2 <Title>Release Notes</Title>
7 <Title>Release 6.3.2</Title>
12 <firstname>Bruce</firstname>
13 <surname>Momjian</surname>
16 <date>Tue Apr 7 16:53:16 EDT 1998</date>
21 This is a bugfix release for 6.3.x.
22 Refer to the release notes for v6.3 for a more complete summary of new features.
27 <itemizedlist spacing="compact">
30 Repairs automatic configuration support for some platforms, including Linux,
31 from breakage inadvertently introduced in v6.3.1.
35 Correctly handles function calls on the left side of BETWEEN and LIKE clauses.
40 A dump/restore is NOT required for those running 6.3 or 6.3.1. A
41 'make distclean', 'make', and 'make install' is all that is required.
42 This last step should be performed while the postmaster is not running.
43 You should re-link any custom applications that use <productname>Postgres</productname> libraries.
46 For upgrades from pre-v6.3 installations,
47 refer to the installation and migration instructions for v6.3.
50 <title>Detailed Change List</title>
56 Configure detection improvements for tcl/tk(Brook Milligan, Alvin)
57 Manual page improvements(Bruce)
58 BETWEEN and LIKE fix(Thomas)
59 fix for psql \connect used by pg_dump(Oliver Elphick)
61 pgaccess, version 0.86
62 qsort removed, now uses libc version, cleanups(Jeroen)
63 fix for buffer over-runs detected(Maurice Gittens)
64 fix for buffer overrun in libpgtcl(Randy Kunkee)
65 fix for UNION with DISTINCT or ORDER BY(Bruce)
66 gettimeofday configure check(Doug Winterburn)
67 Fix "indexes not used" bug(Vadim)
68 docs additions(Thomas)
69 Fix for backend memory leak(Bruce)
70 libreadline cleanup(Erwan MAS)
72 Makefile dependency cleanup(Jeroen van Vianen)
78 <Title>Release 6.3.1</Title>
83 <firstname>Bruce</firstname>
84 <surname>Momjian</surname>
88 Mon Mar 23 10:21:52 EST 1998
96 <itemizedlist spacing="compact">
99 Additional support for multi-byte character sets.
103 Repair byte ordering for mixed-endian clients and servers.
107 Minor updates to allowed SQL syntax.
111 Improvements to the configuration autodetection for installation.
119 A dump/restore is NOT required for those running 6.3. A
120 'make distclean', 'make', and 'make install' is all that is required.
121 This last step should be performed while the postmaster is not running.
122 You should re-link any custom applications that use <productname>Postgres</productname> libraries.
125 For upgrades from pre-v6.3 installations,
126 refer to the installation and migration instructions for v6.3.
129 <title>Detailed Change List</title>
135 ecpg cleanup/fixes, now version 1.1(Michael Meskes)
136 pg_user cleanup(Bruce)
137 large object fix for pg_dump and tclsh (alvin)
138 LIKE fix for multiple adjacent underscores
139 fix for redefining builtin functions(Thomas)
141 upgrade to pg_access 0.83
142 updated CLUSTER manual page
143 multi-byte character set support, see doc/README.mb(Tatsuo)
144 configure --with-pgport fix
146 big-endian fix for backend communications(Kataoka)
147 SUBSTR() and substring() fix(Jan)
148 several jdbc fixes(Peter)
149 libpgtcl improvements, see libptcl/README(Randy Kunkee)
150 Fix for "Datasize = 0" error(Vadim)
151 Prevent \do from wrapping(Bruce)
152 Remove duplicate Russian character set entries
154 Allow optional TABLE keyword in LOCK and SELECT INTO(Thomas)
155 CREATE SEQUENCE options to allow a negative integer(Thomas)
156 Add "PASSWORD" as an allowed column identifier(Thomas)
157 Add checks for UNION target fields(Bruce)
158 Fix Alpha port(Dwayne Bailey)
159 Fix for text arrays containing quotes(Doug Gibson)
160 Solaris compile fix(Albert Chin-A-Young)
161 Better identify tcl and tk libs and includes(Bruce)
166 <Title>Release 6.3</Title>
171 <firstname>Bruce</firstname>
172 <surname>Momjian</surname>
176 Sun Mar 1 14:57:30 EST 1998
182 There are <emphasis>many</emphasis> new features and improvements in this release.
183 Here is a brief, incomplete summary:
185 <itemizedlist spacing="compact">
188 Many new SQL features, including
189 full <acronym>SQL92</acronym> subselect capability
190 (everything is here but target-list subselects).
194 Support for client-side environment variables to specify time zone and date style.
198 Socket interface for client/server connection. This is the default now
199 so you may need to start <application>postmaster</application> with the
200 <quote>-i</quote> flag.
204 Better password authorization mechanisms. Default table permissions have changed.
208 Old-style <quote>time travel</quote> has been removed. Performance has been improved.
214 Bruce Momjian wrote the following notes to introduce the new release.
218 There are some general 6.3 issues that I want to mention. These are
219 only the big items that can not be described in one sentence. A review
220 of the detailed changes list is still needed.
223 First, we now have subselects. Now that we have them, I would like to
224 mention that without subselects, SQL is a very limited language.
225 Subselects are a major feature, and you should review your code for
226 places where subselects provide a better solution for your queries. I
227 think you will find that there are more uses for subselects than you may
228 think. Vadim has put us on the big SQL map with subselects, and fully
229 functional ones too. The only thing you can't do with subselects is to
230 use them in the target list.
233 Second, 6.3 uses unix domain sockets rather than TCP/IP by default. To
234 enable connections from other machines, you have to use the new
235 postmaster -i option, and of course edit pg_hba.conf. Also, for this
236 reason, the format of pg_hba.conf has changed.
239 Third, char() fields will now allow faster access than varchar() or
240 text. Specifically, the text and varchar() have a penalty for access to
241 any columns after the first column of this type. char() used to also
242 have this access penalty, but it no longer does. This may suggest that
243 you redesign some of your tables, especially if you have short character
244 columns that you have defined as varchar() or text. This and other
245 changes make 6.3 even faster than earlier releases.
248 We now have passwords definable independent of any Unix file. There are
249 new SQL USER commands. See the pg_hba.conf manual page for more
250 information. There is a new table, pg_shadow, which is used to store
251 user information and user passwords, and it by default only SELECT-able
252 by the postgres super-user. pg_user is now a view of pg_shadow, and is
253 SELECT-able by PUBLIC. You should keep using pg_user in your
254 application without changes.
257 User-created tables now no longer have SELECT permission to PUBLIC by
258 default. This was done because the ANSI standard requires it. You can
259 of course GRANT any permissions you want after the table is created.
260 System tables continue to be SELECT-able by PUBLIC.
263 We also have real deadlock detection code. No more sixty-second
264 timeouts. And the new locking code implements a FIFO better, so there
265 should be less resource starvation during heavy use.
268 Many complaints have been made about inadequate documenation in previous
269 releases. Thomas has put much effort into many new manuals for this
270 release. Check out the doc/ directory.
273 For performance reasons, time travel is gone, but can be implemented
274 using triggers (see pgsql/contrib/spi/README). Please check out the new
275 \d command for types, operators, etc. Also, views have their own
276 permissions now, not based on the underlying tables, so permissions on
277 them have to be set separately. Check /pgsql/interfaces for some new
278 ways to talk to <productname>Postgres</productname>.
281 This is the first release that really required an explanation for
282 existing users. In many ways, this was necessary because the new
283 release removes many limitations, and the work-arounds people were using
284 are no longer needed.
288 <title>Migration to v6.3</title>
291 A dump/restore using <application>pg_dump</application>
292 or <application>pg_dumpall</application>
293 is required for those wishing to migrate data from any
294 previous release of <productname>Postgres</productname>.
