]> granicus.if.org Git - postgresql/log
postgresql
21 years agoRelease note improvements (Neil, Tom)
Tom Lane [Sat, 23 Nov 2002 02:41:03 +0000 (02:41 +0000)]
Release note improvements (Neil, Tom)

21 years agoRemove inappropriate double-quoting in connectby() code; adjust
Tom Lane [Sat, 23 Nov 2002 01:54:09 +0000 (01:54 +0000)]
Remove inappropriate double-quoting in connectby() code; adjust
regression test to avoid using VALUE as a name.  From Joe Conway.

21 years agoRedesign internal logic of nodeLimit so that it does not need to fetch
Tom Lane [Fri, 22 Nov 2002 22:10:01 +0000 (22:10 +0000)]
Redesign internal logic of nodeLimit so that it does not need to fetch
one more row from the subplan than the COUNT would appear to require.
This costs a little more logic but a number of people have complained
about the old implementation.

21 years agoFix a dozen or so places that were passing unpredictable data strings
Tom Lane [Fri, 22 Nov 2002 16:25:32 +0000 (16:25 +0000)]
Fix a dozen or so places that were passing unpredictable data strings
as elog format strings.  Hai-Chen Tu pointed out the problem in
contrib/dbmirror, but it wasn't the only such error.

21 years agoFix elog in dbmirror to use %s.
Bruce Momjian [Fri, 22 Nov 2002 16:04:41 +0000 (16:04 +0000)]
Fix elog in dbmirror to use %s.

Hai-Chen Tu

21 years agoQuote database name properly when invoking pg_dump. Per report from
Tom Lane [Fri, 22 Nov 2002 03:09:43 +0000 (03:09 +0000)]
Quote database name properly when invoking pg_dump.  Per report from
Christopher Kings-Lynne.

21 years agoTAS code originally written for s390 (32-bit) does not work for s390x
Tom Lane [Fri, 22 Nov 2002 01:13:16 +0000 (01:13 +0000)]
TAS code originally written for s390 (32-bit) does not work for s390x
(64-bit).  Fix it.  Per report from Permaine Cheung.

21 years agoFixups for man pages
Peter Eisentraut [Thu, 21 Nov 2002 23:34:43 +0000 (23:34 +0000)]
Fixups for man pages

21 years agoRegenerate
Peter Eisentraut [Thu, 21 Nov 2002 23:33:22 +0000 (23:33 +0000)]
Regenerate

21 years agoFix breakage in new-in-7.3 timetz_zone() function: was giving random
Tom Lane [Thu, 21 Nov 2002 23:31:20 +0000 (23:31 +0000)]
Fix breakage in new-in-7.3 timetz_zone() function: was giving random
results due to doing arithmetic on uninitialized values.  Add some
documentation about the AT TIME ZONE construct.  Update some other
date/time documentation that seemed out of date for 7.3.

21 years agoWups, stuck ORDER BY on wrong query while trying to eliminate random
Tom Lane [Thu, 21 Nov 2002 22:26:02 +0000 (22:26 +0000)]
Wups, stuck ORDER BY on wrong query while trying to eliminate random
ordering of GROUP BY output due to hash-based aggregation.

21 years agoAdd:
Bruce Momjian [Thu, 21 Nov 2002 19:15:04 +0000 (19:15 +0000)]
Add:

> * Have sequence dependency track use of DEFAULT sequences, seqname.nextval

21 years agoAdd mention of kern.ipc.shm_use_phys for locking shared memory.
Bruce Momjian [Thu, 21 Nov 2002 18:19:51 +0000 (18:19 +0000)]
Add mention of kern.ipc.shm_use_phys for locking shared memory.

21 years agoCode review for superuser_reserved_connections patch. Don't try to do
Tom Lane [Thu, 21 Nov 2002 06:36:08 +0000 (06:36 +0000)]
Code review for superuser_reserved_connections patch.  Don't try to do
database access outside a transaction; revert bogus performance improvement
in SIBackendInit(); improve comments; add documentation (this part courtesy
Neil Conway).

