2 Chapter 0. Release Notes
31 This is basically a cleanup release for 6.5.2. We have added a new pgaccess
32 that was missing in 6.5.2, and installed an NT-specific fix.
36 A dump/restore is not required for those running 6.5.*.
40 Updated version of pgaccess 0.98
43 ------------------------------------------------------------------------
46 ------------------------------------------------------------------------
50 This is basically a cleanup release for 6.5.1. We have fixed a variety of
51 problems reported by 6.5.1 users.
55 A dump/restore is not required for those running 6.5.*.
59 subselect+CASE fixes(Tom)
60 Add SHLIB_LINK setting for solaris_i386 and solaris_sparc ports(Daren Sefcik)
61 Fixes for CASE in WHERE join clauses(Tom)
63 Repair the check for redundant UNIQUE and PRIMARY KEY indices(Thomas)
64 Improve it so that it checks for multi-column constraints(Thomas)
65 Fix for Win32 making problem with MB enabled(Hiroki Kataoka)
66 Allow BSD yacc and bison to compile pl code(Bruce)
69 Fix vacuum's memory consumption(Hiroshi,Tatsuo)
70 Reduce the total memory consumption of vacuum(Tom)
71 Fix for timestamp(datetime)
72 Rule deparsing bugfixes(Tom)
73 Fix quoting problems in mkMakefile.tcldefs.sh.in and mkMakefile.tkdefs.sh.in(Tom)
74 This is to re-use space on index pages freed by vacuum(Vadim)
75 document -x for pg_dump(Bruce)
76 Fix for unary operators in rule deparser(Tom)
77 Comment out FileUnlink of excess segments during mdtruncate()(Tom)
78 Irix linking fix from Yu Cao yucao@falcon.kla-tencor.com
79 Repair logic error in LIKE: should not return LIKE_ABORT
80 when reach end of pattern before end of text(Tom)
81 Repair incorrect cleanup of heap memory allocation during transaction abort(Tom)
82 Updated version of pgaccess 0.98
84 ------------------------------------------------------------------------
85 Release Notes Release 6.5.1
87 ------------------------------------------------------------------------
91 This is basically a cleanup release for 6.5. We have fixed a variety of
92 problems reported by 6.5 users.
96 A dump/restore is not required for those running 6.5.
101 Portability fixes for linux_ppc, Irix, linux_alpha, OpenBSD, alpha
102 Remove QUERY_LIMIT, use SELECT...LIMIT
103 Fix for EXPLAIN on inheritance(Tom)
104 Patch to allow vacuum on multi-segment tables(Hiroshi)
105 R-Tree optimizer selectivity fix(Tom)
106 ACL file descriptor leak fix(Atsushi Ogawa)
107 New expresssion subtree code(Tom)
108 Avoid disk writes for read-only transactions(Vadim)
109 Fix for removal of temp tables if last transaction was aborted(Bruce)
110 Fix to prevent too large tuple from being created(Bruce)
112 Allow port numbers 32k - 64k(Bruce)
113 Add ^ precidence(Bruce)
114 Rename sort files called pg_temp to pg_sorttemp(Bruce)
115 Fix for microseconds in time values(Tom)
116 Tutorial source cleanup
118 Fix for sorting of NULL's in some cases(Tom)
119 Shared library dependencies fixed (Tom)
120 Fixed glitches affecting GROUP BY in subselects(Tom)
121 Fix some compiler warnings (Tomoaki Nishiyama)
122 Add Win1250 (Czech) support (Pavel Behal)
124 ------------------------------------------------------------------------
125 Release 6.5.2 Release 6.5
127 ------------------------------------------------------------------------
131 This release marks a major step in the development team's mastery of the
132 source code we inherited from Berkeley. You will see we are now easily
133 adding major features, thanks to the increasing size and experience of our
134 world-wide development team.
136 Here is a brief summary of the more notable changes:
138 Multi-version concurrency control(MVCC)
140 This removes our old table-level locking, and replaces it with a
141 locking system that is superior to most commercial database systems. In
142 a traditional system, each row that is modified is locked until
143 committed, preventing reads by other users. MVCC uses the natural
144 multi-version nature of PostgreSQL to allow readers to continue reading
145 consistent data during writer activity. Writers continue to use the
146 compact pg_log transaction system. This is all performed without having
147 to allocate a lock for every row like traditional database systems. So,
148 basically, we no longer are restricted by simple table-level locking;
149 we have something better than row-level locking.
151 Hot backups from pg_dump
153 pg_dump takes advantage of the new MVCC features to give a consistant
154 database dump/backup while the database stays online and available for
159 We now have a true numeric data type, with user-specified precision.
163 Temporary tables are guaranteed to have unique names within a database
164 session, and are destroyed on session exit.
168 We now have CASE, INTERSECT, and EXCEPT statement support. We have new
169 LIMIT/OFFSET, SET TRANSACTION ISOLATION LEVEL, SELECT ... FOR UPDATE,
170 and an improved LOCK TABLE command.
174 We continue to speed up PostgreSQL, thanks to the variety of talents
175 within our team. We have sped up memory allocation, optimization, table
176 joins, and row transfer routines.
180 We continue to expand our port list, this time including WinNT/ix86 and
185 Most interfaces have new versions, and existing functionality has been
190 New and updated material is present throughout the documentation. New
191 FAQs have been contributed for SGI and AIX platforms. The Tutorial has
192 introductory information on SQL from Stefan Simkovics. For the User's
193 Guide, there are reference pages covering the postmaster and more
194 utility programs, and a new appendix contains details on date/time
195 behavior. The Administrator's Guide has a new chapter on
196 troubleshooting from Tom Lane. And the Programmer's Guide has a
197 description of query processing, also from Stefan, and details on
198 obtaining the Postgres source tree via anonymous CVS and CVSup.
202 A dump/restore using pg_dump is required for those wishing to migrate data
203 from any previous release of Postgres. pg_upgrade can not be used to upgrade
204 to this release because the on-disk structure of the tables has changed
205 compared to previous releases.
207 The new Multi-Version Concurrency Control (MVCC) features can give somewhat
208 different behaviors in multi-user environments. Read and understand the
209 following section to ensure that your existing applications will give you
210 the behavior you need.
212 Multi-Version Concurrency Control
214 Because readers in 6.5 don't lock data, regardless of transaction isolation
215 level, data read by one transaction can be overwritten by another. In other
216 words, if a row is returned by SELECT it doesn't mean that this row really
217 exists at the time it is returned (i.e. sometime after the statement or
218 transaction began) nor that the row is protected from being deleted or
219 updated by concurrent transactions before the current transaction does a
222 To ensure the actual existence of a row and protect it against concurrent
223 updates one must use SELECT FOR UPDATE or an appropriate LOCK TABLE
224 statement. This should be taken into account when porting applications from
225 previous releases of Postgres and other environments.
227 Keep the above in mind if you are using contrib/refint.* triggers for
228 referential integrity. Additional technics are required now. One way is to
229 use LOCK parent_table IN SHARE ROW EXCLUSIVE MODE command if a transaction
230 is going to update/delete a primary key and use LOCK parent_table IN SHARE
231 MODE command if a transaction is going to update/insert a foreign key.
233 Note: Note that if you run a transaction in SERIALIZABLE mode then
234 you must execute the LOCK commands above before execution of any
235 DML statement (SELECT/INSERT/DELETE/UPDATE/FETCH/COPY_TO) in the
238 These inconveniences will disappear in the future when the ability to read
239 dirty (uncommitted) data (regardless of isolation level) and true
240 referential integrity will be implemented.
246 Fix text<->float8 and text<->float4 conversion functions(Thomas)
247 Fix for creating tables with mixed-case constraints(Billy)
248 Change exp()/pow() behavior to generate error on underflow/overflow(Jan)
249 Fix bug in pg_dump -z
250 Memory overrun cleanups(Tatsuo)
251 Fix for lo_import crash(Tatsuo)
252 Adjust handling of data type names to suppress double quotes(Thomas)
253 Use type coersion for matching columns and DEFAULT(Thomas)
254 Fix deadlock so it only checks once after one second of sleep(Bruce)
255 Fixes for aggregates and PL/pgsql(Hiroshi)
256 Fix for subquery crash(Vadim)
257 Fix for libpq function PQfnumber and case-insensitive names(Bahman Rafatjoo)
258 Fix for large object write-in-middle, no extra block, memory consumption(Tatsuo)
259 Fix for pg_dump -d or -D and quote special characters in INSERT
260 Repair serious problems with dynahash(Tom)
261 Fix INET/CIDR portability problems
262 Fix problem with selectivity error in ALTER TABLE ADD COLUMN(Bruce)
263 Fix executor so mergejoin of different column types works(Tom)
264 Fix for Alpha OR selectivity bug
265 Fix OR index selectivity problem(Bruce)
266 Fix so \d shows proper length for char()/varchar()(Ryan)
267 Fix tutorial code(Clark)
268 Improve destroyuser checking(Oliver)
269 Fix for Kerberos(Rodney McDuff)
270 Fix for dropping database while dirty buffers(Bruce)
271 Fix so sequence nextval() can be case-sensitive(Bruce)
273 Drop buffers before destroying database files(Bruce)
274 Fix case where executor evaluates functions twice(Tatsuo)
275 Allow sequence nextval actions to be case-sensitive(Bruce)
276 Fix optimizer indexing not working for negative numbers(Bruce)
277 Fix for memory leak in executor with fjIsNull
278 Fix for aggregate memory leaks(Erik Riedel)
279 Allow username containing a dash GRANT permissions
280 Cleanup of NULL in inet types
281 Clean up system table bugs(Tom)
282 Fix problems of PAGER and \? command(Masaaki Sakaida)
283 Reduce default multi-segment file size limit to 1GB(Peter)
284 Fix for dumping of CREATE OPERATOR(Tom)
285 Fix for backward scanning of cursors(Hiroshi Inoue)
286 Fix for COPY FROM STDIN when using \i(Tom)