298 <title>Detailed Change List</title>
304 Fix binary cursors broken by MOVE implementation(Vadim)
305 Fix for tcl library crash(Jan)
306 Fix for array handling, from Gerhard Hintermayer
307 Fix acl error, and remove duplicate pqtrace(Bruce)
308 Fix psql \e for empty file(Bruce)
309 Fix for textcat on varchar() fields(Bruce)
310 Fix for DBT Sendproc (Zeugswetter Andres)
311 Fix vacuum analyze syntax problem(Bruce)
312 Fix for international identifiers(Tatsuo)
313 Fix aggregates on inherited tables(Bruce)
314 Fix substr() for out-of-bounds data
315 Fix for select 1=1 or 2=2, select 1=1 and 2=2, and select sum(2+2)(Bruce)
316 Fix notty output to show status result. -q option still turns it off(Bruce)
317 Fix for count(*), aggs with views and multiple tables and sum(3)(Bruce)
319 Fix for PQtrace start/stop several times(Bruce)
320 Fix a variety of locking problems like newer lock waiters getting
321 lock before older waiters, and having readlock people not share
322 locks if a writer is waiting for a lock, and waiting writers not
323 getting priority over waiting readers(Bruce)
324 Fix crashes in psql when executing queries from external files(James)
325 Fix problem with multiple order by columns, with the first one having
327 Use correct hash table support functions for float8 and int4(Thomas)
328 Re-enable JOIN= option in CREATE OPERATOR statement (Thomas)
329 Change precedence for boolean operators to match expected behavior(Thomas)
330 Generate elog(ERROR) on over-large integer(Bruce)
331 Allow multiple-argument functions in constraint clauses(Thomas)
332 Check boolean input literals for 'true','false','yes','no','1','0'
333 and throw elog(ERROR) if unrecognized(Thomas)
334 Major large objects fix
335 Fix for GROUP BY showing duplicates(Vadim)
336 Fix for index scans in MergeJion(Vadim)
340 Subselects with EXISTS, IN, ALL, ANY keywords (Vadim, Bruce, Thomas)
341 New User Manual(Thomas, others)
342 Speedup by inlining some frequently-called functions
343 Real deadlock detection, no more timeouts(Bruce)
344 Add SQL92 "constants" CURRENT_DATE, CURRENT_TIME, CURRENT_TIMESTAMP,
346 Modify constraint syntax to be SQL92-compliant(Thomas)
347 Implement SQL92 PRIMARY KEY and UNIQUE clauses using indices(Thomas)
348 Recognize SQL92 syntax for FOREIGN KEY. Throw elog notice(Thomas)
349 Allow NOT NULL UNIQUE constraint clause (each allowed separately before)(Thomas)
350 Allow Postgres-style casting ("::") of non-constants(Thomas)
351 Add support for SQL3 TRUE and FALSE boolean constants(Thomas)
352 Support SQL92 syntax for IS TRUE/IS FALSE/IS NOT TRUE/IS NOT FALSE(Thomas)
353 Allow shorter strings for boolean literals (e.g. "t", "tr", "tru")(Thomas)
354 Allow SQL92 delimited identifiers(Thomas)
355 Implement SQL92 binary and hexadecimal string decoding (b'10' and x'1F')(Thomas)
356 Support SQL92 syntax for type coercion of literal strings
357 (e.g. "DATETIME 'now'")(Thomas)
358 Add conversions for int2, int4, and OID types to and from text(Thomas)
359 Use shared lock when building indices(Vadim)
360 Free memory allocated for an user query inside transaction block after
361 this query is done, was turned off in <= 6.2.1(Vadim)
362 New SQL statement CREATE PROCEDURAL LANGUAGE(Jan)
363 New <productname>Postgres</productname> Procedural Language (PL) backend interface(Jan)
364 Rename pg_dump -H option to -h(Bruce)
365 Add Java support for passwords, European dates(Peter)
366 Use indices for LIKE and ~, !~ operations(Bruce)
367 Add hash functions for datetime and timespan(Thomas)
368 Time Travel removed(Vadim, Bruce)
369 Add paging for \d and \z, and fix \i(Bruce)
370 Add Unix domain socket support to backend and to frontend library(Goran)
371 Implement CREATE DATABASE/WITH LOCATION and initlocation utility(Thomas)
372 Allow more SQL92 and/or <productname>Postgres</productname> reserved words as column identifiers(Thomas)
373 Augment support for SQL92 SET TIME ZONE...(Thomas)
374 SET/SHOW/RESET TIME ZONE uses TZ backend environment variable(Thomas)
375 Implement SET keyword = DEFAULT and SET TIME ZONE DEFAULT(Thomas)
376 Enable SET TIME ZONE using TZ environment variable(Thomas)
377 Add PGDATESTYLE environment variable to frontend and backend initialization(Thomas)
378 Add PGTZ, PGCOSTHEAP, PGCOSTINDEX, PGRPLANS, PGGEQO
379 frontend library initialization environment variables(Thomas)
380 Regression tests time zone automatically set with "setenv PGTZ PST8PDT"(Thomas)
381 Add pg_description table for info on tables, columns, operators, types, and
383 Increase 16 char limit on system table/index names to 32 characters(Bruce)
384 Rename system indices(Bruce)
385 Add 'GERMAN' option to SET DATESTYLE(Thomas)
386 Define an "ISO-style" timespan output format with "hh:mm:ss" fields(Thomas)
387 Allow fractional values for delta times (e.g. '2.5 days')(Thomas)
388 Validate numeric input more carefully for delta times(Thomas)
389 Implement day of year as possible input to date_part()(Thomas)
390 Define timespan_finite() and text_timespan() functions(Thomas)
391 Remove archive stuff(Bruce)
392 Allow for a pg_password authentication database that is separate from
393 the system password file(Todd)
394 Dump ACLs, GRANT, REVOKE permissions(Matt)
395 Define text, varchar, and bpchar string length functions(Thomas)
396 Fix Query handling for inheritance, and cost computations(Bruce)
397 Implement CREATE TABLE/AS SELECT (alternative to SELECT/INTO)(Thomas)
398 Allow NOT, IS NULL, IS NOT NULL in constraints(Thomas)
399 Implement UNIONs for SELECT(Bruce)
400 Add UNION, GROUP, DISTINCT to INSERT(Bruce)
401 varchar() stores only necessary bytes on disk(Bruce)
403 Mega-Patch for JDBC...see README_6.3 for list of changes(Peter)
404 Remove unused "option" from PQconnectdb()
405 New LOCK command and lock manual page describing deadlocks(Bruce)
406 Add new psql \da, \dd, \df, \do, \dS, and \dT commands(Bruce)
407 Enhance psql \z to show sequences(Bruce)
408 Show NOT NULL and DEFAULT in psql \d table(Bruce)
409 New psql .psqlrc file startup(Andrew)
410 Modify sample startup script in contrib/linux to show syslog(Thomas)
411 New types for IP and MAC addresses in contrib/ip_and_mac(TomH)
412 Unix system time conversions with date/time types in contrib/unixdate(Thomas)
413 Update of contrib stuff(Massimo)
414 Add Unix socket support to DBD::Pg(Goran)
415 New python interface (PyGreSQL 2.0)(D'Arcy)
416 New frontend/backend protocol has a version number, network byte order(Phil)
417 Security features in pg_hba.conf enhanced and documented, many cleanups(Phil)
418 CHAR() now faster access than VARCHAR() or TEXT
419 ecpg embedded SQL preprocessor
420 Reduce system column overhead(Vadmin)
421 Remove pg_time table(Vadim)
422 Add pg_type attribute to identify types that need length (bpchar, varchar)
423 Add report of offending line when COPY command fails
424 Allow VIEW permissions to be set separately from the underlying tables.
425 For security, use GRANT/REVOKE on views as appropriate(Jan)
426 Tables now have no default GRANT SELECT TO PUBLIC. You must
427 explicitly grant such permissions.
