2 $Header: /cvsroot/pgsql/doc/src/sgml/release.sgml,v 1.170 2002/12/14 00:24:23 petere Exp $
5 <appendix id="release">
6 <title>Release Notes</title>
8 <sect1 id="release-devel">
9 <title>7.4 Development Branch</title>
12 Below is a subset of the changes that have gone into the
13 development branch of PostgreSQL since version 7.3. For a complete
14 list of changes, consult the CVS logs.
18 Developers: When you add a feature, mention it here. This avoids
19 lossiness when digging out the information from the CVS logs, and
20 furthermore it advertises your feature to external parties at the
21 earliest possible moment.
23 CDATA means the content is "SGML-free", so you can write without
24 worries about funny characters.
26 <literallayout><![CDATA[
28 Domains now support CHECK constraints
33 <sect1 id="release-7-3">
34 <title>Release 7.3</title>
37 <title>Release date</title>
38 <simpara>2002-11-27</simpara>
42 <title>Overview</title>
45 Major changes in this release:
52 Schemas allow users to create objects in separate namespaces,
53 so two people or applications can have tables with the same
54 name. There is also a public schema for shared tables.
55 Table/index creation can be restricted by removing permissions
62 <term>Drop Column</term>
65 PostgreSQL now supports the <literal>ALTER TABLE ... DROP
66 COLUMN</literal> functionality.
72 <term>Table Functions</term>
75 Functions returning multiple rows and/or multiple columns are
76 now much easier to use than before. You can call such a
77 <quote>table function</quote> in the <literal>SELECT</literal>
78 <literal>FROM</literal> clause, treating its output like a
79 table. Also, <application>PL/pgSQL</application> functions can
86 <term>Prepared Queries</term>
89 PostgreSQL now supports prepared queries, for improved
96 <term>Dependency Tracking</term>
99 PostgreSQL now records object dependencies, which allows
100 improvements in many areas. <command>DROP</command>
101 statements now take either <literal>CASCADE</> or
102 <literal>RESTRICT</> to control whether dependent objects are
109 <term>Privileges</term>
112 Functions and procedural languages now have privileges, and
113 functions can be defined to run with the privileges of their
120 <term>Internationalization</term>
123 Both multibyte and locale support are now always enabled.
132 A variety of logging options have been enhanced.
138 <term>Interfaces</term>
141 A large number of interfaces have been moved to <ulink
142 url="http://gborg.postgresql.org">http://gborg.postgresql.org</>
143 where they can be developed and released independently.
149 <term>Functions/Identifiers</term>
152 By default, functions can now take up to 32 parameters, and
153 identifiers can be up to 63 bytes long. Also, <literal>OPAQUE</>
154 is now deprecated: there are specific <quote>pseudo-datatypes</>
155 to represent each of the former meanings of <literal>OPAQUE</>
156 in function argument and result types.
166 <title>Migration to version 7.3</title>
169 A dump/restore using <application>pg_dump</> is required for those
170 wishing to migrate data from any previous release. If your
171 application examines the system catalogs, additional changes will
172 be required due to the introduction of schemas in 7.3; for more
173 information, see: <ulink
174 url="http://www.ca.postgresql.org/docs/momjian/upgrade_tips_7.3">
175 http://www.ca.postgresql.org/docs/momjian/upgrade_tips_7.3</>.
179 Observe the following incompatibilities:
184 Pre-6.3 clients are no longer supported.
190 <filename>pg_hba.conf</filename> now has a column for the user
191 name and additional features. Existing files need to be
198 Several <filename>postgresql.conf</filename> logging parameters
205 <literal>LIMIT #,#</literal> has been disabled; use
206 <literal>LIMIT # OFFSET #</literal>.
212 <command>INSERT</command> statements with column lists must
213 specify a value for each specified column. For example,
214 <literal>INSERT INTO tab (col1, col2) VALUES ('val1')</literal>
215 is now invalid. It's still allowed to supply fewer columns than
216 expected if the <command>INSERT</command> does not have a column list.
222 <type>serial</type> columns are no longer automatically
223 <literal>UNIQUE</>; thus, an index will not automatically be
230 A <command>SET</command> command inside an aborted transaction
237 <command>COPY</command> no longer considers missing trailing
238 columns to be null. All columns need to be specified.
239 (However, one may achieve a similar effect by specifying a
240 column list in the <command>COPY</command> command.)
246 The data type <type>timestamp</type> is now equivalent to
247 <type>timestamp without time zone</type>, instead of
248 <type>timestamp with time zone</type>.
254 Pre-7.3 databases loaded into 7.3 will not have the new object
255 dependencies for <type>serial</type> columns, unique
256 constraints, and foreign keys. See the directory
257 <filename>contrib/adddepend/</filename> for a detailed
258 description and a script that will add such dependencies.
264 An empty string (<literal>''</literal>) is no longer allowed as
265 the input into an integer field. Formerly, it was silently
276 <title>Changes</title>
279 <title>Server Operation</title>
281 Add pg_locks view to show locks (Neil)
282 Security fixes for password negotiation memory allocation (Neil)
283 Remove support for version 0 FE/BE protocol (PostgreSQL 6.2 and
285 Reserve the last few backend slots for superusers, add parameter
286 superuser_reserved_connections to control this (Nigel J. Andrews)
291 <title>Performance</title>
293 Improve startup by calling localtime() only once (Tom)
294 Cache system catalog information in flat files for faster startup
296 Improve caching of index information (Tom)
297 Optimizer improvements (Tom, Fernando Nasser)
298 Catalog caches now store failed lookups (Tom)
299 Hash function improvements (Neil)
300 Improve performance of query tokenization and network handling (Peter)
301 Speed improvement for large object restore (Mario Weilguni)
302 Mark expired index entries on first lookup, saving later heap fetches
304 Avoid excessive NULL bitmap padding (Manfred Koizar)
305 Add BSD-licensed qsort() for Solaris, for performance (Bruce)
306 Reduce per-row overhead by four bytes (Manfred Koizar)
307 Fix GEQO optimizer bug (Neil Conway)
308 Make WITHOUT OID actually save four bytes per row (Manfred Koizar)
309 Add default_statistics_target variable to specify ANALYZE buckets
311 Use local buffer cache for temporary tables so no WAL overhead (Tom)
312 Improve free space map performance on large tables (Stephen Marshall,
314 Improved WAL write concurrency (Tom)
319 <title>Privileges</title>
321 Add privileges on functions and procedural languages (Peter)
322 Add OWNER to CREATE DATABASE so superusers can create databases
323 on behalf of unprivileged users. (Gavin Sherry, Tom)
324 Add new object permission bits EXECUTE and USAGE (Tom)
325 Add SET SESSION AUTHORIZATION DEFAULT and RESET SESSION AUTHORIZATION
327 Allow functions to be executed with the privilege of the function
333 <title>Server Configuration</title>
335 Server log messages now tagged with LOG, not DEBUG (Bruce)
336 Add user column to pg_hba.conf (Bruce)
337 Have log_connections output two lines in log file (Tom)
338 Remove debug_level from postgresql.conf, now server_min_messages
340 New ALTER DATABASE/USER ... SET command for per-user/database
341 initialization (Peter)
342 New parameters server_min_messages and client_min_messages to
343 control which messages are sent to the server logs or client
345 Allow pg_hba.conf to specify lists of users/databases separated by
346 commas, group names prepended with +, and file names prepended
348 Remove secondary password file capability and pg_password utility
350 Add variable db_user_namespace for database-local user names (Bruce)
351 SSL improvements (Bear Giles)
352 Make encryption of stored passwords the default (Bruce)
353 Allow pg_statistics to be reset by calling pg_stat_reset()
355 Add log_duration parameter (Bruce)
356 Rename debug_print_query to log_statement (Bruce)
357 Rename show_query_stats to show_statement_stats (Bruce)
358 Add param log_min_error_statement to print commands to logs on error
364 <title>Queries</title>
366 Make cursors insensitive, meaning their contents do not change (Tom)
367 Disable LIMIT #,# syntax; now only LIMIT # OFFSET # supported (Bruce)
368 Increase identifier length to 63 (Neil, Bruce)
369 UNION fixes for merging >= 3 columns of different lengths (Tom)
370 Add DEFAULT keyword to INSERT, e.g., INSERT ... (..., DEFAULT, ...)
372 Allow views to have default values using ALTER COLUMN ... SET DEFAULT
374 Fail on INSERTs with column lists that don't supply all column
375 values, e.g., INSERT INTO tab (col1, col2) VALUES ('val1'); (Rod)
376 Fix for join aliases (Tom)
377 Fix for FULL OUTER JOINs (Tom)
378 Improve reporting of invalid identifier and location (Tom, Gavin)
379 Fix OPEN cursor(args) (Tom)
380 Allow 'ctid' to be used in a view and currtid(viewname) (Hiroshi)
381 Fix for CREATE TABLE AS with UNION (Tom)
382 SQL99 syntax improvements (Thomas)
383 Add statement_timeout variable to cancel queries (Bruce)
384 Allow prepared queries with PREPARE/EXECUTE (Neil)
385 Allow FOR UPDATE to appear after LIMIT/OFFSET (Bruce)
386 Add variable autocommit (Tom, David Van Wie)
391 <title>Object Manipulation</title>
393 Make equals signs optional in CREATE DATABASE (Gavin Sherry)
394 Make ALTER TABLE OWNER change index ownership too (Neil)
395 New ALTER TABLE tabname ALTER COLUMN colname SET STORAGE controls
396 TOAST storage, compression (John Gray)
397 Add schema support, CREATE/DROP SCHEMA (Tom)
398 Create schema for temporary tables (Tom)
399 Add variable search_path for schema search (Tom)
400 Add ALTER TABLE SET/DROP NOT NULL (Christopher)
401 New CREATE FUNCTION volatility levels (Tom)
402 Make rule names unique only per table (Tom)
403 Add 'ON tablename' clause to DROP RULE and COMMENT ON RULE (Tom)
404 Add ALTER TRIGGER RENAME (Joe)
405 New current_schema() and current_schemas() inquiry functions (Tom)
406 Allow functions to return multiple rows (table functions) (Joe)
407 Make WITH optional in CREATE DATABASE, for consistency (Bruce)
408 Add object dependency tracking (Rod, Tom)
409 Add RESTRICT/CASCADE to DROP commands (Rod)
410 Add ALTER TABLE DROP for non-CHECK CONSTRAINT (Rod)
411 Autodestroy sequence on DROP of table with SERIAL (Rod)
412 Prevent column dropping if column is used by foreign key (Rod)
413 Automatically drop constraints/functions when object is dropped (Rod)
414 Add CREATE/DROP OPERATOR CLASS (Bill Studenmund, Tom)
415 Add ALTER TABLE DROP COLUMN (Christopher, Tom, Hiroshi)
416 Prevent inherited columns from being removed or renamed (Alvaro
418 Fix foreign key constraints to not error on intermediate database
420 Propagate column or table renaming to foreign key constraints
421 Add CREATE OR REPLACE VIEW (Gavin, Neil, Tom)
422 Add CREATE OR REPLACE RULE (Gavin, Neil, Tom)
423 Have rules execute alphabetically, returning more predictable values
425 Triggers are now fired in alphabetical order (Tom)
426 Add /contrib/adddepend to handle pre-7.3 object dependencies (Rod)
427 Allow better casting when inserting/updating values (Tom)
432 <title>Utility Commands</title>
434 Have COPY TO output embedded carriage returns and newlines as \r and
436 Allow DELIMITER in COPY FROM to be 8-bit clean (Tatsuo)
437 Make pg_dump use ALTER TABLE ADD PRIMARY KEY, for performance (Neil)
438 Disable brackets in multistatement rules (Bruce)
439 Disable VACUUM from being called inside a function (Bruce)
440 Allow dropdb and other scripts to use identifiers with spaces (Bruce)
441 Restrict database comment changes to the current database
442 Allow comments on operators, independent of the underlying function
444 Rollback SET commands in aborted transactions (Tom)
445 EXPLAIN now outputs as a query (Tom)
446 Display condition expressions and sort keys in EXPLAIN (Tom)
447 Add 'SET LOCAL var = value' to set configuration variables for a
448 single transaction (Tom)
449 Allow ANALYZE to run in a transaction (Bruce)
450 Improve COPY syntax using new WITH clauses, keep backward
451 compatibility (Bruce)
452 Fix pg_dump to consistently output tags in non-ASCII dumps (Bruce)
453 Make foreign key constraints clearer in dump file (Rod)
454 Add COMMENT ON CONSTRAINT (Rod)
455 Allow COPY TO/FROM to specify column names (Brent Verner)
456 Dump UNIQUE and PRIMARY KEY contraints as ALTER TABLE (Rod)
457 Have SHOW output a query result (Joe)
458 Generate failure on short COPY lines rather than pad NULLs (Neil)
459 Fix CLUSTER to preserve all table attributes (Alvaro Herrera)
460 New pg_settings table to view/modify GUC settings (Joe)
461 Add smart quoting, portability improvements to pg_dump output (Peter)
462 Dump serial columns out as SERIAL (Tom)
463 Enable large file support, >2G for pg_dump (Peter, Philip Warner,
465 Disallow TRUNCATE on tables that are involved in referential
467 Have TRUNCATE also auto-truncate the toast table of the relation (Tom)
468 Add clusterdb utility that will auto-cluster an entire database
469 based on previous CLUSTER operations (Alvaro Herrera)
470 Overhaul pg_dumpall (Peter)
471 Allow REINDEX of TOAST tables (Tom)
472 Implemented START TRANSACTION, per SQL99 (Neil)
473 Fix rare index corruption when a page split affects bulk delete (Tom)
474 Fix ALTER TABLE ... ADD COLUMN for inheritance (Alvaro Herrera)
479 <title>Data Types and Functions</title>
481 Fix factorial(0) to return 1 (Bruce)
482 Date/time/timezone improvements (Thomas)
483 Fix for array slice extraction (Tom)
484 Fix extract/date_part to report proper microseconds for timestamp
486 Allow text_substr() and bytea_substr() to read TOAST values more
487 efficiently (John Gray)
488 Add domain support (Rod)
489 Make WITHOUT TIME ZONE the default for TIMESTAMP and TIME data types
491 Allow alternate storage scheme of 64-bit integers for date/time types
492 using --enable-integer-datetimes in configure (Thomas)
493 Make timezone(timestamptz) return timestamp rather than a string
495 Allow fractional seconds in date/time types for dates prior to 1BC
497 Limit timestamp data types to 6 decimal places of precision (Thomas)
498 Change timezone conversion functions from timetz() to timezone()
500 Add configuration variables datestyle and timezone (Tom)
501 Add OVERLAY(), which allows substitution of a substring in a string
503 Add SIMILAR TO (Thomas, Tom)
504 Add regular expression SUBSTRING(string FROM pat FOR escape) (Thomas)
505 Add LOCALTIME and LOCALTIMESTAMP functions (Thomas)
506 Add named composite types using CREATE TYPE typename AS (column)
508 Allow composite type definition in the table alias clause (Joe)
509 Add new API to simplify creation of C language table functions (Joe)
510 Remove ODBC-compatible empty parentheses from calls to SQL99
511 functions for which these parentheses do not match the standard
513 Allow macaddr data type to accept 12 hex digits with no separators
515 Add CREATE/DROP CAST (Peter)
516 Add IS DISTINCT FROM operator (Thomas)
517 Add SQL99 TREAT() function, synonym for CAST() (Thomas)
518 Add pg_backend_pid() to output backend pid (Bruce)
519 Add IS OF / IS NOT OF type predicate (Thomas)
520 Allow bit string constants without fully-specified length (Thomas)
521 Allow conversion between 8-byte integers and bit strings (Thomas)
522 Implement hex literal conversion to bit string literal (Thomas)
523 Allow table functions to appear in the FROM clause (Joe)
524 Increase maximum number of function parameters to 32 (Bruce)
525 No longer automatically create index for SERIAL column (Tom)
526 Add current_database() (Rod)
527 Fix cash_words() to not overflow buffer (Tom)
528 Add functions replace(), split_part(), to_hex() (Joe)
529 Fix LIKE for bytea as a right-hand argument (Joe)
530 Prevent crashes caused by SELECT cash_out(2) (Tom)
531 Fix to_char(1,'FM999.99') to return a period (Karel)
532 Fix trigger/type/language functions returning OPAQUE to return
538 <title>Internationalization</title>
540 Add additional encodings: Korean (JOHAB), Thai (WIN874), Vietnamese
541 (TCVN), Arabic (WIN1256), Simplified Chinese (GBK), Korean (UHC)
543 Enable locale support by default (Peter)
544 Add locale variables (Peter)
545 Escape byes >= 0x7f for multibyte in PQescapeBytea/PQunescapeBytea
547 Add locale awareness to regular expression character classes
548 Enable multibyte support by default (Tatso)
549 Add GB18030 multibyte support (Bill Huang)
550 Add CREATE/DROP CONVERSION, allowing loadable encodings (Tatsuo,
552 Add pg_conversion table (Tatsuo)
553 Add SQL99 CONVERT() function (Tatsuo)
554 pg_dumpall, pg_controldata, and pg_resetxlog now national-language
556 New and updated translations
561 <title>Server-side Languages</title>
563 Allow recursive SQL function (Peter)
564 Change PL/Tcl build to use configured compiler and Makefile.shlib
566 Overhaul the PL/pgSQL FOUND variable to be more Oracle-compatible
568 Allow PL/pgSQL to handle quoted identifiers (Tom)
569 Allow set-returning PL/pgSQL functions (Neil)
570 Make PL/pgSQL schema-aware (Joe)
571 Remove some memory leaks (Nigel J. Andrews, Tom)
578 Don't lowercase psql \connect database name for 7.2.0 compatibility
580 Add psql \timing to time user queries (Greg Sabino Mullane)
581 Have psql \d show index information (Greg Sabino Mullane)
582 New psql \dD shows domains (Jonathan Eisler)
583 Allow psql to show rules on views (Paul ?)
584 Fix for psql variable substitution (Tom)
585 Allow psql \d to show temporary table structure (Tom)
586 Allow psql \d to show foreign keys (Rod)
587 Fix \? to honor \pset pager (Bruce)
588 Have psql reports its version number on startup (Tom)
589 Allow \copy to specify column names (Tom)
596 Add $HOME/.pgpass to store host/user password combinations (Alvaro
598 Add PQunescapeBytea() function to libpq (Patrick Welche)
599 Fix for sending large queries over non-blocking connections
601 Fix for libpq using timers on Win9X (David Ford)
602 Allow libpq notify to handle servers with different-length
604 Add libpq PQescapeString() and PQescapeBytea() to Win32 (Bruce)
605 Fix for SSL with non-blocking connections (Jack Bates)
606 Add libpq connection timeout parameter (Denis A Ustimenko)
613 Allow JDBC to compile with JDK 1.4 (Dave)
614 Add JDBC 3 support (Barry)
615 Allows JDBC to set loglevel by adding ?loglevel=X to the connection
617 Add Driver.info() message that prints out the version number (Barry)
618 Add updateable result sets (Raghu Nidagal, Dave)
619 Add support for callable statements (Paul Bethe)
620 Add query cancel capability
621 Add refresh row (Dave)
622 Fix MD5 encryption handling for multibyte servers (Jun Kawai)
623 Add support for prepared statements (Barry)
628 <title>Miscellaneous Interfaces</title>
630 Fixed ECPG bug concerning octal numbers in single quotes (Michael)
631 Move src/interfaces/libpgeasy to http://gborg.postgresql.org (Marc,
633 Improve Python interface (Elliot Lee, Andrew Johnson, Greg Copeland)
634 Add libpgtcl connection close event (Gerhard Hintermayer)
635 Move src/interfaces/libpq++ to http://gborg.postgresql.org (Marc,
637 Move src/interfaces/odbc to http://gborg.postgresql.org (Marc)
638 Move src/interfaces/libpgeasy to http://gborg.postgresql.org (Marc,
640 Move src/interfaces/perl5 to http://gborg.postgresql.org (Marc,
642 Remove src/bin/pgaccess from main tree, now at
643 http://www.pgaccess.org (Bruce)
644 Add pg_on_connection_loss command to libpgtcl (Gerhard Hintermayer,
650 <title>Source Code</title>
652 Fix for parallel make (Peter)
653 AIX fixes for linking Tcl (Andreas Zeugswetter)
654 Allow PL/Perl to build under Cygwin (Jason Tishler)
655 Improve MIPS compiles (Peter, Oliver Elphick)
656 Require Autoconf version 2.53 (Peter)
657 Require readline and zlib by default in configure (Peter)
658 Allow Solaris to use Intimate Shared Memory (ISM), for performance
659 (Scott Brunza, P.J. Josh Rovero)