21 years agoNetBSD versions 1.6 and (we hope) later can use the standard geometry
Tom Lane [Thu, 21 Nov 2002 04:40:26 +0000 (04:40 +0000)]
NetBSD versions 1.6 and (we hope) later can use the standard geometry
comparison file, instead of geometry-positive-zeros.  Per report from
Patrick Welche.

21 years agoDone:
Bruce Momjian [Thu, 21 Nov 2002 01:02:34 +0000 (01:02 +0000)]
Done:

> * -Add hash for evaluating GROUP BY aggregates (Tom)

21 years agoFinish implementation of hashed aggregation. Add enable_hashagg GUC
Tom Lane [Thu, 21 Nov 2002 00:42:20 +0000 (00:42 +0000)]
Finish implementation of hashed aggregation.  Add enable_hashagg GUC
parameter to allow it to be forced off for comparison purposes.
Add ORDER BY clauses to a bunch of regression test queries that will
otherwise produce randomly-ordered output in the new regime.

21 years agoUpdate ports list.
Bruce Momjian [Wed, 20 Nov 2002 22:01:57 +0000 (22:01 +0000)]
Update ports list.

21 years agoFix a dumb cut and paste error from my last commit and update some of the
Barry Lind [Wed, 20 Nov 2002 20:37:53 +0000 (20:37 +0000)]
Fix a dumb cut and paste error from my last commit and update some of the
jdbc3 metadata responses

 Modified Files:
  jdbc/org/postgresql/jdbc2/AbstractJdbc2Statement.java
  jdbc/org/postgresql/jdbc3/AbstractJdbc3DatabaseMetaData.java

21 years agoFixed bug reported by Marko Strukelj and Keith Wannamaker. Using executeBatch
Barry Lind [Wed, 20 Nov 2002 07:34:32 +0000 (07:34 +0000)]
Fixed bug reported by Marko Strukelj and Keith Wannamaker.  Using executeBatch
on a preparedStatement would reset the prepared statment causing subsequent
uses of the preparedStatement to fail (i.e. the following series of calls
would fail: addBatch() executeBatch() addBatch() executBatch()).  This is
a regression from 7.2 where this worked correctly.  The regression test has
also been modified to explicitly test for this case.

 Modified Files:
  jdbc/org/postgresql/jdbc1/AbstractJdbc1Statement.java
  jdbc/org/postgresql/jdbc2/AbstractJdbc2Statement.java
  jdbc/org/postgresql/test/jdbc2/BatchExecuteTest.java

21 years agoAdd an at-least-marginally-plausible method of estimating the number
Tom Lane [Tue, 19 Nov 2002 23:22:00 +0000 (23:22 +0000)]
Add an at-least-marginally-plausible method of estimating the number
of groups produced by GROUP BY.  This improves the accuracy of planning
estimates for grouped subselects, and is needed to check whether a
hashed aggregation plan risks memory overflow.

21 years agoAdd my name:
Bruce Momjian [Tue, 19 Nov 2002 20:35:13 +0000 (20:35 +0000)]
Add my name:

> * Allow savepoints / nested transactions [transactions] (Bruce)

21 years agoUpdate regression for ALL missing from CLUSTER.
Bruce Momjian [Tue, 19 Nov 2002 17:39:00 +0000 (17:39 +0000)]
Update regression for ALL missing from CLUSTER.

21 years agoMost of the code follows the American spelling of the word, which is
Bruce Momjian [Tue, 19 Nov 2002 17:21:40 +0000 (17:21 +0000)]
Most of the code follows the American spelling of the word, which is
"canceled", so I changed the one remaining usage of the British
spelling ("cancelled") over to the former, and updated the translation
files appropriately.

Neil Conway

21 years agoBack out NetBSD/arm32 port.
Bruce Momjian [Tue, 19 Nov 2002 17:17:48 +0000 (17:17 +0000)]
Back out NetBSD/arm32 port.