287 Fix for subselect is compared inside an expression(Jan)
288 Fix handling of error reporting while returning rows(Tom)
289 Fix problems with reference to array types(Tom,Jan)
290 Prevent UPDATE SET oid(Jan)
291 Fix pg_dump so -t option can handle case-sensitive tablenames
292 Fixes for GROUP BY in special cases(Tom, Jan)
293 Fix for memory leak in failed queries(Tom)
294 DEFAULT now supports mixed-case identifiers(Tom)
295 Fix for multi-segment uses of DROP/RENAME table, indexes(Ole Gjerde)
296 Disable use of pg_dump with both -o and -d options(Bruce)
297 Allow pg_dump to properly dump GROUP permissions(Bruce)
298 Fix GROUP BY in INSERT INTO table SELECT * FROM table2(Jan)
299 Fix for computations in views(Jan)
300 Fix for aggregates on array indexes(Tom)
301 Fix for DEFAULT handles single quotes in value requiring too many quotes
302 Fix security problem with non-super users importing/exporting large objects(Tom)
303 Rollback of transaction that creates table cleaned up properly(Tom)
304 Fix to allow long table and column names to generate proper serial names(Tom)
308 Add "vacuumdb" utility
309 Speed up libpq by allocating memory better(Tom)
310 EXPLAIN all indices used(Tom)
311 Implement CASE, COALESCE, NULLIF expression(Thomas)
312 New pg_dump table output format(Constantin)
313 Add string min()/max() functions(Thomas)
314 Extend new type coersion techniques to aggregates(Thomas)
315 New moddatetime contrib(Terry)
316 Update to pgaccess 0.96(Constantin)
317 Add routines for single-byte "char" type(Thomas)
318 Improved substr() function(Thomas)
319 Improved multi-byte handling(Tatsuo)
320 Multi-version concurrency control/MVCC(Vadim)
321 New Serialized mode(Vadim)
322 Fix for tables over 2gigs(Peter)
323 New SET TRANSACTION ISOLATION LEVEL(Vadim)
324 New LOCK TABLE IN ... MODE(Vadim)
325 Update ODBC driver(Byron)
326 New NUMERIC data type(Jan)
327 New SELECT FOR UPDATE(Vadim)
328 Handle "NaN" and "Infinity" for input values(Jan)
329 Improved date/year handling(Thomas)
330 Improved handling of backend connections(Magnus)
331 New options ELOG_TIMESTAMPS and USE_SYSLOG options for log files(Massimo)
332 New TCL_ARRAYS option(Massimo)
333 New INTERSECT and EXCEPT(Stefan)
334 New pg_index.indisprimary for primary key tracking(D'Arcy)
335 New pg_dump option to allow dropping of tables before creation(Brook)
336 Speedup of row output routines(Tom)
337 New READ COMMITTED isolation level(Vadim)
338 New TEMP tables/indexes(Bruce)
339 Prevent sorting if result is already sorted(Jan)
340 New memory allocation optimization(Jan)
341 Allow psql to do \p\g(Bruce)
342 Allow multiple rule actions(Jan)
343 Added LIMIT/OFFSET functionality(Jan)
344 Improve optimizer when joining a large number of tables(Bruce)
345 New intro to SQL from S. Simkovics' Master's Thesis (Stefan, Thomas)
346 New intro to backend processing from S. Simkovics' Master's Thesis (Stefan)
347 Improved int8 support(Ryan Bradetich, Thomas, Tom)
348 New routines to convert between int8 and text/varchar types(Thomas)
349 New bushy plans, where meta-tables are joined(Bruce)
350 Enable right-hand queries by default(Bruce)
351 Allow reliable maximum number of backends to be set at configure time
352 (--with-maxbackends and postmaster switch (-N backends))(Tom)
353 GEQO default now 10 tables because of optimizer speedups(Tom)
354 Allow NULL=Var for MS-SQL portability(Michael, Bruce)
355 Modify contrib check_primary_key() so either "automatic" or "dependent"(Anand)
356 Allow psql \d on a view show query(Ryan)
357 Speedup for LIKE(Bruce)
358 Ecpg fixes/features, see src/interfaces/ecpg/ChangeLog file(Michael)
359 JDBC fixes/features, see src/interfaces/jdbc/CHANGELOG(Peter)
360 Make % operator have precedence like /(Bruce)
361 Add new postgres -O option to allow system table structure changes(Bruce)
362 Update contrib/pginterface/findoidjoins script(Tom)
363 Major speedup in vacuum of deleted rows with indexes(Vadim)
364 Allow non-SQL functions to run different versions based on arguments(Tom)
365 Add -E option that shows actual queries sent by \dt and friends(Masaaki Sakaida)
366 Add version number in startup banners for psql(Masaaki Sakaida)
367 New contrib/vacuumlo removes large objects not referenced(Peter)
368 New initialization for table sizes so non-vacuumed tables perform better(Tom)
369 Improve error messages when a connection is rejected(Tom)
370 Support for arrays of char() and varchar() fields(Massimo)
371 Overhaul of hash code to increase reliability and performance(Tom)
372 Update to PyGreSQL 2.4(D'Arcy)
373 Changed debug options so -d4 and -d5 produce different node displays(Jan)
374 New pg_options: pretty_plan, pretty_parse, pretty_rewritten(Jan)
375 Better optimization statistics for system table access(Tom)
376 Better handling of non-default block sizes(Massimo)
377 Improve GEQO optimizer memory consumption(Tom)
378 UNION now suppports ORDER BY of columns not in target list(Jan)
379 Major libpq++ improvements(Vince Vielhaber)
380 pg_dump now uses -z(ACL's) as default(Bruce)
381 backend cache, memory speedups(Tom)
382 have pg_dump do everything in one snapshot transaction(Vadim)
383 fix for large object memory leakage, fix for pg_dumping(Tom)
384 INET type now respects netmask for comparisons
385 Make VACUUM ANALYZE only use a readlock(Vadim)
386 Allow VIEWs on UNIONS(Jan)
387 pg_dump now can generate consistent snapshots on active databases(Vadim)
391 Improve port matching(Tom)
392 Portability fixes for SunOS
393 Add NT/Win32 backend port and enable dynamic loading(Magnus and Daniel Horak)
394 New port to Cobalt Qube(Mips) running Linux(Tatsuo)
395 Port to NetBSD/m68k(Mr. Mutsuki Nakajima)
396 Port to NetBSD/sun3(Mr. Mutsuki Nakajima)
397 Port to NetBSD/macppc(Toshimi Aoki)
398 Fix for tcl/tk configuration(Vince)
399 Removed CURRENT keyword for rule queries(Jan)
400 NT dynamic loading now works(Daniel Horak)
401 Add ARM32 support(Andrew McMurry)
402 Better support for HPUX 11 and Unixware
403 Improve file handling to be more uniform, prevent file descriptor leak(Tom)
404 New install commands for plpgsql(Jan)
407 ------------------------------------------------------------------------
408 Release 6.5.1 Release 6.4.2
410 ------------------------------------------------------------------------
414 The 6.4.1 release was improperly packaged. This also has one additional bug
419 A dump/restore is not required for those running 6.4.*.
423 Fix for datetime constant problem on some platforms(Thomas)
425 ------------------------------------------------------------------------
426 Release 6.5 Release 6.4.1
428 ------------------------------------------------------------------------
432 This is basically a cleanup release for 6.4. We have fixed a variety of
433 problems reported by 6.4 users.
437 A dump/restore is not required for those running 6.4.
441 Add pg_dump -N flag to force double quotes around identifiers. This is
443 Fix for NOT in where clause causing crash(Bruce)
444 EXPLAIN VERBOSE coredump fix(Vadim)
445 Fix shared-library problems on Linux
446 Fix test for table existance to allow mixed-case and whitespace in
447 the table name(Thomas)
448 Fix a couple of pg_dump bugs
449 Configure matches template/.similar entries better(Tom)
450 Change builtin function names from SPI_* to spi_*
451 OR WHERE clause fix(Vadim)
452 Fixes for mixed-case table names(Billy)
453 contrib/linux/postgres.init.csh/sh fix(Thomas)
454 libpq memory overrun fix
456 Change exp() behavior to generate error on underflow(Thomas)
457 pg_dump fixes for memory leak, inheritance constraints, layout change
458 update pgaccess to 0.93
459 Fix prototype for 64-bit platforms
460 Multi-byte fixes(Tatsuo)
462 Fix memory overruns(Tatsuo)
463 Fix for lo_import() crash(Bruce)
464 Better search for install program(Tom)
467 Use implicit type coersion for matching DEFAULT values(Thomas)
468 Add routines to help with single-byte (internal) character type(Thomas)
469 Compilation of libpq for Win32 fixes(Magnus)
470 Upgrade to PyGreSQL 2.2(D'Arcy)
472 ------------------------------------------------------------------------
473 Release 6.4.2 Release 6.4
475 ------------------------------------------------------------------------
479 There are many new features and improvements in this release. Thanks to our
480 developers and maintainers, nearly every aspect of the system has received
481 some attention since the previous release. Here is a brief, incomplete
484 * Views and rules are now functional thanks to extensive new code in the
485 rewrite rules system from Jan Wieck. He also wrote a chapter on it for
486 the Programmer's Guide.
488 * Jan also contributed a second procedural language, PL/pgSQL, to go with
489 the original PL/pgTCL procedural language he contributed last release.
491 * We have optional multiple-byte character set support from Tatsuo Iishi
492 to complement our existing locale support.
494 * Client/server communications has been cleaned up, with better support
495 for asynchronous messages and interrupts thanks to Tom Lane.
497 * The parser will now perform automatic type coersion to match arguments
498 to available operators and functions, and to match columns and
499 expressions with target columns. This uses a generic mechanism which
500 supports the type extensibility features of Postgres. There is a new
501 chapter in the User's Guide which covers this topic.
503 * Three new data types have been added. Two types, inet and cidr, support
504 various forms of IP network, subnet, and machine addressing. There is
505 now an 8-byte integer type available on some platforms. See the chapter
506 on data types in the User's Guide for details. A fourth type, serial,
507 is now supported by the parser as an amalgam of the int4 type, a
508 sequence, and a unique index.
510 * Several more SQL92-compatible syntax features have been added,
511 including INSERT DEFAULT VALUES
513 * The automatic configuration and installation system has received some
514 attention, and should be more robust for more platforms than it has
519 A dump/restore using pg_dump or pg_dumpall is required for those wishing to
520 migrate data from any previous release of Postgres.