660 Always enable syslog in compile, remove --enable-syslog option
662 Always enable multibyte in compile, remove --enable-multibyte option
664 Always enable locale in compile, remove --enable-locale option
666 Fix for Win9x DLL creation (Magnus Naeslund)
667 Fix for link() usage by WAL code on Win32, BeOS (Jason Tishler)
668 Add sys/types.h to c.h, remove from main files (Peter, Bruce)
669 Fix AIX hang on SMP machines (Tomoyuki Niijima)
670 AIX SMP hang fix (Tomoyuki Niijima)
671 Fix pre-1970 date handling on newer glibc libraries (Tom)
672 Fix PowerPC SMP locking (Tom)
673 Prevent gcc -ffast-math from being used (Peter, Tom)
674 Bison >= 1.50 now required for developer builds
675 Kerberos 5 support now builds with Heimdal (Peter)
676 Add appendix in the User's Guide which lists SQL features (Thomas)
677 Improve loadable module linking to use RTLD_NOW (Tom)
678 New error levels WARNING, INFO, LOG, DEBUG[1-5] (Bruce)
679 New src/port directory holds replaced libc functions (Peter, Bruce)
680 New pg_namespace system catalog for schemas (Tom)
681 Add pg_class.relnamespace for schemas (Tom)
682 Add pg_type.typnamespace for schemas (Tom)
683 Add pg_proc.pronamespace for schemas (Tom)
684 Restructure aggregates to have pg_proc entries (Tom)
685 System relations now have their own namespace, pg_* test not required
687 Rename TOAST index names to be *_index rather than *_idx (Neil)
688 Add namespaces for operators, opclasses (Tom)
689 Add additional checks to server control file (Thomas)
690 New Polish FAQ (Marcin Mazurek)
691 Add Posix semaphore support (Tom)
692 Document need for reindex (Bruce)
693 Rename some internal identifiers to simplify Win32 compile (Jan,
695 Add documentation on computing disk space (Bruce)
696 Remove KSQO from GUC (Bruce)
697 Fix memory leak in rtree (Kenneth Been)
698 Modify a few error messages for consistency (Bruce)
699 Remove unused system table columns (Peter)
700 Make system columns NOT NULL where appropriate (Tom)
701 Clean up use of sprintf in favor of snprintf() (Neil, Jukka Holappa)
702 Remove OPAQUE and create specific subtypes (Tom)
703 Cleanups in array internal handling (Joe, Tom)
704 Disallow pg_atoi('') (Bruce)
705 Remove parameter wal_files because WAL files are now recycled (Bruce)
706 Add version numbers to heap pages (Tom)
711 <title>Contrib</title>
713 Allow inet arrays in /contrib/array (Neil)
714 Gist fixes (Teodor Sigaev, Neil)
715 Upgrade /contrib/mysql
716 Add /contrib/dbsize which shows table sizes without vacuum (Peter)
717 Add /contrib/intagg, integer aggregator routines (mlw)
718 Improve /contrib/oid2name (Neil, Bruce)
719 Improve /contrib/tsearch (Oleg, Teodor Sigaev)
720 Cleanups of /contrib/rserver (Alexey V. Borzov)
721 Update /contrib/oracle conversion utility (Gilles Darold)
722 Update /contrib/dblink (Joe)
723 Improve options supported by /contrib/vacuumlo (Mario Weilguni)
724 Improvements to /contrib/intarray (Oleg, Teodor Sigaev, Andrey
726 Add /contrib/reindexdb utility (Shaun Thomas)
727 Add indexing to /contrib/isbn_issn (Dan Weston)
728 Add /contrib/dbmirror (Steven Singer)
729 Improve /contrib/pgbench (Neil)
730 Add /contrib/tablefunc table function examples (Joe)
731 Add /contrib/ltree data type for tree structures (Teodor Sigaev,
733 Move /contrib/pg_controldata, pg_resetxlog into main tree (Bruce)
734 Fixes to /contrib/cube (Bruno Wolff)
735 Improve /contrib/fulltextindex (Christopher)
743 <sect1 id="release-7-2-3">
744 <title>Release 7.2.3</title>
747 <title>Release date</title>
748 <simpara>2002-10-01</simpara>
752 This has a variety of fixes from 7.2.2, including fixes to prevent
758 <title>Migration to version 7.2.3</title>
761 A dump/restore is <emphasis>not</emphasis> required for those running
767 <title>Changes</title>
771 Prevent possible compressed transaction log loss (Tom)
772 Prevent non-superuser from increasing most recent vacuum info (Tom)
773 Handle pre-1970 date values in newer versions of glibc (Tom)
774 Fix possible hang during server shutdown
775 Prevent spinlock hangs on SMP PPC machines (Tomoyuki Niijima)
776 Fix pg_dump to properly dump FULL JOIN USING (Tom)
783 <sect1 id="release-7-2-2">
784 <title>Release 7.2.2</title>
787 <title>Release date</title>
788 <simpara>2002-08-23</simpara>
792 This has a variety of fixes from 7.2.1.
797 <title>Migration to version 7.2.2</title>
800 A dump/restore is <emphasis>not</emphasis> required for those running
806 <title>Changes</title>
810 Allow EXECUTE of "CREATE TABLE AS ... SELECT" in PL/pgSQL (Tom)
811 Fix for compressed transaction log id wraparound (Tom)
812 Fix PQescapeBytea/PQunescapeBytea so that they handle bytes > 0x7f (Tatsuo)
813 Fix for psql and pg_dump crashing when invoked with non-existand long
815 Fix crash when invoking geometric operators (Tom)
816 Allow OPEN cursor(args) (Tom)
817 Fix for rtree_gist index build (Teodor)
818 Fix for dumping user-defined aggregates (Tom)
819 Contrib/intarray fixes (Oleg)
820 Fix for complex UNION/EXCEPT/INTERSECT queries using parens (Tom)
821 Fix to pg_convert (Tatsuo)
822 Fix for crash with long DATA strings (Thomes, Neil)
823 Fix for repeat(), lpad(), rpad() and long strings (Neil)
830 <sect1 id="release-7-2-1">
831 <title>Release 7.2.1</title>
834 <title>Release date</title>
835 <simpara>2002-03-21</simpara>
839 This has a variety of fixes from 7.2.
844 <title>Migration to version 7.2.1</title>
847 A dump/restore is <emphasis>not</emphasis> required for those running
853 <title>Changes</title>
857 Ensure that sequence counters do not go backwards after a crash (Tom)
858 Fix pgaccess kanji-coversion key binding (Tatsuo)
859 Optimizer improvements (Tom)
860 cash I/O improvements (Tom)
862 Compile fix for missing AuthBlockSig (Heiko)
863 Additional time zones and time zone fixes (Thomas)
864 Allow psql \connect to handle mixed case database and user names (Tom)
865 Return proper OID on command completion even with ON INSERT rules (Tom)
866 Allow COPY FROM to use 8-bit DELIMITERS (Tatsuo)
867 Fix bug in extract/date_part for milliseconds/microseconds (Tatsuo)
868 Improve handling of multiple UNIONs with different lengths (Tom)
869 contrib/btree_gist improvements (Teodor Sigaev)
870 contrib/tsearch dictionary improvements, see README.tsearch for
871 an additional installation step (Thomas T. Thai, Teodor Sigaev)
872 Fix for array subscripts handling (Tom)
873 Allow EXECUTE of "CREATE TABLE AS ... SELECT" in PL/pgSQL (Tom)
880 <sect1 id="release-7-2">
881 <title>Release 7.2</title>
884 <title>Release date</title>
885 <simpara>2002-02-04</simpara>
889 <title>Overview</title>
892 This release improves <productname>PostgreSQL</> for use in
893 high-volume applications.
897 Major changes in this release:
904 Vacuuming no longer locks tables, thus allowing normal user
905 access during the vacuum. A new <command>VACUUM FULL</>
906 command does old-style vacuum by locking the table and
907 shrinking the on-disk copy of the table.
913 <term>Transactions</term>
916 There is no longer a problem with installations that exceed
917 four billion transactions.
926 OIDs are now optional. Users can now create tables without
927 OIDs for cases where OID usage is excessive.
933 <term>Optimizer</term>
936 The system now computes histogram column statistics during
937 <command>ANALYZE</>, allowing much better optimizer choices.
943 <term>Security</term>
946 A new MD5 encryption option allows more secure storage and
947 transfer of passwords. A new Unix-domain socket
948 authentication option is available on Linux and BSD systems.
954 <term>Statistics</term>
957 Administrators can use the new table access statistics module
958 to get fine-grained information about table and index usage.
964 <term>Internationalization</term>
967 Program and library messages can now be displayed in several
977 <title>Migration to version 7.2</title>
980 A dump/restore using <command>pg_dump</command> is required for
981 those wishing to migrate data from any previous release.
985 Observe the following incompatibilities:
990 The semantics of the <command>VACUUM</command> command have
991 changed in this release. You may wish to update your
992 maintenance procedures accordingly.
998 In this release, comparisons using <literal>= NULL</literal>
999 will always return false (or NULL, more precisely). Previous
1000 releases automatically transformed this syntax to <literal>IS
1001 NULL</literal>. The old behavior can be re-enabled using a
1002 <filename>postgresql.conf</filename> parameter.
1008 The <filename>pg_hba.conf</> and <filename>pg_ident.conf</>
1009 configuration is now only reloaded after receiving a
1010 <systemitem>SIGHUP</> signal, not with each connection.
1016 The function <filename>octet_length()</> now returns the uncompressed data length.
1022 The date/time value <literal>'current'</literal> is no longer
1023 available. You will need to rewrite your applications.
1029 The <literal>timestamp()</literal>, <literal>time()</literal>,
1030 and <literal>interval()</literal> functions are no longer
1031 available. Instead of <literal>timestamp()</literal>, use
1032 <literal>timestamp 'string'</literal> or <literal>CAST</literal>.
1042 The <literal>SELECT ... LIMIT #,#</literal> syntax will be removed
1043 in the next release. You should change your queries to use
1044 separate LIMIT and OFFSET clauses, e.g. <literal>LIMIT 10 OFFSET
1050 <title>Changes</title>
1053 <title>Server Operation</title>
1055 Create temporary files in a separate directory (Bruce)
1056 Delete orphaned temporary files on postmaster startup (Bruce)
1057 Added unique indexes to some system tables (Tom)
1058 System table operator reorganization (Oleg Bartunov, Teodor Sigaev, Tom)
1059 Renamed pg_log to pg_clog (Tom)
1060 Enable SIGTERM, SIGQUIT to kill backends (Jan)
1061 Removed compile-time limit on number of backends (Tom)
1062 Better cleanup for semaphore resource failure (Tatsuo, Tom)
1063 Allow safe transaction ID wraparound (Tom)
1064 Removed OIDs from some system tables (Tom)
1065 Removed "triggered data change violation" error check (Tom)
1066 SPI portal creation of prepared/saved plans (Jan)
1067 Allow SPI column functions to work for system columns (Tom)
1068 Long value compression improvement (Tom)
1069 Statistics collector for table, index access (Jan)
1070 Truncate extra-long sequence names to a reasonable value (Tom)
1071 Measure transaction times in milliseconds (Thomas)
1072 Fix TID sequential scans (Hiroshi)
1073 Superuser ID now fixed at 1 (Peter E)
1074 New pg_ctl "reload" option (Tom)
1079 <title>Performance</title>
1081 Optimizer improvements (Tom)
1082 New histogram column statistics for optimizer (Tom)
1083 Reuse write-ahead log files rather than discarding them (Tom)
1084 Cache improvements (Tom)
1085 IS NULL, IS NOT NULL optimizer improvement (Tom)
1086 Improve lock manager to reduce lock contention (Tom)
1087 Keep relcache entries for index access support functions (Tom)
1088 Allow better selectivity with NaN and infinities in NUMERIC (Tom)
1089 R-tree performance improvements (Kenneth Been)
1090 B-tree splits more efficient (Tom)
1095 <title>Privileges</title>
1097 Change UPDATE, DELETE permissions to be distinct (Peter E)
1098 New REFERENCES, TRIGGER privileges (Peter E)
1099 Allow GRANT/REVOKE to/from more than one user at a time (Peter E)
1100 New has_table_privilege() function (Joe Conway)
1101 Allow non-superuser to vacuum database (Tom)
1102 New SET SESSION AUTHORIZATION command (Peter E)
1103 Fix bug in privilege modifications on newly created tables (Tom)
1104 Disallow access to pg_statistic for non-superuser, add user-accessible views (Tom)
1109 <title>Client Authentication</title>
1111 Fork postmaster before doing authentication to prevent hangs (Peter E)
1112 Add ident authentication over Unix domain sockets on Linux, *BSD (Helge Bahmann, Oliver Elphick, Teodor Sigaev, Bruce)
1113 Add a password authentication method that uses MD5 encryption (Bruce)
1114 Allow encryption of stored passwords using MD5 (Bruce)
1115 PAM authentication (Dominic J. Eidson)
1116 Load pg_hba.conf and pg_ident.conf only on startup and SIGHUP (Bruce)
1121 <title>Server Configuration</title>
1123 Interpretation of some time zone abbreviations as Australian rather than North American now settable at run time (Bruce)
1124 New parameter to set default transaction isolation level (Peter E)
1125 New parameter to enable conversion of "expr = NULL" into "expr IS NULL", off by default (Peter E)
1126 New parameter to control memory usage by VACUUM (Tom)
1127 New parameter to set client authentication timeout (Tom)
1128 New parameter to set maximum number of open files (Tom)
1133 <title>Queries</title>
1135 Statements added by INSERT rules now execute after the INSERT (Jan)
1136 Prevent unadorned relation names in target list (Bruce)
1137 NULLs now sort after all normal values in ORDER BY (Tom)
1138 New IS UNKNOWN, IS NOT UNKNOWN Boolean tests (Tom)
1139 New SHARE UPDATE EXCLUSIVE lock mode (Tom)
1140 New EXPLAIN ANALYZE command that shows run times and row counts (Martijn van Oosterhout)
1141 Fix problem with LIMIT and subqueries (Tom)
1142 Fix for LIMIT, DISTINCT ON pushed into subqueries (Tom)
1143 Fix nested EXCEPT/INTERSECT (Tom)
1148 <title>Schema Manipulation</title>
1150 Fix SERIAL in temporary tables (Bruce)
1151 Allow temporary sequences (Bruce)
1152 Sequences now use int8 internally (Tom)
1153 New SERIAL8 creates int8 columns with sequences, default still SERIAL4 (Tom)
1154 Make OIDs optional using WITHOUT OIDS (Tom)
1155 Add %TYPE syntax to CREATE TYPE (Ian Lance Taylor)
1156 Add ALTER TABLE / DROP CONSTRAINT for CHECK constraints (Christopher Kings-Lynne)
1157 New CREATE OR REPLACE FUNCTION to alter existing function (preserving the function OID) (Gavin Sherry)
1158 Add ALTER TABLE / ADD [ UNIQUE | PRIMARY ] (Christopher Kings-Lynne)
1159 Allow column renaming in views
1160 Make ALTER TABLE / RENAME COLUMN update column names of indexes (Brent Verner)
1161 Fix for ALTER TABLE / ADD CONSTRAINT ... CHECK with inherited tables (Stephan Szabo)
1162 ALTER TABLE RENAME update foreign-key trigger arguments correctly (Brent Verner)
1163 DROP AGGREGATE and COMMENT ON AGGREGATE now accept an aggtype (Tom)
1164 Add automatic return type data casting for SQL functions (Tom)
1165 Allow GiST indexes to handle NULLs and multikey indexes (Oleg Bartunov, Teodor Sigaev, Tom)
1166 Enable partial indexes (Martijn van Oosterhout)
1171 <title>Utility Commands</title>
1173 Add RESET ALL, SHOW ALL (Marko Kreen)
1174 CREATE/ALTER USER/GROUP now allow options in any order (Vince)
1175 Add LOCK A, B, C functionality (Neil Padgett)
1176 New ENCRYPTED/UNENCRYPTED option to CREATE/ALTER USER (Bruce)
1177 New light-weight VACUUM does not lock table; old semantics are available as VACUUM FULL (Tom)
1178 Disable COPY TO/FROM on views (Bruce)
1179 COPY DELIMITERS string must be exactly one character (Tom)
1180 VACUUM warning about index tuples fewer than heap now only appears when appropriate (Martijn van Oosterhout)
1181 Fix permission checks for CREATE INDEX (Tom)
1182 Disallow inappropriate use of CREATE/DROP INDEX/TRIGGER/VIEW (Tom)
1187 <title>Data Types and Functions</title>
1189 SUM(), AVG(), COUNT() now uses int8 internally for speed (Tom)
1190 Add convert(), convert2() (Tatsuo)
1191 New function bit_length() (Peter E)
1192 Make the "n" in CHAR(n)/VARCHAR(n) represents letters, not bytes (Tatsuo)
1193 CHAR(), VARCHAR() now reject strings that are too long (Peter E)
1194 BIT VARYING now rejects bit strings that are too long (Peter E)
1195 BIT now rejects bit strings that do not match declared size (Peter E)
1196 INET, CIDR text conversion functions (Alex Pilosov)
1197 INET, CIDR operators << and <<= indexable (Alex Pilosov)
1198 Bytea \### now requires valid three digit octal number
1199 Bytea comparison improvements, now supports =, <>, >, >=, <, and <=
1200 Bytea now supports B-tree indexes
1201 Bytea now supports LIKE, LIKE...ESCAPE, NOT LIKE, NOT LIKE...ESCAPE
1202 Bytea now supports concatenation
1203 New bytea functions: position, substring, trim, btrim, and length
1204 New encode() function mode, "escaped", converts minimally escaped bytea to/from text
1205 Add pg_database_encoding_max_length() (Tatsuo)
1206 Add pg_client_encoding() function (Tatsuo)
1207 now() returns time with millisecond precision (Thomas)
1208 New TIMESTAMP WITHOUT TIMEZONE data type (Thomas)
1209 Add ISO date/time specification with "T", yyyy-mm-ddThh:mm:ss (Thomas)
1210 New xid/int comparison functions (Hiroshi)
1211 Add precision to TIME, TIMESTAMP, and INVERVAL data types (Thomas)
1212 Modify type coercion logic to attempt binary-compatible functions first (Tom)
1213 New encode() function installed by default (Marko Kreen)
1214 Improved to_*() conversion functions (Karel Zak)
1215 Optimize LIKE/ILIKE when using single-byte encodings (Tatsuo)
1216 New functions in contrib/pgcrypto: crypt(), hmac(), encrypt(), gen_salt() (Marko Kreen)
1217 Correct description of translate() function (Bruce)
1218 Add INTERVAL argument for SET TIME ZONE (Thomas)
1219 Add INTERVAL YEAR TO MONTH (etc.) syntax (Thomas)
1220 Optimize length functions when using single-byte encodings (Tatsuo)
1221 Fix path_inter, path_distance, path_length, dist_ppath to handle closed paths (Curtis Barrett, Tom)
1222 octet_length(text) now returns non-compressed length (Tatsuo, Bruce)
1223 Handle "July" full name in date/time literals (Greg Sabino Mullane)
1224 Some datatype() function calls now evaluated differently
1225 Add support for Julian and ISO time specifications (Thomas)
1230 <title>Internationalization</title>
1232 National language support in psql, pg_dump, libpq, and server (Peter E)
1233 Message translations in Chinese (simplified, traditional), Czech, French, German, Hungarian, Russian, Swedish (Peter E, Serguei A. Mokhov, Karel Zak, Weiping He, Zhenbang Wei, Kovacs Zoltan)
1234 Make trim, ltrim, rtrim, btrim, lpad, rpad, translate multibyte aware (Tatsuo)
1235 Add LATIN5,6,7,8,9,10 support (Tatsuo)
1236 Add ISO 8859-5,6,7,8 support (Tatsuo)
1237 Correct LATIN5 to mean ISO-8859-9, not ISO-8859-5 (Tatsuo)
1238 Make mic2ascii() non-ASCII aware (Tatsuo)
1239 Reject invalid multibyte character sequences (Tatsuo)
1244 <title><application>PL/pgSQL</></title>
1246 Now uses portals for SELECT loops, allowing huge result sets (Jan)
1247 CURSOR and REFCURSOR support (Jan)
1248 Can now return open cursors (Jan)
1249 Add ELSEIF (Klaus Reger)
1250 Improve PL/pgSQL error reporting, including location of error (Tom)
1251 Allow IS or FOR key words in cursor declaration, for compatibility (Bruce)
1252 Fix for SELECT ... FOR UPDATE (Tom)
1253 Fix for PERFORM returning multiple rows (Tom)
1254 Make PL/pgSQL use the server's type coercion code (Tom)
1255 Memory leak fix (Jan, Tom)
1256 Make trailing semicolon optional (Tom)
1261 <title>PL/Perl</title>
1263 New untrusted PL/Perl (Alex Pilosov)
1264 PL/Perl is now built on some platforms even if libperl is not shared (Peter E)
1269 <title>PL/Tcl</title>
1271 Now reports errorInfo (Vsevolod Lobko)
1272 Add spi_lastoid function (bob@redivi.com)
1277 <title>PL/Python</title>
1279 ...is new (Andrew Bosma)
1284 <title><application>Psql</></title>
1286 \d displays indexes in unique, primary groupings (Christopher Kings-Lynne)
1287 Allow trailing semicolons in backslash commands (Greg Sabino Mullane)
1288 Read password from /dev/tty if possible
1289 Force new password prompt when changing user and database (Tatsuo, Tom)
1290 Format the correct number of columns for Unicode (Patrice)
1295 <title><application>Libpq</></title>
1297 New function PQescapeString() to escape quotes in command strings (Florian Weimer)
1298 New function PQescapeBytea() escapes binary strings for use as SQL string literals
1305 Return OID of INSERT (Ken K)
1306 Handle more data types (Ken K)
1307 Handle single quotes and newlines in strings (Ken K)
1308 Handle NULL variables (Ken K)
1309 Fix for time zone handling (Barry Lind)
1310 Improved Druid support
1311 Allow eight-bit characters with non-multibyte server (Barry Lind)
1312 Support BIT, BINARY types (Ned Wolpert)
1313 Reduce memory usage (Michael Stephens, Dave Cramer)
1314 Update DatabaseMetaData (Peter E)
1315 Add DatabaseMetaData.getCatalogs() (Peter E)
1316 Encoding fixes (Anders Bengtsson)
1317 Get/setCatalog methods (Jason Davies)
1318 DatabaseMetaData.getColumns() now returns column defaults (Jason Davies)
1319 DatabaseMetaData.getColumns() performance improvement (Jeroen van Vianen)
1320 Some JDBC1 and JDBC2 merging (Anders Bengtsson)
1321 Transaction performance improvements (Barry Lind)
1322 Array fixes (Greg Zoller)
1324 Fix batch processing (Rene Pijlman)
1325 ExecSQL method reorganization (Anders Bengtsson)
1326 GetColumn() fixes (Jeroen van Vianen)
1327 Fix isWriteable() function (Rene Pijlman)
1328 Improved passage of JDBC2 conformance tests (Rene Pijlman)
1329 Add bytea type capability (Barry Lind)
1330 Add isNullable() (Rene Pijlman)
1331 JDBC date/time test suite fixes (Liam Stewart)
1332 Fix for SELECT 'id' AS xxx FROM table (Dave Cramer)
1333 Fix DatabaseMetaData to show precision properly (Mark Lillywhite)
1334 New getImported/getExported keys (Jason Davies)
1335 MD5 password encryption support (Jeremy Wohl)
1336 Fix to actually use type cache (Ned Wolpert)
1343 Remove query size limit (Hiroshi)
1344 Remove text field size limit (Hiroshi)
1345 Fix for SQLPrimaryKeys in multibyte mode (Hiroshi)
1346 Allow ODBC procedure calls (Hiroshi)
1347 Improve boolean handing (Aidan Mountford)
1348 Most configuration options on setable via DSN (Hiroshi)
1349 Multibyte, performance fixes (Hiroshi)
1350 Allow driver to be used with iODBC or unixODBC (Peter E)
1351 MD5 password encryption support (Bruce)
1352 Add more compatibility functions to odbc.sql (Peter E)
1357 <title><application>ECPG</></title>
1359 EXECUTE ... INTO implemented (Christof Petig)
1360 Multiple row descriptor support (e.g. CARDINALITY) (Christof Petig)
1361 Fix for GRANT parameters (Lee Kindness)
1362 Fix INITIALLY DEFERRED bug
1363 Various bug fixes (Michael, Christof Petig)
1364 Auto allocation for indicator variable arrays (int *ind_p=NULL)
1365 Auto allocation for string arrays (char **foo_pp=NULL)
1366 ECPGfree_auto_mem fixed
1367 All function names with external linkage are now prefixed by ECPG
1368 Fixes for arrays of structures (Michael)
1373 <title>Misc. Interfaces</title>
1375 Python fix fetchone() (Gerhard Haring)
1376 Use UTF, Unicode in Tcl where appropriate (Vsevolod Lobko, Reinhard Max)
1377 Add Tcl COPY TO/FROM (ljb)
1378 Prevent output of default index op class in pg_dump (Tom)
1379 Fix libpgeasy memory leak (Bruce)
1384 <title>Build and Install</title>
1386 Configure, dynamic loader, and shared library fixes (Peter E)
1387 Fixes in QNX 4 port (Bernd Tegge)
1388 Fixes in Cygwin and Win32 ports (Jason Tishler, Gerhard Haring, Dmitry Yurtaev, Darko Prenosil, Mikhail Terekhov)
1389 Fix for Win32 socket communication failures (Magnus, Mikhail Terekhov)
1390 Hurd compile fix (Oliver Elphick)
1391 BeOS fixes (Cyril Velter)
1392 Remove configure --enable-unicode-conversion, now enabled by multibyte (Tatsuo)
1393 AIX fixes (Tatsuo, Andreas)
1394 Fix parallel make (Peter E)
1395 Install SQL language manual pages into OS-specific directories (Peter E)
1396 Rename config.h to pg_config.h (Peter E)
1397 Reorganize installation layout of header files (Peter E)
1402 <title>Source Code</title>
1404 Remove SEP_CHAR (Bruce)
1406 Merge GUC and command line handling (Marko Kreen)
1407 Remove EXTEND INDEX (Martijn van Oosterhout, Tom)
1408 New pgjindent utility to indent java code (Bruce)
1409 Remove define of true/false when compiling under C++ (Leandro Fanzone, Tom)
1410 pgindent fixes (Bruce, Tom)
1411 Replace strcasecmp() with strcmp() where appropriate (Peter E)
1412 Dynahash portability improvements (Tom)
1413 Add 'volatile' usage in spinlock structures
1414 Improve signal handling logic (Tom)
1419 <title>Contrib</title>
1421 New contrib/rtree_gist (Oleg Bartunov, Teodor Sigaev)
1422 New contrib/tsearch full-text indexing (Oleg, Teodor Sigaev)
1423 Add contrib/dblink for remote database access (Joe Conway)
1424 contrib/ora2pg Oracle conversion utility (Gilles Darold)
1425 contrib/xml XML conversion utility (John Gray)
1426 contrib/fulltextindex fixes (Christopher Kings-Lynne)
1427 New contrib/fuzzystrmatch with levenshtein and metaphone, soundex merged (Joe Conway)
1428 Add contrib/intarray boolean queries, binary search, fixes (Oleg Bartunov)
1429 New pg_upgrade utility (Bruce)
1430 Add new pg_resetxlog options (Bruce, Tom)
1437 <sect1 id="release-7-1-3">
1438 <title>Release 7.1.3</title>
1441 <title>Release date</title>
1442 <simpara>2001-08-15</simpara>
1446 <title>Migration to version 7.1.3</title>
1449 A dump/restore is <emphasis>not</emphasis> required for those running
1455 <title>Changes</title>
1459 Remove unused WAL segements of large transactions (Tom)
1460 Multiaction rule fix (Tom)
1461 PL/pgSQL memory allocation fix (Jan)
1462 VACUUM buffer fix (Tom)
1463 Regression test fixes (Tom)
1464 pg_dump fixes for GRANT/REVOKE/comments on views, user-defined types (Tom)
1465 Fix subselects with DISTINCT ON or LIMIT (Tom)
1467 Disable COPY TO/FROM a view (Tom)
1468 Cygwin build (Jason Tishler)
1475 <sect1 id="release-7-1-2">
1476 <title>Release 7.1.2</title>
1479 <title>Release date</title>
1480 <simpara>2001-05-11</simpara>
1484 This has one fix from 7.1.1.
1489 <title>Migration to version 7.1.2</title>
1492 A dump/restore is <emphasis>not</emphasis> required for those running
1498 <title>Changes</title>
1502 Fix PL/pgSQL SELECTs when returning no rows
1503 Fix for psql backslash core dump
1504 Referential integrity permission fix
1513 <sect1 id="release-7-1-1">
1514 <title>Release 7.1.1</title>
1517 <title>Release date</title>
1518 <simpara>2001-05-05</simpara>
1522 This has a variety of fixes from 7.1.
1527 <title>Migration to version 7.1.1</title>
1530 A dump/restore is <emphasis>not</emphasis> required for those running
1536 <title>Changes</title>
1540 Fix for numeric MODULO operator (Tom)
1541 pg_dump fixes (Philip)
1542 pg_dump can dump 7.0 databases (Philip)
1543 readline 4.2 fixes (Peter E)
1545 AIX, MSWIN, VAX, N32K fixes (Tom)
1546 Multibytes fixes (Tom)
1547 Unicode fixes (Tatsuo)
1548 Optimizer improvements (Tom)
1549 Fix for whole tuples in functions (Tom)
1550 Fix for pg_ctl and option strings with spaces (Peter E)
1551 ODBC fixes (Hiroshi)
1552 EXTRACT can now take string argument (Thomas)
1553 Python fixes (Darcy)
1560 <sect1 id="release-7-1">
1561 <title>Release 7.1</title>
1564 <title>Release date</title>
1565 <simpara>2001-04-13</simpara>
1569 This release focuses on removing limitations that have existed in the
1570 PostgreSQL code for many years.
1574 Major changes in this release:
1579 Write-ahead Log (WAL)
1583 To maintain database consistency in case
1584 of an operating system crash, previous releases of PostgreSQL have
1585 forced all data modifications to disk before each transaction commit.