21 years agoUpdate ports list.
Bruce Momjian [Tue, 19 Nov 2002 16:40:05 +0000 (16:40 +0000)]
Update ports list.

21 years agoFix typo and markup.
Peter Eisentraut [Mon, 18 Nov 2002 18:10:59 +0000 (18:10 +0000)]
Fix typo and markup.

21 years agoUpdate SGML problem in ports.
Bruce Momjian [Mon, 18 Nov 2002 17:15:27 +0000 (17:15 +0000)]
Update SGML problem in ports.

21 years agoRemove ALL from CLUSTER ALL. Use just CLUSTER.
Bruce Momjian [Mon, 18 Nov 2002 17:12:07 +0000 (17:12 +0000)]
Remove ALL from CLUSTER ALL.  Use just CLUSTER.

21 years agoUpdate ports list.
Bruce Momjian [Mon, 18 Nov 2002 05:02:08 +0000 (05:02 +0000)]
Update ports list.

21 years agoPut back error test for DECLARE CURSOR outside a transaction block ...
Tom Lane [Mon, 18 Nov 2002 01:17:39 +0000 (01:17 +0000)]
Put back error test for DECLARE CURSOR outside a transaction block ...
but do it correctly now.

21 years agoSuppress compile warning, avoid possible problems with signed vs. unsigned
Tom Lane [Mon, 18 Nov 2002 00:40:46 +0000 (00:40 +0000)]
Suppress compile warning, avoid possible problems with signed vs. unsigned
comparisons in recently-added CheckPointWarning code.

21 years agoRe-order REINDEX options in manual for consistency. REINDEX
Bruce Momjian [Sun, 17 Nov 2002 23:43:32 +0000 (23:43 +0000)]
Re-order REINDEX options in manual for consistency.  REINDEX
DATABASE/TABLE/INDEX is the proper order.

21 years agoReduce need for palloc/pfree overhead in varstr_cmp() by using fixed-size
Tom Lane [Sun, 17 Nov 2002 23:01:30 +0000 (23:01 +0000)]
Reduce need for palloc/pfree overhead in varstr_cmp() by using fixed-size
buffers on stack for short strings.

21 years agoRevise SQL features list.
Peter Eisentraut [Sun, 17 Nov 2002 13:28:43 +0000 (13:28 +0000)]
Revise SQL features list.

21 years agoFix CLUSTER regression failure caused by new patch, from Neil Conway
Bruce Momjian [Sat, 16 Nov 2002 03:25:22 +0000 (03:25 +0000)]
Fix CLUSTER regression failure caused by new patch, from Neil Conway

21 years agoAdd relcache support for system catalogs with attnotnull columns.
Tom Lane [Fri, 15 Nov 2002 17:18:49 +0000 (17:18 +0000)]
Add relcache support for system catalogs with attnotnull columns.
The code was not making TupleConstr structs for such catalogs in
several places; with the consequence that the not-null constraint
wasn't actually enforced.  With this change,
INSERT INTO pg_proc VALUES('sdf');
generates a 'Fail to add null value' error instead of a core dump.

21 years agoSynced parser.
Michael Meskes [Fri, 15 Nov 2002 15:47:44 +0000 (15:47 +0000)]
Synced parser.

21 years agoSGML markup fixes.
Bruce Momjian [Fri, 15 Nov 2002 03:22:30 +0000 (03:22 +0000)]
SGML markup fixes.

21 years agoSGML improvements.
Bruce Momjian [Fri, 15 Nov 2002 03:11:18 +0000 (03:11 +0000)]
SGML improvements.

Neil Conway

21 years agoNew version attached. The following is implemented:
Bruce Momjian [Fri, 15 Nov 2002 03:09:39 +0000 (03:09 +0000)]
New version attached.  The following is implemented:

- CLUSTER ALL clusters all the tables that have some index with
  indisclustered set and the calling user owns.
- CLUSTER tablename clusters the named table, using the index with
  indisclustered set.  If no index has the bit set, throws elog(ERROR).
- The multi-relation version (CLUSTER ALL) uses a multitransaction
  approach, similar to what VACUUM does.

