OSDN Git Service

pg-rex/syncrep.git
17 years agoUpdate release notes to be current as of CVS today.
Bruce Momjian [Fri, 22 Sep 2006 03:04:23 +0000 (03:04 +0000)]
Update release notes to be current as of CVS today.

17 years agoUpdate major release note items.
Bruce Momjian [Fri, 22 Sep 2006 01:52:27 +0000 (01:52 +0000)]
Update major release note items.

17 years agoFirst try at a major features list for 8.2.
Bruce Momjian [Fri, 22 Sep 2006 01:39:33 +0000 (01:39 +0000)]
First try at a major features list for 8.2.

17 years agoFix free space map to correctly track the total amount of FSM space needed
Tom Lane [Thu, 21 Sep 2006 20:31:22 +0000 (20:31 +0000)]
Fix free space map to correctly track the total amount of FSM space needed
even when a single relation requires more than max_fsm_pages pages.  Also,
make VACUUM emit a warning in this case, since it likely means that VACUUM
FULL or other drastic corrective measure is needed.  Per reports from Jeff
Frost and others of unexpected changes in the claimed max_fsm_pages need.

17 years agoAdd comments about STORAGE option for GIN
Teodor Sigaev [Thu, 21 Sep 2006 15:09:38 +0000 (15:09 +0000)]
Add comments about STORAGE option for GIN

17 years agoFix table's caption
Teodor Sigaev [Thu, 21 Sep 2006 15:03:53 +0000 (15:03 +0000)]
Fix table's caption

17 years agoUpdated several parts in particular variable handling. This is work that was part...
Michael Meskes [Thu, 21 Sep 2006 09:10:27 +0000 (09:10 +0000)]
Updated several parts in particular variable handling. This is work that was part of SoC.

17 years agoWording changes for continuous archving, per Simon.
Bruce Momjian [Thu, 21 Sep 2006 03:12:58 +0000 (03:12 +0000)]
Wording changes for continuous archving, per Simon.

17 years agoAdd documentation for new in-core advisory lock functions. Merlin Moncure
Tom Lane [Wed, 20 Sep 2006 23:43:22 +0000 (23:43 +0000)]
Add documentation for new in-core advisory lock functions.  Merlin Moncure

17 years agoAdd URL for domain casting:
Bruce Momjian [Wed, 20 Sep 2006 23:01:55 +0000 (23:01 +0000)]
Add URL for domain casting:

>   http://archives.postgresql.org/pgsql-hackers/2006-09/msg01681.php

17 years agoUpdate release notes to use scaron, rather than numeric code for entity.
Bruce Momjian [Wed, 20 Sep 2006 22:48:47 +0000 (22:48 +0000)]
Update release notes to use scaron, rather than numeric code for entity.

17 years agoFix bogus markup.
Tom Lane [Wed, 20 Sep 2006 21:30:20 +0000 (21:30 +0000)]
Fix bogus markup.

17 years agoAdd AIX shared library support item to release notes.
Bruce Momjian [Wed, 20 Sep 2006 19:55:42 +0000 (19:55 +0000)]
Add AIX shared library support item to release notes.

17 years agoChange patternsel (LIKE/regex selectivity estimation) so that if there
Tom Lane [Wed, 20 Sep 2006 19:50:21 +0000 (19:50 +0000)]
Change patternsel (LIKE/regex selectivity estimation) so that if there
is a large enough histogram, it will use the number of matches in the
histogram to derive a selectivity estimate, rather than the admittedly
pretty bogus heuristics involving examining the pattern contents.  I set
'large enough' at 100, but perhaps we should change that later.  Also
apply the same technique in contrib/ltree's <@ and @> estimator.  Per
discussion with Stefan Kaltenbrunner and Matteo Beccati.

17 years agoRelease note wording improvement.
Bruce Momjian [Wed, 20 Sep 2006 11:54:23 +0000 (11:54 +0000)]
Release note wording improvement.

Alvaro.

