2708 think you will find that there are more uses for subselects than you
2709 may think. Vadim has put us on the big SQL map with subselects, and
2710 fully functional ones too. The only thing you can't do with subselects
2711 is to use them in the target list.
2713 Second, 6.3 uses Unix domain sockets rather than TCP/IP by default. To
2714 enable connections from other machines, you have to use the new
2715 postmaster -i option, and of course edit "pg_hba.conf". Also, for this
2716 reason, the format of "pg_hba.conf" has changed.
2718 Third, char() fields will now allow faster access than varchar() or
2719 text. Specifically, the text and varchar() have a penalty for access
2720 to any columns after the first column of this type. char() used to
2721 also have this access penalty, but it no longer does. This may suggest
2722 that you redesign some of your tables, especially if you have short
2723 character columns that you have defined as varchar() or text. This and
2724 other changes make 6.3 even faster than earlier releases.
2726 We now have passwords definable independent of any Unix file. There
2727 are new SQL USER commands. See the Administrator's Guide for more
2728 information. There is a new table, pg_shadow, which is used to store
2729 user information and user passwords, and it by default only
2730 SELECT-able by the postgres super-user. pg_user is now a view of
2731 pg_shadow, and is SELECT-able by PUBLIC. You should keep using pg_user
2732 in your application without changes.
2734 User-created tables now no longer have SELECT permission to PUBLIC by
2735 default. This was done because the ANSI standard requires it. You can
2736 of course GRANT any permissions you want after the table is created.
2737 System tables continue to be SELECT-able by PUBLIC.
2739 We also have real deadlock detection code. No more sixty-second
2740 timeouts. And the new locking code implements a FIFO better, so there
2741 should be less resource starvation during heavy use.
2743 Many complaints have been made about inadequate documentation in
2744 previous releases. Thomas has put much effort into many new manuals
2745 for this release. Check out the doc/ directory.
2747 For performance reasons, time travel is gone, but can be implemented
2748 using triggers (see "pgsql/contrib/spi/README"). Please check out the
2749 new \d command for types, operators, etc. Also, views have their own
2750 permissions now, not based on the underlying tables, so permissions on
2751 them have to be set separately. Check "/pgsql/interfaces" for some new
2752 ways to talk to PostgreSQL.
2754 This is the first release that really required an explanation for
2755 existing users. In many ways, this was necessary because the new
2756 release removes many limitations, and the work-arounds people were
2757 using are no longer needed.
2758 _________________________________________________________________
2760 Migration to version 6.3
2762 A dump/restore using pg_dump or pg_dumpall is required for those
2763 wishing to migrate data from any previous release of PostgreSQL.
2764 _________________________________________________________________
2770 Fix binary cursors broken by MOVE implementation(Vadim)
2771 Fix for tcl library crash(Jan)
2772 Fix for array handling, from Gerhard Hintermayer
2773 Fix acl error, and remove duplicate pqtrace(Bruce)
2774 Fix psql \e for empty file(Bruce)
2775 Fix for textcat on varchar() fields(Bruce)
2776 Fix for DBT Sendproc (Zeugswetter Andres)
2777 Fix vacuum analyze syntax problem(Bruce)
2778 Fix for international identifiers(Tatsuo)
2779 Fix aggregates on inherited tables(Bruce)
2780 Fix substr() for out-of-bounds data
2781 Fix for select 1=1 or 2=2, select 1=1 and 2=2, and select sum(2+2)(Bruce)
2782 Fix notty output to show status result. -q option still turns it off(Bruce)
2783 Fix for count(*), aggs with views and multiple tables and sum(3)(Bruce)
2785 Fix for PQtrace start/stop several times(Bruce)
2786 Fix a variety of locking problems like newer lock waiters getting
2787 lock before older waiters, and having readlock people not share
2788 locks if a writer is waiting for a lock, and waiting writers not
2789 getting priority over waiting readers(Bruce)
2790 Fix crashes in psql when executing queries from external files(James)
2791 Fix problem with multiple order by columns, with the first one having
2793 Use correct hash table support functions for float8 and int4(Thomas)
2794 Re-enable JOIN= option in CREATE OPERATOR statement (Thomas)
2795 Change precedence for boolean operators to match expected behavior(Thomas)
2796 Generate elog(ERROR) on over-large integer(Bruce)
2797 Allow multiple-argument functions in constraint clauses(Thomas)
2798 Check boolean input literals for 'true','false','yes','no','1','0'
2799 and throw elog(ERROR) if unrecognized(Thomas)
2800 Major large objects fix
2801 Fix for GROUP BY showing duplicates(Vadim)
2802 Fix for index scans in MergeJion(Vadim)
2806 Subselects with EXISTS, IN, ALL, ANY keywords (Vadim, Bruce, Thomas)
2807 New User Manual(Thomas, others)
2808 Speedup by inlining some frequently-called functions
2809 Real deadlock detection, no more timeouts(Bruce)
2810 Add SQL92 "constants" CURRENT_DATE, CURRENT_TIME, CURRENT_TIMESTAMP,
2811 CURRENT_USER(Thomas)
2812 Modify constraint syntax to be SQL92-compliant(Thomas)
2813 Implement SQL92 PRIMARY KEY and UNIQUE clauses using indexes(Thomas)
2814 Recognize SQL92 syntax for FOREIGN KEY. Throw elog notice(Thomas)
2815 Allow NOT NULL UNIQUE constraint clause (each allowed separately before)(Thomas
2817 Allow PostgreSQL-style casting ("::") of non-constants(Thomas)
2818 Add support for SQL3 TRUE and FALSE boolean constants(Thomas)
2819 Support SQL92 syntax for IS TRUE/IS FALSE/IS NOT TRUE/IS NOT FALSE(Thomas)
2820 Allow shorter strings for boolean literals (e.g. "t", "tr", "tru")(Thomas)
2821 Allow SQL92 delimited identifiers(Thomas)
2822 Implement SQL92 binary and hexadecimal string decoding (b'10' and x'1F')(Thomas
2824 Support SQL92 syntax for type coercion of literal strings
2825 (e.g. "DATETIME 'now'")(Thomas)
2826 Add conversions for int2, int4, and OID types to and from text(Thomas)
2827 Use shared lock when building indexes(Vadim)
2828 Free memory allocated for an user query inside transaction block after
2829 this query is done, was turned off in <= 6.2.1(Vadim)
2830 New SQL statement CREATE PROCEDURAL LANGUAGE(Jan)