Alvaro Herrera

21 years agoClean up psql:
Bruce Momjian [Fri, 15 Nov 2002 03:07:52 +0000 (03:07 +0000)]
Clean up psql:

Add simple ALTER DATABASE, ALTER TRIGGER, CHECK POINT, CREATE
CONVERSION, CREATE DOMAIN, CREATE LANGUAGE, DEALLOCATE, DROP CONVERSION,
DROP DOMAIN, DROP LANGUAGE, EXECUTE, PREPARE

Complete CAST in CREATE CAST and DROP CAST but doesn't suggest what
should follow.

Add many more SET / SHOW variables to the list.  Taken from SHOW ALL
output.

Complete a case sensitive search to allow \dD, \dd, \dS, \ds, \h, \H to
complete properly.  But there are no matches, then try a case
insensitive search to allow case conversion.  Add all missing help
options.

\Q<tab>    -> \q
\dD<tab>   -> \dD
\dd<tab>   -> \dd
\D<tab><tab><tab> -> \d (with listing of \d? commands)
sel<tab>   -> SELECT

Rod Taylor

21 years agoOpenSSL 0.9.6g in Debian/unstable stopped working with pgcrypto. This
Bruce Momjian [Fri, 15 Nov 2002 02:54:44 +0000 (02:54 +0000)]
OpenSSL 0.9.6g in Debian/unstable stopped working with pgcrypto.  This
is pgcrypto bug as it assumed too much about inner workings of OpenSSL.

Following patch stops pgcrypto using EVP* functions for ciphers and lets
it manage ciphers itself.

This patch supports Blowfish, DES and CAST5 algorithms.

Marko Kreen

21 years agopg_dump for domain constraints.
Bruce Momjian [Fri, 15 Nov 2002 02:52:18 +0000 (02:52 +0000)]
pg_dump for domain constraints.

Rod Taylor

21 years agoAdd DOMAIN check constraints.
Bruce Momjian [Fri, 15 Nov 2002 02:50:21 +0000 (02:50 +0000)]
Add DOMAIN check constraints.

Rod Taylor

21 years agoAdd checkpoint_warning to warn of excessive checkpoints caused by too
Bruce Momjian [Fri, 15 Nov 2002 02:44:57 +0000 (02:44 +0000)]
Add checkpoint_warning to warn of excessive checkpoints caused by too
few WAL files.

21 years agoPush qual clauses containing subplans to the back of the qual list
Tom Lane [Fri, 15 Nov 2002 02:36:53 +0000 (02:36 +0000)]
Push qual clauses containing subplans to the back of the qual list
at each plan node.  Per gripe from Ross Reedstrom.

21 years agoRename hostname_lookup to log_hostname.
Bruce Momjian [Fri, 15 Nov 2002 01:57:28 +0000 (01:57 +0000)]
Rename hostname_lookup to log_hostname.

21 years agoRename show_source_port to log_source_port.
Bruce Momjian [Fri, 15 Nov 2002 01:40:20 +0000 (01:40 +0000)]
Rename show_source_port to log_source_port.

21 years agoRename show_btree_build_stats to log_btree_build_stats
Bruce Momjian [Fri, 15 Nov 2002 01:26:09 +0000 (01:26 +0000)]
Rename show_btree_build_stats to log_btree_build_stats

21 years agoRename:
Bruce Momjian [Fri, 15 Nov 2002 00:47:22 +0000 (00:47 +0000)]
Rename:

! #show_parser_stats = false
! #show_planner_stats = false
! #show_executor_stats = false
! #show_statement_stats = false

TO:

! #log_parser_stats = false
! #log_planner_stats = false
! #log_executor_stats = false
! #log_statement_stats = false

21 years agoRename server_min_messages to log_min_messages. Part of consolidation
Bruce Momjian [Thu, 14 Nov 2002 23:53:27 +0000 (23:53 +0000)]
Rename server_min_messages to log_min_messages.  Part of consolidation
of log_ settings.