428 Clean up tutorial examples(Darren)
432 Add new html development tools, and flow chart in /tools/backend
434 Stratus computer port Robert Gillies
435 Added support for shlib for BSD44_derived & i386_solaris
436 Make configure more automated(Brook)
437 Add script to check regression test results
438 Break parser functions into smaller files, group together(Bruce)
439 Rename heap_create to heap_create_and_catalog, rename heap_creatr
440 to heap_create()(Bruce)
441 Sparc/Linux patch for locking(TomS)
442 Remove PORTNAME and reorganize port-specific stuff(Marc)
443 Add optimizer README file(Bruce)
444 Remove some recursion in optimizer and clean up some code there(Bruce)
445 Fix for NetBSD locking(Henry)
446 Fix for libptcl make(Tatsuo)
448 Change IS TRUE, IS FALSE, ... to expressions using "=" rather than
449 function calls to istrue() or isfalse() to allow optimization(Thomas)
450 Various fixes NetBSD/Sparc related(TomH)
451 Alpha linux locking(Travis,Ryan)
452 Change elog(WARN) to elog(ERROR)(Bruce)
453 FAQ for FreeBSD(Marc)
454 Bring in the PostODBC source tree as part of our standard distribution(Marc)
455 A minor patch for HP/UX 10 vs 9(Stan)
456 New pg_attribute.atttypmod for type-specific info like varchar length(Bruce)
457 Unixware patches(Billy)
458 New i386 'lock' for spin lock asm(Billy)
459 Support for multiplexed backends is removed
460 Start an OpenBSD port
463 Add string functions to regression suite(Thomas)
464 Expand a few function names formerly truncated to 16 characters(Thomas)
465 Remove un-needed malloc() calls and replace with palloc()(Bruce)
472 <Title>Release 6.2.1</Title>
477 <firstname>Bruce</firstname>
478 <surname>Momjian</surname>
482 Fri Oct 17 00:01:27 EDT 1997
488 v6.2.1 is a bug-fix and usability release on v6.2.
493 <itemizedlist spacing="compact">
496 Allow strings to span lines, per <acronym>SQL92</acronym>.
500 Include example trigger function for inserting user names on table updates.
505 This is a minor bug-fix release on v6.2.
506 For upgrades from pre-v6.2 systems, a full dump/reload is required.
507 Refer to the v6.2 release notes for instructions.
510 <title>Migration from v6.2 to v6.2.1</title>
513 This is a minor bug-fix release. A dump/reload is not required from v6.2,
514 but is required from any release prior to v6.2.
517 In upgrading from v6.2, if you choose to dump/reload you will find that
518 avg(money) is now calculated correctly. All other bug fixes take effect
519 upon updating the executables.
522 Another way to avoid dump/reload is to use the following SQL command
523 from psql to update the existing system table:
526 update pg_aggregate set aggfinalfn = 'cash_div_flt8'
527 where aggname = 'avg' and aggbasetype = 790;
531 This will need to be done to every existing database, including template1.
534 <title>Detailed Change List</title>
538 Changes in this release
539 -----------------------
540 Allow TIME and TYPE column names(Thomas)
541 Allow larger range of true/false as boolean values(Thomas)
542 Support output of "now" and "current"(Thomas)
543 Handle DEFAULT with INSERT of NULL properly(Vadim)
544 Fix for relation reference counts problem in buffer manager(Vadim)
545 Allow strings to span lines, like ANSI(Thomas)
546 Fix for backward cursor with ORDER BY(Vadim)
547 Fix avg(cash) computation(Thomas)
548 Fix for specifying a column twice in ORDER/GROUP BY(Vadim)
549 Documented new libpq function to return affected rows, PQcmdTuples(Bruce)
550 Trigger function for inserting user names for INSERT/UPDATE(Brook Milligan)
555 <Title>Release 6.2</Title>
560 <firstname>Bruce</firstname>
561 <surname>Momjian</surname>
565 Thu Oct 02 12:53:46 EDT 1997
571 A dump/restore is required for those wishing to migrate data from
572 previous releases of <productname>Postgres</productname>.
575 <title>Migration from v6.1 to v6.2</title>
578 This migration requires a complete dump of the 6.1 database and a
579 restore of the database in 6.2.
582 Note that the pg_dump and pg_dumpall utility from 6.2 should be used
583 to dump the 6.1 database.
586 <title>Migration from v1.x to v6.2</title>
589 Those migrating from earlier 1.* releases should first upgrade to 1.09
590 because the COPY output format was improved from the 1.02 release.
593 <title>Detailed Change List</title>
599 Fix problems with pg_dump for inheritance, sequences, archive tables(Bruce)
600 Fix compile errors on overflow due to shifts, unsigned, and bad prototypes
601 from Solaris(Diab Jerius)
602 Fix bugs in geometric line arithmetic (bad intersection calculations)(Thomas)
603 Check for geometric intersections at endpoints to avoid rounding ugliness(Thomas)
604 Catch non-functional delete attempts(Vadim)
605 Change time function names to be more consistent(Michael Reifenberg)
606 Check for zero divides(Michael Reifenberg)
607 Fix very old bug which made tuples changed/inserted by a commnd
608 visible to the command itself (so we had multiple update of
609 updated tuples, etc)(Vadim)
610 Fix for SELECT null, 'fail' FROM pg_am (Patrick)
611 SELECT NULL as EMPTY_FIELD now allowed(Patrick)
612 Remove un-needed signal stuff from contrib/pginterface
613 Fix OR (where x <> 1 or x isnull didn't return tuples with x NULL) (Vadim)
614 Fix time_cmp function (Vadim)
615 Fix handling of functions with non-attribute first argument in
616 WHERE clauses (Vadim)
617 Fix GROUP BY when order of entries is different from order
618 in target list (Vadim)
619 Fix pg_dump for aggregates without sfunc1 (Vadim)
623 Default genetic optimizer GEQO parameter is now 8(Bruce)
624 Allow use parameters in target list having aggregates in functions(Vadim)
625 Added JDBC driver as an interface(Adrian & Peter)
627 Return number of tuples inserted/affected by INSERT/UPDATE/DELETE etc.(Vadim)
628 Triggers implemented with CREATE TRIGGER (SQL3)(Vadim)
629 SPI (Server Programming Interface) allows execution of queries inside
631 NOT NULL implemented (SQL92)(Robson Paniago de Miranda)
632 Include reserved words for string handling, outer joins, and unions(Thomas)
633 Implement extended comments ("/* ... */") using exclusive states(Thomas)
634 Add "//" single-line comments(Bruce)
635 Remove some restrictions on characters in operator names(Thomas)
636 DEFAULT and CONSTRAINT for tables implemented (SQL92)(Vadim & Thomas)
637 Add text concatenation operator and function (SQL92)(Thomas)
638 Support WITH TIME ZONE syntax (SQL92)(Thomas)
639 Support INTERVAL unit TO unit syntax (SQL92)(Thomas)
640 Define types DOUBLE PRECISION, INTERVAL, CHARACTER,
641 and CHARACTER VARYING (SQL92)(Thomas)
642 Define type FLOAT(p) and rudimentary DECIMAL(p,s), NUMERIC(p,s) (SQL92)(Thomas)
643 Define EXTRACT(), POSITION(), SUBSTRING(), and TRIM() (SQL92)(Thomas)
644 Define CURRENT_DATE, CURRENT_TIME, CURRENT_TIMESTAMP (SQL92)(Thomas)
645 Add syntax and warnings for UNION, HAVING, INNER and OUTER JOIN (SQL92)(Thomas)