2831 New PostgreSQL Procedural Language (PL) backend interface(Jan)
2832 Rename pg_dump -H option to -h(Bruce)
2833 Add Java support for passwords, European dates(Peter)
2834 Use indexes for LIKE and ~, !~ operations(Bruce)
2835 Add hash functions for datetime and timespan(Thomas)
2836 Time Travel removed(Vadim, Bruce)
2837 Add paging for \d and \z, and fix \i(Bruce)
2838 Add Unix domain socket support to backend and to frontend library(Goran)
2839 Implement CREATE DATABASE/WITH LOCATION and initlocation utility(Thomas)
2840 Allow more SQL92 and/or PostgreSQL reserved words as column identifiers(Thomas)
2841 Augment support for SQL92 SET TIME ZONE...(Thomas)
2842 SET/SHOW/RESET TIME ZONE uses TZ backend environment variable(Thomas)
2843 Implement SET keyword = DEFAULT and SET TIME ZONE DEFAULT(Thomas)
2844 Enable SET TIME ZONE using TZ environment variable(Thomas)
2845 Add PGDATESTYLE environment variable to frontend and backend initialization(Tho
2847 Add PGTZ, PGCOSTHEAP, PGCOSTINDEX, PGRPLANS, PGGEQO
2848 frontend library initialization environment variables(Thomas)
2849 Regression tests time zone automatically set with "setenv PGTZ PST8PDT"(Thomas)
2850 Add pg_description table for info on tables, columns, operators, types, and
2852 Increase 16 char limit on system table/index names to 32 characters(Bruce)
2853 Rename system indexes(Bruce)
2854 Add 'GERMAN' option to SET DATESTYLE(Thomas)
2855 Define an "ISO-style" timespan output format with "hh:mm:ss" fields(Thomas)
2856 Allow fractional values for delta times (e.g. '2.5 days')(Thomas)
2857 Validate numeric input more carefully for delta times(Thomas)
2858 Implement day of year as possible input to date_part()(Thomas)
2859 Define timespan_finite() and text_timespan() functions(Thomas)
2860 Remove archive stuff(Bruce)
2861 Allow for a pg_password authentication database that is separate from
2862 the system password file(Todd)
2863 Dump ACLs, GRANT, REVOKE permissions(Matt)
2864 Define text, varchar, and bpchar string length functions(Thomas)
2865 Fix Query handling for inheritance, and cost computations(Bruce)
2866 Implement CREATE TABLE/AS SELECT (alternative to SELECT/INTO)(Thomas)
2867 Allow NOT, IS NULL, IS NOT NULL in constraints(Thomas)
2868 Implement UNIONs for SELECT(Bruce)
2869 Add UNION, GROUP, DISTINCT to INSERT(Bruce)
2870 varchar() stores only necessary bytes on disk(Bruce)
2871 Fix for BLOBs(Peter)
2872 Mega-Patch for JDBC...see README_6.3 for list of changes(Peter)
2873 Remove unused "option" from PQconnectdb()
2874 New LOCK command and lock manual page describing deadlocks(Bruce)
2875 Add new psql \da, \dd, \df, \do, \dS, and \dT commands(Bruce)
2876 Enhance psql \z to show sequences(Bruce)
2877 Show NOT NULL and DEFAULT in psql \d table(Bruce)
2878 New psql .psqlrc file start-up(Andrew)
2879 Modify sample start-up script in contrib/linux to show syslog(Thomas)
2880 New types for IP and MAC addresses in contrib/ip_and_mac(TomH)
2881 Unix system time conversions with date/time types in contrib/unixdate(Thomas)
2882 Update of contrib stuff(Massimo)
2883 Add Unix socket support to DBD::Pg(Goran)
2884 New python interface (PyGreSQL 2.0)(D'Arcy)
2885 New frontend/backend protocol has a version number, network byte order(Phil)
2886 Security features in pg_hba.conf enhanced and documented, many cleanups(Phil)
2887 CHAR() now faster access than VARCHAR() or TEXT
2888 ecpg embedded SQL preprocessor
2889 Reduce system column overhead(Vadmin)
2890 Remove pg_time table(Vadim)
2891 Add pg_type attribute to identify types that need length (bpchar, varchar)
2892 Add report of offending line when COPY command fails
2893 Allow VIEW permissions to be set separately from the underlying tables.
2894 For security, use GRANT/REVOKE on views as appropriate(Jan)
2895 Tables now have no default GRANT SELECT TO PUBLIC. You must
2896 explicitly grant such permissions.
2897 Clean up tutorial examples(Darren)
2901 Add new html development tools, and flow chart in /tools/backend
2902 Fix for SCO compiles
2903 Stratus computer port Robert Gillies
2904 Added support for shlib for BSD44_derived & i386_solaris
2905 Make configure more automated(Brook)
2906 Add script to check regression test results
2907 Break parser functions into smaller files, group together(Bruce)
2908 Rename heap_create to heap_create_and_catalog, rename heap_creatr
2909 to heap_create()(Bruce)
2910 Sparc/Linux patch for locking(TomS)
2911 Remove PORTNAME and reorganize port-specific stuff(Marc)
2912 Add optimizer README file(Bruce)
2913 Remove some recursion in optimizer and clean up some code there(Bruce)
2914 Fix for NetBSD locking(Henry)
2915 Fix for libptcl make(Tatsuo)
2917 Change IS TRUE, IS FALSE, ... to expressions using "=" rather than
2918 function calls to istrue() or isfalse() to allow optimization(Thomas)
2919 Various fixes NetBSD/Sparc related(TomH)
2920 Alpha linux locking(Travis,Ryan)
2921 Change elog(WARN) to elog(ERROR)(Bruce)
2922 FAQ for FreeBSD(Marc)
2923 Bring in the PostODBC source tree as part of our standard distribution(Marc)
2924 A minor patch for HP/UX 10 vs 9(Stan)
2925 New pg_attribute.atttypmod for type-specific info like varchar length(Bruce)
2926 UnixWare patches(Billy)
2927 New i386 'lock' for spin lock asm(Billy)
2928 Support for multiplexed backends is removed
2929 Start an OpenBSD port
2932 Add string functions to regression suite(Thomas)
2933 Expand a few function names formerly truncated to 16 characters(Thomas)
2934 Remove un-needed malloc() calls and replace with palloc()(Bruce)
2935 _________________________________________________________________
2939 Release date: 1997-10-17
2941 6.2.1 is a bug-fix and usability release on 6.2.
2945 * Allow strings to span lines, per SQL92.
2946 * Include example trigger function for inserting user names on table
2949 This is a minor bug-fix release on 6.2. For upgrades from pre-6.2
2950 systems, a full dump/reload is required. Refer to the 6.2 release
2951 notes for instructions.
2952 _________________________________________________________________
2954 Migration from version 6.2 to version 6.2.1
2956 This is a minor bug-fix release. A dump/reload is not required from
2957 version 6.2, but is required from any release prior to 6.2.
2959 In upgrading from version 6.2, if you choose to dump/reload you will
2960 find that avg(money) is now calculated correctly. All other bug fixes
2961 take effect upon updating the executables.