21 years agoImprove warning messages from tsearch trigger function; clean up some
Tom Lane [Thu, 14 Nov 2002 20:10:00 +0000 (20:10 +0000)]
Improve warning messages from tsearch trigger function; clean up some
casting infelicities.  Allow char(n) fields to be indexed.  Per
Bjoern Metzdorf.

21 years agoFix planning bug introduced in recent code reorganization for hashed
Tom Lane [Thu, 14 Nov 2002 19:00:36 +0000 (19:00 +0000)]
Fix planning bug introduced in recent code reorganization for hashed
aggregates: tuple_fraction has to be adjusted before passing it to
compare_fractional_path_costs().

21 years agoUpdate ports list.
Bruce Momjian [Thu, 14 Nov 2002 18:39:43 +0000 (18:39 +0000)]
Update ports list.

21 years agoFix Magnus's name.
Bruce Momjian [Thu, 14 Nov 2002 14:53:29 +0000 (14:53 +0000)]
Fix Magnus's name.

21 years agoFix version number for FreeBSD/alpha on ports list.
Bruce Momjian [Thu, 14 Nov 2002 14:51:38 +0000 (14:51 +0000)]
Fix version number for FreeBSD/alpha on ports list.

21 years agoUpdate ports list.
Bruce Momjian [Thu, 14 Nov 2002 14:23:59 +0000 (14:23 +0000)]
Update ports list.

21 years agoApplied patch submitted by Mats Lofkvist fixing serious threading problem introduced...
Barry Lind [Thu, 14 Nov 2002 05:35:45 +0000 (05:35 +0000)]
Applied patch submitted by Mats Lofkvist fixing serious threading problem introduced in beta3.
Fixed bug with using setNull()(or setXXX(x, null))  and serverside prepare statements.
Improved error message when using a connection object that has already been closed.

 Modified Files:
  jdbc/org/postgresql/errors.properties
  jdbc/org/postgresql/core/Encoding.java
  jdbc/org/postgresql/core/QueryExecutor.java
  jdbc/org/postgresql/jdbc1/AbstractJdbc1Connection.java
  jdbc/org/postgresql/jdbc1/AbstractJdbc1Statement.java

21 years agoUpdate ports list.
Bruce Momjian [Thu, 14 Nov 2002 04:59:15 +0000 (04:59 +0000)]
Update ports list.

21 years agoThere was an error in the gcc portion of the patch I sent you. Here is
Bruce Momjian [Thu, 14 Nov 2002 04:07:27 +0000 (04:07 +0000)]
There was an error in the gcc portion of the patch I sent you.  Here is
the  fix.

Billy G. Allie

21 years agoUpdate ports list.
Bruce Momjian [Thu, 14 Nov 2002 03:59:46 +0000 (03:59 +0000)]
Update ports list.

21 years agoAllow +1300 as a numeric timezone specifier; we already accept FJST as meaning +1300.
Tom Lane [Wed, 13 Nov 2002 17:24:05 +0000 (17:24 +0000)]
Allow +1300 as a numeric timezone specifier; we already accept FJST as meaning +1300.

21 years agoUse cat instead of echo or awk, to avoid portability problems with backslashes.
Tom Lane [Wed, 13 Nov 2002 16:40:23 +0000 (16:40 +0000)]
Use cat instead of echo or awk, to avoid portability problems with backslashes.

21 years agoUpdate xact.c comments for clarity.
Bruce Momjian [Wed, 13 Nov 2002 03:12:05 +0000 (03:12 +0000)]
Update xact.c comments for clarity.