646 Add more reserved words, mostly for SQL92 compliance(Thomas)
647 Allow hh:mm:ss time entry for timespan/reltime types(Thomas)
648 Add center() routines for lseg, path, polygon(Thomas)
649 Add distance() routines for circle-polygon, polygon-polygon(Thomas)
650 Check explicitly for points and polygons contained within polygons
651 using an axis-crossing algorithm(Thomas)
652 Add routine to convert circle-box(Thomas)
653 Merge conflicting operators for different geometric data types(Thomas)
654 Replace distance operator "<===>" with "<->"(Thomas)
655 Replace "above" operator "!^" with ">^" and "below" operator "!|" with "<^"(Thomas)
656 Add routines for text trimming on both ends, substring, and string position(Thomas)
657 Added conversion routines circle(box) and poly(circle)(Thomas)
658 Allow internal sorts to be stored in memory rather than in files(Bruce & Vadim)
659 Allow functions and operators on internally-identical types to succeed(Bruce)
660 Speed up backend startup after profiling analysis(Bruce)
661 Inline frequently called functions for performance(Bruce)
662 Reduce open() calls(Bruce)
663 psql: Add PAGER for \h and \?,\C fix
664 Fix for psql pager when no tty(Bruce)
665 New entab utility(Bruce)
666 General trigger functions for referential integrity (Vadim)
667 General trigger functions for time travel (Vadim)
668 General trigger functions for AUTOINCREMENT/IDENTITY feature (Vadim)
669 MOVE implementation (Vadim)
673 HPUX 10 patches (Vladimir Turin)
674 Added SCO support, (Daniel Harris)
675 mkLinux patches (Tatsuo Ishii)
676 Change geometric box terminology from "length" to "width"(Thomas)
677 Deprecate temporary unstored slope fields in geometric code(Thomas)
678 Remove restart instructions from INSTALL(Bruce)
679 Look in /usr/ucb first for install(Bruce)
680 Fix c++ copy example code(Thomas)
681 Add -o to psql manual page(Bruce)
682 Prevent relname unallocated string length from being copied into database(Bruce)
683 Cleanup for NAMEDATALEN use(Bruce)
684 Fix pg_proc names over 15 chars in output(Bruce)
685 Add strNcpy() function(Bruce)
686 remove some (void) casts that are unnecessary(Bruce)
687 new interfaces directory(Marc)
688 Replace fopen() calls with calls to fd.c functions(Bruce)
689 Make functions static where possible(Bruce)
690 enclose unused functions in #ifdef NOT_USED(Bruce)
691 Remove call to difftime() in timestamp support to fix SunOS(Bruce & Thomas)
692 Changes for Digital Unix
693 Portability fix for pg_dumpall(Bruce)
694 Rename pg_attribute.attnvals to attdisbursion(Bruce)
695 "intro/unix" manual page now "pgintro"(Bruce)
696 "built-in" manual page now "pgbuiltin"(Bruce)
697 "drop" manual page now "drop_table"(Bruce)
698 Add "create_trigger", "drop_trigger" manual pages(Thomas)
699 Add constraints regression test(Vadim & Thomas)
700 Add comments syntax regression test(Thomas)
701 Add PGINDENT and support program(Bruce)
702 Massive commit to run PGINDENT on all *.c and *.h files(Bruce)
703 Files moved to /src/tools directory(Bruce)
704 SPI and Trigger programming guides (Vadim & D'Arcy)
709 <Title>Release 6.1.1</Title>
714 <firstname>Bruce</firstname>
715 <surname>Momjian</surname>
719 Mon Jul 22 18:04:49 EDT 1997
726 <title>Migration from v6.1 to v6.1.1</title>
729 This is a minor bug-fix release. A dump/reload is not required from v6.1,
730 but is required from any release prior to v6.1.
731 Refer to the release notes for v6.1 for more details.
734 <title>Detailed Change List</title>
738 Changes in this release
739 -----------------------
740 fix for SET with options (Thomas)
741 allow pg_dump/pg_dumpall to preserve ownership of all tables/objects(Bruce)
742 new psql \connect option allows changing usernames without chaning databases
743 fix for initdb --debug option(Yoshihiko Ichikawa))
744 lextest cleanup(Bruce)
746 fix date/time month boundary arithmetic(Thomas)
747 fix timezone daylight handling for some ports(Thomas, Bruce, Tatsuo)
748 timestamp overhauled to use standard functions(Thomas)
749 other code cleanup in date/time routines(Thomas)
750 psql's \d now case-insensitive(Bruce)
751 psql's backslash commands can now have trailing semicolon(Bruce)
752 fix memory leak in psql when using \g(Bruce)
753 major fix for endian handling of communication to server(Thomas, Tatsuo)
754 Fix for Solaris assembler and include files(Yoshihiko Ichikawa)
755 allow underscores in usernames(Bruce)
756 pg_dumpall now returns proper status, portability fix(Bruce)
762 <Title>Release 6.1</Title>
767 <firstname>Bruce</firstname>
768 <surname>Momjian</surname>
772 Sun Jun 8 14:41:13 EDT 1997
778 The regression tests have been adapted and extensively modified for the
779 v6.1 release of <productname>Postgres</productname>.
783 Three new data types (datetime, timespan, and circle) have been added to
784 the native set of <productname>Postgres</productname> types. Points, boxes, paths, and polygons
785 have had their output formats made consistant across the data types.
786 The polygon output in misc.out has only been spot-checked for correctness
787 relative to the original regression output.
791 <productname>Postgres</productname> v6.1 introduces a new, alternate optimizer which uses <FirstTerm>genetic</FirstTerm>
792 algorithms. These algorithms introduce a random behavior in the ordering
793 of query results when the query contains multiple qualifiers or multiple
794 tables (giving the optimizer a choice on order of evaluation). Several
795 regression tests have been modified to explicitly order the results, and
796 hence are insensitive to optimizer choices. A few regression tests are
797 for data types which are inherently unordered (e.g. points and time
798 intervals) and tests involving those types are explicitly bracketed with
799 <Command>set geqo to 'off'</Command> and <Command>reset geqo</Command>.
803 The interpretation of array specifiers (the curly braces around atomic
804 values) appears to have changed sometime after the original regression
805 tests were generated. The current <FileName>./expected/*.out</FileName> files reflect this
806 new interpretation, which may not be correct!
810 The float8 regression test fails on at least some platforms. This is due
811 to differences in implementations of pow() and exp() and the signaling
812 mechanisms used for overflow and underflow conditions.
816 The "random" results in the random test should cause the "random" test
817 to be "failed", since the regression tests are evaluated using a simple
818 diff. However, "random" does not seem to produce random results on my
819 test machine (Linux/gcc/i686).
823 <title>Migration to v6.1</title>
826 This migration requires a complete dump of the 6.0 database and a
827 restore of the database in 6.1.
830 Those migrating from earlier 1.* releases should first upgrade to 1.09
831 because the COPY output format was improved from the 1.02 release.