2963 Another way to avoid dump/reload is to use the following SQL command
2964 from "psql" to update the existing system table:
2965 update pg_aggregate set aggfinalfn = 'cash_div_flt8'
2966 where aggname = 'avg' and aggbasetype = 790;
2968 This will need to be done to every existing database, including
2970 _________________________________________________________________
2974 Allow TIME and TYPE column names(Thomas)
2975 Allow larger range of true/false as boolean values(Thomas)
2976 Support output of "now" and "current"(Thomas)
2977 Handle DEFAULT with INSERT of NULL properly(Vadim)
2978 Fix for relation reference counts problem in buffer manager(Vadim)
2979 Allow strings to span lines, like ANSI(Thomas)
2980 Fix for backward cursor with ORDER BY(Vadim)
2981 Fix avg(cash) computation(Thomas)
2982 Fix for specifying a column twice in ORDER/GROUP BY(Vadim)
2983 Documented new libpq function to return affected rows, PQcmdTuples(Bruce)
2984 Trigger function for inserting user names for INSERT/UPDATE(Brook Milligan)
2986 _________________________________________________________________
2990 Release date: 1997-10-02
2992 A dump/restore is required for those wishing to migrate data from
2993 previous releases of PostgreSQL.
2994 _________________________________________________________________
2996 Migration from version 6.1 to version 6.2
2998 This migration requires a complete dump of the 6.1 database and a
2999 restore of the database in 6.2.
3001 Note that the "pg_dump" and "pg_dumpall" utility from 6.2 should be
3002 used to dump the 6.1 database.
3003 _________________________________________________________________
3005 Migration from version 1.x to version 6.2
3007 Those migrating from earlier 1.* releases should first upgrade to 1.09
3008 because the COPY output format was improved from the 1.02 release.
3009 _________________________________________________________________
3015 Fix problems with pg_dump for inheritance, sequences, archive tables(Bruce)
3016 Fix compile errors on overflow due to shifts, unsigned, and bad prototypes
3017 from Solaris(Diab Jerius)
3018 Fix bugs in geometric line arithmetic (bad intersection calculations)(Thomas)
3019 Check for geometric intersections at endpoints to avoid rounding ugliness(Thoma
3021 Catch non-functional delete attempts(Vadim)
3022 Change time function names to be more consistent(Michael Reifenberg)
3023 Check for zero divides(Michael Reifenberg)
3024 Fix very old bug which made tuples changed/inserted by a commnd
3025 visible to the command itself (so we had multiple update of
3026 updated tuples, etc)(Vadim)
3027 Fix for SELECT null, 'fail' FROM pg_am (Patrick)
3028 SELECT NULL as EMPTY_FIELD now allowed(Patrick)
3029 Remove un-needed signal stuff from contrib/pginterface
3030 Fix OR (where x != 1 or x isnull didn't return tuples with x NULL) (Vadim)
3031 Fix time_cmp function (Vadim)
3032 Fix handling of functions with non-attribute first argument in
3033 WHERE clauses (Vadim)
3034 Fix GROUP BY when order of entries is different from order
3035 in target list (Vadim)
3036 Fix pg_dump for aggregates without sfunc1 (Vadim)
3040 Default genetic optimizer GEQO parameter is now 8(Bruce)
3041 Allow use parameters in target list having aggregates in functions(Vadim)
3042 Added JDBC driver as an interface(Adrian & Peter)
3044 Return number of tuples inserted/affected by INSERT/UPDATE/DELETE etc.(Vadim)
3045 Triggers implemented with CREATE TRIGGER (SQL3)(Vadim)
3046 SPI (Server Programming Interface) allows execution of queries inside
3048 NOT NULL implemented (SQL92)(Robson Paniago de Miranda)
3049 Include reserved words for string handling, outer joins, and unions(Thomas)
3050 Implement extended comments ("/* ... */") using exclusive states(Thomas)
3051 Add "//" single-line comments(Bruce)
3052 Remove some restrictions on characters in operator names(Thomas)
3053 DEFAULT and CONSTRAINT for tables implemented (SQL92)(Vadim & Thomas)
3054 Add text concatenation operator and function (SQL92)(Thomas)
3055 Support WITH TIME ZONE syntax (SQL92)(Thomas)
3056 Support INTERVAL unit TO unit syntax (SQL92)(Thomas)
3057 Define types DOUBLE PRECISION, INTERVAL, CHARACTER,
3058 and CHARACTER VARYING (SQL92)(Thomas)
3059 Define type FLOAT(p) and rudimentary DECIMAL(p,s), NUMERIC(p,s) (SQL92)(Thomas)
3060 Define EXTRACT(), POSITION(), SUBSTRING(), and TRIM() (SQL92)(Thomas)
3061 Define CURRENT_DATE, CURRENT_TIME, CURRENT_TIMESTAMP (SQL92)(Thomas)
3062 Add syntax and warnings for UNION, HAVING, INNER and OUTER JOIN (SQL92)(Thomas)
3063 Add more reserved words, mostly for SQL92 compliance(Thomas)
3064 Allow hh:mm:ss time entry for timespan/reltime types(Thomas)
3065 Add center() routines for lseg, path, polygon(Thomas)
3066 Add distance() routines for circle-polygon, polygon-polygon(Thomas)
3067 Check explicitly for points and polygons contained within polygons
3068 using an axis-crossing algorithm(Thomas)
3069 Add routine to convert circle-box(Thomas)
3070 Merge conflicting operators for different geometric data types(Thomas)
3071 Replace distance operator "<===>" with "<->"(Thomas)
3072 Replace "above" operator "!^" with ">^" and "below" operator "!|" with "<^"(Tho
3074 Add routines for text trimming on both ends, substring, and string position(Tho
3076 Added conversion routines circle(box) and poly(circle)(Thomas)
3077 Allow internal sorts to be stored in memory rather than in files(Bruce & Vadim)
3078 Allow functions and operators on internally-identical types to succeed(Bruce)
3079 Speed up backend start-up after profiling analysis(Bruce)
3080 Inline frequently called functions for performance(Bruce)
3081 Reduce open() calls(Bruce)
3082 psql: Add PAGER for \h and \?,\C fix
3083 Fix for psql pager when no tty(Bruce)
3084 New entab utility(Bruce)
3085 General trigger functions for referential integrity (Vadim)
3086 General trigger functions for time travel (Vadim)
3087 General trigger functions for AUTOINCREMENT/IDENTITY feature (Vadim)
3088 MOVE implementation (Vadim)
3092 HP-UX 10 patches (Vladimir Turin)
3093 Added SCO support, (Daniel Harris)
3094 MkLinux patches (Tatsuo Ishii)
3095 Change geometric box terminology from "length" to "width"(Thomas)
3096 Deprecate temporary unstored slope fields in geometric code(Thomas)
3097 Remove restart instructions from INSTALL(Bruce)
3098 Look in /usr/ucb first for install(Bruce)
3099 Fix c++ copy example code(Thomas)
3100 Add -o to psql manual page(Bruce)
3101 Prevent relname unallocated string length from being copied into database(Bruce
3103 Cleanup for NAMEDATALEN use(Bruce)
3104 Fix pg_proc names over 15 chars in output(Bruce)
3105 Add strNcpy() function(Bruce)
3106 remove some (void) casts that are unnecessary(Bruce)
3107 new interfaces directory(Marc)
3108 Replace fopen() calls with calls to fd.c functions(Bruce)
3109 Make functions static where possible(Bruce)
3110 enclose unused functions in #ifdef NOT_USED(Bruce)
3111 Remove call to difftime() in timestamp support to fix SunOS(Bruce & Thomas)
3112 Changes for Digital Unix
3113 Portability fix for pg_dumpall(Bruce)
3114 Rename pg_attribute.attnvals to attdispersion(Bruce)
3115 "intro/unix" manual page now "pgintro"(Bruce)
3116 "built-in" manual page now "pgbuiltin"(Bruce)
3117 "drop" manual page now "drop_table"(Bruce)
3118 Add "create_trigger", "drop_trigger" manual pages(Thomas)
3119 Add constraints regression test(Vadim & Thomas)
3120 Add comments syntax regression test(Thomas)
3121 Add PGINDENT and support program(Bruce)
3122 Massive commit to run PGINDENT on all *.c and *.h files(Bruce)
3123 Files moved to /src/tools directory(Bruce)
3124 SPI and Trigger programming guides (Vadim & D'Arcy)
3125 _________________________________________________________________
3129 Release date: 1997-07-22
3130 _________________________________________________________________
3132 Migration from version 6.1 to version 6.1.1
3134 This is a minor bug-fix release. A dump/reload is not required from
3135 version 6.1, but is required from any release prior to 6.1. Refer to
3136 the release notes for 6.1 for more details.