21 years agoAdd:
Bruce Momjian [Wed, 13 Nov 2002 02:10:26 +0000 (02:10 +0000)]
Add:

> * Fix SELECT ... INTO and CREATE TABLE AS to have appopriate OID column
>  o -MOVE 0 should not move to end of cursor (Bruce)

21 years agoMake MOVE/FETCH 0 actually move/fetch 0. Add MOVE LAST to move to end
Bruce Momjian [Wed, 13 Nov 2002 00:44:09 +0000 (00:44 +0000)]
Make MOVE/FETCH 0 actually move/fetch 0.  Add MOVE LAST to move to end
of cursor.

21 years agoAdd new palloc0 call as merge of palloc and MemSet(0).
Bruce Momjian [Wed, 13 Nov 2002 00:39:48 +0000 (00:39 +0000)]
Add new palloc0 call as merge of palloc and MemSet(0).

21 years agoSplit MemSet into three parts to constant comparisons can be optimized
Bruce Momjian [Wed, 13 Nov 2002 00:37:06 +0000 (00:37 +0000)]
Split MemSet into three parts to constant comparisons can be optimized
away by the compiler;  used by palloc0.

21 years agoAdd AIX as supported platform.
Bruce Momjian [Tue, 12 Nov 2002 20:19:55 +0000 (20:19 +0000)]
Add AIX as supported platform.

21 years agoUpdate AIX resultmap for horology tests, per Andreas.
Tom Lane [Tue, 12 Nov 2002 20:02:32 +0000 (20:02 +0000)]
Update AIX resultmap for horology tests, per Andreas.

21 years agoImprove comment: add note that grotty special case in mdread() is
Tom Lane [Tue, 12 Nov 2002 15:26:30 +0000 (15:26 +0000)]
Improve comment: add note that grotty special case in mdread() is
required by hash index implementation.

22 years agoFix broken GB18030 <--> UTF-8 conversion map
Tatsuo Ishii [Tue, 12 Nov 2002 11:33:40 +0000 (11:33 +0000)]
Fix broken GB18030 <--> UTF-8 conversion map

22 years agoRemove NO_MKTIME_BEFORE_1970. I had speculated that it was not needed
Tom Lane [Tue, 12 Nov 2002 00:39:08 +0000 (00:39 +0000)]
Remove NO_MKTIME_BEFORE_1970.  I had speculated that it was not needed
anymore given the mktime() workaround now done in DetermineLocalTimeZone.
This has now been confirmed by Robert Bruccoleri for Irix, and I'm going
to extrapolate to AIX as well.

22 years agoCode review for ON COMMIT patch. Make the actual on-commit action happen
Tom Lane [Mon, 11 Nov 2002 22:19:25 +0000 (22:19 +0000)]
Code review for ON COMMIT patch.  Make the actual on-commit action happen
before commit, not after :-( --- the original coding is not only unsafe
if an error occurs while it's processing, but it generates an invalid
sequence of WAL entries.  Resurrect 7.2 logic for deleting items when
no longer needed.  Use an enum instead of random macros.  Editorialize
on names used for routines and constants.  Teach backend/nodes routines
about new field in CreateTable struct.  Add a regression test.

22 years agoMerge documentation updates from 7.3 branch.
Peter Eisentraut [Mon, 11 Nov 2002 20:14:04 +0000 (20:14 +0000)]
Merge documentation updates from 7.3 branch.

22 years agoMerge release note updates from 7.3 branch.
Peter Eisentraut [Mon, 11 Nov 2002 20:06:00 +0000 (20:06 +0000)]
Merge release note updates from 7.3 branch.

22 years agoRevise platform FAQs.
Peter Eisentraut [Mon, 11 Nov 2002 20:04:05 +0000 (20:04 +0000)]
Revise platform FAQs.

22 years agoImprove wording.
Peter Eisentraut [Mon, 11 Nov 2002 20:03:40 +0000 (20:03 +0000)]
Improve wording.

22 years agoUpdate jdbc doc to reflect correct name of the PGConnection class.
Barry Lind [Mon, 11 Nov 2002 07:31:28 +0000 (07:31 +0000)]
Update jdbc doc to reflect correct name of the PGConnection class.