834 <title>Detailed Change List</title>
840 packet length checking in library routines
841 lock manager priority patch
842 check for under/over flow of float8(Bruce)
843 multi-table join fix(Vadim)
844 SIGPIPE crash fix(Darren)
845 large object fixes(Sven)
846 allow btree indexes to handle NULLs(Vadim)
847 timezone fixes(D'Arcy)
848 select SUM(x) can return NULL on no rows(Thomas)
849 internal optimizer, executor bug fixes(Vadim)
850 fix problem where inner loop in < or <= has no rows(Vadim)
851 prevent re-commuting join index clauses(Vadim)
852 fix join clauses for multiple tables(Vadim)
853 fix hash, hashjoin for arrays(Vadim)
854 fix btree for abstime type(Vadim)
855 large object fixes(Raymond)
856 fix buffer leak in hash indices (Vadim)
857 fix rtree for use in inner scan (Vadim)
858 fix gist for use in inner scan, cleanups (Vadim, Andrea)
859 avoid unnecessary local buffers allocation (Vadim, Massimo)
860 fix local buffers leak in transaction aborts (Vadim)
861 fix file manager memmory leaks, cleanups (Vadim, Massimo)
862 fix storage manager memmory leaks (Vadim)
863 fix btree duplicates handling (Vadim)
864 fix deleted tuples re-incarnation caused by vacuum (Vadim)
865 fix SELECT varchar()/char() INTO TABLE made zero-length fields(Bruce)
866 many psql, pg_dump, and libpq memory leaks fixed using Purify (Igor)
870 attribute optimization statistics(Bruce)
871 much faster new btree bulk load code(Paul)
872 BTREE UNIQUE added to bulk load code(Vadim)
873 new lock debug code(Massimo)
874 massive changes to libpg++(Leo)
875 new GEQO optimizer speeds table multi-table optimization(Martin)
876 new WARN message for non-unique insert into unique key(Marc)
877 update x=-3, no spaces, now valid(Bruce)
878 remove case-sensitive identifier handling(Bruce,Thomas,Dan)
879 debug backend now pretty-prints tree(Darren)
880 new Oracle character functions(Edmund)
881 new plaintext password functions(Dan)
882 no such class or insufficient privilege changed to distinct messages(Dan)
883 new ANSI timestamp function(Dan)
884 new ANSI Time and Date types (Thomas)
885 move large chunks of data in backend(Martin)
886 multi-column btree indexes(Vadim)
887 new SET var TO value command(Martin)
888 update transaction status on reads(Dan)
889 new locale settings for character types(Oleg)
890 new SEQUENCE serial number generator(Vadim)
891 GROUP BY function now possible(Vadim)
892 re-organize regression test(Thomas,Marc)
893 new optimizer operation weights(Vadim)
894 new psql \z grant/permit option(Marc)
895 new MONEY data type(D'Arcy,Thomas)
896 tcp socket communication speed improved(Vadim)
897 new VACUUM option for attribute statistics, and for certain columns (Vadim)
898 many geometric type improvements(Thomas,Keith)
899 additional regression tests(Thomas)
900 new datestyle variable(Thomas,Vadim,Martin)
901 more comparison operators for sorting types(Thomas)
902 new conversion functions(Thomas)
903 new more compact btree format(Vadim)
904 allow pg_dumpall to preserve database ownership(Bruce)
905 new SET GEQO=# and R_PLANS variable(Vadim)
906 old (!GEQO) optimizer can use right-sided plans (Vadim)
907 typechecking improvement in SQL parser(Bruce)
908 new SET, SHOW, RESET commands(Thomas,Vadim)
909 new \connect database USER option
910 new destroydb -i option (Igor)
911 new \dt and \di psql commands (Darren)
912 SELECT "\n" now escapes newline (A. Duursma)
913 new geometry conversion functions from old format (Thomas)
917 new configuration script(Marc)
918 readline configuration option added(Marc)
919 OS-specific configuration options removed(Marc)
920 new OS-specific template files(Marc)
921 no more need to edit Makefile.global(Marc)
922 re-arrange include files(Marc)
923 nextstep patches (Gregor Hoffleit)
924 removed WIN32-specific code(Bruce)
925 removed postmaster -e option, now only postgres -e option (Bruce)
926 merge duplicate library code in front/backends(Martin)
927 now works with eBones, international Kerberos(Jun)
928 more shared library support
929 c++ include file cleanup(Bruce)
930 warn about buggy flex(Bruce)
931 DG-UX, Ultrix, Irix, AIX portability fixes
936 <title>Release v6.0</title>
941 <firstname>Bruce</firstname>
942 <surname>Momjian</surname>
946 Wed Jan 29 00:19:54 EST 1997
952 A dump/restore is required for those wishing to migrate data from
953 previous releases of <productname>Postgres</productname>.
956 <title>Migration from v1.09 to v6.0</title>
959 This migration requires a complete dump of the 1.09 database and a
960 restore of the database in 6.0.
963 <title>Migration from pre-v1.09 to v6.0</title>
966 Those migrating from earlier 1.* releases should first upgrade to 1.09
967 because the COPY output format was improved from the 1.02 release.