17 years agoAdd shared linking information to AIX FAQ.
Bruce Momjian [Wed, 20 Sep 2006 02:10:11 +0000 (02:10 +0000)]
Add shared linking information to AIX FAQ.

Laurenz Albe

17 years agoImprovements to the partitioning documentation. Original patch from Greg
Neil Conway [Wed, 20 Sep 2006 01:20:38 +0000 (01:20 +0000)]
Improvements to the partitioning documentation. Original patch from Greg
Stark, additional fixes and editorialization by Neil Conway. Greg's patch
description:

    Some more doc patches for partitioned tables. In particular replace
    the caveat that INCLUDING CONSTRAINTS doesn't exist and replace it
    with documentation of, well, INCLUDING CONSTRAINTS.

    Also, there was an instance of "LIKE WITH DEFAULTS" which is
    actually spelled "LIKE INCLUDING DEFAULTS".

17 years agoImprove usage of effective_cache_size parameter by assuming that all the
Tom Lane [Tue, 19 Sep 2006 22:49:53 +0000 (22:49 +0000)]
Improve usage of effective_cache_size parameter by assuming that all the
tables in the query compete for cache space, not just the one we are
currently costing an indexscan for.  This seems more realistic, and it
definitely will help in examples recently exhibited by Stefan
Kaltenbrunner.  To get the total size of all the tables involved, we must
tweak the handling of 'append relations' a bit --- formerly we looked up
information about the child tables on-the-fly during set_append_rel_pathlist,
but it needs to be done before we start doing any cost estimation, so
push it into the add_base_rels_to_query scan.

17 years agoDo a round of copy-editing for the release notes: fix some typos and
Neil Conway [Tue, 19 Sep 2006 19:54:05 +0000 (19:54 +0000)]
Do a round of copy-editing for the release notes: fix some typos and
grammatical errors, improve the description of some new features.

17 years agoDashed items are now for 8.3:
Bruce Momjian [Tue, 19 Sep 2006 19:20:21 +0000 (19:20 +0000)]
Dashed items are now for 8.3:

< #A hyphen, "-", marks changes that will appear in the upcoming 8.2 release.#
> #A hyphen, "-", marks changes that will appear in the upcoming 8.3 release.#

17 years agoRemove completed TODO items:
Bruce Momjian [Tue, 19 Sep 2006 19:20:00 +0000 (19:20 +0000)]
Remove completed TODO items:

< * -Make postmater and postgres options distinct so the postmaster -o
<   option is no longer needed
< * -Allow pooled connections to list all prepared statements
<
<   This would allow an application inheriting a pooled connection to know
<   the statements prepared in the current session.
<
< * -Re-enable the GUC full_page_writes in 8.2 when reliability issues have
<   been addressed
<  o -Add "include file" functionality in postgresql.conf
<  o -Allow per-database permissions to be set via GRANT
<
<    Allow database connection checks based on GRANT rules in
<    addition to the existing access checks in pg_hba.conf.
<
<  o -Issue a warning if a change-on-restart-only postgresql.conf value
>  o Issue a warning if a change-on-restart-only postgresql.conf value
<    o -Automatically force archiving of partially-filled WAL files when
<      pg_stop_backup() is called or the server is stopped
<    o -Add reporting of the current WAL file and offset, perhaps as
<      part of partial log file archiving
< * -Allow server logs to be remotely read and removed using SQL commands
< * -Allow protocol-level BIND parameter values to be logged
< * -Zero umasked bits in conversion from INET cast to CIDR
< * -Prevent INET cast to CIDR from dropping netmask, SELECT '1.1.1.1'::inet::cidr
< * -Allow INET + INT8 to increment the host part of the address or
<   throw an error on overflow
< * -Add 'tid != tid ' operator for use in corruption recovery
<  o -Allow customization of the known set of TZ names (generalize the
<    present australian_timezones hack)
<  o -Allow timezone names in SQL strings, '2006-05-24 21:11
<     Americas/New_York'::timestamptz
<  o -Add support for day-time syntax, INTERVAL '1 2:03:04' DAY TO
<    SECOND
<  o -Allow NULLs in arrays
< * -Add transaction_timestamp(), statement_timestamp(), clock_timestamp()
<   functionality
<
<   Current CURRENT_TIMESTAMP returns the start time of the current
<   transaction, and gettimeofday() returns the wallclock time. This will
<   make time reporting more consistent and will allow reporting of
<   the statement start time.
<
< * -Allow to_char() to print localized month names
< * -Add sleep() function, remove from regress.c
< * -Allow user-defined functions retuning a domain value to enforce domain
<   constraints
< * -Allow TRUNCATE ... CASCADE/RESTRICT
<
<   This is like DELETE CASCADE, but truncates.
<
< * -Add COMMENT ON for all cluster global objects (roles, databases
<   and tablespaces)
< * -Make row-wise comparisons work per SQL spec
<
<   Right now, '(a, b) < (1, 2)' is processed as 'a < 1 and b < 2', but
<   the SQL standard requires it to be processed as a column-by-column
<   comparison, so the proper comparison is '(a < 1) OR (a = 1 AND b < 2)'.
<
< * -Enable escape_string_warning and standard_conforming_strings
> * Enable standard_conforming_strings
<  o -Add ON COMMIT capability to CREATE TABLE AS ... SELECT
<  o -Allow an alias to be provided for the target table in
<    UPDATE/DELETE (Neil)
<  o -Allow UPDATE tab SET ROW (col, ...) = (val, ...) for updating
<    multiple columns
<  o -Add ALTER TABLE tab INHERIT / NO INHERIT  parent
<  o -Have COPY return the number of rows loaded/unloaded?
<  o -Allow COPY (SELECT ...) TO 'filename'
<
<  o -Allow pooled connections to list all open WITH HOLD cursors
<
<    Because WITH HOLD cursors exist outside transactions, this allows
<    them to be listed so they can be closed.
<
<  o -Allow INSERT INTO tab (col1, ..) VALUES (val1, ..), (val2, ..)
<  o -Allow INSERT/UPDATE ... RETURNING new.col or old.col
<  o -Allow PL/python to return composite types and result sets
<
< * -Have initdb set the input DateStyle (MDY or DMY) based on locale
<  o -Improve psql's handling of multi-line statements
<
<    Currently, while \e saves a single statement as one entry, interactive
<    statements are saved one line at a time.  Ideally all statements
<    would be saved like \e does.
<
<  o -Allow multi-line column values to align in the proper columns
<
<    If the second output column value is 'a\nb', the 'b' should appear
<    in the second display column, rather than the first column as it
<    does now.
<
<  o -Display IN, INOUT, and OUT parameters in \df
<  o -Allow pg_dump to use multiple -t and -n switches, exclusion
<     ability, and regular expression object matching
<  o -Update pg_dump and psql to use the new COPY libpq API (Christopher)
<  o -Add a function to support Parse/DescribeStatement capability
< * -Add fillfactor to control reserved free space during index creation
< * -Add system view to show free space map contents
< * -Allow installing to directories containing spaces
<
<   This is possible if proper quoting is added to the makefiles for the
<   install targets.  Because PostgreSQL supports relocatable installs, it
<   is already possible to install into a directory that doesn't contain
<   spaces and then copy the install to a directory with spaces.
<
< * -%Clean up compiler warnings (especially with gcc version 4)
< * -Remove BeOS and QNX-specific code
<  o -Port contrib/xml2

17 years agoDocument how to use psql's --single-transaction option to rollback restoring
Neil Conway [Tue, 19 Sep 2006 19:04:51 +0000 (19:04 +0000)]
Document how to use psql's --single-transaction option to rollback restoring
an SQL dump if an error occurs. Along the way, make some improvements and
copy-edits to the surrounding text. Patch from Simon Riggs, additional
fixes by Neil Conway.

17 years agoFix broken markup.
Tom Lane [Tue, 19 Sep 2006 17:22:42 +0000 (17:22 +0000)]
Fix broken markup.