3137 _________________________________________________________________
3141 fix for SET with options (Thomas)
3142 allow pg_dump/pg_dumpall to preserve ownership of all tables/objects(Bruce)
3143 new psql \connect option allows changing usernames without changing databases
3144 fix for initdb --debug option(Yoshihiko Ichikawa))
3145 lextest cleanup(Bruce)
3147 fix date/time month boundary arithmetic(Thomas)
3148 fix timezone daylight handling for some ports(Thomas, Bruce, Tatsuo)
3149 timestamp overhauled to use standard functions(Thomas)
3150 other code cleanup in date/time routines(Thomas)
3151 psql's \d now case-insensitive(Bruce)
3152 psql's backslash commands can now have trailing semicolon(Bruce)
3153 fix memory leak in psql when using \g(Bruce)
3154 major fix for endian handling of communication to server(Thomas, Tatsuo)
3155 Fix for Solaris assembler and include files(Yoshihiko Ichikawa)
3156 allow underscores in usernames(Bruce)
3157 pg_dumpall now returns proper status, portability fix(Bruce)
3159 _________________________________________________________________
3163 Release date: 1997-06-08
3165 The regression tests have been adapted and extensively modified for
3166 the 6.1 release of PostgreSQL.
3168 Three new data types (datetime, timespan, and circle) have been added
3169 to the native set of PostgreSQL types. Points, boxes, paths, and
3170 polygons have had their output formats made consistent across the data
3171 types. The polygon output in misc.out has only been spot-checked for
3172 correctness relative to the original regression output.
3174 PostgreSQL 6.1 introduces a new, alternate optimizer which uses
3175 genetic algorithms. These algorithms introduce a random behavior in
3176 the ordering of query results when the query contains multiple
3177 qualifiers or multiple tables (giving the optimizer a choice on order
3178 of evaluation). Several regression tests have been modified to
3179 explicitly order the results, and hence are insensitive to optimizer
3180 choices. A few regression tests are for data types which are
3181 inherently unordered (e.g. points and time intervals) and tests
3182 involving those types are explicitly bracketed with "set geqo to
3183 'off'" and "reset geqo".
3185 The interpretation of array specifiers (the curly braces around atomic
3186 values) appears to have changed sometime after the original regression
3187 tests were generated. The current "./expected/*.out" files reflect
3188 this new interpretation, which may not be correct!
3190 The float8 regression test fails on at least some platforms. This is
3191 due to differences in implementations of pow() and exp() and the
3192 signaling mechanisms used for overflow and underflow conditions.
3194 The "random" results in the random test should cause the "random" test
3195 to be "failed", since the regression tests are evaluated using a
3196 simple diff. However, "random" does not seem to produce random results
3197 on my test machine (Linux/gcc/i686).
3198 _________________________________________________________________
3200 Migration to version 6.1
3202 This migration requires a complete dump of the 6.0 database and a
3203 restore of the database in 6.1.
3205 Those migrating from earlier 1.* releases should first upgrade to 1.09
3206 because the COPY output format was improved from the 1.02 release.