1586 With WAL, only one log file must be flushed to disk, greatly improving
1587 performance. If you have been using -F in previous releases to disable
1588 disk flushes, you may want to consider discontinuing its use.
1599 TOAST - Previous releases had a compiled-in row length limit,
1600 typically 8k - 32k. This limit made storage of long text fields
1601 difficult. With TOAST, long rows of any length can be stored with good
1613 We now support outer joins. The UNION/NOT IN
1614 workaround for outer joins is no longer required. We use the SQL92
1626 The previous C function manager did not
1627 handle null values properly, nor did it support 64-bit <acronym>CPU</acronym>'s (Alpha). The new
1628 function manager does. You can continue using your old custom
1629 functions, but you may want to rewrite them in the future to use the new
1630 function manager call interface.
1641 A large number of complex queries that were
1642 unsupported in previous releases now work. Many combinations of views,
1643 aggregates, UNION, LIMIT, cursors, subqueries, and inherited tables
1644 now work properly. Inherited tables are now accessed by default.
1645 Subqueries in FROM are now supported.
1654 <title>Migration to version 7.1</title>
1657 A dump/restore using pg_dump is required for those wishing to migrate
1658 data from any previous release.
1663 <title>Changes</title>
1669 Many multibyte/Unicode/locale fixes (Tatsuo and others)
1670 More reliable ALTER TABLE RENAME (Tom)
1671 Kerberos V fixes (David Wragg)
1672 Fix for INSERT INTO...SELECT where targetlist has subqueries (Tom)
1673 Prompt username/password on standard error (Bruce)
1674 Large objects inv_read/inv_write fixes (Tom)
1675 Fixes for to_char(), to_date(), to_ascii(), and to_timestamp() (Karel,
1677 Prevent query expressions from leaking memory (Tom)
1678 Allow UPDATE of arrays elements (Tom)
1679 Wake up lock waiters during cancel (Hiroshi)
1680 Fix rare cursor crash when using hash join (Tom)
1681 Fix for DROP TABLE/INDEX in rolled-back transaction (Hiroshi)
1682 Fix psql crash from \l+ if MULTIBYTE enabled (Peter E)
1683 Fix truncation of rule names during CREATE VIEW (Ross Reedstrom)
1684 Fix PL/perl (Alex Kapranoff)
1685 Disallow LOCK on views (Mark Hollomon)
1686 Disallow INSERT/UPDATE/DELETE on views (Mark Hollomon)
1687 Disallow DROP RULE, CREATE INDEX, TRUNCATE on views (Mark Hollomon)
1688 Allow PL/pgSQL accept non-ASCII identifiers (Tatsuo)
1689 Allow views to proper handle GROUP BY, aggregates, DISTINCT (Tom)
1690 Fix rare failure with TRUNCATE command (Tom)
1691 Allow UNION/INTERSECT/EXCEPT to be used with ALL, subqueries, views,
1692 DISTINCT, ORDER BY, SELECT...INTO (Tom)
1693 Fix parser failures during aborted transactions (Tom)
1694 Allow temporary relations to properly clean up indexes (Bruce)
1695 Fix VACUUM problem with moving rows in same page (Tom)
1696 Modify pg_dump to better handle user-defined items in template1 (Philip)
1697 Allow LIMIT in VIEW (Tom)
1698 Require cursor FETCH to honor LIMIT (Tom)
1699 Allow PRIMARY/FOREIGN Key definitions on inherited columns (Stephan)
1700 Allow ORDER BY, LIMIT in sub-selects (Tom)
1701 Allow UNION in CREATE RULE (Tom)
1702 Make ALTER/DROP TABLE rollback-able (Vadim, Tom)
1703 Store initdb collation in pg_control so collation cannot be changed (Tom)
1704 Fix INSERT...SELECT with rules (Tom)
1705 Fix FOR UPDATE inside views and subselects (Tom)
1706 Fix OVERLAPS operators conform to SQL92 spec regarding NULLs (Tom)
1707 Fix lpad() and rpad() to handle length less than input string (Tom)
1708 Fix use of NOTIFY in some rules (Tom)
1709 Overhaul btree code (Tom)
1710 Fix NOT NULL use in Pl/pgSQL variables (Tom)
1711 Overhaul GIST code (Oleg)
1712 Fix CLUSTER to preserve constraints and column default (Tom)
1713 Improved deadlock detection handling (Tom)
1714 Allow multiple SERIAL columns in a table (Tom)
1715 Prevent occasional index corruption (Vadim)
1719 Add OUTER JOINs (Tom)
1720 Function manager overhaul (Tom)
1721 Allow ALTER TABLE RENAME on indexes (Tom)
1722 Improve CLUSTER (Tom)
1723 Improve ps status display for more platforms (Peter E, Marc)
1724 Improve CREATE FUNCTION failure message (Ross)
1725 JDBC improvements (Peter, Travis Bauer, Christopher Cain, William Webber,
1727 Grand Unified Configuration scheme/GUC. Many options can now be set in
1728 data/postgresql.conf, postmaster/postgres flags, or SET commands (Peter E)
1729 Improved handling of file descriptor cache (Tom)
1730 New warning code about auto-created table alias entries (Bruce)
1731 Overhaul initdb process (Tom, Peter E)
1732 Overhaul of inherited tables; inherited tables now accessed by default;
1733 new ONLY keyword prevents it (Chris Bitmead, Tom)
1734 ODBC cleanups/improvements (Nick Gorham, Stephan Szabo, Zoltan Kovacs,
1736 Allow renaming of temp tables (Tom)
1737 Overhaul memory manager contexts (Tom)
1738 pg_dumpall uses CREATE USER or CREATE GROUP rather using COPY (Peter E)
1739 Overhaul pg_dump (Philip Warner)
1740 Allow pg_hba.conf secondary password file to specify only username (Peter E)
1741 Allow TEMPORARY or TEMP keyword when creating temporary tables (Bruce)
1742 New memory leak checker (Karel)
1743 New SET SESSION CHARACTERISTICS (Thomas)
1744 Allow nested block comments (Thomas)
1745 Add WITHOUT TIME ZONE type qualifier (Thomas)
1746 New ALTER TABLE ADD CONSTRAINT (Stephan)
1747 Use NUMERIC accumulators for INTEGER aggregates (Tom)
1748 Overhaul aggregate code (Tom)
1749 New VARIANCE and STDDEV() aggregates
1750 Improve dependency ordering of pg_dump (Philip)
1751 New pg_restore command (Philip)
1752 New pg_dump tar output option (Philip)
1753 New pg_dump of large objects (Philip)
1754 New ESCAPE option to LIKE (Thomas)
1755 New case-insensitive LIKE - ILIKE (Thomas)
1756 Allow functional indexes to use binary-compatible type (Tom)
1757 Allow SQL functions to be used in more contexts (Tom)
1758 New pg_config utility (Peter E)
1759 New PL/pgSQL EXECUTE command which allows dynamic SQL and utility statements
1761 New PL/pgSQL GET DIAGNOSTICS statement for SPI value access (Jan)
1762 New quote_identifiers() and quote_literal() functions (Jan)
1763 New ALTER TABLE table OWNER TO user command (Mark Hollomon)
1764 Allow subselects in FROM, i.e. FROM (SELECT ...) [AS] alias (Tom)
1765 Update PyGreSQL to version 3.1 (D'Arcy)
1766 Store tables as files named by OID (Vadim)
1767 New SQL function setval(seq,val,bool) for use in pg_dump (Philip)
1768 Require DROP VIEW to remove views, no DROP TABLE (Mark)
1769 Allow DROP VIEW view1, view2 (Mark)
1770 Allow multiple objects in DROP INDEX, DROP RULE, and DROP TYPE (Tom)
1771 Allow automatic conversion to/from Unicode (Tatsuo, Eiji)
1772 New /contrib/pgcrypto hashing functions (Marko Kreen)
1773 New pg_dumpall --globals-only option (Peter E)
1774 New CHECKPOINT command for WAL which creates new WAL log file (Vadim)
1775 New AT TIME ZONE syntax (Thomas)
1776 Allow location of Unix domain socket to be configurable (David J. MacKenzie)
1777 Allow postmaster to listen on a specific IP address (David J. MacKenzie)
1778 Allow socket path name to be specified in hostname by using leading slash
1779 (David J. MacKenzie)
1780 Allow CREATE DATABASE to specify template database (Tom)
1781 New utility to convert MySQL schema dumps to SQL92 and PostgreSQL (Thomas)
1782 New /contrib/rserv replication toolkit (Vadim)
1783 New file format for COPY BINARY (Tom)
1784 New /contrib/oid2name to map numeric files to table names (B Palmer)
1785 New "idle in transaction" ps status message (Marc)
1786 Update to pgaccess 0.98.7 (Constantin Teodorescu)
1787 pg_ctl now defaults to -w (wait) on shutdown, new -l (log) option
1788 Add rudimentary dependency checking to pg_dump (Philip)
1792 Fix INET/CIDR type ordering and add new functions (Tom)
1793 Make OID behave as an unsigned type (Tom)
1794 Allow BIGINT as synonym for INT8 (Peter E)
1795 New int2 and int8 comparison operators (Tom)
1796 New BIT and BIT VARYING types (Adriaan Joubert, Tom, Peter E)
1797 CHAR() no longer faster than VARCHAR() because of TOAST (Tom)
1798 New GIST seg/cube examples (Gene Selkov)
1799 Improved round(numeric) handling (Tom)
1800 Fix CIDR output formatting (Tom)
1801 New CIDR abbrev() function (Tom)
1805 Write-Ahead Log (WAL) to provide crash recovery with less performance
1807 ANALYZE stage of VACUUM no longer exclusively locks table (Bruce)
1808 Reduced file seeks (Denis Perchine)
1809 Improve BTREE code for duplicate keys (Tom)
1810 Store all large objects in a single table (Denis Perchine, Tom)
1811 Improve memory allocation performance (Karel, Tom)
1815 New function manager call conventions (Tom)
1816 SGI portability fixes (David Kaelbling)
1817 New configure --enable-syslog option (Peter E)
1818 New BSDI README (Bruce)
1819 configure script moved to top level, not /src (Peter E)
1820 Makefile/configuration/compilation overhaul (Peter E)
1821 New configure --with-python option (Peter E)
1822 Solaris cleanups (Peter E)
1823 Overhaul /contrib Makefiles (Karel)
1824 New OpenSSL configuration option (Magnus, Peter E)
1826 QNX fixes (Maurizio)
1827 New heap_open(), heap_openr() API (Tom)
1828 Remove colon and semi-colon operators (Thomas)
1829 New pg_class.relkind value for views (Mark Hollomon)
1830 Rename ichar() to chr() (Karel)
1831 New documentation for btrim(), ascii(), chr(), repeat() (Karel)
1832 Fixes for NT/Cygwin (Pete Forman)
1833 AIX port fixes (Andreas)
1834 New BeOS port (David Reid, Cyril Velter)
1835 Add proofreader's changes to docs (Addison-Wesley, Bruce)
1836 New Alpha spinlock code (Adriaan Joubert, Compaq)
1837 UnixWare port overhaul (Peter E)
1838 New Darwin/MacOS X port (Peter Bierman, Bruce Hartzler)
1839 New FreeBSD Alpha port (Alfred)
1840 Overhaul shared memory segments (Tom)
1841 Add IBM S/390 support (Neale Ferguson)
1842 Moved macmanuf to /contrib (Larry Rosenman)
1843 Syslog improvements (Larry Rosenman)
1844 New template0 database that contains no user additions (Tom)
1845 New /contrib/cube and /contrib/seg GIST sample code (Gene Selkov)
1846 Allow NetBSD's libedit instead of readline (Peter)
1847 Improved assembly language source code format (Bruce)
1848 New contrib/pg_logger
1849 New --template option to createdb
1850 New contrib/pg_control utility (Oliver)
1851 New FreeBSD tools ipc_check, start-scripts/freebsd
1858 <sect1 id="release-7-0-3">
1859 <title>Release 7.0.3</title>
1862 <title>Release date</title>
1863 <simpara>2000-11-11</simpara>
1867 This has a variety of fixes from 7.0.2.
1872 <title>Migration to version 7.0.3</title>
1875 A dump/restore is <emphasis>not</emphasis> required for those running
1881 <title>Changes</title>
1886 Large object fix (Tom)
1887 Fix lean in COPY WITH OIDS leak (Tom)
1888 Fix backwards-index-scan (Tom)
1889 Fix SELECT ... FOR UPDATE so it checks for duplicate keys (Hiroshi)
1890 Add --enable-syslog to configure (Marc)
1891 Fix abort transaction at backend exit in rare cases (Tom)
1892 Fix for psql \l+ when multibyte enabled (Tatsuo)
1893 Allow PL/pgSQL to accept non ascii identifiers (Tatsuo)
1894 Make vacuum always flush buffers (Tom)
1895 Fix to allow cancel while waiting for a lock (Hiroshi)
1896 Fix for memory aloocation problem in user authentication code (Tom)
1897 Remove bogus use of int4out() (Tom)
1898 Fixes for multiple subqueries in COALESCE or BETWEEN (Tom)
1899 Fix for failure of triggers on heap open in certain cases (Jeroen van
1901 Fix for erroneous selectivity of not-equals (Tom)
1902 Fix for erroneous use of strcmp() (Tom)
1903 Fix for bug where storage manager accesses items beyond end of file
1905 Fix to include kernel errno message in all smgr elog messages (Tom)
1906 Fix for '.' not in PATH at build time (SL Baur)
1907 Fix for out-of-file-descriptors error (Tom)
1908 Fix to make pg_dump dump 'iscachable' flag for functions (Tom)
1909 Fix for subselect in targetlist of Append node (Tom)
1910 Fix for mergejoin plans (Tom)
1911 Fix TRUNCATE failure on relations with indexes (Tom)
1912 Avoid database-wide restart on write error (Hiroshi)
1913 Fix nodeMaterial to honor chgParam by recomputing its output (Tom)
1914 Fix VACUUM problem with moving chain of update tuples when source and
1915 destination of a tuple lie on the same page (Tom)
1916 Fix user.c CommandCounterIncrement (Tom)
1917 Fix for AM/PM boundary problem in to_char() (Karel Zak)
1918 Fix TIME aggregate handling (Tom)
1919 Fix to_char() to avoid coredump on NULL input (Tom)
1921 Fix for inserting/copying longer multibyte strings into char() data
1923 Fix for crash of backend, on abort (Tom)
1930 <sect1 id="release-7-0-2">
1931 <title>Release 7.0.2</title>
1934 <title>Release date</title>
1935 <simpara>2000-06-05</simpara>
1939 This is a repackaging of 7.0.1 with added documentation.
1944 <title>Migration to version 7.0.2</title>
1947 A dump/restore is <emphasis>not</emphasis> required for those running
1953 <title>Changes</title>
1957 Added documentation to tarball.
1964 <sect1 id="release-7-0-1">
1965 <title>Release 7.0.1</title>
1968 <title>Release date</title>
1969 <simpara>2000-06-01</simpara>
1973 This is a cleanup release for 7.0.
1977 <title>Migration to version 7.0.1</title>
1980 A dump/restore is <emphasis>not</emphasis> required for those running
1986 <title>Changes</title>
1990 Fix many CLUSTER failures (Tom)
1991 Allow ALTER TABLE RENAME works on indexes (Tom)
1992 Fix plpgsql to handle datetime->timestamp and timespan->interval (Bruce)
1993 New configure --with-setproctitle switch to use setproctitle() (Marc, Bruce)
1994 Fix the off by one errors in ResultSet from 6.5.3, and more.
1995 jdbc ResultSet fixes (Joseph Shraibman)
1996 optimizer tunings (Tom)
1997 Fix create user for pgaccess
1998 Fix for UNLISTEN failure
1999 IRIX fixes (David Kaelbling)
2000 QNX fixes (Andreas Kardos)
2001 Reduce COPY IN lock level (Tom)
2002 Change libpqeasy to use PQconnectdb() style parameters (Bruce)
2003 Fix pg_dump to handle OID indexes (Tom)
2004 Fix small memory leak (Tom)
2005 Solaris fix for createdb/dropdb (Tatsuo)
2006 Fix for non-blocking connections (Alfred Perlstein)
2007 Fix improper recovery after RENAME TABLE failures (Tom)
2008 Copy pg_ident.conf.sample into /lib directory in install (Bruce)
2009 Add SJIS UDC (NEC selection IBM kanji) support (Eiji Tokuya)
2010 Fix too long syslog message (Tatsuo)
2011 Fix problem with quoted indexes that are too long (Tom)
2012 JDBC ResultSet.getTimestamp() fix (Gregory Krasnow & Floyd Marinescu)
2013 ecpg changes (Michael)
2019 <sect1 id="release-7-0">
2020 <title>Release 7.0</title>
2023 <title>Release date</title>
2024 <simpara>2000-05-08</simpara>
2028 This release contains improvements in many areas, demonstrating
2029 the continued growth of <productname>PostgreSQL</productname>.
2030 There are more improvements and fixes in 7.0 than in any previous
2031 release. The developers have confidence that this is the best
2032 release yet; we do our best to put out only solid releases, and
2033 this one is no exception.
2037 Major changes in this release:
2046 Foreign keys are now implemented, with the exception of PARTIAL MATCH
2047 foreign keys. Many users have been asking for this feature, and we are
2048 pleased to offer it.
2059 Continuing on work started a year ago, the optimizer has been
2060 improved, allowing better query plan selection and faster performance
2061 with less memory usage.
2068 Updated <application>psql</application>
2072 <application>psql</application>, our interactive terminal monitor, has been
2073 updated with a variety of new features. See the <application>psql</application> manual page for details.
2084 SQL92 join syntax is now supported, though only as
2085 <literal>INNER JOIN</> for this release. <literal>JOIN</>,
2086 <literal>NATURAL JOIN</>, <literal>JOIN</>/<literal>USING</>,
2087 and <literal>JOIN</>/<literal>ON</> are available, as are
2088 column correlation names.
2100 In 7.1 or 7.2, we plan to have outer joins, storage for very long rows, and a
2101 write-ahead logging system.
2110 <title>Migration to version 7.0</title>
2113 A dump/restore using <application>pg_dump</application>
2114 is required for those wishing to migrate data from any
2115 previous release of <productname>PostgreSQL</productname>.
2116 For those upgrading from 6.5.*, you may instead use
2117 <application>pg_upgrade</application> to upgrade to this
2118 release; however, a full dump/reload installation is always the
2119 most robust method for upgrades.
2123 Interface and compatibility issues to consider for the new
2129 The date/time types <type>datetime</type> and
2130 <type>timespan</type> have been superseded by the
2131 SQL92-defined types <type>timestamp</type> and
2132 <type>interval</type>. Although there has been some effort to
2133 ease the transition by allowing
2134 <productname>PostgreSQL</productname> to recognize
2135 the deprecated type names and translate them to the new type
2136 names, this mechanism may not be completely transparent to
2137 your existing application.
2143 The optimizer has been substantially improved in the area of
2144 query cost estimation. In some cases, this will result in
2145 decreased query times as the optimizer makes a better choice
2146 for the preferred plan. However, in a small number of cases,
2147 usually involving pathological distributions of data, your
2148 query times may go up. If you are dealing with large amounts
2149 of data, you may want to check your queries to verify
2156 The <acronym>JDBC</acronym> and <acronym>ODBC</acronym>
2157 interfaces have been upgraded and extended.
2163 The string function <function>CHAR_LENGTH</function> is now a
2164 native function. Previous versions translated this into a call
2165 to <function>LENGTH</function>, which could result in
2166 ambiguity with other types implementing
2167 <function>LENGTH</function> such as the geometric types.