22 years agoFixes bug where join to pg_description was incorrect. Also modifies the
Barry Lind [Mon, 11 Nov 2002 07:11:12 +0000 (07:11 +0000)]
Fixes bug where join to pg_description was incorrect.  Also modifies the
regression test to test for this case.  Patch submitted by Kris Jurka.

 Modified Files:
  jdbc/org/postgresql/jdbc1/AbstractJdbc1DatabaseMetaData.java
  jdbc/org/postgresql/test/jdbc2/DatabaseMetaDataTest.java

22 years agoAdd cast to suppress compile warning on Alphas.
Tom Lane [Mon, 11 Nov 2002 03:33:38 +0000 (03:33 +0000)]
Add cast to suppress compile warning on Alphas.

22 years agoBack out use of palloc0 in place if palloc/MemSet. Seems constant len
Bruce Momjian [Mon, 11 Nov 2002 03:02:20 +0000 (03:02 +0000)]
Back out use of palloc0 in place if palloc/MemSet.  Seems constant len
to MemSet is a performance boost.

22 years agoUpdate russian FAQ, from Viktor Vislobokov.
Bruce Momjian [Sun, 10 Nov 2002 19:48:41 +0000 (19:48 +0000)]
Update russian FAQ, from Viktor Vislobokov.

22 years agoMerge palloc()/MemSet(0) calls into a single palloc0() call.
Bruce Momjian [Sun, 10 Nov 2002 07:25:14 +0000 (07:25 +0000)]
Merge palloc()/MemSet(0) calls into a single palloc0() call.

22 years agoAdd palloc0 function to inline MemSet for newNode call.
Bruce Momjian [Sun, 10 Nov 2002 02:17:25 +0000 (02:17 +0000)]
Add palloc0 function to inline MemSet for newNode call.

22 years agoThis trivial patch removes 2 unused defines from pg_config.h.in:
Bruce Momjian [Sun, 10 Nov 2002 00:38:21 +0000 (00:38 +0000)]
This trivial patch removes 2 unused defines from pg_config.h.in:
FASTBUILD and DATEDEBUG weren't referenced by any other code.

Neil Conway

22 years agoI am attaching a patch file that will replace the previous one. It only
Bruce Momjian [Sun, 10 Nov 2002 00:37:28 +0000 (00:37 +0000)]
I am attaching a patch file that will replace the previous one.  It only
contains the patches to Makefile.global.in and Makefile.unixware.  The
Makefile.unixware patch has been updated to include the contents of
LD_LIBRARY_PATH, if present, to the -rpath (-R) option.  This change
will  simplify configuring and building PostgreSQL on systems that
support  LD_LIBRARY_PATH.  You can set LD_LIBRARY_PATH to include all
the directorys  you want to have searched for additional libraries, run
configure, then run  make.  The paths in LD_LIBRARY_PATH will then be
embedded in the executables  via the -rpath (-R) option to the linker,
and so will not require  LD_LIBRARY_PATH in order to run.

Billy G. Allie

22 years agoThis patch makes a minor cleanup to the implementation of PERFORM in
Bruce Momjian [Sun, 10 Nov 2002 00:35:58 +0000 (00:35 +0000)]
This patch makes a minor cleanup to the implementation of PERFORM in
PL/PgSQL. Previously, it had been bundled together with the assign
statement implementation, for some reason that wasn't clear to me
(they certainly don't share any code with one another). So I separated
them and made PERFORM a statement like any other. No changes in
functionality.

Along the way, I added some regression tests for PERFORM, added a
bunch more SGML tags to the PL/PgSQL docs, and removed an obsolete
comment relating to the implementation of RETURN NEXT.