3207 _________________________________________________________________
3213 packet length checking in library routines
3214 lock manager priority patch
3215 check for under/over flow of float8(Bruce)
3216 multitable join fix(Vadim)
3217 SIGPIPE crash fix(Darren)
3218 large object fixes(Sven)
3219 allow btree indexes to handle NULLs(Vadim)
3220 timezone fixes(D'Arcy)
3221 select SUM(x) can return NULL on no rows(Thomas)
3222 internal optimizer, executor bug fixes(Vadim)
3223 fix problem where inner loop in < or <= has no rows(Vadim)
3224 prevent re-commuting join index clauses(Vadim)
3225 fix join clauses for multiple tables(Vadim)
3226 fix hash, hashjoin for arrays(Vadim)
3227 fix btree for abstime type(Vadim)
3228 large object fixes(Raymond)
3229 fix buffer leak in hash indexes (Vadim)
3230 fix rtree for use in inner scan (Vadim)
3231 fix gist for use in inner scan, cleanups (Vadim, Andrea)
3232 avoid unnecessary local buffers allocation (Vadim, Massimo)
3233 fix local buffers leak in transaction aborts (Vadim)
3234 fix file manager memmory leaks, cleanups (Vadim, Massimo)
3235 fix storage manager memmory leaks (Vadim)
3236 fix btree duplicates handling (Vadim)
3237 fix deleted tuples re-incarnation caused by vacuum (Vadim)
3238 fix SELECT varchar()/char() INTO TABLE made zero-length fields(Bruce)
3239 many psql, pg_dump, and libpq memory leaks fixed using Purify (Igor)
3243 attribute optimization statistics(Bruce)
3244 much faster new btree bulk load code(Paul)
3245 BTREE UNIQUE added to bulk load code(Vadim)
3246 new lock debug code(Massimo)
3247 massive changes to libpg++(Leo)
3248 new GEQO optimizer speeds table multitable optimization(Martin)
3249 new WARN message for non-unique insert into unique key(Marc)
3250 update x=-3, no spaces, now valid(Bruce)
3251 remove case-sensitive identifier handling(Bruce,Thomas,Dan)
3252 debug backend now pretty-prints tree(Darren)
3253 new Oracle character functions(Edmund)
3254 new plaintext password functions(Dan)
3255 no such class or insufficient privilege changed to distinct messages(Dan)
3256 new ANSI timestamp function(Dan)
3257 new ANSI Time and Date types (Thomas)
3258 move large chunks of data in backend(Martin)
3259 multicolumn btree indexes(Vadim)
3260 new SET var TO value command(Martin)
3261 update transaction status on reads(Dan)
3262 new locale settings for character types(Oleg)
3263 new SEQUENCE serial number generator(Vadim)
3264 GROUP BY function now possible(Vadim)
3265 re-organize regression test(Thomas,Marc)
3266 new optimizer operation weights(Vadim)
3267 new psql \z grant/permit option(Marc)
3268 new MONEY data type(D'Arcy,Thomas)
3269 tcp socket communication speed improved(Vadim)
3270 new VACUUM option for attribute statistics, and for certain columns (Vadim)
3271 many geometric type improvements(Thomas,Keith)
3272 additional regression tests(Thomas)
3273 new datestyle variable(Thomas,Vadim,Martin)
3274 more comparison operators for sorting types(Thomas)
3275 new conversion functions(Thomas)
3276 new more compact btree format(Vadim)
3277 allow pg_dumpall to preserve database ownership(Bruce)
3278 new SET GEQO=# and R_PLANS variable(Vadim)
3279 old (!GEQO) optimizer can use right-sided plans (Vadim)
3280 typechecking improvement in SQL parser(Bruce)
3281 new SET, SHOW, RESET commands(Thomas,Vadim)
3282 new \connect database USER option
3283 new destroydb -i option (Igor)
3284 new \dt and \di psql commands (Darren)
3285 SELECT "\n" now escapes newline (A. Duursma)
3286 new geometry conversion functions from old format (Thomas)
3290 new configuration script(Marc)
3291 readline configuration option added(Marc)
3292 OS-specific configuration options removed(Marc)
3293 new OS-specific template files(Marc)
3294 no more need to edit Makefile.global(Marc)
3295 re-arrange include files(Marc)
3296 nextstep patches (Gregor Hoffleit)
3297 removed WIN32-specific code(Bruce)
3298 removed postmaster -e option, now only postgres -e option (Bruce)
3299 merge duplicate library code in front/backends(Martin)
3300 now works with eBones, international Kerberos(Jun)
3301 more shared library support
3302 c++ include file cleanup(Bruce)
3303 warn about buggy flex(Bruce)
3304 DG/UX, Ultrix, IRIX, AIX portability fixes
3305 _________________________________________________________________
3309 Release date: 1997-01-29
3311 A dump/restore is required for those wishing to migrate data from
3312 previous releases of PostgreSQL.
3313 _________________________________________________________________
3315 Migration from version 1.09 to version 6.0
3317 This migration requires a complete dump of the 1.09 database and a
3318 restore of the database in 6.0.
3319 _________________________________________________________________
3321 Migration from pre-1.09 to version 6.0
3323 Those migrating from earlier 1.* releases should first upgrade to 1.09
3324 because the COPY output format was improved from the 1.02 release.