2176 <title>Changes</title>
2182 Prevent function calls exceeding maximum number of arguments (Tom)
2183 Improve CASE construct (Tom)
2184 Fix SELECT coalesce(f1,0) FROM int4_tbl GROUP BY f1 (Tom)
2185 Fix SELECT sentence.words[0] FROM sentence GROUP BY sentence.words[0] (Tom)
2186 Fix GROUP BY scan bug (Tom)
2187 Improvements in SQL grammar processing (Tom)
2188 Fix for views involved in INSERT ... SELECT ... (Tom)
2189 Fix for SELECT a/2, a/2 FROM test_missing_target GROUP BY a/2 (Tom)
2190 Fix for subselects in INSERT ... SELECT (Tom)
2191 Prevent INSERT ... SELECT ... ORDER BY (Tom)
2192 Fixes for relations greater than 2GB, including vacuum
2193 Improve propagating system table changes to other backends (Tom)
2194 Improve propagating user table changes to other backends (Tom)
2195 Fix handling of temp tables in complex situations (Bruce, Tom)
2196 Allow table locking at table open, improving concurrent reliability (Tom)
2197 Properly quote sequence names in pg_dump (Ross J. Reedstrom)
2198 Prevent DROP DATABASE while others accessing
2199 Prevent any rows from being returned by GROUP BY if no rows processed (Tom)
2200 Fix SELECT COUNT(1) FROM table WHERE ...' if no rows matching WHERE (Tom)
2201 Fix pg_upgrade so it works for MVCC (Tom)
2202 Fix for SELECT ... WHERE x IN (SELECT ... HAVING SUM(x) > 1) (Tom)
2203 Fix for "f1 datetime DEFAULT 'now'" (Tom)
2204 Fix problems with CURRENT_DATE used in DEFAULT (Tom)
2205 Allow comment-only lines, and ;;; lines too. (Tom)
2206 Improve recovery after failed disk writes, disk full (Hiroshi)
2207 Fix cases where table is mentioned in FROM but not joined (Tom)
2208 Allow HAVING clause without aggregate functions (Tom)
2209 Fix for "--" comment and no trailing newline, as seen in perl interface
2210 Improve pg_dump failure error reports (Bruce)
2211 Allow sorts and hashes to exceed 2GB file sizes (Tom)
2212 Fix for pg_dump dumping of inherited rules (Tom)
2213 Fix for NULL handling comparisons (Tom)
2214 Fix inconsistent state caused by failed CREATE/DROP commands (Hiroshi)
2215 Fix for dbname with dash
2216 Prevent DROP INDEX from interfering with other backends (Tom)
2217 Fix file descriptor leak in verify_password()
2218 Fix for "Unable to identify an operator =$" problem
2219 Fix ODBC so no segfault if CommLog and Debug enabled (Dirk Niggemann)
2220 Fix for recursive exit call (Massimo)
2221 Fix for extra-long timezones (Jeroen van Vianen)
2222 Make pg_dump preserve primary key information (Peter E)
2223 Prevent databases with single quotes (Peter E)
2224 Prevent DROP DATABASE inside transaction (Peter E)
2225 ecpg memory leak fixes (Stephen Birch)
2226 Fix for SELECT null::text, SELECT int4fac(null) and SELECT 2 + (null) (Tom)
2227 Y2K timestamp fix (Massimo)
2228 Fix for VACUUM 'HEAP_MOVED_IN was not expected' errors (Tom)
2229 Fix for views with tables/columns containing spaces (Tom)
2230 Prevent permissions on indexes (Peter E)
2231 Fix for spinlock stuck problem when error is generated (Hiroshi)
2232 Fix ipcclean on Linux
2233 Fix handling of NULL constraint conditions (Tom)
2234 Fix memory leak in odbc driver (Nick Gorham)
2235 Fix for permission check on UNION tables (Tom)
2236 Fix to allow SELECT 'a' LIKE 'a' (Tom)
2237 Fix for SELECT 1 + NULL (Tom)
2239 Fix log() on numeric type (Tom)
2240 Deprecate ':' and ';' operators
2241 Allow vacuum of temporary tables
2242 Disallow inherited columns with the same name as new columns
2243 Recover or force failure when disk space is exhausted (Hiroshi)
2244 Fix INSERT INTO ... SELECT with AS columns matching result columns
2245 Fix INSERT ... SELECT ... GROUP BY groups by target columns not source columns (Tom)
2246 Fix CREATE TABLE test (a char(5) DEFAULT text '', b int4) with INSERT (Tom)
2247 Fix UNION with LIMIT
2248 Fix CREATE TABLE x AS SELECT 1 UNION SELECT 2
2249 Fix CREATE TABLE test(col char(2) DEFAULT user)
2250 Fix mismatched types in CREATE TABLE ... DEFAULT
2251 Fix SELECT * FROM pg_class where oid in (0,-1)
2252 Fix SELECT COUNT('asdf') FROM pg_class WHERE oid=12
2253 Prevent user who can create databases can modifying pg_database table (Peter E)
2254 Fix btree to give a useful elog when key > 1/2 (page - overhead) (Tom)
2255 Fix INSERT of 0.0 into DECIMAL(4,4) field (Tom)
2259 New CLI interface include file sqlcli.h, based on SQL3/SQL98
2260 Remove all limits on query length, row length limit still exists (Tom)
2261 Update jdbc protocol to 2.0 (Jens Glaser <email>jens@jens.de</email>)
2262 Add TRUNCATE command to quickly truncate relation (Mike Mascari)
2263 Fix to give super user and createdb user proper update catalog rights (Peter E)
2264 Allow ecpg bool variables to have NULL values (Christof)
2265 Issue ecpg error if NULL value for variable with no NULL indicator (Christof)
2266 Allow ^C to cancel COPY command (Massimo)
2267 Add SET FSYNC and SHOW PG_OPTIONS commands(Massimo)
2268 Function name overloading for dynamically-loaded C functions (Frankpitt)
2269 Add CmdTuples() to libpq++(Vince)
2270 New CREATE CONSTRAINT TRIGGER and SET CONSTRAINTS commands(Jan)
2271 Allow CREATE FUNCTION/WITH clause to be used for all language types
2272 configure --enable-debug adds -g (Peter E)
2273 configure --disable-debug removes -g (Peter E)
2274 Allow more complex default expressions (Tom)
2275 First real FOREIGN KEY constraint trigger functionality (Jan)
2276 Add FOREIGN KEY ... MATCH FULL ... ON DELETE CASCADE (Jan)
2277 Add FOREIGN KEY ... MATCH <unspecified> referential actions (Don Baccus)
2278 Allow WHERE restriction on ctid (physical heap location) (Hiroshi)
2279 Move pginterface from contrib to interface directory, rename to pgeasy (Bruce)
2280 Change pgeasy connectdb() parameter ordering (Bruce)
2281 Require SELECT DISTINCT target list to have all ORDER BY columns (Tom)
2282 Add Oracle's COMMENT ON command (Mike Mascari <email>mascarim@yahoo.com</email>)
2283 libpq's PQsetNoticeProcessor function now returns previous hook(Peter E)
2284 Prevent PQsetNoticeProcessor from being set to NULL (Peter E)
2285 Make USING in COPY optional (Bruce)
2286 Allow subselects in the target list (Tom)
2287 Allow subselects on the left side of comparison operators (Tom)
2288 New parallel regression test (Jan)
2289 Change backend-side COPY to write files with permissions 644 not 666 (Tom)
2290 Force permissions on PGDATA directory to be secure, even if it exists (Tom)
2291 Added psql LASTOID variable to return last inserted oid (Peter E)
2292 Allow concurrent vacuum and remove pg_vlock vacuum lock file (Tom)
2293 Add permissions check for vacuum (Peter E)
2294 New libpq functions to allow asynchronous connections: PQconnectStart(),
2295 PQconnectPoll(), PQresetStart(), PQresetPoll(), PQsetenvStart(),
2296 PQsetenvPoll(), PQsetenvAbort (Ewan Mellor)
2297 New libpq PQsetenv() function (Ewan Mellor)
2298 create/alter user extension (Peter E)
2299 New postmaster.pid and postmaster.opts under $PGDATA (Tatsuo)
2300 New scripts for create/drop user/db (Peter E)
2301 Major psql overhaul (Peter E)
2302 Add const to libpq interface (Peter E)
2303 New libpq function PQoidValue (Peter E)
2304 Show specific non-aggregate causing problem with GROUP BY (Tom)
2305 Make changes to pg_shadow recreate pg_pwd file (Peter E)
2306 Add aggregate(DISTINCT ...) (Tom)
2307 Allow flag to control COPY input/output of NULLs (Peter E)
2308 Make postgres user have a password by default (Peter E)
2309 Add CREATE/ALTER/DROP GROUP (Peter E)
2310 All administration scripts now support --long options (Peter E, Karel)
2311 Vacuumdb script now supports --all option (Peter E)
2312 ecpg new portable FETCH syntax
2313 Add ecpg EXEC SQL IFDEF, EXEC SQL IFNDEF, EXEC SQL ELSE, EXEC SQL ELIF
2314 and EXEC SQL ENDIF directives
2315 Add pg_ctl script to control backend start-up (Tatsuo)
2316 Add postmaster.opts.default file to store start-up flags (Tatsuo)
2317 Allow --with-mb=SQL_ASCII
2318 Increase maximum number of index keys to 16 (Bruce)
2319 Increase maximum number of function arguments to 16 (Bruce)
2320 Allow configuration of maximum number of index keys and arguments (Bruce)
2321 Allow unprivileged users to change their passwords (Peter E)
2322 Password authentication enabled; required for new users (Peter E)
2323 Disallow dropping a user who owns a database (Peter E)
2324 Change initdb option --with-mb to --enable-multibyte
2325 Add option for initdb to prompts for superuser password (Peter E)
2326 Allow complex type casts like col::numeric(9,2) and col::int2::float8 (Tom)
2327 Updated user interfaces on initdb, initlocation, pg_dump, ipcclean (Peter E)
2328 New pg_char_to_encoding() and pg_encoding_to_char() functions (Tatsuo)
2329 Libpq non-blocking mode (Alfred Perlstein)
2330 Improve conversion of types in casts that don't specify a length
2331 New plperl internal programming language (Mark Hollomon)
2332 Allow COPY IN to read file that do not end with a newline (Tom)
2333 Indicate when long identifiers are truncated (Tom)
2334 Allow aggregates to use type equivalency (Peter E)
2335 Add Oracle's to_char(), to_date(), to_datetime(), to_timestamp(), to_number()
2336 conversion functions (Karel Zak <zakkr@zf.jcu.cz>)
2337 Add SELECT DISTINCT ON (expr [, expr ...]) targetlist ... (Tom)
2338 Check to be sure ORDER BY is compatible with the DISTINCT operation (Tom)
2339 Add NUMERIC and int8 types to ODBC
2340 Improve EXPLAIN results for Append, Group, Agg, Unique (Tom)
2341 Add ALTER TABLE ... ADD FOREIGN KEY (Stephan Szabo)
2342 Allow SELECT .. FOR UPDATE in PL/pgSQL (Hiroshi)
2343 Enable backward sequential scan even after reaching EOF (Hiroshi)
2344 Add btree indexing of boolean values, >= and <= (Don Baccus)
2345 Print current line number when COPY FROM fails (Massimo)
2346 Recognize POSIX time zone e.g. "PST+8" and "GMT-8" (Thomas)
2347 Add DEC as synonym for DECIMAL (Thomas)
2348 Add SESSION_USER as SQL92 keyword, same as CURRENT_USER (Thomas)
2349 Implement SQL92 column aliases (aka correlation names) (Thomas)
2350 Implement SQL92 join syntax (Thomas)
2351 Make INTERVAL reserved word allowed as a column identifier (Thomas)
2352 Implement REINDEX command (Hiroshi)
2353 Accept ALL in aggregate function SUM(ALL col) (Tom)
2354 Prevent GROUP BY from using column aliases (Tom)
2355 New psql \encoding option (Tatsuo)
2356 Allow PQrequestCancel() to terminate when in waiting-for-lock state (Hiroshi)
2357 Allow negation of a negative number in all cases
2358 Add ecpg descriptors (Christof, Michael)
2359 Allow CREATE VIEW v AS SELECT f1::char(8) FROM tbl
2360 Allow casts with length, like foo::char(8)
2361 New libpq functions PQsetClientEncoding(), PQclientEncoding() (Tatsuo)
2362 Add support for SJIS user defined characters (Tatsuo)
2363 Larger views/rules supported
2364 Make libpq's PQconndefaults() thread-safe (Tom)
2365 Disable // as comment to be ANSI conforming, should use -- (Tom)
2366 Allow column aliases on views CREATE VIEW name (collist)
2367 Fixes for views with subqueries (Tom)
2368 Allow UPDATE table SET fld = (SELECT ...) (Tom)
2369 SET command options no longer require quotes
2370 Update pgaccess to 0.98.6
2371 New SET SEED command
2372 New pg_options.sample file
2373 New SET FSYNC command (Massimo)
2374 Allow pg_descriptions when creating tables
2375 Allow pg_descriptions when creating types, columns, and functions
2376 Allow psql \copy to allow delimiters (Peter E)
2377 Allow psql to print nulls as distinct from "" [null] (Peter E)
2381 Many array fixes (Tom)
2382 Allow bare column names to be subscripted as arrays (Tom)
2383 Improve type casting of int and float constants (Tom)
2384 Cleanups for int8 inputs, range checking, and type conversion (Tom)
2385 Fix for SELECT timespan('21:11:26'::time) (Tom)
2386 netmask('x.x.x.x/0') is 255.255.255.255 instead of 0.0.0.0 (Oleg Sharoiko)
2387 Add btree index on NUMERIC (Jan)
2388 Perl fix for large objects containing NUL characters (Douglas Thomson)
2389 ODBC fix for for large objects (free)
2390 Fix indexing of cidr data type
2391 Fix for Ethernet MAC addresses (macaddr type) comparisons
2392 Fix for date/time types when overflows happened in computations (Tom)
2393 Allow array on int8 (Peter E)
2394 Fix for rounding/overflow of NUMERIC type, like NUMERIC(4,4) (Tom)
2395 Allow NUMERIC arrays
2396 Fix bugs in NUMERIC ceil() and floor() functions (Tom)
2397 Make char_length()/octet_length including trailing blanks (Tom)
2398 Made abstime/reltime use int4 instead of time_t (Peter E)
2399 New lztext data type for compressed text fields
2400 Revise code to handle coercion of int and float constants (Tom)
2401 Start at new code to implement a BIT and BIT VARYING type (Adriaan Joubert)
2402 NUMERIC now accepts scientific notation (Tom)
2403 NUMERIC to int4 rounds (Tom)
2404 Convert float4/8 to NUMERIC properly (Tom)
2405 Allow type conversion with NUMERIC (Thomas)
2406 Make ISO date style (2000-02-16 09:33) the default (Thomas)
2407 Add NATIONAL CHAR [ VARYING ] (Thomas)
2408 Allow NUMERIC round and trunc to accept negative scales (Tom)
2409 New TIME WITH TIME ZONE type (Thomas)
2410 Add MAX()/MIN() on time type (Thomas)
2411 Add abs(), mod(), fac() for int8 (Thomas)
2412 Rename functions to round(), sqrt(), cbrt(), pow() for float8 (Thomas)
2413 Add transcendental math functions (e.g. sin(), acos()) for float8 (Thomas)
2414 Add exp() and ln() for NUMERIC type
2415 Rename NUMERIC power() to pow() (Thomas)
2416 Improved TRANSLATE() function (Edwin Ramirez, Tom)
2417 Allow X=-Y operators (Tom)
2418 Allow SELECT float8(COUNT(*))/(SELECT COUNT(*) FROM t) FROM t GROUP BY f1; (Tom)
2419 Allow LOCALE to use indexes in regular expression searches (Tom)
2420 Allow creation of functional indexes to use default types
2424 Prevent exponential space consumption with many AND's and OR's (Tom)
2425 Collect attribute selectivity values for system columns (Tom)
2426 Reduce memory usage of aggregates (Tom)
2427 Fix for LIKE optimization to use indexes with multibyte encodings (Tom)
2428 Fix r-tree index optimizer selectivity (Thomas)
2429 Improve optimizer selectivity computations and functions (Tom)
2430 Optimize btree searching for cases where many equal keys exist (Tom)
2431 Enable fast LIKE index processing only if index present (Tom)
2432 Re-use free space on index pages with duplicates (Tom)
2433 Improve hash join processing (Tom)
2434 Prevent descending sort if result is already sorted(Hiroshi)
2435 Allow commuting of index scan query qualifications (Tom)
2436 Prefer index scans in cases where ORDER BY/GROUP BY is required (Tom)
2437 Allocate large memory requests in fix-sized chunks for performance (Tom)
2438 Fix vacuum's performance by reducing memory allocation requests (Tom)
2439 Implement constant-expression simplification (Bernard Frankpitt, Tom)
2440 Use secondary columns to be used to determine start of index scan (Hiroshi)
2441 Prevent quadruple use of disk space when doing internal sorting (Tom)
2442 Faster sorting by calling fewer functions (Tom)
2443 Create system indexes to match all system caches (Bruce, Hiroshi)
2444 Make system caches use system indexes (Bruce)
2445 Make all system indexes unique (Bruce)
2446 Improve pg_statistics management for VACUUM speed improvement (Tom)
2447 Flush backend cache less frequently (Tom, Hiroshi)
2448 COPY now reuses previous memory allocation, improving performance (Tom)
2449 Improve optimization cost estimation (Tom)
2450 Improve optimizer estimate of range queries x > lowbound AND x < highbound (Tom)
2451 Use DNF instead of CNF where appropriate (Tom, Taral)
2452 Further cleanup for OR-of-AND WHERE-clauses (Tom)
2453 Make use of index in OR clauses (x = 1 AND y = 2) OR (x = 2 AND y = 4) (Tom)
2454 Smarter optimizer computations for random index page access (Tom)
2455 New SET variable to control optimizer costs (Tom)
2456 Optimizer queries based on LIMIT, OFFSET, and EXISTS qualifications (Tom)
2457 Reduce optimizer internal housekeeping of join paths for speedup (Tom)
2458 Major subquery speedup (Tom)
2459 Fewer fsync writes when fsync is not disabled (Tom)
2460 Improved LIKE optimizer estimates (Tom)
2461 Prevent fsync in SELECT-only queries (Vadim)
2462 Make index creation use psort code, because it is now faster (Tom)
2463 Allow creation of sort temp tables > 1 Gig
2467 Fix for linux PPC compile
2468 New generic expression-tree-walker subroutine (Tom)
2469 Change form() to varargform() to prevent portability problems
2470 Improved range checking for large integers on Alphas
2471 Clean up #include in /include directory (Bruce)
2472 Add scripts for checking includes (Bruce)
2473 Remove un-needed #include's from *.c files (Bruce)
2474 Change #include's to use <> and "" as appropriate (Bruce)
2475 Enable WIN32 compilation of libpq
2476 Alpha spinlock fix from Uncle George <email>gatgul@voicenet.com</email>
2477 Overhaul of optimizer data structures (Tom)
2478 Fix to cygipc library (Yutaka Tanida)
2479 Allow pgsql to work on newer Cygwin snapshots (Dan)
2480 New catalog version number (Tom)
2482 Rename heap_replace to heap_update
2483 Update for QNX (Dr. Andreas Kardos)
2484 New platform-specific regression handling (Tom)
2485 Rename oid8 -> oidvector and int28 -> int2vector (Bruce)
2486 Included all yacc and lex files into the distribution (Peter E.)
2487 Remove lextest, no longer needed (Peter E)
2488 Fix for libpq and psql on Win32 (Magnus)
2489 Internally change datetime and timespan into timestamp and interval (Thomas)
2490 Fix for plpgsql on BSD/OS
2491 Add SQL_ASCII test case to the regression test (Tatsuo)
2492 configure --with-mb now deprecated (Tatsuo)
2494 NetBSD fixes (Johnny C. Lam <email>lamj@stat.cmu.edu</email>)
2495 Fixes for Alpha compiles
2496 New multibyte encodings
2502 <sect1 id="release-6-5-3">
2503 <title>Release 6.5.3</title>
2506 <title>Release date</title>
2507 <simpara>1999-10-13</simpara>
2511 This is basically a cleanup release for 6.5.2. We have added a new
2512 <application>PgAccess</> that was missing in 6.5.2, and installed an NT-specific fix.
2517 <title>Migration to version 6.5.3</title>
2520 A dump/restore is <emphasis>not</emphasis> required for those running
2525 <title>Changes</title>
2529 Updated version of pgaccess 0.98
2531 Fix dumping rules on inherited tables
2538 <sect1 id="release-6-5-2">
2539 <title>Release 6.5.2</title>
2542 <title>Release date</title>
2543 <simpara>1999-09-15</simpara>
2547 This is basically a cleanup release for 6.5.1. We have fixed a variety of
2548 problems reported by 6.5.1 users.
2553 <title>Migration to version 6.5.2</title>
2556 A dump/restore is <emphasis>not</emphasis> required for those running
2562 <title>Changes</title>
2566 subselect+CASE fixes(Tom)
2567 Add SHLIB_LINK setting for solaris_i386 and solaris_sparc ports(Daren Sefcik)
2568 Fixes for CASE in WHERE join clauses(Tom)
2569 Fix BTScan abort(Tom)
2570 Repair the check for redundant UNIQUE and PRIMARY KEY indexes(Thomas)
2571 Improve it so that it checks for multicolumn constraints(Thomas)
2572 Fix for Win32 making problem with MB enabled(Hiroki Kataoka)
2573 Allow BSD yacc and bison to compile pl code(Bruce)
2574 Fix SET NAMES working
2576 Fix vacuum's memory consumption(Hiroshi,Tatsuo)
2577 Reduce the total memory consumption of vacuum(Tom)
2578 Fix for timestamp(datetime)
2579 Rule deparsing bugfixes(Tom)
2580 Fix quoting problems in mkMakefile.tcldefs.sh.in and mkMakefile.tkdefs.sh.in(Tom)
2581 This is to re-use space on index pages freed by vacuum(Vadim)
2582 document -x for pg_dump(Bruce)
2583 Fix for unary operators in rule deparser(Tom)
2584 Comment out FileUnlink of excess segments during mdtruncate()(Tom)
2585 IRIX linking fix from Yu Cao >yucao@falcon.kla-tencor.com<
2586 Repair logic error in LIKE: should not return LIKE_ABORT
2587 when reach end of pattern before end of text(Tom)
2588 Repair incorrect cleanup of heap memory allocation during transaction abort(Tom)
2589 Updated version of pgaccess 0.98
2595 <sect1 id="release-6-5-1">
2596 <title>Release 6.5.1</title>
2599 <title>Release date</title>
2600 <simpara>1999-07-15</simpara>
2604 This is basically a cleanup release for 6.5. We have fixed a variety of
2605 problems reported by 6.5 users.
2609 <title>Migration to version 6.5.1</title>
2612 A dump/restore is <emphasis>not</emphasis> required for those running
2618 <title>Changes</title>
2623 Portability fixes for linux_ppc, IRIX, linux_alpha, OpenBSD, alpha
2624 Remove QUERY_LIMIT, use SELECT...LIMIT
2625 Fix for EXPLAIN on inheritance(Tom)
2626 Patch to allow vacuum on multisegment tables(Hiroshi)
2627 R-Tree optimizer selectivity fix(Tom)
2628 ACL file descriptor leak fix(Atsushi Ogawa)
2629 New expresssion subtree code(Tom)
2630 Avoid disk writes for read-only transactions(Vadim)
2631 Fix for removal of temp tables if last transaction was aborted(Bruce)
2632 Fix to prevent too large tuple from being created(Bruce)
2634 Allow port numbers 32k - 64k(Bruce)
2635 Add ^ precidence(Bruce)
2636 Rename sort files called pg_temp to pg_sorttemp(Bruce)
2637 Fix for microseconds in time values(Tom)
2638 Tutorial source cleanup
2640 Fix for sorting of NULL's in some cases(Tom)
2641 Shared library dependencies fixed (Tom)
2642 Fixed glitches affecting GROUP BY in subselects(Tom)
2643 Fix some compiler warnings (Tomoaki Nishiyama)
2644 Add Win1250 (Czech) support (Pavel Behal)
2650 <sect1 id="release-6-5">
2651 <title>Release 6.5</title>
2654 <title>Release date</title>
2655 <simpara>1999-06-09</simpara>
2659 This release marks a major step in the development team's mastery of the source
2660 code we inherited from Berkeley. You will see we are now easily adding
2661 major features, thanks to the increasing size and experience of our
2662 world-wide development team.