Neil Conway

22 years ago> > I'll re-check that with the ppc architecture guy here.
Bruce Momjian [Sun, 10 Nov 2002 00:33:43 +0000 (00:33 +0000)]
> > I'll re-check that with the ppc architecture guy here.
>
> ... he is now about to write an inlined version that can go into
> s_lock.h . I'll send the new patch later on...

OK, here it comes:

An inlined version of tas(), that works for both, powerpc and
powerpc64. The patch is against 7.3b5 and passes the test suite on
both architectures.

Reinhard Max

22 years agoAdd documentation for array_upper/lower().
Bruce Momjian [Sun, 10 Nov 2002 00:32:16 +0000 (00:32 +0000)]
Add documentation for array_upper/lower().

22 years agoI just discovered, that there is missing a const when passing a buffer
Bruce Momjian [Sun, 10 Nov 2002 00:14:22 +0000 (00:14 +0000)]
I just discovered, that there is missing a const when passing a buffer
to  PQescapeBytea and PQunescapeBytea. I fixed it and tried to create a
usable  diff (I'm not so familar to diff).

Tommi M?kitalo

22 years agoTweak CREATE SEQUENCE grammar to be more SQL1999 standards compliant.
Bruce Momjian [Sun, 10 Nov 2002 00:10:20 +0000 (00:10 +0000)]
Tweak CREATE SEQUENCE grammar to be more SQL1999 standards compliant.

Neil Conway

22 years agoAdd code to handle [ON COMMIT { PRESERVE ROWS | DELETE ROWS | DROP }]
Bruce Momjian [Sat, 9 Nov 2002 23:56:39 +0000 (23:56 +0000)]
Add code to handle [ON COMMIT { PRESERVE ROWS | DELETE ROWS | DROP }]
for temp tables.

Gavin Sherry

22 years agoUpdate discussion of floating-point problems in regression tests.
Tom Lane [Fri, 8 Nov 2002 20:26:12 +0000 (20:26 +0000)]
Update discussion of floating-point problems in regression tests.

22 years agoThis patch removes a bunch of superfluous #include directives: if
Bruce Momjian [Fri, 8 Nov 2002 20:23:57 +0000 (20:23 +0000)]
This patch removes a bunch of superfluous #include directives: if
postgres.h or c.h includes a system header (such as stdio.h or
stdlib.h), there's no need to specifically include it in any of the .c
files in the backend.

Neil Conway

22 years agoClean up format of SQL.
Bruce Momjian [Fri, 8 Nov 2002 20:22:12 +0000 (20:22 +0000)]
Clean up format of SQL.

22 years agoThe attached patch defines functions for getting distances between
Bruce Momjian [Fri, 8 Nov 2002 20:20:22 +0000 (20:20 +0000)]
The attached patch defines functions for getting distances between
points on the surface of the earth and locating points within a
specified distance using an index based on the contrib/cube package. The
new functions are all of language type sql. A couple of bugs in the old
earthdistance function based on the point datatype are fixed. A
regression test has been added for both sets of functions. The README
file has been updated to include documentation on the new stuff. There
are comments about how this package is also useful for Astronomers.

Bruno Wolff III

22 years agoRevise geometry regression testing to eliminate most cross-platform
Tom Lane [Fri, 8 Nov 2002 20:09:55 +0000 (20:09 +0000)]
Revise geometry regression testing to eliminate most cross-platform
variation.  To do this, set extra_float_digits to -3 in the geometry
test, and tweak the CIRCLE_TBL dataset to avoid values that suffer
from severe cancellation error (eg, circles that just touch an axis).
We still need two geometry 'expected' files to account for the
difference between platforms that display minus zero as '-0' and those
that just say '0', but with luck that's all we'll need.

22 years agoHere is a patch that does just that, while maintaining the
Bruce Momjian [Fri, 8 Nov 2002 19:12:21 +0000 (19:12 +0000)]
Here is a patch that does just that, while maintaining the
"traditional" behavior, so the change should be transparent. Use the
command "\pset pager always" to turn it on. Anything else does the
normal toggle between "on" and "off"

Greg Sabino Mullane