17 years agoFix shared library creation to work properly on AIX. Albe Laurenz
Tom Lane [Tue, 19 Sep 2006 15:36:08 +0000 (15:36 +0000)]
Fix shared library creation to work properly on AIX.  Albe Laurenz

17 years agoMinor additions and typo fixes for the backup documentation. Patch from
Neil Conway [Tue, 19 Sep 2006 15:18:41 +0000 (15:18 +0000)]
Minor additions and typo fixes for the backup documentation. Patch from
Simon Riggs, minor editorialization by Neil Conway.

17 years agoMore 8.2 release item reordering.
Bruce Momjian [Tue, 19 Sep 2006 02:25:13 +0000 (02:25 +0000)]
More 8.2 release item reordering.

17 years agoReorder 8.2 release note items more logically.
Bruce Momjian [Tue, 19 Sep 2006 02:08:30 +0000 (02:08 +0000)]
Reorder 8.2 release note items more logically.

17 years agoUpdate migration changes, more cleanups.
Bruce Momjian [Tue, 19 Sep 2006 00:57:07 +0000 (00:57 +0000)]
Update migration changes, more cleanups.

17 years agoProperly mention in the release notes that fillfactor controls both heap
Bruce Momjian [Mon, 18 Sep 2006 23:19:50 +0000 (23:19 +0000)]
Properly mention in the release notes that fillfactor controls both heap
and indexes.  Other cleanups.

17 years agoWrap long lines in 8.2 release notes.
Bruce Momjian [Mon, 18 Sep 2006 23:01:43 +0000 (23:01 +0000)]
Wrap long lines in 8.2 release notes.

17 years agoUpdates from Jim Nasby.
Bruce Momjian [Mon, 18 Sep 2006 22:43:16 +0000 (22:43 +0000)]
Updates from Jim Nasby.

17 years agoAdd built-in userlock manipulation functions to replace the former
Tom Lane [Mon, 18 Sep 2006 22:40:40 +0000 (22:40 +0000)]
Add built-in userlock manipulation functions to replace the former
contrib functionality.  Along the way, remove the USER_LOCKS configuration
symbol, since it no longer makes any sense to try to compile that out.
No user documentation yet ... mmoncure has promised to write some.
Thanks to Abhijit Menon-Sen for creating a first draft to work from.

17 years agoMove 8.2 release documentation into SGML.
Bruce Momjian [Mon, 18 Sep 2006 21:45:10 +0000 (21:45 +0000)]
Move 8.2 release documentation into SGML.

17 years agoMake the order of the CASCADE and RESTRICT keywords in the DROP OWNED
Neil Conway [Mon, 18 Sep 2006 21:19:29 +0000 (21:19 +0000)]
Make the order of the CASCADE and RESTRICT keywords in the DROP OWNED
syntax summary consistent with the other SQL reference pages. Patch
from Euler Taveira de Oliveira.

17 years agoDocumentation for VALUES lists. Joe Conway and Tom Lane
Tom Lane [Mon, 18 Sep 2006 19:54:01 +0000 (19:54 +0000)]
Documentation for VALUES lists.  Joe Conway and Tom Lane

17 years agoAdd URL for commenting postgresql.conf:
Bruce Momjian [Mon, 18 Sep 2006 16:24:07 +0000 (16:24 +0000)]
Add URL for commenting postgresql.conf:

>    http://archives.postgresql.org/pgsql-hackers/2006-09/msg01481.php

17 years agoAdd URL for UUID:
Bruce Momjian [Mon, 18 Sep 2006 16:22:32 +0000 (16:22 +0000)]
Add URL for UUID:

>
>   http://archives.postgresql.org/pgsql-patches/2006-09/msg00209.php
>

17 years agoFix problems with column name list of CREATE TABLE AS being applied to
Tom Lane [Mon, 18 Sep 2006 16:04:04 +0000 (16:04 +0000)]
Fix problems with column name list of CREATE TABLE AS being applied to
the input query's target list too soon, causing it to affect processing
of ORDER BY in the input query.