970 <title>Detailed Change List</title>
976 ALTER TABLE bug - running postgress process needs to re-read table definition
977 Allow vacuum to be run on one table or entire database(Bruce)
979 Fix array over-runs of memory writes(Kurt)
980 Fix elusive btree range/non-range bug(Dan)
981 Fix for hash indexes on some types like time and date
982 Fix for pg_log size explosion
983 Fix permissions on lo_export()(Bruce)
984 Fix unitialized reads of memory(Kurt)
985 Fixed ALTER TABLE ... char(3) bug(Bruce)
986 Fixed a few small memory leaks
987 Fixed EXPLAIN handling of options and changed full_path option name
988 Fixed output of group acl permissions
989 Memory leaks (hunt and destroy with tools like Purify(Kurt)
990 Minor improvements to rules system
992 New asserts for run-checking
993 Overhauled parser/analyze code to properly report errors and increase speed
994 Pg_dump -d now handles NULL's properly(Bruce)
995 Prevent SELECT NULL from crashing server (Bruce)
996 Properly report errors when INSERT ... SELECT columns did not match
997 Properly report errors when insert column names were not correct
998 Psql \g filename now works(Bruce)
999 Psql fixed problem with multiple statements on one line with multiple outputs
1000 Removed duplicate system oid's
1001 SELECT * INTO TABLE . GROUP/ORDER BY gives unlink error if table exists(Bruce)
1002 Several fixes for queries that crashed the backend
1003 Starting quote in insert string errors(Bruce)
1004 Submitting an empty query now returns empty status, not just " " query(Bruce)
1008 Add EXPLAIN manual page(Bruce)
1009 Add UNIQUE index capability(Dan)
1010 Add hostname/user level access control rather than just hostname and user
1011 Add synonym of != for <>(Bruce)
1012 Allow "select oid,* from table"
1013 Allow BY,ORDER BY to specify columns by number, or by non-alias table.column(Bruce)
1014 Allow COPY from the frontend(Bryan)
1015 Allow GROUP BY to use alias column name(Bruce)
1016 Allow actual compression, not just reuse on the same page(Vadim)
1017 Allow installation-configuration option to auto-add all local users(Bryan)
1018 Allow libpq to distinguish between text value '' and null(Bruce)
1019 Allow non-postgres users with createdb privs to destroydb's
1020 Allow restriction on who can create C functions(Bryan)
1021 Allow restriction on who can do backend COPY(Bryan)
1022 Can shrink tables, pg_time and pg_log(Vadim & Erich)
1023 Change debug level 2 to print queries only, changed debug heading layout(Bruce)
1024 Change default decimal constant representation from float4 to float8(Bruce)
1025 European date format now set when postmaster is started
1026 Execute lowercase function names if not found with exact case
1027 Fixes for aggregate/GROUP processing, allow 'select sum(func(x),sum(x+y) from z'
1028 Gist now included in the distrubution(Marc)
1029 Idend authentication of local users(Bryan)
1030 Implement BETWEEN qualifier(Bruce)
1031 Implement IN qualifier(Bruce)
1032 Libpq has PQgetisnull()(Bruce)
1033 Libpq++ improvements
1034 New options to initdb(Bryan)
1035 Pg_dump allow dump of oid's(Bruce)
1036 Pg_dump create indexes after tables are loaded for speed(Bruce)
1037 Pg_dumpall dumps all databases, and the user table
1038 Pginterface additions for NULL values(Bruce)
1039 Prevent postmaster from being run as root
1040 Psql \h and \? is now readable(Bruce)
1041 Psql allow backslashed, semicolons anywhere on the line(Bruce)
1042 Psql changed command prompt for lines in query or in quotes(Bruce)
1043 Psql char(3) now displays as (bp)char in \d output(Bruce)
1044 Psql return code now more accurate(Bryan?)
1045 Psql updated help syntax(Bruce)
1046 Re-visit and fix vacuum(Vadim)
1047 Reduce size of regression diffs, remove timezone name difference(Bruce)
1048 Remove compile-time parameters to enable binary distributions(Bryan)
1049 Reverse meaning of HBA masks(Bryan)
1050 Secure Authentication of local users(Bryan)
1051 Speed up vacuum(Vadim)
1052 Vacuum now had VERBOSE option(Bruce)
1056 All functions now have prototypes that are compared against the calls
1057 Allow asserts to be disabled easly from Makefile.global(Bruce)
1058 Change oid constants used in code to #define names
1059 Decoupled sparc and solaris defines(Kurt)
1060 Gcc -Wall compiles cleanly with warnings only from unfixable constructs
1061 Major include file reorganization/reduction(Marc)
1062 Make now stops on compile failure(Bryan)
1063 Makefile restructuring(Bryan, Marc)
1064 Merge bsdi_2_1 to bsdi(Bruce)
1065 Monitor program removed
1066 Name change from Postgres95 to PostgreSQL
1067 New config.h file(Marc, Bryan)
1068 PG_VERSION now set to 6.0 and used by postmaster
1069 Portability additions, including Ultrix, DG/UX, AIX, and Solaris
1070 Reduced the number of #define's, centeralized #define's
1071 Remove duplicate OIDS in system tables(Dan)
1072 Remove duplicate system catalog info or report mismatches(Dan)
1073 Removed many os-specific #define's
1074 Restructured object file generation/location(Bryan, Marc)
1075 Restructured port-specific file locations(Bryan, Marc)
1076 Unused/uninialized variables corrected
1081 <title>Release v1.09</title>
1086 <firstname>Bruce</firstname>
1087 <surname>Momjian</surname>
1097 Sorry, we stopped keeping track of changes from 1.02 to 1.09. Some of
1098 the changes listed in 6.0 were actually included in the 1.02.1 to 1.09
1103 <title>Release v1.02</title>
1108 <firstname>Bruce</firstname>
1109 <surname>Momjian</surname>
1113 Thu Aug 1 18:00:00 EDT 1996
1119 <title>Migration from v1.02 to v1.02.1</title>
1122 Here is a new migration file for 1.02.1. It includes the 'copy' change
1123 and a script to convert old ascii files.
1127 The following notes are for the benefit of users who want to migrate
1128 databases from postgres95 1.01 and 1.02 to postgres95 1.02.1.
1131 If you are starting afresh with postgres95 1.02.1 and do not need
1132 to migrate old databases, you do not need to read any further.
1136 In order to upgrade older postgres95 version 1.01 or 1.02 databases to
1137 version 1.02.1, the following steps are required:
1142 Start up a new 1.02.1 postmaster
1146 Add the new built-in functions and operators of 1.02.1 to 1.01 or 1.02
1147 databases. This is done by running the new 1.02.1 server against
1148 your own 1.01 or 1.02 database and applying the queries attached at
1149 the end of thie file. This can be done easily through psql. If your
1150 1.01 or 1.02 database is named "testdb" and you have cut the commands
1151 from the end of this file and saved them in addfunc.sql:
1153 % psql testdb -f addfunc.sql
1156 Those upgrading 1.02 databases will get a warning when executing the
1157 last two statements in the file because they are already present in 1.02. This is
1158 not a cause for concern.
1162 <title>Dump/Reload Procedure</title>
1165 If you are trying to reload a pg_dump or text-mode 'copy tablename to
1166 stdout' generated with a previous version, you will need to run the
1167 attached sed script on the ASCII file before loading it into the
1168 database. The old format used '.' as end-of-data, while '\.' is now the
1169 end-of-data marker. Also, empty strings are now loaded in as '' rather
1170 than NULL. See the copy manual page for full details.
1173 sed 's/^\.$/\\./g' <in_file >out_file
1177 If you are loading an older binary copy or non-stdout copy, there is no
1178 end-of-data character, and hence no conversion necessary.
1181 -- following lines added by agc to reflect the case-insensitive
1182 -- regexp searching for varchar (in 1.02), and bpchar (in 1.02.1)
1183 create operator ~* (leftarg = bpchar, rightarg = text, procedure = texticregexeq);
1184 create operator !~* (leftarg = bpchar, rightarg = text, procedure = texticregexne);
1185 create operator ~* (leftarg = varchar, rightarg = text, procedure = texticregexeq);
1186 create operator !~* (leftarg = varchar, rightarg = text, procedure = texticregexne);
1191 <title>Detailed Change List</title>
1195 Source code maintenance and development
1196 * worldwide team of volunteers
1197 * the source tree now in CVS at ftp.ki.net
1200 * psql (and underlying libpq library) now has many more options for
1201 formatting output, including HTML
1202 * pg_dump now output the schema and/or the data, with many fixes to
1203 enhance completeness.
1204 * psql used in place of monitor in administration shell scripts.
1205 monitor to be depreciated in next release.
1206 * date/time functions enhanced
1207 * NULL insert/update/comparison fixed/enhanced
1208 * TCL/TK lib and shell fixed to work with both tck7.4/tk4.0 and tcl7.5/tk4.1
1210 Bug Fixes (almost too numerous to mention)
1212 * storage management
1213 * check for NULL pointer before dereferencing
1217 * added SolarisX86 port
1218 * added BSDI 2.1 port
1223 Contributors (appologies to any missed)
1224 * Kurt J. Lidl <lidl@va.pubnix.com>
1225 (missed in first run, but no less important)
1226 * Erich Stamberger <eberger@gewi.kfunigraz.ac.at>
1227 * Jason Wright <jason@shiloh.vnet.net>
1228 * Cees de Groot <C.deGroot@inter.NL.net>
1229 * ernst.molitor@uni-bonn.de
1230 * michael.siebenborn@ae3.Hypo.DE (Michael Siebenborn (6929))
1231 * Brian E. Gallew <geek+@cmu.edu>
1232 * Vadim B. Mikheev <vadim@sable.krasnoyarsk.su>
1233 * Adam Sussman <myddryn@vidya.com>
1234 * Chris Dunlop <chris@onthe.net.au>
1235 * Marc G. Fournier <scrappy@ki.net>
1236 * Dan McGuirk <mcguirk@indirect.com>
1237 * Dr_George_D_Detlefsen <drgeorge@ilt.com>
1238 * Erich Stamberger <eberger@gewi.kfunigraz.ac.at>
1239 * Massimo Dal Zotto <dz@cs.unitn.it>
1240 * Randy Kunkee <kunkee@Starbase.NeoSoft.COM>
1241 * Rick Weldon <rick@wisetech.com>
1242 * Thomas van Reimersdahl <reimersd@dali.techinfo.rwth-aachen.de>
1243 * david bennett <dave@bensoft.com>
1244 * ernst.molitor@uni-bonn.de
1245 * Julian Assange <proff@suburbia.net>
1246 * Bruce Momjian <maillist@candle.pha.pa.us>
1247 * Paul "Shag" Walmsley <ccshag@cclabs.missouri.edu>
1248 * "Alistair G. Crooks" <azcb0@sde.uts.amdahl.com>
1252 <title>Release v1.01</title>
1257 <firstname>Bruce</firstname>
1258 <surname>Momjian</surname>
1262 Fri Feb 23 18:20:36 PST 1996
1268 <title>Migration from v1.0 to v1.01</title>
1271 The following notes are for the benefit of users who want to migrate
1272 databases from postgres95 1.0 to postgres95 1.01.
1275 If you are starting afresh with postgres95 1.01 and do not need
1276 to migrate old databases, you do not need to read any further.
1279 In order to postgres95 version 1.01 with databases created with
1280 postgres95 version 1.0, the following steps are required:
1285 Set the definition of NAMEDATALEN in src/Makefile.global to 16
1286 and OIDNAMELEN to 20.
1290 Decide whether you want to use Host based authentication.
1295 If you do, you must create a file name "pg_hba" in your top-level data
1296 directory (typically the value of your $PGDATA). src/libpq/pg_hba
1297 shows an example syntax.