526 Fix for a tiny memory leak in PQsetdb/PQfinish(Bryan)
527 Remove char2-16 data types, use char/varchar(Darren)
528 Pqfn not handles a NOTICE message(Anders)
529 Reduced busywaiting overhead for spinlocks with many backends (dg)
530 Stuck spinlock detection (dg)
531 Fix up "ISO-style" timespan decoding and encoding(Thomas)
532 Fix problem with table drop after rollback of transaction(Vadim)
533 Change error message and remove non-functional update message(Vadim)
534 Fix for COPY array checking
535 Fix for SELECT 1 UNION SELECT NULL
536 Fix for buffer leaks in large object calls(Pascal)
537 Change owner from oid to int4 type(Bruce)
538 Fix a bug in the oracle compatibility functions btrim() ltrim() and rtrim()
539 Fix for shared invalidation cache overflow(Massimo)
540 Prevent file descriptor leaks in failed COPY's(Bruce)
541 Fix memory leak in libpgtcl's pg_select(Constantin)
542 Fix problems with username/passwords over 8 characters(Tom)
543 Fix problems with handling of asynchronous NOTIFY in backend(Tom)
544 Fix of many bad system table entries(Tom)
548 Upgrade ecpg and ecpglib,see src/interfaces/ecpc/ChangeLog(Michael)
549 Show the index used in an EXPLAIN(Zeugswetter)
550 EXPLAIN invokes rule system and shows plan(s) for rewritten queries(Jan)
551 Multi-byte awareness of many data types and functions, via configure(Tatsuo)
552 New configure --with-mb option(Tatsuo)
553 New initdb --pgencoding option(Tatsuo)
554 New createdb -E multibyte option(Tatsuo)
555 Select version(); now returns PostgreSQL version(Jeroen)
556 Libpq now allows asynchronous clients(Tom)
557 Allow cancel from client of backend query(Tom)
558 Psql now cancels query with Control-C(Tom)
559 Libpq users need not issue dummy queries to get NOTIFY messages(Tom)
560 NOTIFY now sends sender's PID, so you can tell whether it was your own(Tom)
561 PGresult struct now includes associated error message, if any(Tom)
562 Define "tz_hour" and "tz_minute" arguments to date_part()(Thomas)
563 Add routines to convert between varchar and bpchar(Thomas)
564 Add routines to allow sizing of varchar and bpchar into target columns(Thomas)
565 Add bit flags to support timezonehour and minute in data retrieval(Thomas)
566 Allow more variations on valid floating point numbers (e.g. ".1", "1e6")(Thomas)
567 Fixes for unary minus parsing with leading spaces(Thomas)
568 Implement TIMEZONE_HOUR, TIMEZONE_MINUTE per SQL92 specs(Thomas)
569 Check for and properly ignore FOREIGN KEY column constraints(Thomas)
570 Define USER as synonym for CURRENT_USER per SQL92 specs(Thomas)
571 Enable HAVING clause but no fixes elsewhere yet.
572 Make "char" type a synonym for "char(1)" (actually implemented as bpchar)(Thomas)
573 Save string type if specified for DEFAULT clause handling(Thomas)
574 Coerce operations involving different data types(Thomas)
575 Allow some index use for columns of different types(Thomas)
576 Add capabilities for automatic type conversion(Thomas)
577 Cleanups for large objects, so file is truncated on open(Peter)
578 Readline cleanups(Tom)
579 Allow psql \f \ to make spaces as delimiter(Bruce)
580 Pass pg_attribute.atttypmod to the frontend for column field lengths(Tom,Bruce)
581 Msql compatibility library in /contrib(Aldrin)
582 Remove the requirement that ORDER/GROUP BY clause identifiers be
583 included in the target list(David)
584 Convert columns to match columns in UNION clauses(Thomas)
585 Remove fork()/exec() and only do fork()(Bruce)
587 Show backend status on ps command line(only works on some platforms)(Bruce)
588 Pg_hba.conf now has a sameuser option in the database field
589 Make lo_unlink take oid param, not int4
590 New DISABLE_COMPLEX_MACRO for compilers that can't handle our macros(Bruce)
591 Libpgtcl now handles NOTIFY as a Tcl event, need not send dummy queries(Tom)
592 libpgtcl cleanups(Tom)
593 Add -error option to libpgtcl's pg_result command(Tom)
594 New locale patch, see docs/README/locale(Oleg)
595 Fix for pg_dump so CONSTRAINT and CHECK syntax is correct(ccb)
596 New contrib/lo code for large object orphan removal(Peter)
597 New psql command "SET CLIENT_ENCODING TO 'encoding'" for multi-bytes
598 feature, see /doc/README.mb(Tatsuo)
599 /contrib/noupdate code to revoke update permission on a column
600 Libpq can now be compiled on win32(Magnus)
601 Add PQsetdbLogin() in libpq
602 New 8-byte integer type, checked by configure for OS support(Thomas)
603 Better support for quoted table/column names(Thomas)
604 Surround table and column names with double-quotes in pg_dump(Thomas)
605 PQreset() now works with passwords(Tom)
606 Handle case of GROUP BY target list column number out of range(David)
607 Allow UNION in subselects
608 Add auto-size to screen to \d? commands(Bruce)
609 Use UNION to show all \d? results in one query(Bruce)
610 Add \d? field search feature(Bruce)
611 Pg_dump issues fewer \connect requests(Tom)
612 Make pg_dump -z flag work better, document it in manual page(Tom)
613 Add HAVING clause with full support for subselects and unions(Stephan)
614 Full text indexing routines in contrib/fulltextindex(Maarten)
615 Transaction ids now stored in shared memory(Vadim)
616 New PGCLIENTENCODING when issuing COPY command(Tatsuo)
617 Support for SQL92 syntax "SET NAMES"(Tatsuo)
618 Support for LATIN2-5(Tatsuo)
619 Add UNICODE regression test case(Tatsuo)
620 Lock manager cleanup, new locking modes for LLL(Vadim)
621 Allow index use with OR clauses(Bruce)
622 Allows "SELECT NULL ORDER BY 1;"
623 Explain VERBOSE prints the plan, and now pretty-prints the plan to
624 the postmaster log file(Bruce)
625 Add Indices display to \d command(Bruce)
626 Allow GROUP BY on functions(David)
627 New pg_class.relkind for large objects(Bruce)
628 New way to send libpq NOTICE messages to a different location(Tom)
629 New \w write command to psql(Bruce)
630 New /contrib/findoidjoins scans oid columns to find join relationships(Bruce)
631 Allow binary-compatible indices to be considered when checking for valid
632 indices for restriction clauses containing a constant(Thomas)
633 New ISBN/ISSN code in /contrib/isbn_issn
634 Allow NOT LIKE, IN, NOT IN, BETWEEN, and NOT BETWEEN constraint(Thomas)
635 New rewrite system fixes many problems with rules and views(Jan)
636 * Rules on relations work
637 * Event qualifications on insert/update/delete work
638 * New OLD variable to reference CURRENT, CURRENT will be remove in future
639 * Update rules can reference NEW and OLD in rule qualifications/actions
640 * Insert/update/delete rules on views work
641 * Multiple rule actions are now supported, surrounded by parentheses
642 * Regular users can create views/rules on tables they have RULE permits
643 * Rules and views inherit the permissions on the creator
644 * No rules at the column level
645 * No UPDATE NEW/OLD rules
646 * New pg_tables, pg_indexes, pg_rules and pg_views system views
647 * Only a single action on SELECT rules
648 * Total rewrite overhaul, perhaps for 6.5
650 * handle aggregates on views
651 * handle insert into select from view works
652 System indexes are now multi-key(Bruce)
653 Oidint2, oidint4, and oidname types are removed(Bruce)
654 Use system cache for more system table lookups(Bruce)
655 New backend programming language PL/pgSQL in backend/pl(Jan)
656 New SERIAL data type, auto-creates sequence/index(Thomas)
657 Enable assert checking without a recompile(Massimo)
658 User lock enhancements(Massimo)
659 New setval() command to set sequence value(Massimo)
660 Auto-remove unix socket file on startup if no postmaster running(Massimo)
661 Conditional trace package(Massimo)
662 New UNLISTEN command(Massimo)
663 Psql and libpq now compile under win32 using win32.mak(Magnus)
664 Lo_read no longer stores trailing NULL(Bruce)
665 Identifiers are now truncated to 31 characters internally(Bruce)
666 Createuser options now availble on the command line
667 Code for 64-bit integer supported added, configure tested, int8 type(Thomas)
668 Prevent file descriptor leaf from failed COPY(Bruce)
669 New pg_upgrade command(Bruce)
670 Updated /contrib directories(Massimo)
671 New CREATE TABLE DEFAULT VALUES statement available(Thomas)
672 New INSERT INTO TABLE DEFAULT VALUES statement available(Thomas)
673 New DECLARE and FETCH feature(Thomas)
674 libpq's internal structures now not exported(Tom)
675 Allow up to 8 key indexes(Bruce)
676 Remove ARCHIVE keyword, that is no longer used(Thomas)
677 pg_dump -n flag to supress quotes around indentifiers
678 disable system columns for views(Jan)
679 new INET and CIDR types for network addresses(TomH, Paul)
680 no more double quotes in psql output
681 pg_dump now dumps views(Terry)
682 new SET QUERY_LIMIT(Tatsuo,Jan)
686 /contrib cleanup(Jun)
687 Inline some small functions called for every row(Bruce)
690 Multi-byte regression tests(Soonmyung.)
691 Remove --disabled options from configure
692 Define PGDOC to use POSTGRESDIR by default
693 Make regression optional
694 Remove extra braces code to pgindent(Bruce)
695 Add bsdi shared library support(Bruce)
696 New --without-CXX support configure option(Brook)
698 Update backend flowchart in tools/backend(Bruce)
699 Change atttypmod from int16 to int32(Bruce, Tom)
700 Getrusage() fix for platforms that do not have it(Tom)
701 Add PQconnectdb, PGUSER, PGPASSWORD to libpq man page
702 NS32K platform fixes(Phil Nelson, John Buller)
703 Sco 7/UnixWare 2.x fixes(Billy,others)
704 Sparc/Solaris 2.5 fixes(Ryan)
705 Pgbuiltin.3 is obsolete, move to doc files(Thomas)
706 Even more documention(Thomas)
707 Nextstep support(Jacek)
709 pginterface manual page(Bruce)
710 shared libraries all have version numbers
711 merged all OS-specific shared library defines into one file
712 smarter TCL/TK configuration checking(Billy)
713 smarter perl configuration(Brook)
714 configure uses supplied install-sh if no install script found(Tom)
715 new Makefile.shlib for shared library configuration(Tom)
717 ------------------------------------------------------------------------
718 Release 6.4.1 Release 6.3.2
720 ------------------------------------------------------------------------
724 This is a bugfix release for 6.3.x. Refer to the release notes for v6.3 for
725 a more complete summary of new features.
729 * Repairs automatic configuration support for some platforms, including
730 Linux, from breakage inadvertently introduced in v6.3.1.
732 * Correctly handles function calls on the left side of BETWEEN and LIKE
735 A dump/restore is NOT required for those running 6.3 or 6.3.1. A 'make
736 distclean', 'make', and 'make install' is all that is required. This last
737 step should be performed while the postmaster is not running. You should
738 re-link any custom applications that use Postgres libraries.
740 For upgrades from pre-v6.3 installations, refer to the installation and
741 migration instructions for v6.3.
747 Configure detection improvements for tcl/tk(Brook Milligan, Alvin)
748 Manual page improvements(Bruce)
749 BETWEEN and LIKE fix(Thomas)
750 fix for psql \connect used by pg_dump(Oliver Elphick)
752 pgaccess, version 0.86
753 qsort removed, now uses libc version, cleanups(Jeroen)
754 fix for buffer over-runs detected(Maurice Gittens)
755 fix for buffer overrun in libpgtcl(Randy Kunkee)
756 fix for UNION with DISTINCT or ORDER BY(Bruce)
757 gettimeofday configure check(Doug Winterburn)
758 Fix "indexes not used" bug(Vadim)
759 docs additions(Thomas)
760 Fix for backend memory leak(Bruce)
761 libreadline cleanup(Erwan MAS)
762 Remove DISTDIR(Bruce)
763 Makefile dependency cleanup(Jeroen van Vianen)
766 ------------------------------------------------------------------------
767 Release 6.4 Release 6.3.1
769 ------------------------------------------------------------------------
775 * Additional support for multi-byte character sets.
777 * Repair byte ordering for mixed-endian clients and servers.
779 * Minor updates to allowed SQL syntax.
781 * Improvements to the configuration autodetection for installation.
783 A dump/restore is NOT required for those running 6.3. A 'make distclean',
784 'make', and 'make install' is all that is required. This last step should be
785 performed while the postmaster is not running. You should re-link any custom
786 applications that use Postgres libraries.
788 For upgrades from pre-v6.3 installations, refer to the installation and
789 migration instructions for v6.3.
795 ecpg cleanup/fixes, now version 1.1(Michael Meskes)
796 pg_user cleanup(Bruce)
797 large object fix for pg_dump and tclsh (alvin)
798 LIKE fix for multiple adjacent underscores
799 fix for redefining builtin functions(Thomas)
801 upgrade to pg_access 0.83
802 updated CLUSTER manual page
803 multi-byte character set support, see doc/README.mb(Tatsuo)
804 configure --with-pgport fix
806 big-endian fix for backend communications(Kataoka)
807 SUBSTR() and substring() fix(Jan)
808 several jdbc fixes(Peter)
809 libpgtcl improvements, see libptcl/README(Randy Kunkee)
810 Fix for "Datasize = 0" error(Vadim)
811 Prevent \do from wrapping(Bruce)
812 Remove duplicate Russian character set entries
814 Allow optional TABLE keyword in LOCK and SELECT INTO(Thomas)
815 CREATE SEQUENCE options to allow a negative integer(Thomas)
816 Add "PASSWORD" as an allowed column identifier(Thomas)
817 Add checks for UNION target fields(Bruce)
818 Fix Alpha port(Dwayne Bailey)
819 Fix for text arrays containing quotes(Doug Gibson)
820 Solaris compile fix(Albert Chin-A-Young)
821 Better identify tcl and tk libs and includes(Bruce)
823 ------------------------------------------------------------------------
824 Release 6.3.2 Release 6.3
826 ------------------------------------------------------------------------
830 There are many new features and improvements in this release. Here is a
831 brief, incomplete summary:
833 * Many new SQL features, including full SQL92 subselect capability
834 (everything is here but target-list subselects).