17 years agoAdd item:
Bruce Momjian [Mon, 18 Sep 2006 15:52:23 +0000 (15:52 +0000)]
Add item:

>
> * Set client encoding based on the client operating system encoding
>
>   Currently client_encoding is set in postgresql.conf, which
>   defaults to the server encoding.

17 years agoAdd URL to UPDATE unique index case x=x+1:
Bruce Momjian [Mon, 18 Sep 2006 14:56:19 +0000 (14:56 +0000)]
Add URL to UPDATE unique index case x=x+1:

<
>    http://archives.postgresql.org/pgsql-hackers/2006-09/msg01458.php

17 years agoImprove wordings by David Fuhry <dfuhry@cs.kent.edu>
Teodor Sigaev [Mon, 18 Sep 2006 12:11:36 +0000 (12:11 +0000)]
Improve wordings by David Fuhry <dfuhry@cs.kent.edu>

17 years agoFix CREATE TABLE ... AS VALUES ... to work rather than Assert'ing;
Tom Lane [Mon, 18 Sep 2006 00:52:14 +0000 (00:52 +0000)]
Fix CREATE TABLE ... AS VALUES ... to work rather than Assert'ing;
oversight in original implementation of VALUES.  Also fix an oversight
in recent addition of options to CREATE TABLE AS: they weren't getting
propagated if the query was a set-operation such as UNION.

17 years agoChange ANALYZE to take ShareUpdateExclusiveLock not AccessShareLock on
Tom Lane [Sun, 17 Sep 2006 22:50:31 +0000 (22:50 +0000)]
Change ANALYZE to take ShareUpdateExclusiveLock not AccessShareLock on
the table being analyzed.  This prevents two ANALYZEs from running
concurrently on the same table and possibly suffering concurrent-update
failures while trying to store their results into pg_statistic.  The
downside is that a database-wide ANALYZE executed within a transaction
block will hold ShareUpdateExclusiveLock on many tables simultaneously,
which could lead to concurrency issues or even deadlock against another
such ANALYZE.  However, this seems a corner case of less importance
than getting unexpected errors from a foreground ANALYZE when autovacuum
elects to analyze the same table concurrently.  Per discussion.

17 years agoMarginal cleanup in arrangements for ensuring StrategyHintVacuum is cleared
Tom Lane [Sun, 17 Sep 2006 22:16:22 +0000 (22:16 +0000)]
Marginal cleanup in arrangements for ensuring StrategyHintVacuum is cleared
after an error during VACUUM.  We have a PG_TRY block anyway around the only
call sites, so just reset it in the CATCH clause instead of having
AtEOXact_Buffers blindly do it during xact end.  I think the old code was
actively wrong for the case of a failure during ANALYZE inside a
subtransaction --- the flag wouldn't get cleared until main transaction end.
Probably not worth back-patching though.

17 years agoUpdate TODO for short header versions:
Bruce Momjian [Sat, 16 Sep 2006 22:06:54 +0000 (22:06 +0000)]
Update TODO for short header versions:

<  o Reorder physical storage order to reduce padding?
<
<    This involves having the user-specified order of columns
<    be different from the physical order.  SELECT * would
<    need to reorder the physical values to match the
<    user-specified ordering.
<
<  o Store disk pages with no alignment/padding?
<
<    This necessitates adding CPU-required padding when moving
<    rows from disk to memory.
<
<    One idea is to store the header in network byte order (high bits
<    first), and read the high bits to determine the header length.
<    http://archives.postgresql.org/pgsql-hackers/2006-09/msg00848.php
>    One idea is to create zero-or-one-byte-header versions
>    of varlena data types.  In involves setting the high-bit and
>    0-127 length in the single-byte header, or clear the high bit
>    and store the 7-bit ASCII value in the rest of the byte.
>    The small-header versions have no alignment requirements.
>    http://archives.postgresql.org/pgsql-hackers/2006-09/msg01372.php

17 years agoRename the recently-added pg_timezonenames view to pg_timezone_abbrevs,
Tom Lane [Sat, 16 Sep 2006 20:14:34 +0000 (20:14 +0000)]
Rename the recently-added pg_timezonenames view to pg_timezone_abbrevs,
and create a new view pg_timezone_names that provides information about
the zones known in the 'zic' database.  Magnus Hagander, with some
additional work by Tom Lane.