2666 Here is a brief summary of the more notable changes:
2671 Multiversion concurrency control(MVCC)
2675 This removes our old
2676 table-level locking, and replaces it with a locking system that is
2677 superior to most commercial database systems. In a traditional system,
2678 each row that is modified is locked until committed, preventing reads by
2679 other users. MVCC uses the natural multiversion nature of PostgreSQL
2680 to allow readers to continue reading consistent data during writer
2681 activity. Writers continue to use the compact pg_log transaction
2682 system. This is all performed without having to allocate a lock for
2683 every row like traditional database systems. So, basically, we no
2684 longer are restricted by simple table-level locking;
2685 we have something better than row-level locking.
2692 Hot backups from <application>pg_dump</application>
2696 <application>pg_dump</application> takes advantage of the new
2697 MVCC features to give a consistent database dump/backup while
2698 the database stays online and available for queries.
2709 We now have a true numeric data type, with
2710 user-specified precision.
2721 Temporary tables are guaranteed to have unique names
2722 within a database session, and are destroyed on session exit.
2733 We now have CASE, INTERSECT, and EXCEPT statement
2734 support. We have new LIMIT/OFFSET, SET TRANSACTION ISOLATION LEVEL,
2735 SELECT ... FOR UPDATE, and an improved LOCK TABLE command.
2746 We continue to speed up PostgreSQL, thanks to the variety of
2747 talents within our team. We have sped up memory allocation,
2748 optimization, table joins, and row transfer routines.
2759 We continue to expand our port list, this time including
2760 <systemitem class="osname">Windows NT</>/<systemitem>ix86</> and <systemitem class="osname">NetBSD</>/<systemitem>arm32</>.
2771 Most interfaces have new versions, and existing functionality
2783 New and updated material is present throughout the
2784 documentation. New <acronym>FAQ</acronym>s have been
2785 contributed for <systemitem class="osname">SGI</> and <systemitem class="osname">AIX</> platforms.
2786 The <citetitle>Tutorial</citetitle> has introductory information
2787 on <acronym>SQL</acronym> from Stefan Simkovics.
2788 For the <citetitle>User's Guide</citetitle>, there are
2789 reference pages covering the postmaster and more utility
2790 programs, and a new appendix
2791 contains details on date/time behavior.
2792 The <citetitle>Administrator's Guide</citetitle> has a new
2793 chapter on troubleshooting from Tom Lane.
2794 And the <citetitle>Programmer's Guide</citetitle> has a
2795 description of query processing, also from Stefan, and details
2796 on obtaining the <productname>PostgreSQL</productname> source
2797 tree via anonymous <productname>CVS</productname> and
2798 <productname>CVSup</productname>.
2806 <title>Migration to version 6.5</title>
2809 A dump/restore using <application>pg_dump</application>
2810 is required for those wishing to migrate data from any
2811 previous release of <productname>PostgreSQL</productname>.
2812 <application>pg_upgrade</application> can <emphasis>not</emphasis>
2813 be used to upgrade to this release because the on-disk structure
2814 of the tables has changed compared to previous releases.
2818 The new Multiversion Concurrency Control (MVCC) features can
2819 give somewhat different behaviors in multiuser
2820 environments. <emphasis>Read and understand the following section
2821 to ensure that your existing applications will give you the
2822 behavior you need.</emphasis>
2826 <title>Multiversion Concurrency Control</title>
2829 Because readers in 6.5 don't lock data, regardless of transaction
2830 isolation level, data read by one transaction can be overwritten by
2831 another. In other words, if a row is returned by
2832 <command>SELECT</command> it doesn't mean that this row really exists
2833 at the time it is returned (i.e. sometime after the statement or
2834 transaction began) nor that the row is protected from being deleted or
2835 updated by concurrent transactions before the current transaction does
2836 a commit or rollback.
2840 To ensure the actual existence of a row and protect it against
2841 concurrent updates one must use <command>SELECT FOR UPDATE</command> or
2842 an appropriate <command>LOCK TABLE</command> statement. This should be
2843 taken into account when porting applications from previous releases of
2844 <productname>PostgreSQL</productname> and other environments.
2848 Keep the above in mind if you are using
2849 <filename>contrib/refint.*</filename> triggers for
2850 referential integrity. Additional techniques are required now. One way is
2851 to use <command>LOCK parent_table IN SHARE ROW EXCLUSIVE MODE</command>
2852 command if a transaction is going to update/delete a primary key and
2853 use <command>LOCK parent_table IN SHARE MODE</command> command if a
2854 transaction is going to update/insert a foreign key.
2858 Note that if you run a transaction in SERIALIZABLE mode then you must
2859 execute the <command>LOCK</command> commands above before execution of any
2860 <acronym>DML</acronym> statement
2861 (<command>SELECT/INSERT/DELETE/UPDATE/FETCH/COPY_TO</command>) in the
2868 These inconveniences will disappear in the future
2869 when the ability to read dirty
2870 (uncommitted) data (regardless of isolation level) and true referential
2871 integrity will be implemented.
2877 <title>Changes</title>
2883 Fix text<->float8 and text<->float4 conversion functions(Thomas)
2884 Fix for creating tables with mixed-case constraints(Billy)
2885 Change exp()/pow() behavior to generate error on underflow/overflow(Jan)
2886 Fix bug in pg_dump -z
2887 Memory overrun cleanups(Tatsuo)
2888 Fix for lo_import crash(Tatsuo)
2889 Adjust handling of data type names to suppress double quotes(Thomas)
2890 Use type coercion for matching columns and DEFAULT(Thomas)
2891 Fix deadlock so it only checks once after one second of sleep(Bruce)
2892 Fixes for aggregates and PL/pgsql(Hiroshi)
2893 Fix for subquery crash(Vadim)
2894 Fix for libpq function PQfnumber and case-insensitive names(Bahman Rafatjoo)
2895 Fix for large object write-in-middle, no extra block, memory consumption(Tatsuo)
2896 Fix for pg_dump -d or -D and quote special characters in INSERT
2897 Repair serious problems with dynahash(Tom)
2898 Fix INET/CIDR portability problems
2899 Fix problem with selectivity error in ALTER TABLE ADD COLUMN(Bruce)
2900 Fix executor so mergejoin of different column types works(Tom)
2901 Fix for Alpha OR selectivity bug
2902 Fix OR index selectivity problem(Bruce)
2903 Fix so \d shows proper length for char()/varchar()(Ryan)
2904 Fix tutorial code(Clark)
2905 Improve destroyuser checking(Oliver)
2906 Fix for Kerberos(Rodney McDuff)
2907 Fix for dropping database while dirty buffers(Bruce)
2908 Fix so sequence nextval() can be case-sensitive(Bruce)
2910 Drop buffers before destroying database files(Bruce)
2911 Fix case where executor evaluates functions twice(Tatsuo)
2912 Allow sequence nextval actions to be case-sensitive(Bruce)
2913 Fix optimizer indexing not working for negative numbers(Bruce)
2914 Fix for memory leak in executor with fjIsNull
2915 Fix for aggregate memory leaks(Erik Riedel)
2916 Allow username containing a dash GRANT permissions
2917 Cleanup of NULL in inet types
2918 Clean up system table bugs(Tom)
2919 Fix problems of PAGER and \? command(Masaaki Sakaida)
2920 Reduce default multisegment file size limit to 1GB(Peter)
2921 Fix for dumping of CREATE OPERATOR(Tom)
2922 Fix for backward scanning of cursors(Hiroshi Inoue)
2923 Fix for COPY FROM STDIN when using \i(Tom)
2924 Fix for subselect is compared inside an expression(Jan)
2925 Fix handling of error reporting while returning rows(Tom)
2926 Fix problems with reference to array types(Tom,Jan)
2927 Prevent UPDATE SET oid(Jan)
2928 Fix pg_dump so -t option can handle case-sensitive tablenames
2929 Fixes for GROUP BY in special cases(Tom, Jan)
2930 Fix for memory leak in failed queries(Tom)
2931 DEFAULT now supports mixed-case identifiers(Tom)
2932 Fix for multisegment uses of DROP/RENAME table, indexes(Ole Gjerde)
2933 Disable use of pg_dump with both -o and -d options(Bruce)
2934 Allow pg_dump to properly dump GROUP permissions(Bruce)
2935 Fix GROUP BY in INSERT INTO table SELECT * FROM table2(Jan)
2936 Fix for computations in views(Jan)
2937 Fix for aggregates on array indexes(Tom)
2938 Fix for DEFAULT handles single quotes in value requiring too many quotes
2939 Fix security problem with non-super users importing/exporting large objects(Tom)
2940 Rollback of transaction that creates table cleaned up properly(Tom)
2941 Fix to allow long table and column names to generate proper serial names(Tom)
2945 Add "vacuumdb" utility
2946 Speed up libpq by allocating memory better(Tom)
2947 EXPLAIN all indexes used(Tom)
2948 Implement CASE, COALESCE, NULLIF expression(Thomas)
2949 New pg_dump table output format(Constantin)
2950 Add string min()/max() functions(Thomas)
2951 Extend new type coercion techniques to aggregates(Thomas)
2952 New moddatetime contrib(Terry)
2953 Update to pgaccess 0.96(Constantin)
2954 Add routines for single-byte "char" type(Thomas)
2955 Improved substr() function(Thomas)
2956 Improved multibyte handling(Tatsuo)
2957 Multiversion concurrency control/MVCC(Vadim)
2958 New Serialized mode(Vadim)
2959 Fix for tables over 2gigs(Peter)
2960 New SET TRANSACTION ISOLATION LEVEL(Vadim)
2961 New LOCK TABLE IN ... MODE(Vadim)
2962 Update ODBC driver(Byron)
2963 New NUMERIC data type(Jan)
2964 New SELECT FOR UPDATE(Vadim)
2965 Handle "NaN" and "Infinity" for input values(Jan)
2966 Improved date/year handling(Thomas)
2967 Improved handling of backend connections(Magnus)
2968 New options ELOG_TIMESTAMPS and USE_SYSLOG options for log files(Massimo)
2969 New TCL_ARRAYS option(Massimo)
2970 New INTERSECT and EXCEPT(Stefan)
2971 New pg_index.indisprimary for primary key tracking(D'Arcy)
2972 New pg_dump option to allow dropping of tables before creation(Brook)
2973 Speedup of row output routines(Tom)
2974 New READ COMMITTED isolation level(Vadim)
2975 New TEMP tables/indexes(Bruce)
2976 Prevent sorting if result is already sorted(Jan)
2977 New memory allocation optimization(Jan)
2978 Allow psql to do \p\g(Bruce)
2979 Allow multiple rule actions(Jan)
2980 Added LIMIT/OFFSET functionality(Jan)
2981 Improve optimizer when joining a large number of tables(Bruce)
2982 New intro to SQL from S. Simkovics' Master's Thesis (Stefan, Thomas)
2983 New intro to backend processing from S. Simkovics' Master's Thesis (Stefan)
2984 Improved int8 support(Ryan Bradetich, Thomas, Tom)
2985 New routines to convert between int8 and text/varchar types(Thomas)
2986 New bushy plans, where meta-tables are joined(Bruce)
2987 Enable right-hand queries by default(Bruce)
2988 Allow reliable maximum number of backends to be set at configure time
2989 (--with-maxbackends and postmaster switch (-N backends))(Tom)
2990 GEQO default now 10 tables because of optimizer speedups(Tom)
2991 Allow NULL=Var for MS-SQL portability(Michael, Bruce)
2992 Modify contrib check_primary_key() so either "automatic" or "dependent"(Anand)
2993 Allow psql \d on a view show query(Ryan)
2994 Speedup for LIKE(Bruce)
2995 Ecpg fixes/features, see src/interfaces/ecpg/ChangeLog file(Michael)
2996 JDBC fixes/features, see src/interfaces/jdbc/CHANGELOG(Peter)
2997 Make % operator have precedence like /(Bruce)
2998 Add new postgres -O option to allow system table structure changes(Bruce)
2999 Update contrib/pginterface/findoidjoins script(Tom)
3000 Major speedup in vacuum of deleted rows with indexes(Vadim)
3001 Allow non-SQL functions to run different versions based on arguments(Tom)
3002 Add -E option that shows actual queries sent by \dt and friends(Masaaki Sakaida)
3003 Add version number in start-up banners for psql(Masaaki Sakaida)
3004 New contrib/vacuumlo removes large objects not referenced(Peter)
3005 New initialization for table sizes so non-vacuumed tables perform better(Tom)
3006 Improve error messages when a connection is rejected(Tom)
3007 Support for arrays of char() and varchar() fields(Massimo)
3008 Overhaul of hash code to increase reliability and performance(Tom)
3009 Update to PyGreSQL 2.4(D'Arcy)
3010 Changed debug options so -d4 and -d5 produce different node displays(Jan)
3011 New pg_options: pretty_plan, pretty_parse, pretty_rewritten(Jan)
3012 Better optimization statistics for system table access(Tom)
3013 Better handling of non-default block sizes(Massimo)
3014 Improve GEQO optimizer memory consumption(Tom)
3015 UNION now suppports ORDER BY of columns not in target list(Jan)
3016 Major libpq++ improvements(Vince Vielhaber)
3017 pg_dump now uses -z(ACL's) as default(Bruce)
3018 backend cache, memory speedups(Tom)
3019 have pg_dump do everything in one snapshot transaction(Vadim)
3020 fix for large object memory leakage, fix for pg_dumping(Tom)
3021 INET type now respects netmask for comparisons
3022 Make VACUUM ANALYZE only use a readlock(Vadim)
3023 Allow VIEWs on UNIONS(Jan)
3024 pg_dump now can generate consistent snapshots on active databases(Vadim)
3028 Improve port matching(Tom)
3029 Portability fixes for SunOS
3030 Add NT/Win32 backend port and enable dynamic loading(Magnus and Daniel Horak)
3031 New port to Cobalt Qube(Mips) running Linux(Tatsuo)
3032 Port to NetBSD/m68k(Mr. Mutsuki Nakajima)
3033 Port to NetBSD/sun3(Mr. Mutsuki Nakajima)
3034 Port to NetBSD/macppc(Toshimi Aoki)
3035 Fix for tcl/tk configuration(Vince)
3036 Removed CURRENT keyword for rule queries(Jan)
3037 NT dynamic loading now works(Daniel Horak)
3038 Add ARM32 support(Andrew McMurry)
3039 Better support for HP-UX 11 and UnixWare
3040 Improve file handling to be more uniform, prevent file descriptor leak(Tom)
3041 New install commands for plpgsql(Jan)
3048 <sect1 id="release-6-4-2">
3049 <title>Release 6.4.2</title>
3052 <title>Release date</title>
3053 <simpara>1998-12-20</simpara>
3057 The 6.4.1 release was improperly packaged. This also has one additional
3063 <title>Migration to version 6.4.2</title>
3066 A dump/restore is <emphasis>not</emphasis> required for those running
3071 <title>Changes</title>
3075 Fix for datetime constant problem on some platforms(Thomas)
3083 <sect1 id="release-6-4-1">
3084 <title>Release 6.4.1</title>
3087 <title>Release date</title>
3088 <simpara>1998-12-18</simpara>
3092 This is basically a cleanup release for 6.4. We have fixed a variety of
3093 problems reported by 6.4 users.
3098 <title>Migration to version 6.4.1</title>
3101 A dump/restore is <emphasis>not</emphasis> required for those running
3106 <title>Changes</title>
3110 Add pg_dump -N flag to force double quotes around identifiers. This is
3112 Fix for NOT in where clause causing crash(Bruce)
3113 EXPLAIN VERBOSE coredump fix(Vadim)
3114 Fix shared-library problems on Linux
3115 Fix test for table existance to allow mixed-case and whitespace in
3116 the table name(Thomas)
3117 Fix a couple of pg_dump bugs
3118 Configure matches template/.similar entries better(Tom)
3119 Change builtin function names from SPI_* to spi_*
3120 OR WHERE clause fix(Vadim)
3121 Fixes for mixed-case table names(Billy)
3122 contrib/linux/postgres.init.csh/sh fix(Thomas)
3123 libpq memory overrun fix
3125 Change exp() behavior to generate error on underflow(Thomas)
3126 pg_dump fixes for memory leak, inheritance constraints, layout change
3127 update pgaccess to 0.93
3128 Fix prototype for 64-bit platforms
3129 Multibyte fixes(Tatsuo)
3131 Fix memory overruns(Tatsuo)
3132 Fix for lo_import() crash(Bruce)
3133 Better search for install program(Tom)
3136 Use implicit type coercion for matching DEFAULT values(Thomas)
3137 Add routines to help with single-byte (internal) character type(Thomas)
3138 Compilation of libpq for Win32 fixes(Magnus)
3139 Upgrade to PyGreSQL 2.2(D'Arcy)
3147 <sect1 id="release-6-4">
3148 <title>Release 6.4</title>
3151 <title>Release date</title>
3152 <simpara>1998-10-30</simpara>
3156 There are <emphasis>many</emphasis> new features and improvements in this release.
3157 Thanks to our developers and maintainers, nearly every aspect of the system
3158 has received some attention since the previous release.
3159 Here is a brief, incomplete summary:
3164 Views and rules are now functional thanks to extensive new code in the
3165 rewrite rules system from Jan Wieck. He also wrote a chapter on it
3166 for the <citetitle>Programmer's Guide</citetitle>.
3171 Jan also contributed a second procedural language, <application>PL/pgSQL</application>, to go with the
3172 original <application>PL/pgTCL</application> procedural language he contributed last release.
3178 We have optional multiple-byte character set support from Tatsuo Ishii
3179 to complement our existing locale support.
3185 Client/server communications has been cleaned up, with better support for
3186 asynchronous messages and interrupts thanks to Tom Lane.
3192 The parser will now perform automatic type coercion to match arguments
3193 to available operators and functions, and to match columns and expressions
3194 with target columns. This uses a generic mechanism which supports
3195 the type extensibility features of <productname>PostgreSQL</productname>.
3196 There is a new chapter in the <citetitle>User's Guide</citetitle>
3197 which covers this topic.
3203 Three new data types have been added.
3204 Two types, <type>inet</type> and <type>cidr</type>, support various forms
3205 of IP network, subnet, and machine addressing. There is now an 8-byte integer
3206 type available on some platforms. See the chapter on data types
3207 in the <citetitle>User's Guide</citetitle> for details.
3208 A fourth type, <type>serial</type>, is now supported by the parser as an
3209 amalgam of the <type>int4</type> type, a sequence, and a unique index.
3215 Several more <acronym>SQL92</acronym>-compatible syntax features have been
3216 added, including <command>INSERT DEFAULT VALUES</command>
3222 The automatic configuration and installation system has received some
3223 attention, and should be more robust for more platforms than it has ever
3232 <title>Migration to version 6.4</title>
3235 A dump/restore using <application>pg_dump</application>
3236 or <application>pg_dumpall</application>
3237 is required for those wishing to migrate data from any
3238 previous release of <productname>PostgreSQL</productname>.