836 * Support for client-side environment variables to specify time zone and
839 * Socket interface for client/server connection. This is the default now
840 so you may need to start postmaster with the "-i" flag.
842 * Better password authorization mechanisms. Default table permissions
845 * Old-style "time travel" has been removed. Performance has been
848 Note: Bruce Momjian wrote the following notes to introduce the new
851 There are some general 6.3 issues that I want to mention. These are only the
852 big items that can not be described in one sentence. A review of the
853 detailed changes list is still needed.
855 First, we now have subselects. Now that we have them, I would like to
856 mention that without subselects, SQL is a very limited language. Subselects
857 are a major feature, and you should review your code for places where
858 subselects provide a better solution for your queries. I think you will find
859 that there are more uses for subselects than you may think. Vadim has put us
860 on the big SQL map with subselects, and fully functional ones too. The only
861 thing you can't do with subselects is to use them in the target list.
863 Second, 6.3 uses unix domain sockets rather than TCP/IP by default. To
864 enable connections from other machines, you have to use the new postmaster
865 -i option, and of course edit pg_hba.conf. Also, for this reason, the format
866 of pg_hba.conf has changed.
868 Third, char() fields will now allow faster access than varchar() or text.
869 Specifically, the text and varchar() have a penalty for access to any
870 columns after the first column of this type. char() used to also have this
871 access penalty, but it no longer does. This may suggest that you redesign
872 some of your tables, especially if you have short character columns that you
873 have defined as varchar() or text. This and other changes make 6.3 even
874 faster than earlier releases.
876 We now have passwords definable independent of any Unix file. There are new
877 SQL USER commands. See the pg_hba.conf manual page for more information.
878 There is a new table, pg_shadow, which is used to store user information and
879 user passwords, and it by default only SELECT-able by the postgres
880 super-user. pg_user is now a view of pg_shadow, and is SELECT-able by
881 PUBLIC. You should keep using pg_user in your application without changes.
883 User-created tables now no longer have SELECT permission to PUBLIC by
884 default. This was done because the ANSI standard requires it. You can of
885 course GRANT any permissions you want after the table is created. System
886 tables continue to be SELECT-able by PUBLIC.
888 We also have real deadlock detection code. No more sixty-second timeouts.
889 And the new locking code implements a FIFO better, so there should be less
890 resource starvation during heavy use.
892 Many complaints have been made about inadequate documenation in previous
893 releases. Thomas has put much effort into many new manuals for this release.
894 Check out the doc/ directory.
896 For performance reasons, time travel is gone, but can be implemented using
897 triggers (see pgsql/contrib/spi/README). Please check out the new \d command
898 for types, operators, etc. Also, views have their own permissions now, not
899 based on the underlying tables, so permissions on them have to be set
900 separately. Check /pgsql/interfaces for some new ways to talk to Postgres.
902 This is the first release that really required an explanation for existing
903 users. In many ways, this was necessary because the new release removes many
904 limitations, and the work-arounds people were using are no longer needed.
908 A dump/restore using pg_dump or pg_dumpall is required for those wishing to
909 migrate data from any previous release of Postgres.
915 Fix binary cursors broken by MOVE implementation(Vadim)
916 Fix for tcl library crash(Jan)
917 Fix for array handling, from Gerhard Hintermayer
918 Fix acl error, and remove duplicate pqtrace(Bruce)
919 Fix psql \e for empty file(Bruce)
920 Fix for textcat on varchar() fields(Bruce)
921 Fix for DBT Sendproc (Zeugswetter Andres)
922 Fix vacuum analyze syntax problem(Bruce)
923 Fix for international identifiers(Tatsuo)
924 Fix aggregates on inherited tables(Bruce)
925 Fix substr() for out-of-bounds data
926 Fix for select 1=1 or 2=2, select 1=1 and 2=2, and select sum(2+2)(Bruce)
927 Fix notty output to show status result. -q option still turns it off(Bruce)
928 Fix for count(*), aggs with views and multiple tables and sum(3)(Bruce)
930 Fix for PQtrace start/stop several times(Bruce)
931 Fix a variety of locking problems like newer lock waiters getting
932 lock before older waiters, and having readlock people not share
933 locks if a writer is waiting for a lock, and waiting writers not
934 getting priority over waiting readers(Bruce)
935 Fix crashes in psql when executing queries from external files(James)
936 Fix problem with multiple order by columns, with the first one having
938 Use correct hash table support functions for float8 and int4(Thomas)
939 Re-enable JOIN= option in CREATE OPERATOR statement (Thomas)
940 Change precedence for boolean operators to match expected behavior(Thomas)
941 Generate elog(ERROR) on over-large integer(Bruce)
942 Allow multiple-argument functions in constraint clauses(Thomas)
943 Check boolean input literals for 'true','false','yes','no','1','0'
944 and throw elog(ERROR) if unrecognized(Thomas)
945 Major large objects fix
946 Fix for GROUP BY showing duplicates(Vadim)
947 Fix for index scans in MergeJion(Vadim)
951 Subselects with EXISTS, IN, ALL, ANY keywords (Vadim, Bruce, Thomas)
952 New User Manual(Thomas, others)
953 Speedup by inlining some frequently-called functions
954 Real deadlock detection, no more timeouts(Bruce)
955 Add SQL92 "constants" CURRENT_DATE, CURRENT_TIME, CURRENT_TIMESTAMP,
957 Modify constraint syntax to be SQL92-compliant(Thomas)
958 Implement SQL92 PRIMARY KEY and UNIQUE clauses using indices(Thomas)
959 Recognize SQL92 syntax for FOREIGN KEY. Throw elog notice(Thomas)
960 Allow NOT NULL UNIQUE constraint clause (each allowed separately before)(Thomas)
961 Allow Postgres-style casting ("::") of non-constants(Thomas)
962 Add support for SQL3 TRUE and FALSE boolean constants(Thomas)
963 Support SQL92 syntax for IS TRUE/IS FALSE/IS NOT TRUE/IS NOT FALSE(Thomas)
964 Allow shorter strings for boolean literals (e.g. "t", "tr", "tru")(Thomas)
965 Allow SQL92 delimited identifiers(Thomas)
966 Implement SQL92 binary and hexadecimal string decoding (b'10' and x'1F')(Thomas)
967 Support SQL92 syntax for type coercion of literal strings
968 (e.g. "DATETIME 'now'")(Thomas)
969 Add conversions for int2, int4, and OID types to and from text(Thomas)
970 Use shared lock when building indices(Vadim)
971 Free memory allocated for an user query inside transaction block after
972 this query is done, was turned off in <= 6.2.1(Vadim)
973 New SQL statement CREATE PROCEDURAL LANGUAGE(Jan)
974 New Postgres Procedural Language (PL) backend interface(Jan)
975 Rename pg_dump -H option to -h(Bruce)
976 Add Java support for passwords, European dates(Peter)
977 Use indices for LIKE and ~, !~ operations(Bruce)
978 Add hash functions for datetime and timespan(Thomas)
979 Time Travel removed(Vadim, Bruce)
980 Add paging for \d and \z, and fix \i(Bruce)
981 Add Unix domain socket support to backend and to frontend library(Goran)
982 Implement CREATE DATABASE/WITH LOCATION and initlocation utility(Thomas)
983 Allow more SQL92 and/or Postgres reserved words as column identifiers(Thomas)
984 Augment support for SQL92 SET TIME ZONE...(Thomas)
985 SET/SHOW/RESET TIME ZONE uses TZ backend environment variable(Thomas)
986 Implement SET keyword = DEFAULT and SET TIME ZONE DEFAULT(Thomas)
987 Enable SET TIME ZONE using TZ environment variable(Thomas)
988 Add PGDATESTYLE environment variable to frontend and backend initialization(Thomas)
989 Add PGTZ, PGCOSTHEAP, PGCOSTINDEX, PGRPLANS, PGGEQO
990 frontend library initialization environment variables(Thomas)
991 Regression tests time zone automatically set with "setenv PGTZ PST8PDT"(Thomas)
992 Add pg_description table for info on tables, columns, operators, types, and
994 Increase 16 char limit on system table/index names to 32 characters(Bruce)
995 Rename system indices(Bruce)
996 Add 'GERMAN' option to SET DATESTYLE(Thomas)
997 Define an "ISO-style" timespan output format with "hh:mm:ss" fields(Thomas)
998 Allow fractional values for delta times (e.g. '2.5 days')(Thomas)
999 Validate numeric input more carefully for delta times(Thomas)
1000 Implement day of year as possible input to date_part()(Thomas)
1001 Define timespan_finite() and text_timespan() functions(Thomas)
1002 Remove archive stuff(Bruce)
1003 Allow for a pg_password authentication database that is separate from
1004 the system password file(Todd)
1005 Dump ACLs, GRANT, REVOKE permissions(Matt)
1006 Define text, varchar, and bpchar string length functions(Thomas)
1007 Fix Query handling for inheritance, and cost computations(Bruce)
1008 Implement CREATE TABLE/AS SELECT (alternative to SELECT/INTO)(Thomas)
1009 Allow NOT, IS NULL, IS NOT NULL in constraints(Thomas)
1010 Implement UNIONs for SELECT(Bruce)
1011 Add UNION, GROUP, DISTINCT to INSERT(Bruce)
1012 varchar() stores only necessary bytes on disk(Bruce)
1013 Fix for BLOBs(Peter)
1014 Mega-Patch for JDBC...see README_6.3 for list of changes(Peter)
1015 Remove unused "option" from PQconnectdb()
1016 New LOCK command and lock manual page describing deadlocks(Bruce)
1017 Add new psql \da, \dd, \df, \do, \dS, and \dT commands(Bruce)
1018 Enhance psql \z to show sequences(Bruce)
1019 Show NOT NULL and DEFAULT in psql \d table(Bruce)
1020 New psql .psqlrc file startup(Andrew)
1021 Modify sample startup script in contrib/linux to show syslog(Thomas)
1022 New types for IP and MAC addresses in contrib/ip_and_mac(TomH)
1023 Unix system time conversions with date/time types in contrib/unixdate(Thomas)
1024 Update of contrib stuff(Massimo)
1025 Add Unix socket support to DBD::Pg(Goran)
1026 New python interface (PyGreSQL 2.0)(D'Arcy)
1027 New frontend/backend protocol has a version number, network byte order(Phil)
1028 Security features in pg_hba.conf enhanced and documented, many cleanups(Phil)
1029 CHAR() now faster access than VARCHAR() or TEXT
1030 ecpg embedded SQL preprocessor
1031 Reduce system column overhead(Vadmin)
1032 Remove pg_time table(Vadim)
1033 Add pg_type attribute to identify types that need length (bpchar, varchar)
1034 Add report of offending line when COPY command fails
1035 Allow VIEW permissions to be set separately from the underlying tables.
1036 For security, use GRANT/REVOKE on views as appropriate(Jan)
1037 Tables now have no default GRANT SELECT TO PUBLIC. You must
1038 explicitly grant such permissions.