3325 _________________________________________________________________
3331 ALTER TABLE bug - running postgress process needs to re-read table definition
3332 Allow vacuum to be run on one table or entire database(Bruce)
3334 Fix array over-runs of memory writes(Kurt)
3335 Fix elusive btree range/non-range bug(Dan)
3336 Fix for hash indexes on some types like time and date
3337 Fix for pg_log size explosion
3338 Fix permissions on lo_export()(Bruce)
3339 Fix unitialized reads of memory(Kurt)
3340 Fixed ALTER TABLE ... char(3) bug(Bruce)
3341 Fixed a few small memory leaks
3342 Fixed EXPLAIN handling of options and changed full_path option name
3343 Fixed output of group acl permissions
3344 Memory leaks (hunt and destroy with tools like Purify(Kurt)
3345 Minor improvements to rules system
3347 New asserts for run-checking
3348 Overhauled parser/analyze code to properly report errors and increase speed
3349 Pg_dump -d now handles NULL's properly(Bruce)
3350 Prevent SELECT NULL from crashing server (Bruce)
3351 Properly report errors when INSERT ... SELECT columns did not match
3352 Properly report errors when insert column names were not correct
3353 Psql \g filename now works(Bruce)
3354 Psql fixed problem with multiple statements on one line with multiple outputs
3355 Removed duplicate system OIDs
3356 SELECT * INTO TABLE . GROUP/ORDER BY gives unlink error if table exists(Bruce)
3357 Several fixes for queries that crashed the backend
3358 Starting quote in insert string errors(Bruce)
3359 Submitting an empty query now returns empty status, not just " " query(Bruce)
3363 Add EXPLAIN manual page(Bruce)
3364 Add UNIQUE index capability(Dan)
3365 Add hostname/user level access control rather than just hostname and user
3366 Add synonym of != for <>(Bruce)
3367 Allow "select oid,* from table"
3368 Allow BY,ORDER BY to specify columns by number, or by non-alias table.column(Br
3370 Allow COPY from the frontend(Bryan)
3371 Allow GROUP BY to use alias column name(Bruce)
3372 Allow actual compression, not just reuse on the same page(Vadim)
3373 Allow installation-configuration option to auto-add all local users(Bryan)
3374 Allow libpq to distinguish between text value '' and null(Bruce)
3375 Allow non-postgres users with createdb privs to destroydb's
3376 Allow restriction on who can create C functions(Bryan)
3377 Allow restriction on who can do backend COPY(Bryan)
3378 Can shrink tables, pg_time and pg_log(Vadim & Erich)
3379 Change debug level 2 to print queries only, changed debug heading layout(Bruce)
3380 Change default decimal constant representation from float4 to float8(Bruce)
3381 European date format now set when postmaster is started
3382 Execute lowercase function names if not found with exact case
3383 Fixes for aggregate/GROUP processing, allow 'select sum(func(x),sum(x+y) from z
3385 Gist now included in the distrubution(Marc)
3386 Idend authentication of local users(Bryan)
3387 Implement BETWEEN qualifier(Bruce)
3388 Implement IN qualifier(Bruce)
3389 Libpq has PQgetisnull()(Bruce)
3390 Libpq++ improvements
3391 New options to initdb(Bryan)
3392 Pg_dump allow dump of OIDs(Bruce)
3393 Pg_dump create indexes after tables are loaded for speed(Bruce)
3394 Pg_dumpall dumps all databases, and the user table
3395 Pginterface additions for NULL values(Bruce)
3396 Prevent postmaster from being run as root
3397 Psql \h and \? is now readable(Bruce)
3398 Psql allow backslashed, semicolons anywhere on the line(Bruce)
3399 Psql changed command prompt for lines in query or in quotes(Bruce)
3400 Psql char(3) now displays as (bp)char in \d output(Bruce)
3401 Psql return code now more accurate(Bryan?)
3402 Psql updated help syntax(Bruce)
3403 Re-visit and fix vacuum(Vadim)
3404 Reduce size of regression diffs, remove timezone name difference(Bruce)
3405 Remove compile-time parameters to enable binary distributions(Bryan)
3406 Reverse meaning of HBA masks(Bryan)
3407 Secure Authentication of local users(Bryan)
3408 Speed up vacuum(Vadim)
3409 Vacuum now had VERBOSE option(Bruce)
3413 All functions now have prototypes that are compared against the calls
3414 Allow asserts to be disabled easly from Makefile.global(Bruce)
3415 Change oid constants used in code to #define names
3416 Decoupled sparc and solaris defines(Kurt)
3417 Gcc -Wall compiles cleanly with warnings only from unfixable constructs
3418 Major include file reorganization/reduction(Marc)
3419 Make now stops on compile failure(Bryan)
3420 Makefile restructuring(Bryan, Marc)
3421 Merge bsdi_2_1 to bsdi(Bruce)
3422 Monitor program removed
3423 Name change from Postgres95 to PostgreSQL
3424 New config.h file(Marc, Bryan)
3425 PG_VERSION now set to 6.0 and used by postmaster
3426 Portability additions, including Ultrix, DG/UX, AIX, and Solaris
3427 Reduced the number of #define's, centeralized #define's
3428 Remove duplicate OIDS in system tables(Dan)
3429 Remove duplicate system catalog info or report mismatches(Dan)
3430 Removed many os-specific #define's
3431 Restructured object file generation/location(Bryan, Marc)
3432 Restructured port-specific file locations(Bryan, Marc)
3433 Unused/uninialized variables corrected
3434 _________________________________________________________________
3438 Release date: 1996-11-04
3440 Sorry, we didn't keep track of changes from 1.02 to 1.09. Some of the
3441 changes listed in 6.0 were actually included in the 1.02.1 to 1.09
3443 _________________________________________________________________
3447 Release date: 1996-08-01
3448 _________________________________________________________________
3450 Migration from version 1.02 to version 1.02.1
3452 Here is a new migration file for 1.02.1. It includes the 'copy' change
3453 and a script to convert old ASCII files.
3455 Note: The following notes are for the benefit of users who want to
3456 migrate databases from Postgres95 1.01 and 1.02 to Postgres95
3459 If you are starting afresh with Postgres95 1.02.1 and do not need
3460 to migrate old databases, you do not need to read any further.
3462 In order to upgrade older Postgres95 version 1.01 or 1.02 databases to
3463 version 1.02.1, the following steps are required:
3464 1. Start up a new 1.02.1 postmaster
3465 2. Add the new built-in functions and operators of 1.02.1 to 1.01 or
3466 1.02 databases. This is done by running the new 1.02.1 server
3467 against your own 1.01 or 1.02 database and applying the queries
3468 attached at the end of the file. This can be done easily through
3469 "psql". If your 1.01 or 1.02 database is named testdb and you have
3470 cut the commands from the end of this file and saved them in
3472 % psql testdb -f addfunc.sql
3473 Those upgrading 1.02 databases will get a warning when executing
3474 the last two statements in the file because they are already
3475 present in 1.02. This is not a cause for concern.
3476 _________________________________________________________________
3478 Dump/Reload Procedure
3480 If you are trying to reload a pg_dump or text-mode, copy tablename to
3481 stdout generated with a previous version, you will need to run the
3482 attached "sed" script on the ASCII file before loading it into the
3483 database. The old format used '.' as end-of-data, while '\.' is now
3484 the end-of-data marker. Also, empty strings are now loaded in as ''
3485 rather than NULL. See the copy manual page for full details.
3486 sed 's/^\.$/\\./g' <in_file >out_file
3488 If you are loading an older binary copy or non-stdout copy, there is
3489 no end-of-data character, and hence no conversion necessary.
3490 -- following lines added by agc to reflect the case-insensitive
3491 -- regexp searching for varchar (in 1.02), and bpchar (in 1.02.1)
3492 create operator ~* (leftarg = bpchar, rightarg = text, procedure = texticregexe
3494 create operator !~* (leftarg = bpchar, rightarg = text, procedure = texticregex
3496 create operator ~* (leftarg = varchar, rightarg = text, procedure = texticregex
3498 create operator !~* (leftarg = varchar, rightarg = text, procedure = texticrege
3500 _________________________________________________________________
3504 Source code maintenance and development
3505 * worldwide team of volunteers
3506 * the source tree now in CVS at ftp.ki.net
3509 * psql (and underlying libpq library) now has many more options for
3510 formatting output, including HTML
3511 * pg_dump now output the schema and/or the data, with many fixes to
3512 enhance completeness.
3513 * psql used in place of monitor in administration shell scripts.
3514 monitor to be depreciated in next release.
3515 * date/time functions enhanced
3516 * NULL insert/update/comparison fixed/enhanced
3517 * TCL/TK lib and shell fixed to work with both tck7.4/tk4.0 and tcl7.5/tk4.1
3519 Bug Fixes (almost too numerous to mention)
3521 * storage management
3522 * check for NULL pointer before dereferencing
3526 * added SolarisX86 port
3527 * added BSD/OS 2.1 port
3529 _________________________________________________________________
3533 Release date: 1996-02-23
3534 _________________________________________________________________
3536 Migration from version 1.0 to version 1.01
3538 The following notes are for the benefit of users who want to migrate
3539 databases from Postgres95 1.0 to Postgres95 1.01.
3541 If you are starting afresh with Postgres95 1.01 and do not need to
3542 migrate old databases, you do not need to read any further.
3544 In order to Postgres95 version 1.01 with databases created with
3545 Postgres95 version 1.0, the following steps are required:
3546 1. Set the definition of NAMEDATALEN in "src/Makefile.global" to 16
3547 and OIDNAMELEN to 20.
3548 2. Decide whether you want to use Host based authentication.
3549 a. If you do, you must create a file name pg_hba in your
3550 top-level data directory (typically the value of your
3551 $PGDATA). "src/libpq/pg_hba" shows an example syntax.
3552 b. If you do not want host-based authentication, you can comment
3555 in "src/Makefile.global"
3556 Note that host-based authentication is turned on by default,
3557 and if you do not take steps A or B above, the out-of-the-box
3558 1.01 will not allow you to connect to 1.0 databases.