3243 <title>Changes</title>
3249 Fix for a tiny memory leak in PQsetdb/PQfinish(Bryan)
3250 Remove char2-16 data types, use char/varchar(Darren)
3251 Pqfn not handles a NOTICE message(Anders)
3252 Reduced busywaiting overhead for spinlocks with many backends (dg)
3253 Stuck spinlock detection (dg)
3254 Fix up "ISO-style" timespan decoding and encoding(Thomas)
3255 Fix problem with table drop after rollback of transaction(Vadim)
3256 Change error message and remove non-functional update message(Vadim)
3257 Fix for COPY array checking
3258 Fix for SELECT 1 UNION SELECT NULL
3259 Fix for buffer leaks in large object calls(Pascal)
3260 Change owner from oid to int4 type(Bruce)
3261 Fix a bug in the oracle compatibility functions btrim() ltrim() and rtrim()
3262 Fix for shared invalidation cache overflow(Massimo)
3263 Prevent file descriptor leaks in failed COPY's(Bruce)
3264 Fix memory leak in libpgtcl's pg_select(Constantin)
3265 Fix problems with username/passwords over 8 characters(Tom)
3266 Fix problems with handling of asynchronous NOTIFY in backend(Tom)
3267 Fix of many bad system table entries(Tom)
3271 Upgrade ecpg and ecpglib,see src/interfaces/ecpc/ChangeLog(Michael)
3272 Show the index used in an EXPLAIN(Zeugswetter)
3273 EXPLAIN invokes rule system and shows plan(s) for rewritten queries(Jan)
3274 Multibyte awareness of many data types and functions, via configure(Tatsuo)
3275 New configure --with-mb option(Tatsuo)
3276 New initdb --pgencoding option(Tatsuo)
3277 New createdb -E multibyte option(Tatsuo)
3278 Select version(); now returns PostgreSQL version(Jeroen)
3279 Libpq now allows asynchronous clients(Tom)
3280 Allow cancel from client of backend query(Tom)
3281 Psql now cancels query with Control-C(Tom)
3282 Libpq users need not issue dummy queries to get NOTIFY messages(Tom)
3283 NOTIFY now sends sender's PID, so you can tell whether it was your own(Tom)
3284 PGresult struct now includes associated error message, if any(Tom)
3285 Define "tz_hour" and "tz_minute" arguments to date_part()(Thomas)
3286 Add routines to convert between varchar and bpchar(Thomas)
3287 Add routines to allow sizing of varchar and bpchar into target columns(Thomas)
3288 Add bit flags to support timezonehour and minute in data retrieval(Thomas)
3289 Allow more variations on valid floating point numbers (e.g. ".1", "1e6")(Thomas)
3290 Fixes for unary minus parsing with leading spaces(Thomas)
3291 Implement TIMEZONE_HOUR, TIMEZONE_MINUTE per SQL92 specs(Thomas)
3292 Check for and properly ignore FOREIGN KEY column constraints(Thomas)
3293 Define USER as synonym for CURRENT_USER per SQL92 specs(Thomas)
3294 Enable HAVING clause but no fixes elsewhere yet.
3295 Make "char" type a synonym for "char(1)" (actually implemented as bpchar)(Thomas)
3296 Save string type if specified for DEFAULT clause handling(Thomas)
3297 Coerce operations involving different data types(Thomas)
3298 Allow some index use for columns of different types(Thomas)
3299 Add capabilities for automatic type conversion(Thomas)
3300 Cleanups for large objects, so file is truncated on open(Peter)
3301 Readline cleanups(Tom)
3302 Allow psql \f \ to make spaces as delimiter(Bruce)
3303 Pass pg_attribute.atttypmod to the frontend for column field lengths(Tom,Bruce)
3304 Msql compatibility library in /contrib(Aldrin)
3305 Remove the requirement that ORDER/GROUP BY clause identifiers be
3306 included in the target list(David)
3307 Convert columns to match columns in UNION clauses(Thomas)
3308 Remove fork()/exec() and only do fork()(Bruce)
3309 Jdbc cleanups(Peter)
3310 Show backend status on ps command line(only works on some platforms)(Bruce)
3311 Pg_hba.conf now has a sameuser option in the database field
3312 Make lo_unlink take oid param, not int4
3313 New DISABLE_COMPLEX_MACRO for compilers that can't handle our macros(Bruce)
3314 Libpgtcl now handles NOTIFY as a Tcl event, need not send dummy queries(Tom)
3315 libpgtcl cleanups(Tom)
3316 Add -error option to libpgtcl's pg_result command(Tom)
3317 New locale patch, see docs/README/locale(Oleg)
3318 Fix for pg_dump so CONSTRAINT and CHECK syntax is correct(ccb)
3319 New contrib/lo code for large object orphan removal(Peter)
3320 New psql command "SET CLIENT_ENCODING TO 'encoding'" for multibytes
3321 feature, see /doc/README.mb(Tatsuo)
3322 /contrib/noupdate code to revoke update permission on a column
3323 Libpq can now be compiled on win32(Magnus)
3324 Add PQsetdbLogin() in libpq
3325 New 8-byte integer type, checked by configure for OS support(Thomas)
3326 Better support for quoted table/column names(Thomas)
3327 Surround table and column names with double-quotes in pg_dump(Thomas)
3328 PQreset() now works with passwords(Tom)
3329 Handle case of GROUP BY target list column number out of range(David)
3330 Allow UNION in subselects
3331 Add auto-size to screen to \d? commands(Bruce)
3332 Use UNION to show all \d? results in one query(Bruce)
3333 Add \d? field search feature(Bruce)
3334 Pg_dump issues fewer \connect requests(Tom)
3335 Make pg_dump -z flag work better, document it in manual page(Tom)
3336 Add HAVING clause with full support for subselects and unions(Stephan)
3337 Full text indexing routines in contrib/fulltextindex(Maarten)
3338 Transaction ids now stored in shared memory(Vadim)
3339 New PGCLIENTENCODING when issuing COPY command(Tatsuo)
3340 Support for SQL92 syntax "SET NAMES"(Tatsuo)
3341 Support for LATIN2-5(Tatsuo)
3342 Add UNICODE regression test case(Tatsuo)
3343 Lock manager cleanup, new locking modes for LLL(Vadim)
3344 Allow index use with OR clauses(Bruce)
3345 Allows "SELECT NULL ORDER BY 1;"
3346 Explain VERBOSE prints the plan, and now pretty-prints the plan to
3347 the postmaster log file(Bruce)
3348 Add indexes display to \d command(Bruce)
3349 Allow GROUP BY on functions(David)
3350 New pg_class.relkind for large objects(Bruce)
3351 New way to send libpq NOTICE messages to a different location(Tom)
3352 New \w write command to psql(Bruce)
3353 New /contrib/findoidjoins scans oid columns to find join relationships(Bruce)
3354 Allow binary-compatible indexes to be considered when checking for valid
3355 Indexes for restriction clauses containing a constant(Thomas)
3356 New ISBN/ISSN code in /contrib/isbn_issn
3357 Allow NOT LIKE, IN, NOT IN, BETWEEN, and NOT BETWEEN constraint(Thomas)
3358 New rewrite system fixes many problems with rules and views(Jan)
3359 * Rules on relations work
3360 * Event qualifications on insert/update/delete work
3361 * New OLD variable to reference CURRENT, CURRENT will be remove in future
3362 * Update rules can reference NEW and OLD in rule qualifications/actions
3363 * Insert/update/delete rules on views work
3364 * Multiple rule actions are now supported, surrounded by parentheses
3365 * Regular users can create views/rules on tables they have RULE permits
3366 * Rules and views inherit the permissions on the creator
3367 * No rules at the column level
3368 * No UPDATE NEW/OLD rules
3369 * New pg_tables, pg_indexes, pg_rules and pg_views system views
3370 * Only a single action on SELECT rules
3371 * Total rewrite overhaul, perhaps for 6.5
3373 * handle aggregates on views
3374 * handle insert into select from view works
3375 System indexes are now multikey(Bruce)
3376 Oidint2, oidint4, and oidname types are removed(Bruce)
3377 Use system cache for more system table lookups(Bruce)
3378 New backend programming language PL/pgSQL in backend/pl(Jan)
3379 New SERIAL data type, auto-creates sequence/index(Thomas)
3380 Enable assert checking without a recompile(Massimo)
3381 User lock enhancements(Massimo)
3382 New setval() command to set sequence value(Massimo)
3383 Auto-remove unix socket file on start-up if no postmaster running(Massimo)
3384 Conditional trace package(Massimo)
3385 New UNLISTEN command(Massimo)
3386 Psql and libpq now compile under win32 using win32.mak(Magnus)
3387 Lo_read no longer stores trailing NULL(Bruce)
3388 Identifiers are now truncated to 31 characters internally(Bruce)
3389 Createuser options now availble on the command line
3390 Code for 64-bit integer supported added, configure tested, int8 type(Thomas)
3391 Prevent file descriptor leaf from failed COPY(Bruce)
3392 New pg_upgrade command(Bruce)
3393 Updated /contrib directories(Massimo)
3394 New CREATE TABLE DEFAULT VALUES statement available(Thomas)
3395 New INSERT INTO TABLE DEFAULT VALUES statement available(Thomas)
3396 New DECLARE and FETCH feature(Thomas)
3397 libpq's internal structures now not exported(Tom)
3398 Allow up to 8 key indexes(Bruce)
3399 Remove ARCHIVE keyword, that is no longer used(Thomas)
3400 pg_dump -n flag to supress quotes around indentifiers
3401 disable system columns for views(Jan)
3402 new INET and CIDR types for network addresses(TomH, Paul)
3403 no more double quotes in psql output
3404 pg_dump now dumps views(Terry)
3405 new SET QUERY_LIMIT(Tatsuo,Jan)
3409 /contrib cleanup(Jun)
3410 Inline some small functions called for every row(Bruce)
3413 Multibyte regression tests(Soonmyung.)
3414 Remove --disabled options from configure
3415 Define PGDOC to use POSTGRESDIR by default
3416 Make regression optional
3417 Remove extra braces code to pgindent(Bruce)
3418 Add bsdi shared library support(Bruce)
3419 New --without-CXX support configure option(Brook)
3421 Update backend flowchart in tools/backend(Bruce)
3422 Change atttypmod from int16 to int32(Bruce, Tom)
3423 Getrusage() fix for platforms that do not have it(Tom)
3424 Add PQconnectdb, PGUSER, PGPASSWORD to libpq man page
3425 NS32K platform fixes(Phil Nelson, John Buller)
3426 SCO 7/UnixWare 2.x fixes(Billy,others)
3427 Sparc/Solaris 2.5 fixes(Ryan)
3428 Pgbuiltin.3 is obsolete, move to doc files(Thomas)
3429 Even more documention(Thomas)
3430 Nextstep support(Jacek)
3432 pginterface manual page(Bruce)
3433 shared libraries all have version numbers
3434 merged all OS-specific shared library defines into one file
3435 smarter TCL/TK configuration checking(Billy)
3436 smarter perl configuration(Brook)
3437 configure uses supplied install-sh if no install script found(Tom)
3438 new Makefile.shlib for shared library configuration(Tom)
3444 <sect1 id="release-6-3-2">
3445 <title>Release 6.3.2</title>
3448 <title>Release date</title>
3449 <simpara>1998-04-07</simpara>
3453 This is a bug-fix release for 6.3.x.
3454 Refer to the release notes for version 6.3 for a more complete summary of new features.
3462 Repairs automatic configuration support for some platforms, including Linux,
3463 from breakage inadvertently introduced in version 6.3.1.
3469 Correctly handles function calls on the left side of BETWEEN and LIKE clauses.
3476 A dump/restore is NOT required for those running 6.3 or 6.3.1. A
3477 <literal>make distclean</>, <literal>make</>, and <literal>make install</> is all that is required.
3478 This last step should be performed while the postmaster is not running.
3479 You should re-link any custom applications that use <productname>PostgreSQL</productname> libraries.
3482 For upgrades from pre-6.3 installations,
3483 refer to the installation and migration instructions for version 6.3.
3487 <title>Changes</title>
3491 Configure detection improvements for tcl/tk(Brook Milligan, Alvin)
3492 Manual page improvements(Bruce)
3493 BETWEEN and LIKE fix(Thomas)
3494 fix for psql \connect used by pg_dump(Oliver Elphick)
3496 pgaccess, version 0.86
3497 qsort removed, now uses libc version, cleanups(Jeroen)
3498 fix for buffer over-runs detected(Maurice Gittens)
3499 fix for buffer overrun in libpgtcl(Randy Kunkee)
3500 fix for UNION with DISTINCT or ORDER BY(Bruce)
3501 gettimeofday configure check(Doug Winterburn)
3502 Fix "indexes not used" bug(Vadim)
3503 docs additions(Thomas)
3504 Fix for backend memory leak(Bruce)
3505 libreadline cleanup(Erwan MAS)
3506 Remove DISTDIR(Bruce)
3507 Makefile dependency cleanup(Jeroen van Vianen)
3514 <sect1 id="release-6-3-1">
3515 <title>Release 6.3.1</title>
3518 <title>Release date</title>
3519 <simpara>1998-03-23</simpara>
3528 Additional support for multibyte character sets.
3534 Repair byte ordering for mixed-endian clients and servers.
3540 Minor updates to allowed SQL syntax.
3546 Improvements to the configuration autodetection for installation.
3553 A dump/restore is NOT required for those running 6.3. A
3554 <literal>make distclean</>, <literal>make</>, and <literal>make install</> is all that is required.
3555 This last step should be performed while the postmaster is not running.
3556 You should re-link any custom applications that use <productname>PostgreSQL</productname> libraries.
3559 For upgrades from pre-6.3 installations,
3560 refer to the installation and migration instructions for version 6.3.
3564 <title>Changes</title>
3568 ecpg cleanup/fixes, now version 1.1(Michael Meskes)
3569 pg_user cleanup(Bruce)
3570 large object fix for pg_dump and tclsh (alvin)
3571 LIKE fix for multiple adjacent underscores
3572 fix for redefining builtin functions(Thomas)
3574 upgrade to pg_access 0.83
3575 updated CLUSTER manual page
3576 multibyte character set support, see doc/README.mb(Tatsuo)
3577 configure --with-pgport fix
3579 big-endian fix for backend communications(Kataoka)
3580 SUBSTR() and substring() fix(Jan)
3581 several jdbc fixes(Peter)
3582 libpgtcl improvements, see libptcl/README(Randy Kunkee)
3583 Fix for "Datasize = 0" error(Vadim)
3584 Prevent \do from wrapping(Bruce)
3585 Remove duplicate Russian character set entries
3587 Allow optional TABLE keyword in LOCK and SELECT INTO(Thomas)
3588 CREATE SEQUENCE options to allow a negative integer(Thomas)
3589 Add "PASSWORD" as an allowed column identifier(Thomas)
3590 Add checks for UNION target fields(Bruce)
3591 Fix Alpha port(Dwayne Bailey)
3592 Fix for text arrays containing quotes(Doug Gibson)
3593 Solaris compile fix(Albert Chin-A-Young)
3594 Better identify tcl and tk libs and includes(Bruce)
3600 <sect1 id="release-6-3">
3601 <title>Release 6.3</title>
3604 <title>Release date</title>
3605 <simpara>1998-03-01</simpara>
3609 There are <emphasis>many</emphasis> new features and improvements in this release.
3610 Here is a brief, incomplete summary:
3615 Many new SQL features, including
3616 full <acronym>SQL92</acronym> subselect capability
3617 (everything is here but target-list subselects).
3623 Support for client-side environment variables to specify time zone and date style.
3629 Socket interface for client/server connection. This is the default now
3630 so you may need to start <application>postmaster</application> with the
3631 <option>-i</option> flag.
3637 Better password authorization mechanisms. Default table permissions have changed.
3643 Old-style <firstterm>time travel</firstterm>
3644 has been removed. Performance has been improved.
3652 Bruce Momjian wrote the following notes to introduce the new release.
3657 There are some general 6.3 issues that I want to mention. These are
3658 only the big items that can not be described in one sentence. A review
3659 of the detailed changes list is still needed.
3662 First, we now have subselects. Now that we have them, I would like to
3663 mention that without subselects, SQL is a very limited language.
3664 Subselects are a major feature, and you should review your code for
3665 places where subselects provide a better solution for your queries. I
3666 think you will find that there are more uses for subselects than you may
3667 think. Vadim has put us on the big SQL map with subselects, and fully
3668 functional ones too. The only thing you can't do with subselects is to
3669 use them in the target list.
3672 Second, 6.3 uses Unix domain sockets rather than TCP/IP by default. To
3673 enable connections from other machines, you have to use the new
3674 postmaster -i option, and of course edit <filename>pg_hba.conf</filename>. Also, for this
3675 reason, the format of <filename>pg_hba.conf</filename> has changed.
3678 Third, <type>char()</type> fields will now allow faster access than <type>varchar()</type> or
3679 <type>text</type>. Specifically, the <type>text</> and <type>varchar()</type> have a penalty for access to
3680 any columns after the first column of this type. <type>char()</type> used to also
3681 have this access penalty, but it no longer does. This may suggest that
3682 you redesign some of your tables, especially if you have short character
3683 columns that you have defined as <type>varchar()</type> or <type>text</type>. This and other
3684 changes make 6.3 even faster than earlier releases.
3687 We now have passwords definable independent of any Unix file. There are
3688 new SQL USER commands.
3689 See the <citetitle>Administrator's Guide</citetitle> for more
3690 information. There is a new table, pg_shadow, which is used to store
3691 user information and user passwords, and it by default only SELECT-able
3692 by the <systemitem>postgres</systemitem> super-user. pg_user is now a view of pg_shadow, and is
3693 SELECT-able by PUBLIC. You should keep using pg_user in your
3694 application without changes.
3697 User-created tables now no longer have SELECT permission to PUBLIC by
3698 default. This was done because the ANSI standard requires it. You can
3699 of course GRANT any permissions you want after the table is created.
3700 System tables continue to be SELECT-able by PUBLIC.
3703 We also have real deadlock detection code. No more sixty-second
3704 timeouts. And the new locking code implements a <acronym>FIFO</acronym> better, so there
3705 should be less resource starvation during heavy use.
3708 Many complaints have been made about inadequate documentation in previous
3709 releases. Thomas has put much effort into many new manuals for this
3710 release. Check out the doc/ directory.
3713 For performance reasons, time travel is gone, but can be implemented
3714 using triggers (see <filename>pgsql/contrib/spi/README</filename>). Please check out the new
3715 \d command for types, operators, etc. Also, views have their own
3716 permissions now, not based on the underlying tables, so permissions on
3717 them have to be set separately. Check <filename>/pgsql/interfaces</filename> for some new
3718 ways to talk to <productname>PostgreSQL</productname>.
3721 This is the first release that really required an explanation for
3722 existing users. In many ways, this was necessary because the new
3723 release removes many limitations, and the work-arounds people were using
3724 are no longer needed.
3728 <title>Migration to version 6.3</title>
3731 A dump/restore using <application>pg_dump</application>
3732 or <application>pg_dumpall</application>
3733 is required for those wishing to migrate data from any
3734 previous release of <productname>PostgreSQL</productname>.
3739 <title>Changes</title>
3745 Fix binary cursors broken by MOVE implementation(Vadim)
3746 Fix for tcl library crash(Jan)
3747 Fix for array handling, from Gerhard Hintermayer
3748 Fix acl error, and remove duplicate pqtrace(Bruce)
3749 Fix psql \e for empty file(Bruce)
3750 Fix for textcat on varchar() fields(Bruce)
3751 Fix for DBT Sendproc (Zeugswetter Andres)
3752 Fix vacuum analyze syntax problem(Bruce)
3753 Fix for international identifiers(Tatsuo)
3754 Fix aggregates on inherited tables(Bruce)
3755 Fix substr() for out-of-bounds data
3756 Fix for select 1=1 or 2=2, select 1=1 and 2=2, and select sum(2+2)(Bruce)
3757 Fix notty output to show status result. -q option still turns it off(Bruce)
3758 Fix for count(*), aggs with views and multiple tables and sum(3)(Bruce)
3760 Fix for PQtrace start/stop several times(Bruce)
3761 Fix a variety of locking problems like newer lock waiters getting
3762 lock before older waiters, and having readlock people not share
3763 locks if a writer is waiting for a lock, and waiting writers not
3764 getting priority over waiting readers(Bruce)
3765 Fix crashes in psql when executing queries from external files(James)
3766 Fix problem with multiple order by columns, with the first one having
3768 Use correct hash table support functions for float8 and int4(Thomas)
3769 Re-enable JOIN= option in CREATE OPERATOR statement (Thomas)
3770 Change precedence for boolean operators to match expected behavior(Thomas)
3771 Generate elog(ERROR) on over-large integer(Bruce)
3772 Allow multiple-argument functions in constraint clauses(Thomas)
3773 Check boolean input literals for 'true','false','yes','no','1','0'
3774 and throw elog(ERROR) if unrecognized(Thomas)
3775 Major large objects fix
3776 Fix for GROUP BY showing duplicates(Vadim)
3777 Fix for index scans in MergeJion(Vadim)
3781 Subselects with EXISTS, IN, ALL, ANY keywords (Vadim, Bruce, Thomas)
3782 New User Manual(Thomas, others)
3783 Speedup by inlining some frequently-called functions
3784 Real deadlock detection, no more timeouts(Bruce)
3785 Add SQL92 "constants" CURRENT_DATE, CURRENT_TIME, CURRENT_TIMESTAMP,
3786 CURRENT_USER(Thomas)
3787 Modify constraint syntax to be SQL92-compliant(Thomas)
3788 Implement SQL92 PRIMARY KEY and UNIQUE clauses using indexes(Thomas)
3789 Recognize SQL92 syntax for FOREIGN KEY. Throw elog notice(Thomas)
3790 Allow NOT NULL UNIQUE constraint clause (each allowed separately before)(Thomas)
3791 Allow PostgreSQL-style casting ("::") of non-constants(Thomas)
3792 Add support for SQL3 TRUE and FALSE boolean constants(Thomas)
3793 Support SQL92 syntax for IS TRUE/IS FALSE/IS NOT TRUE/IS NOT FALSE(Thomas)
3794 Allow shorter strings for boolean literals (e.g. "t", "tr", "tru")(Thomas)
3795 Allow SQL92 delimited identifiers(Thomas)
3796 Implement SQL92 binary and hexadecimal string decoding (b'10' and x'1F')(Thomas)
3797 Support SQL92 syntax for type coercion of literal strings
3798 (e.g. "DATETIME 'now'")(Thomas)
3799 Add conversions for int2, int4, and OID types to and from text(Thomas)
3800 Use shared lock when building indexes(Vadim)
3801 Free memory allocated for an user query inside transaction block after
3802 this query is done, was turned off in <= 6.2.1(Vadim)
3803 New SQL statement CREATE PROCEDURAL LANGUAGE(Jan)
3804 New <productname>PostgreSQL</productname> Procedural Language (PL) backend interface(Jan)
3805 Rename pg_dump -H option to -h(Bruce)
3806 Add Java support for passwords, European dates(Peter)
3807 Use indexes for LIKE and ~, !~ operations(Bruce)
3808 Add hash functions for datetime and timespan(Thomas)
3809 Time Travel removed(Vadim, Bruce)
3810 Add paging for \d and \z, and fix \i(Bruce)
3811 Add Unix domain socket support to backend and to frontend library(Goran)
3812 Implement CREATE DATABASE/WITH LOCATION and initlocation utility(Thomas)
3813 Allow more SQL92 and/or <productname>PostgreSQL</productname> reserved words as column identifiers(Thomas)
3814 Augment support for SQL92 SET TIME ZONE...(Thomas)
3815 SET/SHOW/RESET TIME ZONE uses TZ backend environment variable(Thomas)
3816 Implement SET keyword = DEFAULT and SET TIME ZONE DEFAULT(Thomas)
3817 Enable SET TIME ZONE using TZ environment variable(Thomas)
3818 Add PGDATESTYLE environment variable to frontend and backend initialization(Thomas)
3819 Add PGTZ, PGCOSTHEAP, PGCOSTINDEX, PGRPLANS, PGGEQO
3820 frontend library initialization environment variables(Thomas)
3821 Regression tests time zone automatically set with "setenv PGTZ PST8PDT"(Thomas)
3822 Add pg_description table for info on tables, columns, operators, types, and
3824 Increase 16 char limit on system table/index names to 32 characters(Bruce)
3825 Rename system indexes(Bruce)
3826 Add 'GERMAN' option to SET DATESTYLE(Thomas)
3827 Define an "ISO-style" timespan output format with "hh:mm:ss" fields(Thomas)
3828 Allow fractional values for delta times (e.g. '2.5 days')(Thomas)
3829 Validate numeric input more carefully for delta times(Thomas)
3830 Implement day of year as possible input to date_part()(Thomas)
3831 Define timespan_finite() and text_timespan() functions(Thomas)
3832 Remove archive stuff(Bruce)
3833 Allow for a pg_password authentication database that is separate from
3834 the system password file(Todd)
3835 Dump ACLs, GRANT, REVOKE permissions(Matt)
3836 Define text, varchar, and bpchar string length functions(Thomas)
3837 Fix Query handling for inheritance, and cost computations(Bruce)
3838 Implement CREATE TABLE/AS SELECT (alternative to SELECT/INTO)(Thomas)
3839 Allow NOT, IS NULL, IS NOT NULL in constraints(Thomas)
3840 Implement UNIONs for SELECT(Bruce)
3841 Add UNION, GROUP, DISTINCT to INSERT(Bruce)
3842 varchar() stores only necessary bytes on disk(Bruce)
3843 Fix for BLOBs(Peter)
3844 Mega-Patch for JDBC...see README_6.3 for list of changes(Peter)
3845 Remove unused "option" from PQconnectdb()
3846 New LOCK command and lock manual page describing deadlocks(Bruce)
3847 Add new psql \da, \dd, \df, \do, \dS, and \dT commands(Bruce)
3848 Enhance psql \z to show sequences(Bruce)
3849 Show NOT NULL and DEFAULT in psql \d table(Bruce)
3850 New psql .psqlrc file start-up(Andrew)
3851 Modify sample start-up script in contrib/linux to show syslog(Thomas)
3852 New types for IP and MAC addresses in contrib/ip_and_mac(TomH)
3853 Unix system time conversions with date/time types in contrib/unixdate(Thomas)
3854 Update of contrib stuff(Massimo)
3855 Add Unix socket support to DBD::Pg(Goran)
3856 New python interface (PyGreSQL 2.0)(D'Arcy)
3857 New frontend/backend protocol has a version number, network byte order(Phil)
3858 Security features in pg_hba.conf enhanced and documented, many cleanups(Phil)
3859 CHAR() now faster access than VARCHAR() or TEXT
3860 ecpg embedded SQL preprocessor
3861 Reduce system column overhead(Vadmin)
3862 Remove pg_time table(Vadim)
3863 Add pg_type attribute to identify types that need length (bpchar, varchar)
3864 Add report of offending line when COPY command fails
3865 Allow VIEW permissions to be set separately from the underlying tables.
3866 For security, use GRANT/REVOKE on views as appropriate(Jan)
3867 Tables now have no default GRANT SELECT TO PUBLIC. You must
3868 explicitly grant such permissions.
3869 Clean up tutorial examples(Darren)
3873 Add new html development tools, and flow chart in /tools/backend
3874 Fix for SCO compiles
3875 Stratus computer port Robert Gillies
3876 Added support for shlib for BSD44_derived & i386_solaris
3877 Make configure more automated(Brook)
3878 Add script to check regression test results
3879 Break parser functions into smaller files, group together(Bruce)
3880 Rename heap_create to heap_create_and_catalog, rename heap_creatr
3881 to heap_create()(Bruce)
3882 Sparc/Linux patch for locking(TomS)
3883 Remove PORTNAME and reorganize port-specific stuff(Marc)
3884 Add optimizer README file(Bruce)
3885 Remove some recursion in optimizer and clean up some code there(Bruce)
3886 Fix for NetBSD locking(Henry)
3887 Fix for libptcl make(Tatsuo)
3889 Change IS TRUE, IS FALSE, ... to expressions using "=" rather than
3890 function calls to istrue() or isfalse() to allow optimization(Thomas)
3891 Various fixes NetBSD/Sparc related(TomH)
3892 Alpha linux locking(Travis,Ryan)
3893 Change elog(WARN) to elog(ERROR)(Bruce)
3894 FAQ for FreeBSD(Marc)
3895 Bring in the PostODBC source tree as part of our standard distribution(Marc)
3896 A minor patch for HP/UX 10 vs 9(Stan)
3897 New pg_attribute.atttypmod for type-specific info like varchar length(Bruce)
3898 UnixWare patches(Billy)
3899 New i386 'lock' for spin lock asm(Billy)
3900 Support for multiplexed backends is removed
3901 Start an OpenBSD port
3904 Add string functions to regression suite(Thomas)
3905 Expand a few function names formerly truncated to 16 characters(Thomas)
3906 Remove un-needed malloc() calls and replace with palloc()(Bruce)
3912 <sect1 id="release-6-2-1">
3913 <title>Release 6.2.1</title>
3916 <title>Release date</title>
3917 <simpara>1997-10-17</simpara>
3921 6.2.1 is a bug-fix and usability release on 6.2.