1039 Clean up tutorial examples(Darren)
1043 Add new html development tools, and flow chart in /tools/backend
1044 Fix for SCO compiles
1045 Stratus computer port Robert Gillies
1046 Added support for shlib for BSD44_derived & i386_solaris
1047 Make configure more automated(Brook)
1048 Add script to check regression test results
1049 Break parser functions into smaller files, group together(Bruce)
1050 Rename heap_create to heap_create_and_catalog, rename heap_creatr
1051 to heap_create()(Bruce)
1052 Sparc/Linux patch for locking(TomS)
1053 Remove PORTNAME and reorganize port-specific stuff(Marc)
1054 Add optimizer README file(Bruce)
1055 Remove some recursion in optimizer and clean up some code there(Bruce)
1056 Fix for NetBSD locking(Henry)
1057 Fix for libptcl make(Tatsuo)
1059 Change IS TRUE, IS FALSE, ... to expressions using "=" rather than
1060 function calls to istrue() or isfalse() to allow optimization(Thomas)
1061 Various fixes NetBSD/Sparc related(TomH)
1062 Alpha linux locking(Travis,Ryan)
1063 Change elog(WARN) to elog(ERROR)(Bruce)
1064 FAQ for FreeBSD(Marc)
1065 Bring in the PostODBC source tree as part of our standard distribution(Marc)
1066 A minor patch for HP/UX 10 vs 9(Stan)
1067 New pg_attribute.atttypmod for type-specific info like varchar length(Bruce)
1068 Unixware patches(Billy)
1069 New i386 'lock' for spin lock asm(Billy)
1070 Support for multiplexed backends is removed
1071 Start an OpenBSD port
1074 Add string functions to regression suite(Thomas)
1075 Expand a few function names formerly truncated to 16 characters(Thomas)
1076 Remove un-needed malloc() calls and replace with palloc()(Bruce)
1078 ------------------------------------------------------------------------
1079 Release 6.3.1 Release 6.2.1
1081 ------------------------------------------------------------------------
1085 v6.2.1 is a bug-fix and usability release on v6.2.
1089 * Allow strings to span lines, per SQL92.
1091 * Include example trigger function for inserting user names on table
1094 This is a minor bug-fix release on v6.2. For upgrades from pre-v6.2 systems,
1095 a full dump/reload is required. Refer to the v6.2 release notes for
1098 Migration from v6.2 to v6.2.1
1100 This is a minor bug-fix release. A dump/reload is not required from v6.2,
1101 but is required from any release prior to v6.2.
1103 In upgrading from v6.2, if you choose to dump/reload you will find that
1104 avg(money) is now calculated correctly. All other bug fixes take effect upon
1105 updating the executables.
1107 Another way to avoid dump/reload is to use the following SQL command from
1108 psql to update the existing system table:
1110 update pg_aggregate set aggfinalfn = 'cash_div_flt8'
1111 where aggname = 'avg' and aggbasetype = 790;
1113 This will need to be done to every existing database, including template1.
1115 Detailed Change List
1117 Changes in this release
1118 -----------------------
1119 Allow TIME and TYPE column names(Thomas)
1120 Allow larger range of true/false as boolean values(Thomas)
1121 Support output of "now" and "current"(Thomas)
1122 Handle DEFAULT with INSERT of NULL properly(Vadim)
1123 Fix for relation reference counts problem in buffer manager(Vadim)
1124 Allow strings to span lines, like ANSI(Thomas)
1125 Fix for backward cursor with ORDER BY(Vadim)
1126 Fix avg(cash) computation(Thomas)
1127 Fix for specifying a column twice in ORDER/GROUP BY(Vadim)
1128 Documented new libpq function to return affected rows, PQcmdTuples(Bruce)
1129 Trigger function for inserting user names for INSERT/UPDATE(Brook Milligan)
1131 ------------------------------------------------------------------------
1132 Release 6.3 Release 6.2
1134 ------------------------------------------------------------------------
1138 A dump/restore is required for those wishing to migrate data from previous
1139 releases of Postgres.
1141 Migration from v6.1 to v6.2
1143 This migration requires a complete dump of the 6.1 database and a restore of
1144 the database in 6.2.
1146 Note that the pg_dump and pg_dumpall utility from 6.2 should be used to dump
1149 Migration from v1.x to v6.2
1151 Those migrating from earlier 1.* releases should first upgrade to 1.09
1152 because the COPY output format was improved from the 1.02 release.
1154 Detailed Change List
1158 Fix problems with pg_dump for inheritance, sequences, archive tables(Bruce)
1159 Fix compile errors on overflow due to shifts, unsigned, and bad prototypes
1160 from Solaris(Diab Jerius)
1161 Fix bugs in geometric line arithmetic (bad intersection calculations)(Thomas)
1162 Check for geometric intersections at endpoints to avoid rounding ugliness(Thomas)
1163 Catch non-functional delete attempts(Vadim)
1164 Change time function names to be more consistent(Michael Reifenberg)
1165 Check for zero divides(Michael Reifenberg)
1166 Fix very old bug which made tuples changed/inserted by a commnd
1167 visible to the command itself (so we had multiple update of
1168 updated tuples, etc)(Vadim)
1169 Fix for SELECT null, 'fail' FROM pg_am (Patrick)
1170 SELECT NULL as EMPTY_FIELD now allowed(Patrick)
1171 Remove un-needed signal stuff from contrib/pginterface
1172 Fix OR (where x != 1 or x isnull didn't return tuples with x NULL) (Vadim)
1173 Fix time_cmp function (Vadim)
1174 Fix handling of functions with non-attribute first argument in
1175 WHERE clauses (Vadim)
1176 Fix GROUP BY when order of entries is different from order
1177 in target list (Vadim)
1178 Fix pg_dump for aggregates without sfunc1 (Vadim)
1182 Default genetic optimizer GEQO parameter is now 8(Bruce)
1183 Allow use parameters in target list having aggregates in functions(Vadim)
1184 Added JDBC driver as an interface(Adrian & Peter)
1186 Return number of tuples inserted/affected by INSERT/UPDATE/DELETE etc.(Vadim)
1187 Triggers implemented with CREATE TRIGGER (SQL3)(Vadim)
1188 SPI (Server Programming Interface) allows execution of queries inside
1190 NOT NULL implemented (SQL92)(Robson Paniago de Miranda)
1191 Include reserved words for string handling, outer joins, and unions(Thomas)
1192 Implement extended comments ("/* ... */") using exclusive states(Thomas)
1193 Add "//" single-line comments(Bruce)
1194 Remove some restrictions on characters in operator names(Thomas)
1195 DEFAULT and CONSTRAINT for tables implemented (SQL92)(Vadim & Thomas)
1196 Add text concatenation operator and function (SQL92)(Thomas)
1197 Support WITH TIME ZONE syntax (SQL92)(Thomas)
1198 Support INTERVAL unit TO unit syntax (SQL92)(Thomas)
1199 Define types DOUBLE PRECISION, INTERVAL, CHARACTER,
1200 and CHARACTER VARYING (SQL92)(Thomas)
1201 Define type FLOAT(p) and rudimentary DECIMAL(p,s), NUMERIC(p,s) (SQL92)(Thomas)
1202 Define EXTRACT(), POSITION(), SUBSTRING(), and TRIM() (SQL92)(Thomas)
1203 Define CURRENT_DATE, CURRENT_TIME, CURRENT_TIMESTAMP (SQL92)(Thomas)
1204 Add syntax and warnings for UNION, HAVING, INNER and OUTER JOIN (SQL92)(Thomas)
1205 Add more reserved words, mostly for SQL92 compliance(Thomas)
1206 Allow hh:mm:ss time entry for timespan/reltime types(Thomas)
1207 Add center() routines for lseg, path, polygon(Thomas)
1208 Add distance() routines for circle-polygon, polygon-polygon(Thomas)
1209 Check explicitly for points and polygons contained within polygons
1210 using an axis-crossing algorithm(Thomas)
1211 Add routine to convert circle-box(Thomas)
1212 Merge conflicting operators for different geometric data types(Thomas)
1213 Replace distance operator "<===>" with "<->"(Thomas)
1214 Replace "above" operator "!^" with ">^" and "below" operator "!|" with "<^"(Thomas)
1215 Add routines for text trimming on both ends, substring, and string position(Thomas)
1216 Added conversion routines circle(box) and poly(circle)(Thomas)
1217 Allow internal sorts to be stored in memory rather than in files(Bruce & Vadim)
1218 Allow functions and operators on internally-identical types to succeed(Bruce)
1219 Speed up backend startup after profiling analysis(Bruce)
1220 Inline frequently called functions for performance(Bruce)
1221 Reduce open() calls(Bruce)
1222 psql: Add PAGER for \h and \?,\C fix
1223 Fix for psql pager when no tty(Bruce)
1224 New entab utility(Bruce)
1225 General trigger functions for referential integrity (Vadim)
1226 General trigger functions for time travel (Vadim)
1227 General trigger functions for AUTOINCREMENT/IDENTITY feature (Vadim)
1228 MOVE implementation (Vadim)
1232 HPUX 10 patches (Vladimir Turin)
1233 Added SCO support, (Daniel Harris)
1234 mkLinux patches (Tatsuo Ishii)
1235 Change geometric box terminology from "length" to "width"(Thomas)
1236 Deprecate temporary unstored slope fields in geometric code(Thomas)
1237 Remove restart instructions from INSTALL(Bruce)
1238 Look in /usr/ucb first for install(Bruce)
1239 Fix c++ copy example code(Thomas)
1240 Add -o to psql manual page(Bruce)
1241 Prevent relname unallocated string length from being copied into database(Bruce)
1242 Cleanup for NAMEDATALEN use(Bruce)
1243 Fix pg_proc names over 15 chars in output(Bruce)
1244 Add strNcpy() function(Bruce)
1245 remove some (void) casts that are unnecessary(Bruce)
1246 new interfaces directory(Marc)
1247 Replace fopen() calls with calls to fd.c functions(Bruce)
1248 Make functions static where possible(Bruce)
1249 enclose unused functions in #ifdef NOT_USED(Bruce)
1250 Remove call to difftime() in timestamp support to fix SunOS(Bruce & Thomas)
1251 Changes for Digital Unix
1252 Portability fix for pg_dumpall(Bruce)
1253 Rename pg_attribute.attnvals to attdisbursion(Bruce)
1254 "intro/unix" manual page now "pgintro"(Bruce)
1255 "built-in" manual page now "pgbuiltin"(Bruce)
1256 "drop" manual page now "drop_table"(Bruce)
1257 Add "create_trigger", "drop_trigger" manual pages(Thomas)
1258 Add constraints regression test(Vadim & Thomas)
1259 Add comments syntax regression test(Thomas)
1260 Add PGINDENT and support program(Bruce)
1261 Massive commit to run PGINDENT on all *.c and *.h files(Bruce)
1262 Files moved to /src/tools directory(Bruce)
1263 SPI and Trigger programming guides (Vadim & D'Arcy)
1265 ------------------------------------------------------------------------
1266 Release 6.2.1 Release 6.1.1
1268 ------------------------------------------------------------------------
1272 Migration from v6.1 to v6.1.1
1274 This is a minor bug-fix release. A dump/reload is not required from v6.1,
1275 but is required from any release prior to v6.1. Refer to the release notes
1276 for v6.1 for more details.