1301 If you do not want host-based authentication, you can comment out
1306 in src/Makefile.global
1309 Note that host-based authentication is turned on by default, and if
1310 you do not take steps A or B above, the out-of-the-box 1.01 will
1311 not allow you to connect to 1.0 databases.
1316 Compile and install 1.01, but DO NOT do the initdb step.
1320 Before doing anything else, terminate your 1.0 postmaster, and
1321 backup your existing $PGDATA directory.
1325 Set your PGDATA environment variable to your 1.0 databases, but set up
1326 path up so that 1.01 binaries are being used.
1330 Modify the file $PGDATA/PG_VERSION from 5.0 to 5.1
1334 Start up a new 1.01 postmaster
1338 Add the new built-in functions and operators of 1.01 to 1.0
1339 databases. This is done by running the new 1.01 server against
1340 your own 1.0 database and applying the queries attached and saving
1341 in the file 1.0_to_1.01.sql. This can be done easily through psql.
1342 If your 1.0 database is name "testdb":
1345 % psql testdb -f 1.0_to_1.01.sql
1348 and then execute the following commands (cut and paste from here):
1351 -- add builtin functions that are new to 1.01
1353 create function int4eqoid (int4, oid) returns bool as 'foo'
1354 language 'internal';
1355 create function oideqint4 (oid, int4) returns bool as 'foo'
1356 language 'internal';
1357 create function char2icregexeq (char2, text) returns bool as 'foo'
1358 language 'internal';
1359 create function char2icregexne (char2, text) returns bool as 'foo'
1360 language 'internal';
1361 create function char4icregexeq (char4, text) returns bool as 'foo'
1362 language 'internal';
1363 create function char4icregexne (char4, text) returns bool as 'foo'
1364 language 'internal';
1365 create function char8icregexeq (char8, text) returns bool as 'foo'
1366 language 'internal';
1367 create function char8icregexne (char8, text) returns bool as 'foo'
1368 language 'internal';
1369 create function char16icregexeq (char16, text) returns bool as 'foo'
1370 language 'internal';
1371 create function char16icregexne (char16, text) returns bool as 'foo'
1372 language 'internal';
1373 create function texticregexeq (text, text) returns bool as 'foo'
1374 language 'internal';
1375 create function texticregexne (text, text) returns bool as 'foo'
1376 language 'internal';
1378 -- add builtin functions that are new to 1.01
1380 create operator = (leftarg = int4, rightarg = oid, procedure = int4eqoid);
1381 create operator = (leftarg = oid, rightarg = int4, procedure = oideqint4);
1382 create operator ~* (leftarg = char2, rightarg = text, procedure = char2icregexeq);
1383 create operator !~* (leftarg = char2, rightarg = text, procedure = char2icregexne);
1384 create operator ~* (leftarg = char4, rightarg = text, procedure = char4icregexeq);
1385 create operator !~* (leftarg = char4, rightarg = text, procedure = char4icregexne);
1386 create operator ~* (leftarg = char8, rightarg = text, procedure = char8icregexeq);
1387 create operator !~* (leftarg = char8, rightarg = text, procedure = char8icregexne);
1388 create operator ~* (leftarg = char16, rightarg = text, procedure = char16icregexeq);
1389 create operator !~* (leftarg = char16, rightarg = text, procedure = char16icregexne);
1390 create operator ~* (leftarg = text, rightarg = text, procedure = texticregexeq);
1391 create operator !~* (leftarg = text, rightarg = text, procedure = texticregexne);
1397 <title>Detailed Change List</title>
1402 * 1.01 is backwards compatible with 1.0 database provided the user
1403 follow the steps outlined in the MIGRATION_from_1.0_to_1.01 file.
1404 If those steps are not taken, 1.01 is not compatible with 1.0 database.
1407 * added PQdisplayTuples() to libpq and changed monitor and psql to use it
1408 * added NeXT port (requires SysVIPC implementation)
1409 * added CAST .. AS ... syntax
1410 * added ASC and DESC keywords
1411 * added 'internal' as a possible language for CREATE FUNCTION
1412 internal functions are C functions which have been statically linked
1413 into the postgres backend.
1414 * a new type "name" has been added for system identifiers (table names,
1415 attribute names, etc.) This replaces the old char16 type. The
1416 of name is set by the NAMEDATALEN #define in src/Makefile.global
1417 * a readable reference manual that describes the query language.
1418 * added host-based access control. A configuration file ($PGDATA/pg_hba)
1419 is used to hold the configuration data. If host-based access control
1420 is not desired, comment out HBA=1 in src/Makefile.global.
1421 * changed regex handling to be uniform use of Henry Spencer's regex code
1422 regardless of platform. The regex code is included in the distribution
1423 * added functions and operators for case-insensitive regular expressions.
1424 The operators are ~* and !~*.
1425 * pg_dump uses COPY instead of SELECT loop for better performance
1428 * fixed an optimizer bug that was causing core dumps when
1429 functions calls were used in comparisons in the WHERE clause
1430 * changed all uses of getuid to geteuid so that effective uids are used
1431 * psql now returns non-zero status on errors when using -c
1432 * applied public patches 1-14
1436 <title>Release v1.0</title>
1441 <firstname>Bruce</firstname>
1442 <surname>Momjian</surname>
1446 Tue Sep 5 11:24:11 PDT 1995
1452 <title>Detailed Change List</title>
1457 * The copyright of <productname>Postgres</productname> 1.0 has been loosened to be freely modifiable
1458 and modifiable for any purpose. Please read the COPYRIGHT file.
1459 Thanks to Professor Michael Stonebraker for making this possible.
1462 * date formats have to be MM-DD-YYYY (or DD-MM-YYYY if you're using
1463 EUROPEAN STYLE). This follows SQL-92 specs.
1464 * "delimiters" is now a keyword
1467 * sql LIKE syntax has been added
1468 * copy command now takes an optional USING DELIMITER specification.
1469 delimiters can be any single-character string.
1470 * IRIX 5.3 port has been added.
1471 Thanks to Paul Walmsley and others.
1472 * updated pg_dump to work with new libpq
1473 * \d has been added psql
1474 Thanks to Keith Parks
1475 * regexp performance for architectures that use POSIX regex has been
1476 improved due to caching of precompiled patterns.
1477 Thanks to Alistair Crooks
1478 * a new version of libpq++
1479 Thanks to William Wanders
1482 * arbitrary userids can be specified in the createuser script
1483 * \c to connect to other databases in psql now works.
1484 * bad pg_proc entry for float4inc() is fixed
1485 * users with usecreatedb field set can now create databases without
1486 having to be usesuper
1487 * remove access control entries when the entry no longer has any
1489 * fixed non-portable datetimes implementation
1490 * added kerberos flags to the src/backend/Makefile
1491 * libpq now works with kerberos
1492 * typographic errors in the user manual have been corrected.
1493 * btrees with multiple index never worked, now we tell you they don't
1494 work when you try to use them
1498 <title><productname>Postgres95</productname> Beta 0.03</title>
1503 <firstname>Bruce</firstname>
1504 <surname>Momjian</surname>
1508 Fri Jul 21 14:49:31 PDT 1995
1514 <title>Detailed Change List</title>
1518 Incompatible changes:
1519 * BETA-0.3 IS INCOMPATIBLE WITH DATABASES CREATED WITH PREVIOUS VERSIONS
1520 (due to system catalog changes and indexing structure changes).
1521 * double-quote (") is deprecated as a quoting character for string literals;
1522 you need to convert them to single quotes (').
1523 * name of aggregates (eg. int4sum) are renamed in accordance with the
1524 SQL standard (eg. sum).
1525 * CHANGE ACL syntax is replaced by GRANT/REVOKE syntax.