3929 Allow strings to span lines, per <acronym>SQL92</acronym>.
3935 Include example trigger function for inserting user names on table updates.
3942 This is a minor bug-fix release on 6.2.
3943 For upgrades from pre-6.2 systems, a full dump/reload is required.
3944 Refer to the 6.2 release notes for instructions.
3948 <title>Migration from version 6.2 to version 6.2.1</title>
3951 This is a minor bug-fix release. A dump/reload is not required from version 6.2,
3952 but is required from any release prior to 6.2.
3955 In upgrading from version 6.2, if you choose to dump/reload you will find that
3956 avg(money) is now calculated correctly. All other bug fixes take effect
3957 upon updating the executables.
3960 Another way to avoid dump/reload is to use the following SQL command
3961 from <command>psql</command> to update the existing system table:
3964 update pg_aggregate set aggfinalfn = 'cash_div_flt8'
3965 where aggname = 'avg' and aggbasetype = 790;
3969 This will need to be done to every existing database, including template1.
3974 <title>Changes</title>
3978 Allow TIME and TYPE column names(Thomas)
3979 Allow larger range of true/false as boolean values(Thomas)
3980 Support output of "now" and "current"(Thomas)
3981 Handle DEFAULT with INSERT of NULL properly(Vadim)
3982 Fix for relation reference counts problem in buffer manager(Vadim)
3983 Allow strings to span lines, like ANSI(Thomas)
3984 Fix for backward cursor with ORDER BY(Vadim)
3985 Fix avg(cash) computation(Thomas)
3986 Fix for specifying a column twice in ORDER/GROUP BY(Vadim)
3987 Documented new libpq function to return affected rows, PQcmdTuples(Bruce)
3988 Trigger function for inserting user names for INSERT/UPDATE(Brook Milligan)
3994 <sect1 id="release-6-2">
3995 <title>Release 6.2</title>
3998 <title>Release date</title>
3999 <simpara>1997-10-02</simpara>
4003 A dump/restore is required for those wishing to migrate data from
4004 previous releases of <productname>PostgreSQL</productname>.
4008 <title>Migration from version 6.1 to version 6.2</title>
4011 This migration requires a complete dump of the 6.1 database and a
4012 restore of the database in 6.2.
4015 Note that the <command>pg_dump</command> and <command>pg_dumpall</command> utility from 6.2 should be used
4016 to dump the 6.1 database.
4021 <title>Migration from version 1.<replaceable>x</> to version 6.2</title>
4024 Those migrating from earlier 1.* releases should first upgrade to 1.09
4025 because the COPY output format was improved from the 1.02 release.
4030 <title>Changes</title>
4036 Fix problems with pg_dump for inheritance, sequences, archive tables(Bruce)
4037 Fix compile errors on overflow due to shifts, unsigned, and bad prototypes
4038 from Solaris(Diab Jerius)
4039 Fix bugs in geometric line arithmetic (bad intersection calculations)(Thomas)
4040 Check for geometric intersections at endpoints to avoid rounding ugliness(Thomas)
4041 Catch non-functional delete attempts(Vadim)
4042 Change time function names to be more consistent(Michael Reifenberg)
4043 Check for zero divides(Michael Reifenberg)
4044 Fix very old bug which made tuples changed/inserted by a commnd
4045 visible to the command itself (so we had multiple update of
4046 updated tuples, etc)(Vadim)
4047 Fix for SELECT null, 'fail' FROM pg_am (Patrick)
4048 SELECT NULL as EMPTY_FIELD now allowed(Patrick)
4049 Remove un-needed signal stuff from contrib/pginterface
4050 Fix OR (where x != 1 or x isnull didn't return tuples with x NULL) (Vadim)
4051 Fix time_cmp function (Vadim)
4052 Fix handling of functions with non-attribute first argument in
4053 WHERE clauses (Vadim)
4054 Fix GROUP BY when order of entries is different from order
4055 in target list (Vadim)
4056 Fix pg_dump for aggregates without sfunc1 (Vadim)
4060 Default genetic optimizer GEQO parameter is now 8(Bruce)
4061 Allow use parameters in target list having aggregates in functions(Vadim)
4062 Added JDBC driver as an interface(Adrian & Peter)
4064 Return number of tuples inserted/affected by INSERT/UPDATE/DELETE etc.(Vadim)
4065 Triggers implemented with CREATE TRIGGER (SQL3)(Vadim)
4066 SPI (Server Programming Interface) allows execution of queries inside
4068 NOT NULL implemented (SQL92)(Robson Paniago de Miranda)
4069 Include reserved words for string handling, outer joins, and unions(Thomas)
4070 Implement extended comments ("/* ... */") using exclusive states(Thomas)
4071 Add "//" single-line comments(Bruce)
4072 Remove some restrictions on characters in operator names(Thomas)
4073 DEFAULT and CONSTRAINT for tables implemented (SQL92)(Vadim & Thomas)
4074 Add text concatenation operator and function (SQL92)(Thomas)
4075 Support WITH TIME ZONE syntax (SQL92)(Thomas)
4076 Support INTERVAL unit TO unit syntax (SQL92)(Thomas)
4077 Define types DOUBLE PRECISION, INTERVAL, CHARACTER,
4078 and CHARACTER VARYING (SQL92)(Thomas)
4079 Define type FLOAT(p) and rudimentary DECIMAL(p,s), NUMERIC(p,s) (SQL92)(Thomas)
4080 Define EXTRACT(), POSITION(), SUBSTRING(), and TRIM() (SQL92)(Thomas)
4081 Define CURRENT_DATE, CURRENT_TIME, CURRENT_TIMESTAMP (SQL92)(Thomas)
4082 Add syntax and warnings for UNION, HAVING, INNER and OUTER JOIN (SQL92)(Thomas)
4083 Add more reserved words, mostly for SQL92 compliance(Thomas)
4084 Allow hh:mm:ss time entry for timespan/reltime types(Thomas)
4085 Add center() routines for lseg, path, polygon(Thomas)
4086 Add distance() routines for circle-polygon, polygon-polygon(Thomas)
4087 Check explicitly for points and polygons contained within polygons
4088 using an axis-crossing algorithm(Thomas)
4089 Add routine to convert circle-box(Thomas)
4090 Merge conflicting operators for different geometric data types(Thomas)
4091 Replace distance operator "<===>" with "<->"(Thomas)
4092 Replace "above" operator "!^" with ">^" and "below" operator "!|" with "<^"(Thomas)
4093 Add routines for text trimming on both ends, substring, and string position(Thomas)
4094 Added conversion routines circle(box) and poly(circle)(Thomas)
4095 Allow internal sorts to be stored in memory rather than in files(Bruce & Vadim)
4096 Allow functions and operators on internally-identical types to succeed(Bruce)
4097 Speed up backend start-up after profiling analysis(Bruce)
4098 Inline frequently called functions for performance(Bruce)
4099 Reduce open() calls(Bruce)
4100 psql: Add PAGER for \h and \?,\C fix
4101 Fix for psql pager when no tty(Bruce)
4102 New entab utility(Bruce)
4103 General trigger functions for referential integrity (Vadim)
4104 General trigger functions for time travel (Vadim)
4105 General trigger functions for AUTOINCREMENT/IDENTITY feature (Vadim)
4106 MOVE implementation (Vadim)
4110 HP-UX 10 patches (Vladimir Turin)
4111 Added SCO support, (Daniel Harris)
4112 MkLinux patches (Tatsuo Ishii)
4113 Change geometric box terminology from "length" to "width"(Thomas)
4114 Deprecate temporary unstored slope fields in geometric code(Thomas)
4115 Remove restart instructions from INSTALL(Bruce)
4116 Look in /usr/ucb first for install(Bruce)
4117 Fix c++ copy example code(Thomas)
4118 Add -o to psql manual page(Bruce)
4119 Prevent relname unallocated string length from being copied into database(Bruce)
4120 Cleanup for NAMEDATALEN use(Bruce)
4121 Fix pg_proc names over 15 chars in output(Bruce)
4122 Add strNcpy() function(Bruce)
4123 remove some (void) casts that are unnecessary(Bruce)
4124 new interfaces directory(Marc)
4125 Replace fopen() calls with calls to fd.c functions(Bruce)
4126 Make functions static where possible(Bruce)
4127 enclose unused functions in #ifdef NOT_USED(Bruce)
4128 Remove call to difftime() in timestamp support to fix SunOS(Bruce & Thomas)
4129 Changes for Digital Unix
4130 Portability fix for pg_dumpall(Bruce)
4131 Rename pg_attribute.attnvals to attdispersion(Bruce)
4132 "intro/unix" manual page now "pgintro"(Bruce)
4133 "built-in" manual page now "pgbuiltin"(Bruce)
4134 "drop" manual page now "drop_table"(Bruce)
4135 Add "create_trigger", "drop_trigger" manual pages(Thomas)
4136 Add constraints regression test(Vadim & Thomas)
4137 Add comments syntax regression test(Thomas)
4138 Add PGINDENT and support program(Bruce)
4139 Massive commit to run PGINDENT on all *.c and *.h files(Bruce)
4140 Files moved to /src/tools directory(Bruce)
4141 SPI and Trigger programming guides (Vadim & D'Arcy)
4147 <sect1 id="release-6-1-1">
4148 <title>Release 6.1.1</title>
4151 <title>Release date</title>
4152 <simpara>1997-07-22</simpara>
4156 <title>Migration from version 6.1 to version 6.1.1</title>
4159 This is a minor bug-fix release. A dump/reload is not required from version 6.1,
4160 but is required from any release prior to 6.1.
4161 Refer to the release notes for 6.1 for more details.
4166 <title>Changes</title>
4170 fix for SET with options (Thomas)
4171 allow pg_dump/pg_dumpall to preserve ownership of all tables/objects(Bruce)
4172 new psql \connect option allows changing usernames without changing databases
4173 fix for initdb --debug option(Yoshihiko Ichikawa))
4174 lextest cleanup(Bruce)
4176 fix date/time month boundary arithmetic(Thomas)
4177 fix timezone daylight handling for some ports(Thomas, Bruce, Tatsuo)
4178 timestamp overhauled to use standard functions(Thomas)
4179 other code cleanup in date/time routines(Thomas)
4180 psql's \d now case-insensitive(Bruce)
4181 psql's backslash commands can now have trailing semicolon(Bruce)
4182 fix memory leak in psql when using \g(Bruce)
4183 major fix for endian handling of communication to server(Thomas, Tatsuo)
4184 Fix for Solaris assembler and include files(Yoshihiko Ichikawa)
4185 allow underscores in usernames(Bruce)
4186 pg_dumpall now returns proper status, portability fix(Bruce)
4192 <sect1 id="release-6-1">
4193 <title>Release 6.1</title>
4196 <title>Release date</title>
4197 <simpara>1997-06-08</simpara>
4201 The regression tests have been adapted and extensively modified for the
4202 6.1 release of <productname>PostgreSQL</productname>.
4206 Three new data types (<type>datetime</type>, <type>timespan</type>, and <type>circle</type>) have been added to
4207 the native set of <productname>PostgreSQL</productname> types. Points, boxes, paths, and polygons
4208 have had their output formats made consistent across the data types.
4209 The polygon output in misc.out has only been spot-checked for correctness
4210 relative to the original regression output.
4214 <productname>PostgreSQL</productname> 6.1 introduces a new, alternate
4215 optimizer which uses <firstterm>genetic</firstterm>
4216 algorithms. These algorithms introduce a random behavior in the ordering
4217 of query results when the query contains multiple qualifiers or multiple
4218 tables (giving the optimizer a choice on order of evaluation). Several
4219 regression tests have been modified to explicitly order the results, and
4220 hence are insensitive to optimizer choices. A few regression tests are
4221 for data types which are inherently unordered (e.g. points and time
4222 intervals) and tests involving those types are explicitly bracketed with
4223 <command>set geqo to 'off'</command> and <command>reset geqo</command>.
4227 The interpretation of array specifiers (the curly braces around atomic
4228 values) appears to have changed sometime after the original regression
4229 tests were generated. The current <filename>./expected/*.out</filename> files reflect this
4230 new interpretation, which may not be correct!
4234 The float8 regression test fails on at least some platforms. This is due
4235 to differences in implementations of <function>pow()</function> and <function>exp()</function> and the signaling
4236 mechanisms used for overflow and underflow conditions.
4240 The <quote>random</> results in the random test should cause the
4241 <quote>random</quote> test to be <quote>failed</quote>, since the
4242 regression tests are evaluated using a simple diff. However,
4243 <quote>random</> does not seem to produce random results on my test
4244 machine (Linux/<application>gcc</>/i686).
4248 <title>Migration to version 6.1</title>
4251 This migration requires a complete dump of the 6.0 database and a
4252 restore of the database in 6.1.
4255 Those migrating from earlier 1.* releases should first upgrade to 1.09
4256 because the COPY output format was improved from the 1.02 release.
4261 <title>Changes</title>
4267 packet length checking in library routines
4268 lock manager priority patch
4269 check for under/over flow of float8(Bruce)
4270 multitable join fix(Vadim)
4271 SIGPIPE crash fix(Darren)
4272 large object fixes(Sven)
4273 allow btree indexes to handle NULLs(Vadim)
4274 timezone fixes(D'Arcy)
4275 select SUM(x) can return NULL on no rows(Thomas)
4276 internal optimizer, executor bug fixes(Vadim)
4277 fix problem where inner loop in < or <= has no rows(Vadim)
4278 prevent re-commuting join index clauses(Vadim)
4279 fix join clauses for multiple tables(Vadim)
4280 fix hash, hashjoin for arrays(Vadim)
4281 fix btree for abstime type(Vadim)
4282 large object fixes(Raymond)
4283 fix buffer leak in hash indexes (Vadim)
4284 fix rtree for use in inner scan (Vadim)
4285 fix gist for use in inner scan, cleanups (Vadim, Andrea)
4286 avoid unnecessary local buffers allocation (Vadim, Massimo)
4287 fix local buffers leak in transaction aborts (Vadim)
4288 fix file manager memmory leaks, cleanups (Vadim, Massimo)
4289 fix storage manager memmory leaks (Vadim)
4290 fix btree duplicates handling (Vadim)
4291 fix deleted tuples re-incarnation caused by vacuum (Vadim)
4292 fix SELECT varchar()/char() INTO TABLE made zero-length fields(Bruce)
4293 many psql, pg_dump, and libpq memory leaks fixed using Purify (Igor)
4297 attribute optimization statistics(Bruce)
4298 much faster new btree bulk load code(Paul)
4299 BTREE UNIQUE added to bulk load code(Vadim)
4300 new lock debug code(Massimo)
4301 massive changes to libpg++(Leo)
4302 new GEQO optimizer speeds table multitable optimization(Martin)
4303 new WARN message for non-unique insert into unique key(Marc)
4304 update x=-3, no spaces, now valid(Bruce)
4305 remove case-sensitive identifier handling(Bruce,Thomas,Dan)
4306 debug backend now pretty-prints tree(Darren)
4307 new Oracle character functions(Edmund)
4308 new plaintext password functions(Dan)
4309 no such class or insufficient privilege changed to distinct messages(Dan)
4310 new ANSI timestamp function(Dan)
4311 new ANSI Time and Date types (Thomas)
4312 move large chunks of data in backend(Martin)
4313 multicolumn btree indexes(Vadim)
4314 new SET var TO value command(Martin)
4315 update transaction status on reads(Dan)
4316 new locale settings for character types(Oleg)
4317 new SEQUENCE serial number generator(Vadim)
4318 GROUP BY function now possible(Vadim)
4319 re-organize regression test(Thomas,Marc)
4320 new optimizer operation weights(Vadim)
4321 new psql \z grant/permit option(Marc)
4322 new MONEY data type(D'Arcy,Thomas)
4323 tcp socket communication speed improved(Vadim)
4324 new VACUUM option for attribute statistics, and for certain columns (Vadim)
4325 many geometric type improvements(Thomas,Keith)
4326 additional regression tests(Thomas)
4327 new datestyle variable(Thomas,Vadim,Martin)
4328 more comparison operators for sorting types(Thomas)
4329 new conversion functions(Thomas)
4330 new more compact btree format(Vadim)
4331 allow pg_dumpall to preserve database ownership(Bruce)
4332 new SET GEQO=# and R_PLANS variable(Vadim)
4333 old (!GEQO) optimizer can use right-sided plans (Vadim)
4334 typechecking improvement in SQL parser(Bruce)
4335 new SET, SHOW, RESET commands(Thomas,Vadim)
4336 new \connect database USER option
4337 new destroydb -i option (Igor)
4338 new \dt and \di psql commands (Darren)
4339 SELECT "\n" now escapes newline (A. Duursma)
4340 new geometry conversion functions from old format (Thomas)
4344 new configuration script(Marc)
4345 readline configuration option added(Marc)
4346 OS-specific configuration options removed(Marc)
4347 new OS-specific template files(Marc)
4348 no more need to edit Makefile.global(Marc)
4349 re-arrange include files(Marc)
4350 nextstep patches (Gregor Hoffleit)
4351 removed WIN32-specific code(Bruce)
4352 removed postmaster -e option, now only postgres -e option (Bruce)
4353 merge duplicate library code in front/backends(Martin)
4354 now works with eBones, international Kerberos(Jun)
4355 more shared library support
4356 c++ include file cleanup(Bruce)
4357 warn about buggy flex(Bruce)
4358 DG/UX, Ultrix, IRIX, AIX portability fixes
4364 <sect1 id="release-6-0">
4365 <title>Release 6.0</title>
4368 <title>Release date</title>
4369 <simpara>1997-01-29</simpara>
4373 A dump/restore is required for those wishing to migrate data from
4374 previous releases of <productname>PostgreSQL</productname>.
4378 <title>Migration from version 1.09 to version 6.0</title>
4381 This migration requires a complete dump of the 1.09 database and a
4382 restore of the database in 6.0.
4387 <title>Migration from pre-1.09 to version 6.0</title>
4390 Those migrating from earlier 1.* releases should first upgrade to 1.09
4391 because the COPY output format was improved from the 1.02 release.
4396 <title>Changes</title>
4402 ALTER TABLE bug - running postgress process needs to re-read table definition
4403 Allow vacuum to be run on one table or entire database(Bruce)
4405 Fix array over-runs of memory writes(Kurt)
4406 Fix elusive btree range/non-range bug(Dan)
4407 Fix for hash indexes on some types like time and date
4408 Fix for pg_log size explosion
4409 Fix permissions on lo_export()(Bruce)
4410 Fix unitialized reads of memory(Kurt)
4411 Fixed ALTER TABLE ... char(3) bug(Bruce)
4412 Fixed a few small memory leaks
4413 Fixed EXPLAIN handling of options and changed full_path option name
4414 Fixed output of group acl permissions
4415 Memory leaks (hunt and destroy with tools like Purify(Kurt)
4416 Minor improvements to rules system
4418 New asserts for run-checking
4419 Overhauled parser/analyze code to properly report errors and increase speed
4420 Pg_dump -d now handles NULL's properly(Bruce)
4421 Prevent SELECT NULL from crashing server (Bruce)
4422 Properly report errors when INSERT ... SELECT columns did not match
4423 Properly report errors when insert column names were not correct
4424 Psql \g filename now works(Bruce)
4425 Psql fixed problem with multiple statements on one line with multiple outputs
4426 Removed duplicate system OIDs
4427 SELECT * INTO TABLE . GROUP/ORDER BY gives unlink error if table exists(Bruce)
4428 Several fixes for queries that crashed the backend
4429 Starting quote in insert string errors(Bruce)
4430 Submitting an empty query now returns empty status, not just " " query(Bruce)
4434 Add EXPLAIN manual page(Bruce)
4435 Add UNIQUE index capability(Dan)
4436 Add hostname/user level access control rather than just hostname and user
4437 Add synonym of != for <>(Bruce)
4438 Allow "select oid,* from table"
4439 Allow BY,ORDER BY to specify columns by number, or by non-alias table.column(Bruce)
4440 Allow COPY from the frontend(Bryan)
4441 Allow GROUP BY to use alias column name(Bruce)
4442 Allow actual compression, not just reuse on the same page(Vadim)
4443 Allow installation-configuration option to auto-add all local users(Bryan)
4444 Allow libpq to distinguish between text value '' and null(Bruce)
4445 Allow non-postgres users with createdb privs to destroydb's
4446 Allow restriction on who can create C functions(Bryan)
4447 Allow restriction on who can do backend COPY(Bryan)
4448 Can shrink tables, pg_time and pg_log(Vadim & Erich)
4449 Change debug level 2 to print queries only, changed debug heading layout(Bruce)
4450 Change default decimal constant representation from float4 to float8(Bruce)
4451 European date format now set when postmaster is started
4452 Execute lowercase function names if not found with exact case
4453 Fixes for aggregate/GROUP processing, allow 'select sum(func(x),sum(x+y) from z'
4454 Gist now included in the distrubution(Marc)
4455 Idend authentication of local users(Bryan)
4456 Implement BETWEEN qualifier(Bruce)
4457 Implement IN qualifier(Bruce)
4458 Libpq has PQgetisnull()(Bruce)
4459 Libpq++ improvements
4460 New options to initdb(Bryan)
4461 Pg_dump allow dump of OIDs(Bruce)
4462 Pg_dump create indexes after tables are loaded for speed(Bruce)
4463 Pg_dumpall dumps all databases, and the user table
4464 Pginterface additions for NULL values(Bruce)
4465 Prevent postmaster from being run as root
4466 Psql \h and \? is now readable(Bruce)
4467 Psql allow backslashed, semicolons anywhere on the line(Bruce)
4468 Psql changed command prompt for lines in query or in quotes(Bruce)
4469 Psql char(3) now displays as (bp)char in \d output(Bruce)
4470 Psql return code now more accurate(Bryan?)
4471 Psql updated help syntax(Bruce)
4472 Re-visit and fix vacuum(Vadim)
4473 Reduce size of regression diffs, remove timezone name difference(Bruce)
4474 Remove compile-time parameters to enable binary distributions(Bryan)
4475 Reverse meaning of HBA masks(Bryan)
4476 Secure Authentication of local users(Bryan)
4477 Speed up vacuum(Vadim)
4478 Vacuum now had VERBOSE option(Bruce)
4482 All functions now have prototypes that are compared against the calls
4483 Allow asserts to be disabled easly from Makefile.global(Bruce)
4484 Change oid constants used in code to #define names
4485 Decoupled sparc and solaris defines(Kurt)
4486 Gcc -Wall compiles cleanly with warnings only from unfixable constructs
4487 Major include file reorganization/reduction(Marc)
4488 Make now stops on compile failure(Bryan)
4489 Makefile restructuring(Bryan, Marc)
4490 Merge bsdi_2_1 to bsdi(Bruce)
4491 Monitor program removed
4492 Name change from Postgres95 to PostgreSQL
4493 New config.h file(Marc, Bryan)
4494 PG_VERSION now set to 6.0 and used by postmaster
4495 Portability additions, including Ultrix, DG/UX, AIX, and Solaris
4496 Reduced the number of #define's, centeralized #define's
4497 Remove duplicate OIDS in system tables(Dan)
4498 Remove duplicate system catalog info or report mismatches(Dan)
4499 Removed many os-specific #define's
4500 Restructured object file generation/location(Bryan, Marc)
4501 Restructured port-specific file locations(Bryan, Marc)
4502 Unused/uninialized variables corrected
4508 <sect1 id="release-1-09">
4509 <title>Release 1.09</title>
4512 <title>Release date</title>
4513 <simpara>1996-11-04</simpara>
4517 Sorry, we didn't keep track of changes from 1.02 to 1.09. Some of
4518 the changes listed in 6.0 were actually included in the 1.02.1 to 1.09
4523 <sect1 id="release-1-02">
4524 <title>Release 1.02</title>
4527 <title>Release date</title>
4528 <simpara>1996-08-01</simpara>
4532 <title>Migration from version 1.02 to version 1.02.1</title>
4535 Here is a new migration file for 1.02.1. It includes the 'copy' change
4536 and a script to convert old <acronym>ASCII</acronym> files.