1278 Detailed Change List
1280 Changes in this release
1281 -----------------------
1282 fix for SET with options (Thomas)
1283 allow pg_dump/pg_dumpall to preserve ownership of all tables/objects(Bruce)
1284 new psql \connect option allows changing usernames without changing databases
1285 fix for initdb --debug option(Yoshihiko Ichikawa))
1286 lextest cleanup(Bruce)
1288 fix date/time month boundary arithmetic(Thomas)
1289 fix timezone daylight handling for some ports(Thomas, Bruce, Tatsuo)
1290 timestamp overhauled to use standard functions(Thomas)
1291 other code cleanup in date/time routines(Thomas)
1292 psql's \d now case-insensitive(Bruce)
1293 psql's backslash commands can now have trailing semicolon(Bruce)
1294 fix memory leak in psql when using \g(Bruce)
1295 major fix for endian handling of communication to server(Thomas, Tatsuo)
1296 Fix for Solaris assembler and include files(Yoshihiko Ichikawa)
1297 allow underscores in usernames(Bruce)
1298 pg_dumpall now returns proper status, portability fix(Bruce)
1300 ------------------------------------------------------------------------
1301 Release 6.2 Release 6.1
1303 ------------------------------------------------------------------------
1307 The regression tests have been adapted and extensively modified for the v6.1
1308 release of Postgres.
1310 Three new data types (datetime, timespan, and circle) have been added to the
1311 native set of Postgres types. Points, boxes, paths, and polygons have had
1312 their output formats made consistant across the data types. The polygon
1313 output in misc.out has only been spot-checked for correctness relative to
1314 the original regression output.
1316 Postgres v6.1 introduces a new, alternate optimizer which uses genetic
1317 algorithms. These algorithms introduce a random behavior in the ordering of
1318 query results when the query contains multiple qualifiers or multiple tables
1319 (giving the optimizer a choice on order of evaluation). Several regression
1320 tests have been modified to explicitly order the results, and hence are
1321 insensitive to optimizer choices. A few regression tests are for data types
1322 which are inherently unordered (e.g. points and time intervals) and tests
1323 involving those types are explicitly bracketed with set geqo to 'off' and
1326 The interpretation of array specifiers (the curly braces around atomic
1327 values) appears to have changed sometime after the original regression tests
1328 were generated. The current ./expected/*.out files reflect this new
1329 interpretation, which may not be correct!
1331 The float8 regression test fails on at least some platforms. This is due to
1332 differences in implementations of pow() and exp() and the signaling
1333 mechanisms used for overflow and underflow conditions.
1335 The "random" results in the random test should cause the "random" test to be
1336 "failed", since the regression tests are evaluated using a simple diff.
1337 However, "random" does not seem to produce random results on my test machine
1342 This migration requires a complete dump of the 6.0 database and a restore of
1343 the database in 6.1.
1345 Those migrating from earlier 1.* releases should first upgrade to 1.09
1346 because the COPY output format was improved from the 1.02 release.
1348 Detailed Change List
1352 packet length checking in library routines
1353 lock manager priority patch
1354 check for under/over flow of float8(Bruce)
1355 multi-table join fix(Vadim)
1356 SIGPIPE crash fix(Darren)
1357 large object fixes(Sven)
1358 allow btree indexes to handle NULLs(Vadim)
1359 timezone fixes(D'Arcy)
1360 select SUM(x) can return NULL on no rows(Thomas)
1361 internal optimizer, executor bug fixes(Vadim)
1362 fix problem where inner loop in < or <= has no rows(Vadim)
1363 prevent re-commuting join index clauses(Vadim)
1364 fix join clauses for multiple tables(Vadim)
1365 fix hash, hashjoin for arrays(Vadim)
1366 fix btree for abstime type(Vadim)
1367 large object fixes(Raymond)
1368 fix buffer leak in hash indices (Vadim)
1369 fix rtree for use in inner scan (Vadim)
1370 fix gist for use in inner scan, cleanups (Vadim, Andrea)
1371 avoid unnecessary local buffers allocation (Vadim, Massimo)
1372 fix local buffers leak in transaction aborts (Vadim)
1373 fix file manager memmory leaks, cleanups (Vadim, Massimo)
1374 fix storage manager memmory leaks (Vadim)
1375 fix btree duplicates handling (Vadim)
1376 fix deleted tuples re-incarnation caused by vacuum (Vadim)
1377 fix SELECT varchar()/char() INTO TABLE made zero-length fields(Bruce)
1378 many psql, pg_dump, and libpq memory leaks fixed using Purify (Igor)
1382 attribute optimization statistics(Bruce)
1383 much faster new btree bulk load code(Paul)
1384 BTREE UNIQUE added to bulk load code(Vadim)
1385 new lock debug code(Massimo)
1386 massive changes to libpg++(Leo)
1387 new GEQO optimizer speeds table multi-table optimization(Martin)
1388 new WARN message for non-unique insert into unique key(Marc)
1389 update x=-3, no spaces, now valid(Bruce)
1390 remove case-sensitive identifier handling(Bruce,Thomas,Dan)
1391 debug backend now pretty-prints tree(Darren)
1392 new Oracle character functions(Edmund)
1393 new plaintext password functions(Dan)
1394 no such class or insufficient privilege changed to distinct messages(Dan)
1395 new ANSI timestamp function(Dan)
1396 new ANSI Time and Date types (Thomas)
1397 move large chunks of data in backend(Martin)
1398 multi-column btree indexes(Vadim)
1399 new SET var TO value command(Martin)
1400 update transaction status on reads(Dan)
1401 new locale settings for character types(Oleg)
1402 new SEQUENCE serial number generator(Vadim)
1403 GROUP BY function now possible(Vadim)
1404 re-organize regression test(Thomas,Marc)
1405 new optimizer operation weights(Vadim)
1406 new psql \z grant/permit option(Marc)
1407 new MONEY data type(D'Arcy,Thomas)
1408 tcp socket communication speed improved(Vadim)
1409 new VACUUM option for attribute statistics, and for certain columns (Vadim)
1410 many geometric type improvements(Thomas,Keith)
1411 additional regression tests(Thomas)
1412 new datestyle variable(Thomas,Vadim,Martin)
1413 more comparison operators for sorting types(Thomas)
1414 new conversion functions(Thomas)
1415 new more compact btree format(Vadim)
1416 allow pg_dumpall to preserve database ownership(Bruce)
1417 new SET GEQO=# and R_PLANS variable(Vadim)
1418 old (!GEQO) optimizer can use right-sided plans (Vadim)
1419 typechecking improvement in SQL parser(Bruce)
1420 new SET, SHOW, RESET commands(Thomas,Vadim)
1421 new \connect database USER option
1422 new destroydb -i option (Igor)
1423 new \dt and \di psql commands (Darren)
1424 SELECT "\n" now escapes newline (A. Duursma)
1425 new geometry conversion functions from old format (Thomas)
1429 new configuration script(Marc)
1430 readline configuration option added(Marc)
1431 OS-specific configuration options removed(Marc)
1432 new OS-specific template files(Marc)
1433 no more need to edit Makefile.global(Marc)
1434 re-arrange include files(Marc)
1435 nextstep patches (Gregor Hoffleit)
1436 removed WIN32-specific code(Bruce)
1437 removed postmaster -e option, now only postgres -e option (Bruce)
1438 merge duplicate library code in front/backends(Martin)
1439 now works with eBones, international Kerberos(Jun)
1440 more shared library support
1441 c++ include file cleanup(Bruce)
1442 warn about buggy flex(Bruce)
1443 DG-UX, Ultrix, Irix, AIX portability fixes
1445 ------------------------------------------------------------------------
1446 Release 6.1.1 Release v6.0
1448 ------------------------------------------------------------------------
1452 A dump/restore is required for those wishing to migrate data from previous
1453 releases of Postgres.
1455 Migration from v1.09 to v6.0
1457 This migration requires a complete dump of the 1.09 database and a restore
1458 of the database in 6.0.
1460 Migration from pre-v1.09 to v6.0
1462 Those migrating from earlier 1.* releases should first upgrade to 1.09
1463 because the COPY output format was improved from the 1.02 release.