1526 * float literals (eg. 3.14) are now of type float4 (instead of float8 in
1527 previous releases); you might have to do typecasting if you depend on it
1528 being of type float8. If you neglect to do the typecasting and you assign
1529 a float literal to a field of type float8, you may get incorrect values
1531 * LIBPQ has been totally revamped so that frontend applications
1532 can connect to multiple backends
1533 * the usesysid field in pg_user has been changed from int2 to int4 to
1534 allow wider range of Unix user ids.
1535 * the netbsd/freebsd/bsd o/s ports have been consolidated into a
1536 single BSD44_derived port. (thanks to Alistair Crooks)
1538 SQL standard-compliance (the following details changes that makes postgres95
1539 more compliant to the SQL-92 standard):
1540 * the following SQL types are now built-in: smallint, int(eger), float, real,
1541 char(N), varchar(N), date and time.
1543 The following are aliases to existing postgres types:
1545 integer, int -> int4
1546 float, real -> float4
1547 char(N) and varchar(N) are implemented as truncated text types. In
1548 addition, char(N) does blank-padding.
1549 * single-quote (') is used for quoting string literals; '' (in addition to
1550 \') is supported as means of inserting a single quote in a string
1551 * SQL standard aggregate names (MAX, MIN, AVG, SUM, COUNT) are used
1552 (Also, aggregates can now be overloaded, i.e. you can define your
1553 own MAX aggregate to take in a user-defined type.)
1554 * CHANGE ACL removed. GRANT/REVOKE syntax added.
1555 - Privileges can be given to a group using the "GROUP" keyword.
1557 GRANT SELECT ON foobar TO GROUP my_group;
1558 The keyword 'PUBLIC' is also supported to mean all users.
1560 Privileges can only be granted or revoked to one user or group
1563 "WITH GRANT OPTION" is not supported. Only class owners can change
1565 - The default access control is to to grant users readonly access.
1566 You must explicitly grant insert/update access to users. To change
1567 this, modify the line in
1568 src/backend/utils/acl.h
1569 that defines ACL_WORLD_DEFAULT
1572 * the bug where aggregates of empty tables were not run has been fixed. Now,
1573 aggregates run on empty tables will return the initial conditions of the
1574 aggregates. Thus, COUNT of an empty table will now properly return 0.
1575 MAX/MIN of an empty table will return a tuple of value NULL.
1576 * allow the use of \; inside the monitor
1577 * the LISTEN/NOTIFY asynchronous notification mechanism now work
1578 * NOTIFY in rule action bodies now work
1579 * hash indices work, and access methods in general should perform better.
1580 creation of large btree indices should be much faster. (thanks to Paul
1583 Other changes and enhancements:
1584 * addition of an EXPLAIN statement used for explaining the query execution
1585 plan (eg. "EXPLAIN SELECT * FROM EMP" prints out the execution plan for
1587 * WARN and NOTICE messages no longer have timestamps on them. To turn on
1588 timestamps of error messages, uncomment the line in
1589 src/backend/utils/elog.h:
1590 /* define ELOG_TIMESTAMPS */
1591 * On an access control violation, the message
1592 "Either no such class or insufficient privilege"
1593 will be given. This is the same message that is returned when
1594 a class is not found. This dissuades non-privileged users from
1595 guessing the existence of privileged classes.
1596 * some additional system catalog changes have been made that are not
1597 visible to the user.
1600 * The -oid option has been added to the "pg_result" tcl command.
1601 pg_result -oid returns oid of the last tuple inserted. If the
1602 last command was not an INSERT, then pg_result -oid returns "".
1603 * the large object interface is available as pg_lo* tcl commands:
1604 pg_lo_open, pg_lo_close, pg_lo_creat, etc.
1606 Portability enhancements and New Ports:
1607 * flex/lex problems have been cleared up. Now, you should be able to use
1608 flex instead of lex on any platforms. We no longer make assumptions of
1609 what lexer you use based on the platform you use.
1610 * The Linux-ELF port is now supported. Various configuration have been
1611 tested: The following configuration is known to work:
1612 kernel 1.2.10, gcc 2.6.3, libc 4.7.2, flex 2.5.2, bison 1.24
1613 with everything in ELF format,
1616 * ipcclean added to the distribution
1617 ipcclean usually does not need to be run, but if your backend crashes
1618 and leaves shared memory segments hanging around, ipcclean will
1619 clean them up for you.
1622 * the user manual has been revised and libpq documentation added.
1627 <title><productname>Postgres95</productname> Beta 0.02</title>
1632 <firstname>Bruce</firstname>
1633 <surname>Momjian</surname>
1637 Thu May 25 16:54:46 PDT 1995
1643 <title>Detailed Change List</title>
1647 Incompatible changes:
1648 * The SQL statement for creating a database is 'CREATE DATABASE' instead
1649 of 'CREATEDB'. Similarly, dropping a database is 'DROP DATABASE' instead
1650 of 'DESTROYDB'. However, the names of the executables 'createdb' and
1651 'destroydb' remain the same.
1654 * pgperl - a Perl (4.036) interface to Postgres95
1655 * pg_dump - a utility for dumping out a postgres database into a
1656 script file containing query commands. The script files are in a ASCII
1657 format and can be used to reconstruct the database, even on other
1658 machines and other architectures. (Also good for converting
1659 a Postgres 4.2 database to Postgres95 database.)
1661 The following ports have been incorporated into postgres95-beta-0.02:
1662 * the NetBSD port by Alistair Crooks
1663 * the AIX port by Mike Tung
1664 * the Windows NT port by Jon Forrest (more stuff but not done yet)
1665 * the Linux ELF port by Brian Gallew
1667 The following bugs have been fixed in postgres95-beta-0.02:
1668 * new lines not escaped in COPY OUT and problem with COPY OUT when first
1670 * cannot type return to use the default user id in createuser
1671 * SELECT DISTINCT on big tables crashes
1672 * Linux installation problems
1673 * monitor doesn't allow use of 'localhost' as PGHOST
1674 * psql core dumps when doing \c or \l
1675 * the "pgtclsh" target missing from src/bin/pgtclsh/Makefile
1676 * libpgtcl has a hard-wired default port number
1677 * SELECT DISTINCT INTO TABLE hangs
1678 * CREATE TYPE doesn't accept 'variable' as the internallength
1679 * wrong result using more than 1 aggregate in a SELECT
1684 <title><productname>Postgres95</productname> Beta 0.01</title>
1689 <firstname>Bruce</firstname>
1690 <surname>Momjian</surname>
1694 Mon May 1 19:03:10 PDT 1995
1704 <Title>Timing Results</Title>
1707 These timing results are from running the regression test with the commands
1710 % cd src/test/regress
1716 Timing under Linux 2.0.27 seems to have a roughly 5% variation from run
1717 to run, presumably due to the scheduling vagaries of multitasking systems.
1720 <Title>v6.4beta</Title>
1723 The times for this release are not directly comparable to those for previous releases
1724 since some additional regression tests have been included.
1725 In general, however, v6.4 should be slightly faster than the previous release (thanks, Bruce!).
1730 02:26 Dual Pentium Pro 180, 96MB, UW-SCSI, Linux 2.0.30, gcc 2.7.2.1 -O2 -m486
1737 The times for this release are not directly comparable to those for previous releases
1738 since some additional regression tests have been included and some obsolete tests involving
1739 time travel have been removed.
1740 In general, however, v6.3 is substantially faster than previous releases (thanks, Bruce!).
1745 02:30 Dual Pentium Pro 180, 96MB, UW-SCSI, Linux 2.0.30, gcc 2.7.2.1 -O2 -m486
1746 04:12 Dual Pentium Pro 180, 96MB, EIDE, Linux 2.0.30, gcc 2.7.2.1 -O2 -m486
1755 06:12 Pentium Pro 180, 32MB, EIDE, Linux 2.0.30, gcc 2.7.2 -O2 -m486
1756 12:06 P-100, 48MB, Linux 2.0.29, gcc
1757 39:58 Sparc IPC 32MB, Solaris 2.5, gcc 2.7.2.1 -O -g