4540 The following notes are for the benefit of users who want to migrate
4541 databases from <productname>Postgres95</> 1.01 and 1.02 to <productname>Postgres95</> 1.02.1.
4544 If you are starting afresh with <productname>Postgres95</> 1.02.1 and do not need
4545 to migrate old databases, you do not need to read any further.
4550 In order to upgrade older <productname>Postgres95</> version 1.01 or 1.02 databases to
4551 version 1.02.1, the following steps are required:
4556 Start up a new 1.02.1 postmaster
4561 Add the new built-in functions and operators of 1.02.1 to 1.01 or 1.02
4562 databases. This is done by running the new 1.02.1 server against
4563 your own 1.01 or 1.02 database and applying the queries attached at
4564 the end of the file. This can be done easily through <command>psql</>. If your
4565 1.01 or 1.02 database is named <literal>testdb</literal> and you have cut the commands
4566 from the end of this file and saved them in <filename>addfunc.sql</filename>:
4568 % psql testdb -f addfunc.sql
4571 Those upgrading 1.02 databases will get a warning when executing the
4572 last two statements in the file because they are already present in 1.02. This is
4573 not a cause for concern.
4580 <title>Dump/Reload Procedure</title>
4583 If you are trying to reload a pg_dump or text-mode, <literal>copy tablename to
4584 stdout</literal> generated with a previous version, you will need to run the
4585 attached <command>sed</command> script on the ASCII file before loading it into the
4586 database. The old format used '.' as end-of-data, while '\.' is now the
4587 end-of-data marker. Also, empty strings are now loaded in as '' rather
4588 than NULL. See the copy manual page for full details.
4591 sed 's/^\.$/\\./g' <in_file >out_file
4595 If you are loading an older binary copy or non-<systemitem>stdout</> copy, there is no
4596 end-of-data character, and hence no conversion necessary.
4599 -- following lines added by agc to reflect the case-insensitive
4600 -- regexp searching for varchar (in 1.02), and bpchar (in 1.02.1)
4601 create operator ~* (leftarg = bpchar, rightarg = text, procedure = texticregexeq);
4602 create operator !~* (leftarg = bpchar, rightarg = text, procedure = texticregexne);
4603 create operator ~* (leftarg = varchar, rightarg = text, procedure = texticregexeq);
4604 create operator !~* (leftarg = varchar, rightarg = text, procedure = texticregexne);
4610 <title>Changes</title>
4614 Source code maintenance and development
4615 * worldwide team of volunteers
4616 * the source tree now in CVS at ftp.ki.net
4619 * psql (and underlying libpq library) now has many more options for
4620 formatting output, including HTML
4621 * pg_dump now output the schema and/or the data, with many fixes to
4622 enhance completeness.
4623 * psql used in place of monitor in administration shell scripts.
4624 monitor to be deprecated in next release.
4625 * date/time functions enhanced
4626 * NULL insert/update/comparison fixed/enhanced
4627 * TCL/TK lib and shell fixed to work with both tck7.4/tk4.0 and tcl7.5/tk4.1
4629 Bug Fixes (almost too numerous to mention)
4631 * storage management
4632 * check for NULL pointer before dereferencing
4636 * added SolarisX86 port
4637 * added BSD/OS 2.1 port
4642 Contributors (appologies to any missed)
4643 * Kurt J. Lidl <lidl@va.pubnix.com>
4644 (missed in first run, but no less important)
4645 * Erich Stamberger <eberger@gewi.kfunigraz.ac.at>
4646 * Jason Wright <jason@shiloh.vnet.net>
4647 * Cees de Groot <C.deGroot@inter.NL.net>
4648 * ernst.molitor@uni-bonn.de
4649 * michael.siebenborn@ae3.Hypo.DE (Michael Siebenborn (6929))
4650 * Brian E. Gallew <geek+@cmu.edu>
4651 * Vadim B. Mikheev <vadim@sable.krasnoyarsk.su>
4652 * Adam Sussman <myddryn@vidya.com>
4653 * Chris Dunlop <chris@onthe.net.au>
4654 * Marc G. Fournier <scrappy@ki.net>
4655 * Dan McGuirk <mcguirk@indirect.com>
4656 * Dr_George_D_Detlefsen <drgeorge@ilt.com>
4657 * Erich Stamberger <eberger@gewi.kfunigraz.ac.at>
4658 * Massimo Dal Zotto <dz@cs.unitn.it>
4659 * Randy Kunkee <kunkee@Starbase.NeoSoft.COM>
4660 * Rick Weldon <rick@wisetech.com>
4661 * Thomas van Reimersdahl <reimersd@dali.techinfo.rwth-aachen.de>
4662 * david bennett <dave@bensoft.com>
4663 * ernst.molitor@uni-bonn.de
4664 * Julian Assange <proff@suburbia.net>
4665 * Bruce Momjian <maillist@candle.pha.pa.us>
4666 * Paul "Shag" Walmsley <ccshag@cclabs.missouri.edu>
4667 * "Alistair G. Crooks" <azcb0@sde.uts.amdahl.com>
4672 <sect1 id="release-1-01">
4673 <title>Release 1.01</title>
4676 <title>Release date</title>
4677 <simpara>1996-02-23</simpara>
4682 <title>Migration from version 1.0 to version 1.01</title>
4685 The following notes are for the benefit of users who want to migrate
4686 databases from <productname>Postgres95</> 1.0 to <productname>Postgres95</> 1.01.
4689 If you are starting afresh with <productname>Postgres95</> 1.01 and do not need
4690 to migrate old databases, you do not need to read any further.
4693 In order to <productname>Postgres95</> version 1.01 with databases created with
4694 <productname>Postgres95</> version 1.0, the following steps are required:
4699 Set the definition of <symbol>NAMEDATALEN</symbol> in <filename>src/Makefile.global</filename> to 16
4700 and <symbol>OIDNAMELEN</symbol> to 20.
4705 Decide whether you want to use Host based authentication.
4710 If you do, you must create a file name <literal>pg_hba</literal> in your top-level data
4711 directory (typically the value of your <envar>$PGDATA</envar>). <filename>src/libpq/pg_hba</filename>
4712 shows an example syntax.
4717 If you do not want host-based authentication, you can comment out
4722 in <filename>src/Makefile.global</filename>
4725 Note that host-based authentication is turned on by default, and if
4726 you do not take steps A or B above, the out-of-the-box 1.01 will
4727 not allow you to connect to 1.0 databases.
4735 Compile and install 1.01, but DO NOT do the <command>initdb</command> step.
4740 Before doing anything else, terminate your 1.0 postmaster, and
4741 backup your existing <envar>$PGDATA</envar> directory.
4746 Set your <envar>PGDATA</envar> environment variable to your 1.0 databases, but set up
4747 path up so that 1.01 binaries are being used.
4752 Modify the file <filename><envar>$PGDATA</envar>/PG_VERSION</filename> from 5.0 to 5.1
4757 Start up a new 1.01 postmaster
4762 Add the new built-in functions and operators of 1.01 to 1.0
4763 databases. This is done by running the new 1.01 server against
4764 your own 1.0 database and applying the queries attached and saving
4765 in the file 1.0_to_1.01.sql. This can be done easily through <command>psql</command>.
4766 If your 1.0 database is name <literal>testdb</literal>:
4769 % psql testdb -f 1.0_to_1.01.sql
4772 and then execute the following commands (cut and paste from here):
4775 -- add builtin functions that are new to 1.01
4777 create function int4eqoid (int4, oid) returns bool as 'foo'
4778 language 'internal';
4779 create function oideqint4 (oid, int4) returns bool as 'foo'
4780 language 'internal';
4781 create function char2icregexeq (char2, text) returns bool as 'foo'
4782 language 'internal';
4783 create function char2icregexne (char2, text) returns bool as 'foo'
4784 language 'internal';
4785 create function char4icregexeq (char4, text) returns bool as 'foo'
4786 language 'internal';
4787 create function char4icregexne (char4, text) returns bool as 'foo'
4788 language 'internal';
4789 create function char8icregexeq (char8, text) returns bool as 'foo'
4790 language 'internal';
4791 create function char8icregexne (char8, text) returns bool as 'foo'
4792 language 'internal';
4793 create function char16icregexeq (char16, text) returns bool as 'foo'
4794 language 'internal';
4795 create function char16icregexne (char16, text) returns bool as 'foo'
4796 language 'internal';
4797 create function texticregexeq (text, text) returns bool as 'foo'
4798 language 'internal';
4799 create function texticregexne (text, text) returns bool as 'foo'
4800 language 'internal';
4802 -- add builtin functions that are new to 1.01
4804 create operator = (leftarg = int4, rightarg = oid, procedure = int4eqoid);
4805 create operator = (leftarg = oid, rightarg = int4, procedure = oideqint4);
4806 create operator ~* (leftarg = char2, rightarg = text, procedure = char2icregexeq);
4807 create operator !~* (leftarg = char2, rightarg = text, procedure = char2icregexne);
4808 create operator ~* (leftarg = char4, rightarg = text, procedure = char4icregexeq);
4809 create operator !~* (leftarg = char4, rightarg = text, procedure = char4icregexne);
4810 create operator ~* (leftarg = char8, rightarg = text, procedure = char8icregexeq);
4811 create operator !~* (leftarg = char8, rightarg = text, procedure = char8icregexne);
4812 create operator ~* (leftarg = char16, rightarg = text, procedure = char16icregexeq);
4813 create operator !~* (leftarg = char16, rightarg = text, procedure = char16icregexne);
4814 create operator ~* (leftarg = text, rightarg = text, procedure = texticregexeq);
4815 create operator !~* (leftarg = text, rightarg = text, procedure = texticregexne);
4823 <title>Changes</title>
4828 * 1.01 is backwards compatible with 1.0 database provided the user
4829 follow the steps outlined in the MIGRATION_from_1.0_to_1.01 file.
4830 If those steps are not taken, 1.01 is not compatible with 1.0 database.
4833 * added PQdisplayTuples() to libpq and changed monitor and psql to use it
4834 * added NeXT port (requires SysVIPC implementation)
4835 * added CAST .. AS ... syntax
4836 * added ASC and DESC keywords
4837 * added 'internal' as a possible language for CREATE FUNCTION
4838 internal functions are C functions which have been statically linked
4839 into the postgres backend.
4840 * a new type "name" has been added for system identifiers (table names,
4841 attribute names, etc.) This replaces the old char16 type. The
4842 of name is set by the NAMEDATALEN #define in src/Makefile.global
4843 * a readable reference manual that describes the query language.
4844 * added host-based access control. A configuration file ($PGDATA/pg_hba)
4845 is used to hold the configuration data. If host-based access control
4846 is not desired, comment out HBA=1 in src/Makefile.global.
4847 * changed regex handling to be uniform use of Henry Spencer's regex code
4848 regardless of platform. The regex code is included in the distribution
4849 * added functions and operators for case-insensitive regular expressions.
4850 The operators are ~* and !~*.
4851 * pg_dump uses COPY instead of SELECT loop for better performance
4854 * fixed an optimizer bug that was causing core dumps when
4855 functions calls were used in comparisons in the WHERE clause
4856 * changed all uses of getuid to geteuid so that effective uids are used
4857 * psql now returns non-zero status on errors when using -c
4858 * applied public patches 1-14
4864 <sect1 id="release-1-0">
4865 <title>Release 1.0</title>
4868 <title>Release date</title>
4869 <simpara>1995-09-05</simpara>
4873 <title>Changes</title>
4878 * The copyright of <productname>Postgres</productname> 1.0 has been loosened to be freely modifiable
4879 and modifiable for any purpose. Please read the COPYRIGHT file.
4880 Thanks to Professor Michael Stonebraker for making this possible.
4883 * date formats have to be MM-DD-YYYY (or DD-MM-YYYY if you're using
4884 EUROPEAN STYLE). This follows SQL-92 specs.
4885 * "delimiters" is now a keyword
4888 * sql LIKE syntax has been added
4889 * copy command now takes an optional USING DELIMITER specification.
4890 delimiters can be any single-character string.
4891 * IRIX 5.3 port has been added.
4892 Thanks to Paul Walmsley and others.
4893 * updated pg_dump to work with new libpq
4894 * \d has been added psql
4895 Thanks to Keith Parks
4896 * regexp performance for architectures that use POSIX regex has been
4897 improved due to caching of precompiled patterns.
4898 Thanks to Alistair Crooks
4899 * a new version of libpq++
4900 Thanks to William Wanders
4903 * arbitrary userids can be specified in the createuser script
4904 * \c to connect to other databases in psql now works.
4905 * bad pg_proc entry for float4inc() is fixed
4906 * users with usecreatedb field set can now create databases without
4907 having to be usesuper
4908 * remove access control entries when the entry no longer has any
4910 * fixed non-portable datetimes implementation
4911 * added kerberos flags to the src/backend/Makefile
4912 * libpq now works with kerberos
4913 * typographic errors in the user manual have been corrected.
4914 * btrees with multiple index never worked, now we tell you they don't
4915 work when you try to use them
4921 <sect1 id="release-0-03">
4922 <title><productname>Postgres95</productname> Release 0.03</title>
4925 <title>Release date</title>
4926 <simpara>1995-07-21</simpara>
4930 <title>Changes</title>
4933 Incompatible changes:
4934 * BETA-0.3 IS INCOMPATIBLE WITH DATABASES CREATED WITH PREVIOUS VERSIONS
4935 (due to system catalog changes and indexing structure changes).
4936 * double-quote (") is deprecated as a quoting character for string literals;
4937 you need to convert them to single quotes ('). <!-- " -->
4938 * name of aggregates (eg. int4sum) are renamed in accordance with the
4939 SQL standard (eg. sum).
4940 * CHANGE ACL syntax is replaced by GRANT/REVOKE syntax.
4941 * float literals (eg. 3.14) are now of type float4 (instead of float8 in
4942 previous releases); you might have to do typecasting if you depend on it
4943 being of type float8. If you neglect to do the typecasting and you assign
4944 a float literal to a field of type float8, you may get incorrect values
4946 * LIBPQ has been totally revamped so that frontend applications
4947 can connect to multiple backends
4948 * the usesysid field in pg_user has been changed from int2 to int4 to
4949 allow wider range of Unix user ids.
4950 * the netbsd/freebsd/bsd o/s ports have been consolidated into a
4951 single BSD44_derived port. (thanks to Alistair Crooks)
4953 SQL standard-compliance (the following details changes that makes postgres95
4954 more compliant to the SQL-92 standard):
4955 * the following SQL types are now built-in: smallint, int(eger), float, real,
4956 char(N), varchar(N), date and time.
4958 The following are aliases to existing postgres types:
4960 integer, int -> int4
4961 float, real -> float4
4962 char(N) and varchar(N) are implemented as truncated text types. In
4963 addition, char(N) does blank-padding.
4964 * single-quote (') is used for quoting string literals; '' (in addition to
4965 \') is supported as means of inserting a single quote in a string
4966 * SQL standard aggregate names (MAX, MIN, AVG, SUM, COUNT) are used
4967 (Also, aggregates can now be overloaded, i.e. you can define your
4968 own MAX aggregate to take in a user-defined type.)
4969 * CHANGE ACL removed. GRANT/REVOKE syntax added.
4970 - Privileges can be given to a group using the "GROUP" keyword.
4972 GRANT SELECT ON foobar TO GROUP my_group;
4973 The keyword 'PUBLIC' is also supported to mean all users.
4975 Privileges can only be granted or revoked to one user or group
4978 "WITH GRANT OPTION" is not supported. Only class owners can change
4980 - The default access control is to to grant users readonly access.
4981 You must explicitly grant insert/update access to users. To change
4982 this, modify the line in
4983 src/backend/utils/acl.h
4984 that defines ACL_WORLD_DEFAULT
4987 * the bug where aggregates of empty tables were not run has been fixed. Now,
4988 aggregates run on empty tables will return the initial conditions of the
4989 aggregates. Thus, COUNT of an empty table will now properly return 0.
4990 MAX/MIN of an empty table will return a tuple of value NULL.
4991 * allow the use of \; inside the monitor
4992 * the LISTEN/NOTIFY asynchronous notification mechanism now work
4993 * NOTIFY in rule action bodies now work
4994 * hash indexes work, and access methods in general should perform better.
4995 creation of large btree indexes should be much faster. (thanks to Paul
4998 Other changes and enhancements:
4999 * addition of an EXPLAIN statement used for explaining the query execution
5000 plan (eg. "EXPLAIN SELECT * FROM EMP" prints out the execution plan for
5002 * WARN and NOTICE messages no longer have timestamps on them. To turn on
5003 timestamps of error messages, uncomment the line in
5004 src/backend/utils/elog.h:
5005 /* define ELOG_TIMESTAMPS */
5006 * On an access control violation, the message
5007 "Either no such class or insufficient privilege"
5008 will be given. This is the same message that is returned when
5009 a class is not found. This dissuades non-privileged users from
5010 guessing the existence of privileged classes.
5011 * some additional system catalog changes have been made that are not
5012 visible to the user.
5015 * The -oid option has been added to the "pg_result" tcl command.
5016 pg_result -oid returns oid of the last tuple inserted. If the
5017 last command was not an INSERT, then pg_result -oid returns "".
5018 * the large object interface is available as pg_lo* tcl commands:
5019 pg_lo_open, pg_lo_close, pg_lo_creat, etc.
5021 Portability enhancements and New Ports:
5022 * flex/lex problems have been cleared up. Now, you should be able to use
5023 flex instead of lex on any platforms. We no longer make assumptions of
5024 what lexer you use based on the platform you use.
5025 * The Linux-ELF port is now supported. Various configuration have been
5026 tested: The following configuration is known to work:
5027 kernel 1.2.10, gcc 2.6.3, libc 4.7.2, flex 2.5.2, bison 1.24
5028 with everything in ELF format,
5031 * ipcclean added to the distribution
5032 ipcclean usually does not need to be run, but if your backend crashes
5033 and leaves shared memory segments hanging around, ipcclean will
5034 clean them up for you.
5037 * the user manual has been revised and libpq documentation added.
5043 <sect1 id="release-0-02">
5044 <title><productname>Postgres95</productname> Release 0.02</title>
5047 <title>Release date</title>
5048 <simpara>1995-05-25</simpara>
5052 <title>Changes</title>
5056 Incompatible changes:
5057 * The SQL statement for creating a database is 'CREATE DATABASE' instead
5058 of 'CREATEDB'. Similarly, dropping a database is 'DROP DATABASE' instead
5059 of 'DESTROYDB'. However, the names of the executables 'createdb' and
5060 'destroydb' remain the same.
5063 * pgperl - a Perl (4.036) interface to Postgres95
5064 * pg_dump - a utility for dumping out a postgres database into a
5065 script file containing query commands. The script files are in a ASCII
5066 format and can be used to reconstruct the database, even on other
5067 machines and other architectures. (Also good for converting
5068 a Postgres 4.2 database to Postgres95 database.)
5070 The following ports have been incorporated into postgres95-beta-0.02:
5071 * the NetBSD port by Alistair Crooks
5072 * the AIX port by Mike Tung
5073 * the Windows NT port by Jon Forrest (more stuff but not done yet)
5074 * the Linux ELF port by Brian Gallew
5076 The following bugs have been fixed in postgres95-beta-0.02:
5077 * new lines not escaped in COPY OUT and problem with COPY OUT when first
5079 * cannot type return to use the default user id in createuser
5080 * SELECT DISTINCT on big tables crashes
5081 * Linux installation problems
5082 * monitor doesn't allow use of 'localhost' as PGHOST
5083 * psql core dumps when doing \c or \l
5084 * the "pgtclsh" target missing from src/bin/pgtclsh/Makefile
5085 * libpgtcl has a hard-wired default port number
5086 * SELECT DISTINCT INTO TABLE hangs
5087 * CREATE TYPE doesn't accept 'variable' as the internallength
5088 * wrong result using more than 1 aggregate in a SELECT
5094 <sect1 id="release-0-01">
5095 <title><productname>Postgres95</productname> Release 0.01</title>
5098 <title>Release date</title>
5099 <simpara>1995-05-01</simpara>
5108 <sect1 id="timing-results">
5109 <title>Timing Results</title>
5112 These timing results are from running the regression test with the commands
5115 % cd src/test/regress
5121 Timing under Linux 2.0.27 seems to have a roughly 5% variation from run
5122 to run, presumably due to the scheduling vagaries of multitasking systems.
5126 <title>Version 6.5</title>
5129 As has been the case for previous releases, timing between
5130 releases is not directly comparable since new regression tests
5131 have been added. In general, 6.5 is faster than previous
5136 Timing with <function>fsync()</function> disabled:
5140 02:00 Dual Pentium Pro 180, 224MB, UW-SCSI, Linux 2.0.36, gcc 2.7.2.3 -O2 -m486
5141 04:38 Sparc Ultra 1 143MHz, 64MB, Solaris 2.6
5146 Timing with <function>fsync()</function> enabled:
5150 04:21 Dual Pentium Pro 180, 224MB, UW-SCSI, Linux 2.0.36, gcc 2.7.2.3 -O2 -m486
5153 For the <systemitem class="osname">Linux</systemitem> system above, using <acronym>UW-SCSI</acronym> disks rather than (older) <acronym>IDE</acronym>
5154 disks leads to a 50% improvement in speed on the regression test.
5159 <title>Version 6.4beta</title>
5162 The times for this release are not directly comparable to those for previous releases
5163 since some additional regression tests have been included.
5164 In general, however, 6.4 should be slightly faster than the previous release (thanks, Bruce!).
5169 02:26 Dual Pentium Pro 180, 96MB, UW-SCSI, Linux 2.0.30, gcc 2.7.2.1 -O2 -m486
5175 <title>Version 6.3</title>
5178 The times for this release are not directly comparable to those for previous releases
5179 since some additional regression tests have been included and some obsolete tests involving
5180 time travel have been removed.
5181 In general, however, 6.3 is substantially faster than previous releases (thanks, Bruce!).
5186 02:30 Dual Pentium Pro 180, 96MB, UW-SCSI, Linux 2.0.30, gcc 2.7.2.1 -O2 -m486
5187 04:12 Dual Pentium Pro 180, 96MB, EIDE, Linux 2.0.30, gcc 2.7.2.1 -O2 -m486
5193 <title>Version 6.1</title>
5198 06:12 Pentium Pro 180, 32MB, EIDE, Linux 2.0.30, gcc 2.7.2 -O2 -m486
5199 12:06 P-100, 48MB, Linux 2.0.29, gcc
5200 39:58 Sparc IPC 32MB, Solaris 2.5, gcc 2.7.2.1 -O -g
5208 <!-- Keep this comment at the end of the file
5213 sgml-minimize-attributes:nil
5214 sgml-always-quote-attributes:t
5217 sgml-parent-document:nil
5218 sgml-default-dtd-file:"./reference.ced"
5219 sgml-exposed-tags:nil
5220 sgml-local-catalogs:("/usr/lib/sgml/catalog")
5221 sgml-local-ecat-files:nil
projects / postgresql / blob