1465 Detailed Change List
1469 ALTER TABLE bug - running postgress process needs to re-read table definition
1470 Allow vacuum to be run on one table or entire database(Bruce)
1472 Fix array over-runs of memory writes(Kurt)
1473 Fix elusive btree range/non-range bug(Dan)
1474 Fix for hash indexes on some types like time and date
1475 Fix for pg_log size explosion
1476 Fix permissions on lo_export()(Bruce)
1477 Fix unitialized reads of memory(Kurt)
1478 Fixed ALTER TABLE ... char(3) bug(Bruce)
1479 Fixed a few small memory leaks
1480 Fixed EXPLAIN handling of options and changed full_path option name
1481 Fixed output of group acl permissions
1482 Memory leaks (hunt and destroy with tools like Purify(Kurt)
1483 Minor improvements to rules system
1485 New asserts for run-checking
1486 Overhauled parser/analyze code to properly report errors and increase speed
1487 Pg_dump -d now handles NULL's properly(Bruce)
1488 Prevent SELECT NULL from crashing server (Bruce)
1489 Properly report errors when INSERT ... SELECT columns did not match
1490 Properly report errors when insert column names were not correct
1491 Psql \g filename now works(Bruce)
1492 Psql fixed problem with multiple statements on one line with multiple outputs
1493 Removed duplicate system oid's
1494 SELECT * INTO TABLE . GROUP/ORDER BY gives unlink error if table exists(Bruce)
1495 Several fixes for queries that crashed the backend
1496 Starting quote in insert string errors(Bruce)
1497 Submitting an empty query now returns empty status, not just " " query(Bruce)
1501 Add EXPLAIN manual page(Bruce)
1502 Add UNIQUE index capability(Dan)
1503 Add hostname/user level access control rather than just hostname and user
1504 Add synonym of != for (Bruce)
1505 Allow "select oid,* from table"
1506 Allow BY,ORDER BY to specify columns by number, or by non-alias table.column(Bruce)
1507 Allow COPY from the frontend(Bryan)
1508 Allow GROUP BY to use alias column name(Bruce)
1509 Allow actual compression, not just reuse on the same page(Vadim)
1510 Allow installation-configuration option to auto-add all local users(Bryan)
1511 Allow libpq to distinguish between text value '' and null(Bruce)
1512 Allow non-postgres users with createdb privs to destroydb's
1513 Allow restriction on who can create C functions(Bryan)
1514 Allow restriction on who can do backend COPY(Bryan)
1515 Can shrink tables, pg_time and pg_log(Vadim & Erich)
1516 Change debug level 2 to print queries only, changed debug heading layout(Bruce)
1517 Change default decimal constant representation from float4 to float8(Bruce)
1518 European date format now set when postmaster is started
1519 Execute lowercase function names if not found with exact case
1520 Fixes for aggregate/GROUP processing, allow 'select sum(func(x),sum(x+y) from z'
1521 Gist now included in the distrubution(Marc)
1522 Idend authentication of local users(Bryan)
1523 Implement BETWEEN qualifier(Bruce)
1524 Implement IN qualifier(Bruce)
1525 Libpq has PQgetisnull()(Bruce)
1526 Libpq++ improvements
1527 New options to initdb(Bryan)
1528 Pg_dump allow dump of oid's(Bruce)
1529 Pg_dump create indexes after tables are loaded for speed(Bruce)
1530 Pg_dumpall dumps all databases, and the user table
1531 Pginterface additions for NULL values(Bruce)
1532 Prevent postmaster from being run as root
1533 Psql \h and \? is now readable(Bruce)
1534 Psql allow backslashed, semicolons anywhere on the line(Bruce)
1535 Psql changed command prompt for lines in query or in quotes(Bruce)
1536 Psql char(3) now displays as (bp)char in \d output(Bruce)
1537 Psql return code now more accurate(Bryan?)
1538 Psql updated help syntax(Bruce)
1539 Re-visit and fix vacuum(Vadim)
1540 Reduce size of regression diffs, remove timezone name difference(Bruce)
1541 Remove compile-time parameters to enable binary distributions(Bryan)
1542 Reverse meaning of HBA masks(Bryan)
1543 Secure Authentication of local users(Bryan)
1544 Speed up vacuum(Vadim)
1545 Vacuum now had VERBOSE option(Bruce)
1549 All functions now have prototypes that are compared against the calls
1550 Allow asserts to be disabled easly from Makefile.global(Bruce)
1551 Change oid constants used in code to #define names
1552 Decoupled sparc and solaris defines(Kurt)
1553 Gcc -Wall compiles cleanly with warnings only from unfixable constructs
1554 Major include file reorganization/reduction(Marc)
1555 Make now stops on compile failure(Bryan)
1556 Makefile restructuring(Bryan, Marc)
1557 Merge bsdi_2_1 to bsdi(Bruce)
1558 Monitor program removed
1559 Name change from Postgres95 to PostgreSQL
1560 New config.h file(Marc, Bryan)
1561 PG_VERSION now set to 6.0 and used by postmaster
1562 Portability additions, including Ultrix, DG/UX, AIX, and Solaris
1563 Reduced the number of #define's, centeralized #define's
1564 Remove duplicate OIDS in system tables(Dan)
1565 Remove duplicate system catalog info or report mismatches(Dan)
1566 Removed many os-specific #define's
1567 Restructured object file generation/location(Bryan, Marc)
1568 Restructured port-specific file locations(Bryan, Marc)
1569 Unused/uninialized variables corrected
1571 ------------------------------------------------------------------------
1572 Release 6.1 Release v1.09
1574 ------------------------------------------------------------------------
1578 Sorry, we stopped keeping track of changes from 1.02 to 1.09. Some of the
1579 changes listed in 6.0 were actually included in the 1.02.1 to 1.09 releases.
1581 ------------------------------------------------------------------------
1582 Release v6.0 Release v1.02
1584 ------------------------------------------------------------------------
1588 Migration from v1.02 to v1.02.1
1590 Here is a new migration file for 1.02.1. It includes the 'copy' change and a
1591 script to convert old ascii files.
1593 Note: The following notes are for the benefit of users who want to
1594 migrate databases from postgres95 1.01 and 1.02 to postgres95
1597 If you are starting afresh with postgres95 1.02.1 and do not need
1598 to migrate old databases, you do not need to read any further.
1600 In order to upgrade older postgres95 version 1.01 or 1.02 databases to
1601 version 1.02.1, the following steps are required:
1603 1. Start up a new 1.02.1 postmaster
1605 2. Add the new built-in functions and operators of 1.02.1 to 1.01 or 1.02
1606 databases. This is done by running the new 1.02.1 server against your
1607 own 1.01 or 1.02 database and applying the queries attached at the end
1608 of thie file. This can be done easily through psql. If your 1.01 or
1609 1.02 database is named "testdb" and you have cut the commands from the
1610 end of this file and saved them in addfunc.sql:
1612 % psql testdb -f addfunc.sql
1614 Those upgrading 1.02 databases will get a warning when executing the
1615 last two statements in the file because they are already present in
1616 1.02. This is not a cause for concern.
1618 Dump/Reload Procedure
1620 If you are trying to reload a pg_dump or text-mode 'copy tablename to
1621 stdout' generated with a previous version, you will need to run the attached
1622 sed script on the ASCII file before loading it into the database. The old
1623 format used '.' as end-of-data, while '\.' is now the end-of-data marker.
1624 Also, empty strings are now loaded in as '' rather than NULL. See the copy
1625 manual page for full details.
1627 sed 's/^\.$/\\./g' in_file out_file
1629 If you are loading an older binary copy or non-stdout copy, there is no
1630 end-of-data character, and hence no conversion necessary.
1632 -- following lines added by agc to reflect the case-insensitive
1633 -- regexp searching for varchar (in 1.02), and bpchar (in 1.02.1)
1634 create operator ~* (leftarg = bpchar, rightarg = text, procedure = texticregexeq);
1635 create operator !~* (leftarg = bpchar, rightarg = text, procedure = texticregexne);
1636 create operator ~* (leftarg = varchar, rightarg = text, procedure = texticregexeq);
1637 create operator !~* (leftarg = varchar, rightarg = text, procedure = texticregexne);
1639 Detailed Change List
1641 Source code maintenance and development
1642 * worldwide team of volunteers
1643 * the source tree now in CVS at ftp.ki.net
1646 * psql (and underlying libpq library) now has many more options for
1647 formatting output, including HTML
1648 * pg_dump now output the schema and/or the data, with many fixes to
1649 enhance completeness.
1650 * psql used in place of monitor in administration shell scripts.
1651 monitor to be depreciated in next release.
1652 * date/time functions enhanced
1653 * NULL insert/update/comparison fixed/enhanced
1654 * TCL/TK lib and shell fixed to work with both tck7.4/tk4.0 and tcl7.5/tk4.1
1656 Bug Fixes (almost too numerous to mention)
1658 * storage management
1659 * check for NULL pointer before dereferencing
1663 * added SolarisX86 port
1664 * added BSDI 2.1 port
1667 ------------------------------------------------------------------------
1668 Release v1.09 Release v1.01
1670 ------------------------------------------------------------------------
1674 Migration from v1.0 to v1.01
1676 The following notes are for the benefit of users who want to migrate
1677 databases from postgres95 1.0 to postgres95 1.01.
1679 If you are starting afresh with postgres95 1.01 and do not need to migrate
1680 old databases, you do not need to read any further.
1682 In order to postgres95 version 1.01 with databases created with postgres95
1683 version 1.0, the following steps are required:
1685 1. Set the definition of NAMEDATALEN in src/Makefile.global to 16 and
1688 2. Decide whether you want to use Host based authentication.
1690 a. If you do, you must create a file name "pg_hba" in your top-level
1691 data directory (typically the value of your $PGDATA).
1692 src/libpq/pg_hba shows an example syntax.
1694 b. If you do not want host-based authentication, you can comment out
1699 in src/Makefile.global
1701 Note that host-based authentication is turned on by default, and
1702 if you do not take steps A or B above, the out-of-the-box 1.01
1703 will not allow you to connect to 1.0 databases.
1705 3. Compile and install 1.01, but DO NOT do the initdb step.
1707 4. Before doing anything else, terminate your 1.0 postmaster, and backup
1708 your existing $PGDATA directory.
1710 5. Set your PGDATA environment variable to your 1.0 databases, but set up
1711 path up so that 1.01 binaries are being used.
1713 6. Modify the file $PGDATA/PG_VERSION from 5.0 to 5.1
1715 7. Start up a new 1.01 postmaster
1717 8. Add the new built-in functions and operators of 1.01 to 1.0 databases.
1718 This is done by running the new 1.01 server against your own 1.0
1719 database and applying the queries attached and saving in the file
1720 1.0_to_1.01.sql. This can be done easily through psql. If your 1.0
1721 database is name "testdb":
1723 % psql testdb -f 1.0_to_1.01.sql
1725 and then execute the following commands (cut and paste from here):
1727 -- add builtin functions that are new to 1.01
1729 create function int4eqoid (int4, oid) returns bool as 'foo'
1730 language 'internal';
1731 create function oideqint4 (oid, int4) returns bool as 'foo'
1732 language 'internal';
1733 create function char2icregexeq (char2, text) returns bool as 'foo'
1734 language 'internal';
1735 create function char2icregexne (char2, text) returns bool as 'foo'
1736 language 'internal';
1737 create function char4icregexeq (char4, text) returns bool as 'foo'
1738 language 'internal';
1739 create function char4icregexne (char4, text) returns bool as 'foo'
1740 language 'internal';
1741 create function char8icregexeq (char8, text) returns bool as 'foo'
1742 language 'internal';
1743 create function char8icregexne (char8, text) returns bool as 'foo'
1744 language 'internal';
1745 create function char16icregexeq (char16, text) returns bool as 'foo'
1746 language 'internal';
1747 create function char16icregexne (char16, text) returns bool as 'foo'
1748 language 'internal';
1749 create function texticregexeq (text, text) returns bool as 'foo'
1750 language 'internal';
1751 create function texticregexne (text, text) returns bool as 'foo'
1752 language 'internal';
1754 -- add builtin functions that are new to 1.01
1756 create operator = (leftarg = int4, rightarg = oid, procedure = int4eqoid);
1757 create operator = (leftarg = oid, rightarg = int4, procedure = oideqint4);
1758 create operator ~* (leftarg = char2, rightarg = text, procedure = char2icregexeq);
1759 create operator !~* (leftarg = char2, rightarg = text, procedure = char2icregexne);
1760 create operator ~* (leftarg = char4, rightarg = text, procedure = char4icregexeq);
1761 create operator !~* (leftarg = char4, rightarg = text, procedure = char4icregexne);
1762 create operator ~* (leftarg = char8, rightarg = text, procedure = char8icregexeq);
1763 create operator !~* (leftarg = char8, rightarg = text, procedure = char8icregexne);
1764 create operator ~* (leftarg = char16, rightarg = text, procedure = char16icregexeq);
1765 create operator !~* (leftarg = char16, rightarg = text, procedure = char16icregexne);
1766 create operator ~* (leftarg = text, rightarg = text, procedure = texticregexeq);
1767 create operator !~* (leftarg = text, rightarg = text, procedure = texticregexne);
1769 Detailed Change List
1772 * 1.01 is backwards compatible with 1.0 database provided the user
1773 follow the steps outlined in the MIGRATION_from_1.0_to_1.01 file.
1774 If those steps are not taken, 1.01 is not compatible with 1.0 database.
1777 * added PQdisplayTuples() to libpq and changed monitor and psql to use it
1778 * added NeXT port (requires SysVIPC implementation)
1779 * added CAST .. AS ... syntax
1780 * added ASC and DESC keywords
1781 * added 'internal' as a possible language for CREATE FUNCTION
1782 internal functions are C functions which have been statically linked
1783 into the postgres backend.
1784 * a new type "name" has been added for system identifiers (table names,
1785 attribute names, etc.) This replaces the old char16 type. The
1786 of name is set by the NAMEDATALEN #define in src/Makefile.global
1787 * a readable reference manual that describes the query language.
1788 * added host-based access control. A configuration file ($PGDATA/pg_hba)
1789 is used to hold the configuration data. If host-based access control
1790 is not desired, comment out HBA=1 in src/Makefile.global.
1791 * changed regex handling to be uniform use of Henry Spencer's regex code
1792 regardless of platform. The regex code is included in the distribution
1793 * added functions and operators for case-insensitive regular expressions.