3559 3. Compile and install 1.01, but DO NOT do the "initdb" step.
3560 4. Before doing anything else, terminate your 1.0 postmaster, and
3561 backup your existing $PGDATA directory.
3562 5. Set your PGDATA environment variable to your 1.0 databases, but
3563 set up path up so that 1.01 binaries are being used.
3564 6. Modify the file "$PGDATA/PG_VERSION" from 5.0 to 5.1
3565 7. Start up a new 1.01 postmaster
3566 8. Add the new built-in functions and operators of 1.01 to 1.0
3567 databases. This is done by running the new 1.01 server against
3568 your own 1.0 database and applying the queries attached and saving
3569 in the file 1.0_to_1.01.sql. This can be done easily through
3570 "psql". If your 1.0 database is name testdb:
3571 % psql testdb -f 1.0_to_1.01.sql
3572 and then execute the following commands (cut and paste from here):
3573 -- add builtin functions that are new to 1.01
3575 create function int4eqoid (int4, oid) returns bool as 'foo'
3576 language 'internal';
3577 create function oideqint4 (oid, int4) returns bool as 'foo'
3578 language 'internal';
3579 create function char2icregexeq (char2, text) returns bool as 'foo'
3580 language 'internal';
3581 create function char2icregexne (char2, text) returns bool as 'foo'
3582 language 'internal';
3583 create function char4icregexeq (char4, text) returns bool as 'foo'
3584 language 'internal';
3585 create function char4icregexne (char4, text) returns bool as 'foo'
3586 language 'internal';
3587 create function char8icregexeq (char8, text) returns bool as 'foo'
3588 language 'internal';
3589 create function char8icregexne (char8, text) returns bool as 'foo'
3590 language 'internal';
3591 create function char16icregexeq (char16, text) returns bool as 'foo'
3592 language 'internal';
3593 create function char16icregexne (char16, text) returns bool as 'foo'
3594 language 'internal';
3595 create function texticregexeq (text, text) returns bool as 'foo'
3596 language 'internal';
3597 create function texticregexne (text, text) returns bool as 'foo'
3598 language 'internal';
3600 -- add builtin functions that are new to 1.01
3602 create operator = (leftarg = int4, rightarg = oid, procedure = int4eqoid);
3603 create operator = (leftarg = oid, rightarg = int4, procedure = oideqint4);
3604 create operator ~* (leftarg = char2, rightarg = text, procedure = char2icregexe
3606 create operator !~* (leftarg = char2, rightarg = text, procedure = char2icregex
3608 create operator ~* (leftarg = char4, rightarg = text, procedure = char4icregexe
3610 create operator !~* (leftarg = char4, rightarg = text, procedure = char4icregex
3612 create operator ~* (leftarg = char8, rightarg = text, procedure = char8icregexe
3614 create operator !~* (leftarg = char8, rightarg = text, procedure = char8icregex
3616 create operator ~* (leftarg = char16, rightarg = text, procedure = char16icrege
3618 create operator !~* (leftarg = char16, rightarg = text, procedure = char16icreg
3620 create operator ~* (leftarg = text, rightarg = text, procedure = texticregexeq)
3622 create operator !~* (leftarg = text, rightarg = text, procedure = texticregexne
3624 _________________________________________________________________
3629 * 1.01 is backwards compatible with 1.0 database provided the user
3630 follow the steps outlined in the MIGRATION_from_1.0_to_1.01 file.
3631 If those steps are not taken, 1.01 is not compatible with 1.0 database.
3634 * added PQdisplayTuples() to libpq and changed monitor and psql to use it
3635 * added NeXT port (requires SysVIPC implementation)
3636 * added CAST .. AS ... syntax
3637 * added ASC and DESC keywords
3638 * added 'internal' as a possible language for CREATE FUNCTION
3639 internal functions are C functions which have been statically linked
3640 into the postgres backend.
3641 * a new type "name" has been added for system identifiers (table names,
3642 attribute names, etc.) This replaces the old char16 type. The
3643 of name is set by the NAMEDATALEN #define in src/Makefile.global
3644 * a readable reference manual that describes the query language.
3645 * added host-based access control. A configuration file ($PGDATA/pg_hba)
3646 is used to hold the configuration data. If host-based access control
3647 is not desired, comment out HBA=1 in src/Makefile.global.
3648 * changed regex handling to be uniform use of Henry Spencer's regex code
3649 regardless of platform. The regex code is included in the distribution
3650 * added functions and operators for case-insensitive regular expressions.
3651 The operators are ~* and !~*.
3652 * pg_dump uses COPY instead of SELECT loop for better performance
3655 * fixed an optimizer bug that was causing core dumps when
3656 functions calls were used in comparisons in the WHERE clause
3657 * changed all uses of getuid to geteuid so that effective uids are used
3658 * psql now returns non-zero status on errors when using -c
3659 * applied public patches 1-14
3660 _________________________________________________________________
3664 Release date: 1995-09-05
3665 _________________________________________________________________
3670 * The copyright of Postgres 1.0 has been loosened to be freely modifiable
3671 and modifiable for any purpose. Please read the COPYRIGHT file.
3672 Thanks to Professor Michael Stonebraker for making this possible.
3675 * date formats have to be MM-DD-YYYY (or DD-MM-YYYY if you're using
3676 EUROPEAN STYLE). This follows SQL-92 specs.
3677 * "delimiters" is now a keyword
3680 * sql LIKE syntax has been added
3681 * copy command now takes an optional USING DELIMITER specification.
3682 delimiters can be any single-character string.
3683 * IRIX 5.3 port has been added.
3684 Thanks to Paul Walmsley and others.
3685 * updated pg_dump to work with new libpq
3686 * \d has been added psql
3687 Thanks to Keith Parks
3688 * regexp performance for architectures that use POSIX regex has been
3689 improved due to caching of precompiled patterns.
3690 Thanks to Alistair Crooks
3691 * a new version of libpq++
3692 Thanks to William Wanders
3695 * arbitrary userids can be specified in the createuser script
3696 * \c to connect to other databases in psql now works.
3697 * bad pg_proc entry for float4inc() is fixed
3698 * users with usecreatedb field set can now create databases without
3699 having to be usesuper
3700 * remove access control entries when the entry no longer has any
3702 * fixed non-portable datetimes implementation
3703 * added kerberos flags to the src/backend/Makefile
3704 * libpq now works with kerberos
3705 * typographic errors in the user manual have been corrected.
3706 * btrees with multiple index never worked, now we tell you they don't
3707 work when you try to use them
3708 _________________________________________________________________
3710 Postgres95 Release 0.03
3712 Release date: 1995-07-21
3713 _________________________________________________________________
3717 Incompatible changes:
3718 * BETA-0.3 IS INCOMPATIBLE WITH DATABASES CREATED WITH PREVIOUS VERSIONS
3719 (due to system catalog changes and indexing structure changes).
3720 * double-quote (") is deprecated as a quoting character for string literals;
3721 you need to convert them to single quotes (').
3722 * name of aggregates (eg. int4sum) are renamed in accordance with the
3723 SQL standard (eg. sum).
3724 * CHANGE ACL syntax is replaced by GRANT/REVOKE syntax.
3725 * float literals (eg. 3.14) are now of type float4 (instead of float8 in
3726 previous releases); you might have to do typecasting if you depend on it
3727 being of type float8. If you neglect to do the typecasting and you assign
3728 a float literal to a field of type float8, you may get incorrect values
3730 * LIBPQ has been totally revamped so that frontend applications
3731 can connect to multiple backends
3732 * the usesysid field in pg_user has been changed from int2 to int4 to
3733 allow wider range of Unix user ids.
3734 * the netbsd/freebsd/bsd o/s ports have been consolidated into a
3735 single BSD44_derived port. (thanks to Alistair Crooks)
3737 SQL standard-compliance (the following details changes that makes postgres95
3738 more compliant to the SQL-92 standard):
3739 * the following SQL types are now built-in: smallint, int(eger), float, real,
3740 char(N), varchar(N), date and time.
3742 The following are aliases to existing postgres types:
3744 integer, int -> int4
3745 float, real -> float4
3746 char(N) and varchar(N) are implemented as truncated text types. In
3747 addition, char(N) does blank-padding.
3748 * single-quote (') is used for quoting string literals; '' (in addition to
3749 \') is supported as means of inserting a single quote in a string
3750 * SQL standard aggregate names (MAX, MIN, AVG, SUM, COUNT) are used
3751 (Also, aggregates can now be overloaded, i.e. you can define your
3752 own MAX aggregate to take in a user-defined type.)
3753 * CHANGE ACL removed. GRANT/REVOKE syntax added.
3754 - Privileges can be given to a group using the "GROUP" keyword.
3756 GRANT SELECT ON foobar TO GROUP my_group;
3757 The keyword 'PUBLIC' is also supported to mean all users.
3759 Privileges can only be granted or revoked to one user or group
3762 "WITH GRANT OPTION" is not supported. Only class owners can change
3764 - The default access control is to to grant users readonly access.
3765 You must explicitly grant insert/update access to users. To change
3766 this, modify the line in
3767 src/backend/utils/acl.h
3768 that defines ACL_WORLD_DEFAULT
3771 * the bug where aggregates of empty tables were not run has been fixed. Now,
3772 aggregates run on empty tables will return the initial conditions of the
3773 aggregates. Thus, COUNT of an empty table will now properly return 0.
3774 MAX/MIN of an empty table will return a tuple of value NULL.
3775 * allow the use of \; inside the monitor
3776 * the LISTEN/NOTIFY asynchronous notification mechanism now work
3777 * NOTIFY in rule action bodies now work
3778 * hash indexes work, and access methods in general should perform better.
3779 creation of large btree indexes should be much faster. (thanks to Paul
3782 Other changes and enhancements:
3783 * addition of an EXPLAIN statement used for explaining the query execution
3784 plan (eg. "EXPLAIN SELECT * FROM EMP" prints out the execution plan for
3786 * WARN and NOTICE messages no longer have timestamps on them. To turn on
3787 timestamps of error messages, uncomment the line in
3788 src/backend/utils/elog.h:
3789 /* define ELOG_TIMESTAMPS */
3790 * On an access control violation, the message
3791 "Either no such class or insufficient privilege"
3792 will be given. This is the same message that is returned when
3793 a class is not found. This dissuades non-privileged users from
3794 guessing the existence of privileged classes.
3795 * some additional system catalog changes have been made that are not
3796 visible to the user.
3799 * The -oid option has been added to the "pg_result" tcl command.
3800 pg_result -oid returns oid of the last tuple inserted. If the
3801 last command was not an INSERT, then pg_result -oid returns "".
3802 * the large object interface is available as pg_lo* tcl commands:
3803 pg_lo_open, pg_lo_close, pg_lo_creat, etc.
3805 Portability enhancements and New Ports:
3806 * flex/lex problems have been cleared up. Now, you should be able to use
3807 flex instead of lex on any platforms. We no longer make assumptions of
3808 what lexer you use based on the platform you use.
3809 * The Linux-ELF port is now supported. Various configuration have been
3810 tested: The following configuration is known to work:
3811 kernel 1.2.10, gcc 2.6.3, libc 4.7.2, flex 2.5.2, bison 1.24
3812 with everything in ELF format,
3815 * ipcclean added to the distribution
3816 ipcclean usually does not need to be run, but if your backend crashes
3817 and leaves shared memory segments hanging around, ipcclean will
3818 clean them up for you.
3821 * the user manual has been revised and libpq documentation added.
3822 _________________________________________________________________
3824 Postgres95 Release 0.02
3826 Release date: 1995-05-25
3827 _________________________________________________________________
3831 Incompatible changes:
3832 * The SQL statement for creating a database is 'CREATE DATABASE' instead
3833 of 'CREATEDB'. Similarly, dropping a database is 'DROP DATABASE' instead
3834 of 'DESTROYDB'. However, the names of the executables 'createdb' and
3835 'destroydb' remain the same.
3838 * pgperl - a Perl (4.036) interface to Postgres95
3839 * pg_dump - a utility for dumping out a postgres database into a
3840 script file containing query commands. The script files are in a ASCII
3841 format and can be used to reconstruct the database, even on other
3842 machines and other architectures. (Also good for converting
3843 a Postgres 4.2 database to Postgres95 database.)
3845 The following ports have been incorporated into postgres95-beta-0.02:
3846 * the NetBSD port by Alistair Crooks
3847 * the AIX port by Mike Tung
3848 * the Windows NT port by Jon Forrest (more stuff but not done yet)
3849 * the Linux ELF port by Brian Gallew
3851 The following bugs have been fixed in postgres95-beta-0.02:
3852 * new lines not escaped in COPY OUT and problem with COPY OUT when first
3854 * cannot type return to use the default user id in createuser
3855 * SELECT DISTINCT on big tables crashes
3856 * Linux installation problems
3857 * monitor doesn't allow use of 'localhost' as PGHOST
3858 * psql core dumps when doing \c or \l
3859 * the "pgtclsh" target missing from src/bin/pgtclsh/Makefile
3860 * libpgtcl has a hard-wired default port number
3861 * SELECT DISTINCT INTO TABLE hangs
3862 * CREATE TYPE doesn't accept 'variable' as the internallength
3863 * wrong result using more than 1 aggregate in a SELECT
3864 _________________________________________________________________
3866 Postgres95 Release 0.01
3868 Release date: 1995-05-01