1794 The operators are ~* and !~*.
1795 * pg_dump uses COPY instead of SELECT loop for better performance
1798 * fixed an optimizer bug that was causing core dumps when
1799 functions calls were used in comparisons in the WHERE clause
1800 * changed all uses of getuid to geteuid so that effective uids are used
1801 * psql now returns non-zero status on errors when using -c
1802 * applied public patches 1-14
1804 ------------------------------------------------------------------------
1805 Release v1.02 Release v1.0
1807 ------------------------------------------------------------------------
1811 Detailed Change List
1814 * The copyright of Postgres 1.0 has been loosened to be freely modifiable
1815 and modifiable for any purpose. Please read the COPYRIGHT file.
1816 Thanks to Professor Michael Stonebraker for making this possible.
1819 * date formats have to be MM-DD-YYYY (or DD-MM-YYYY if you're using
1820 EUROPEAN STYLE). This follows SQL-92 specs.
1821 * "delimiters" is now a keyword
1824 * sql LIKE syntax has been added
1825 * copy command now takes an optional USING DELIMITER specification.
1826 delimiters can be any single-character string.
1827 * IRIX 5.3 port has been added.
1828 Thanks to Paul Walmsley and others.
1829 * updated pg_dump to work with new libpq
1830 * \d has been added psql
1831 Thanks to Keith Parks
1832 * regexp performance for architectures that use POSIX regex has been
1833 improved due to caching of precompiled patterns.
1834 Thanks to Alistair Crooks
1835 * a new version of libpq++
1836 Thanks to William Wanders
1839 * arbitrary userids can be specified in the createuser script
1840 * \c to connect to other databases in psql now works.
1841 * bad pg_proc entry for float4inc() is fixed
1842 * users with usecreatedb field set can now create databases without
1843 having to be usesuper
1844 * remove access control entries when the entry no longer has any
1846 * fixed non-portable datetimes implementation
1847 * added kerberos flags to the src/backend/Makefile
1848 * libpq now works with kerberos
1849 * typographic errors in the user manual have been corrected.
1850 * btrees with multiple index never worked, now we tell you they don't
1851 work when you try to use them
1853 ------------------------------------------------------------------------
1854 Release v1.01 Postgres95 Beta 0.03
1856 ------------------------------------------------------------------------
1858 Postgres95 Beta 0.03
1860 Detailed Change List
1862 Incompatible changes:
1863 * BETA-0.3 IS INCOMPATIBLE WITH DATABASES CREATED WITH PREVIOUS VERSIONS
1864 (due to system catalog changes and indexing structure changes).
1865 * double-quote (") is deprecated as a quoting character for string literals;
1866 you need to convert them to single quotes (').
1867 * name of aggregates (eg. int4sum) are renamed in accordance with the
1868 SQL standard (eg. sum).
1869 * CHANGE ACL syntax is replaced by GRANT/REVOKE syntax.
1870 * float literals (eg. 3.14) are now of type float4 (instead of float8 in
1871 previous releases); you might have to do typecasting if you depend on it
1872 being of type float8. If you neglect to do the typecasting and you assign
1873 a float literal to a field of type float8, you may get incorrect values
1875 * LIBPQ has been totally revamped so that frontend applications
1876 can connect to multiple backends
1877 * the usesysid field in pg_user has been changed from int2 to int4 to
1878 allow wider range of Unix user ids.
1879 * the netbsd/freebsd/bsd o/s ports have been consolidated into a
1880 single BSD44_derived port. (thanks to Alistair Crooks)
1882 SQL standard-compliance (the following details changes that makes postgres95
1883 more compliant to the SQL-92 standard):
1884 * the following SQL types are now built-in: smallint, int(eger), float, real,
1885 char(N), varchar(N), date and time.
1887 The following are aliases to existing postgres types:
1889 integer, int -> int4
1890 float, real -> float4
1891 char(N) and varchar(N) are implemented as truncated text types. In
1892 addition, char(N) does blank-padding.
1893 * single-quote (') is used for quoting string literals; '' (in addition to
1894 \') is supported as means of inserting a single quote in a string
1895 * SQL standard aggregate names (MAX, MIN, AVG, SUM, COUNT) are used
1896 (Also, aggregates can now be overloaded, i.e. you can define your
1897 own MAX aggregate to take in a user-defined type.)
1898 * CHANGE ACL removed. GRANT/REVOKE syntax added.
1899 - Privileges can be given to a group using the "GROUP" keyword.
1901 GRANT SELECT ON foobar TO GROUP my_group;
1902 The keyword 'PUBLIC' is also supported to mean all users.
1904 Privileges can only be granted or revoked to one user or group
1907 "WITH GRANT OPTION" is not supported. Only class owners can change
1909 - The default access control is to to grant users readonly access.
1910 You must explicitly grant insert/update access to users. To change
1911 this, modify the line in
1912 src/backend/utils/acl.h
1913 that defines ACL_WORLD_DEFAULT
1916 * the bug where aggregates of empty tables were not run has been fixed. Now,
1917 aggregates run on empty tables will return the initial conditions of the
1918 aggregates. Thus, COUNT of an empty table will now properly return 0.
1919 MAX/MIN of an empty table will return a tuple of value NULL.
1920 * allow the use of \; inside the monitor
1921 * the LISTEN/NOTIFY asynchronous notification mechanism now work
1922 * NOTIFY in rule action bodies now work
1923 * hash indices work, and access methods in general should perform better.
1924 creation of large btree indices should be much faster. (thanks to Paul
1927 Other changes and enhancements:
1928 * addition of an EXPLAIN statement used for explaining the query execution
1929 plan (eg. "EXPLAIN SELECT * FROM EMP" prints out the execution plan for
1931 * WARN and NOTICE messages no longer have timestamps on them. To turn on
1932 timestamps of error messages, uncomment the line in
1933 src/backend/utils/elog.h:
1934 /* define ELOG_TIMESTAMPS */
1935 * On an access control violation, the message
1936 "Either no such class or insufficient privilege"
1937 will be given. This is the same message that is returned when
1938 a class is not found. This dissuades non-privileged users from
1939 guessing the existence of privileged classes.
1940 * some additional system catalog changes have been made that are not
1941 visible to the user.
1944 * The -oid option has been added to the "pg_result" tcl command.
1945 pg_result -oid returns oid of the last tuple inserted. If the
1946 last command was not an INSERT, then pg_result -oid returns "".
1947 * the large object interface is available as pg_lo* tcl commands:
1948 pg_lo_open, pg_lo_close, pg_lo_creat, etc.
1950 Portability enhancements and New Ports:
1951 * flex/lex problems have been cleared up. Now, you should be able to use
1952 flex instead of lex on any platforms. We no longer make assumptions of
1953 what lexer you use based on the platform you use.
1954 * The Linux-ELF port is now supported. Various configuration have been
1955 tested: The following configuration is known to work:
1956 kernel 1.2.10, gcc 2.6.3, libc 4.7.2, flex 2.5.2, bison 1.24
1957 with everything in ELF format,
1960 * ipcclean added to the distribution
1961 ipcclean usually does not need to be run, but if your backend crashes
1962 and leaves shared memory segments hanging around, ipcclean will
1963 clean them up for you.
1966 * the user manual has been revised and libpq documentation added.
1968 ------------------------------------------------------------------------
1969 Release v1.0 Postgres95 Beta 0.02
1971 ------------------------------------------------------------------------
1973 Postgres95 Beta 0.02
1975 Detailed Change List
1977 Incompatible changes:
1978 * The SQL statement for creating a database is 'CREATE DATABASE' instead
1979 of 'CREATEDB'. Similarly, dropping a database is 'DROP DATABASE' instead
1980 of 'DESTROYDB'. However, the names of the executables 'createdb' and
1981 'destroydb' remain the same.
1984 * pgperl - a Perl (4.036) interface to Postgres95
1985 * pg_dump - a utility for dumping out a postgres database into a
1986 script file containing query commands. The script files are in a ASCII
1987 format and can be used to reconstruct the database, even on other
1988 machines and other architectures. (Also good for converting
1989 a Postgres 4.2 database to Postgres95 database.)
1991 The following ports have been incorporated into postgres95-beta-0.02:
1992 * the NetBSD port by Alistair Crooks
1993 * the AIX port by Mike Tung
1994 * the Windows NT port by Jon Forrest (more stuff but not done yet)
1995 * the Linux ELF port by Brian Gallew
1997 The following bugs have been fixed in postgres95-beta-0.02:
1998 * new lines not escaped in COPY OUT and problem with COPY OUT when first
2000 * cannot type return to use the default user id in createuser
2001 * SELECT DISTINCT on big tables crashes
2002 * Linux installation problems
2003 * monitor doesn't allow use of 'localhost' as PGHOST
2004 * psql core dumps when doing \c or \l
2005 * the "pgtclsh" target missing from src/bin/pgtclsh/Makefile
2006 * libpgtcl has a hard-wired default port number
2007 * SELECT DISTINCT INTO TABLE hangs
2008 * CREATE TYPE doesn't accept 'variable' as the internallength
2009 * wrong result using more than 1 aggregate in a SELECT
2011 ------------------------------------------------------------------------
2012 Postgres95 Beta 0.03 Postgres95 Beta 0.01
2014 ------------------------------------------------------------------------
2016 Postgres95 Beta 0.01
2020 ------------------------------------------------------------------------
2021 Postgres95 Beta 0.02 Timing Results
2024 ------------------------------------------------------------------------
2028 These timing results are from running the regression test with the commands
2030 % cd src/test/regress
2035 Timing under Linux 2.0.27 seems to have a roughly 5% variation from run to
2036 run, presumably due to the scheduling vagaries of multitasking systems.
2040 As has been the case for previous releases, timing between releases is not
2041 directly comparable since new regression tests have been added. In general,
2042 v6.5 is faster than previous releases.
2044 Timing with fsync() disabled:
2047 02:00 Dual Pentium Pro 180, 224MB, UW-SCSI, Linux 2.0.36, gcc 2.7.2.3 -O2 -m486
2048 04:38 Sparc Ultra 1 143MHz, 64MB, Solaris 2.6
2051 Timing with fsync() enabled:
2054 04:21 Dual Pentium Pro 180, 224MB, UW-SCSI, Linux 2.0.36, gcc 2.7.2.3 -O2 -m486
2057 For the linux system above, using UW-SCSI disks rather than (older) IDE
2058 disks leads to a 50% improvement in speed on the regression test.
2062 The times for this release are not directly comparable to those for previous
2063 releases since some additional regression tests have been included. In
2064 general, however, v6.4 should be slightly faster than the previous release
2068 02:26 Dual Pentium Pro 180, 96MB, UW-SCSI, Linux 2.0.30, gcc 2.7.2.1 -O2 -m486
2072 The times for this release are not directly comparable to those for previous
2073 releases since some additional regression tests have been included and some
2074 obsolete tests involving time travel have been removed. In general, however,
2075 v6.3 is substantially faster than previous releases (thanks, Bruce!).
2078 02:30 Dual Pentium Pro 180, 96MB, UW-SCSI, Linux 2.0.30, gcc 2.7.2.1 -O2 -m486
2079 04:12 Dual Pentium Pro 180, 96MB, EIDE, Linux 2.0.30, gcc 2.7.2.1 -O2 -m486
2084 06:12 Pentium Pro 180, 32MB, EIDE, Linux 2.0.30, gcc 2.7.2 -O2 -m486
2085 12:06 P-100, 48MB, Linux 2.0.29, gcc
2086 39:58 Sparc IPC 32MB, Solaris 2.5, gcc 2.7.2.1 -O -g
2088 ------------------------------------------------------------------------
2090 Postgres95 Beta 0.01