17 years agoUpdate timezone data files to release 2006k of the zic database.
Tom Lane [Sat, 16 Sep 2006 17:21:03 +0000 (17:21 +0000)]
Update timezone data files to release 2006k of the zic database.

17 years agoRename xml_valid() to xml_is_well_formed(), but provide a temporary
Tom Lane [Sat, 16 Sep 2006 16:18:11 +0000 (16:18 +0000)]
Rename xml_valid() to xml_is_well_formed(), but provide a temporary
alias with the old name for backwards compatibility.  Per discussion,
the old name is actively wrong because validity and well-formedness
have different meanings in XML.

17 years agoDefine errcode as __msvc_errcode not __vc_errcode for MSVC builds,
Tom Lane [Sat, 16 Sep 2006 13:35:49 +0000 (13:35 +0000)]
Define errcode as __msvc_errcode not __vc_errcode for MSVC builds,
per Magnus.

17 years agoFix some more uses of str[n]casecmp that should be pg_str[n]casecmp.
Tom Lane [Sat, 16 Sep 2006 13:31:40 +0000 (13:31 +0000)]
Fix some more uses of str[n]casecmp that should be pg_str[n]casecmp.

17 years agoRemove emacs info from footer of SGML files.
Bruce Momjian [Sat, 16 Sep 2006 00:30:20 +0000 (00:30 +0000)]
Remove emacs info from footer of SGML files.

17 years agoDocument Warm Standby for High Availability
Bruce Momjian [Fri, 15 Sep 2006 22:02:21 +0000 (22:02 +0000)]
Document Warm Standby for High Availability

Includes sample standby script.

Simon Riggs

17 years agoDocument WAL recovery now uses checkpoints.
Bruce Momjian [Fri, 15 Sep 2006 21:55:07 +0000 (21:55 +0000)]
Document WAL recovery now uses checkpoints.

Simon Riggs

17 years agoSeveral fixes for MSVC build scripts, from Magnus.
Tom Lane [Fri, 15 Sep 2006 21:42:02 +0000 (21:42 +0000)]
Several fixes for MSVC build scripts, from Magnus.

17 years agoFix some uses of str[n]casecmp that should be pg_str[n]casecmp.
Tom Lane [Fri, 15 Sep 2006 21:35:30 +0000 (21:35 +0000)]
Fix some uses of str[n]casecmp that should be pg_str[n]casecmp.

17 years agoFix some uses of str[n]casecmp that should be pg_str[n]casecmp,
Tom Lane [Fri, 15 Sep 2006 21:34:23 +0000 (21:34 +0000)]
Fix some uses of str[n]casecmp that should be pg_str[n]casecmp,
per Magnus.

17 years agoRemove WINLDAPAPI decoration from ldap_start_tls_sA typedef, per Magnus.
Tom Lane [Fri, 15 Sep 2006 21:28:08 +0000 (21:28 +0000)]
Remove WINLDAPAPI decoration from ldap_start_tls_sA typedef, per Magnus.

17 years agoImprove confusing comment for HeapTupleSatisfiesNow, per gripe from Greg Stark.
Tom Lane [Fri, 15 Sep 2006 16:39:32 +0000 (16:39 +0000)]
Improve confusing comment for HeapTupleSatisfiesNow, per gripe from Greg Stark.

17 years agoMake postgresql.conf.sample match the initdb defaults. This fixes
Bruce Momjian [Thu, 14 Sep 2006 23:21:47 +0000 (23:21 +0000)]
Make postgresql.conf.sample match the initdb defaults.  This fixes
comment alignment on most systems.

17 years agoAdd a couple of information functions to support direct checks on whether
Tom Lane [Thu, 14 Sep 2006 22:05:06 +0000 (22:05 +0000)]
Add a couple of information functions to support direct checks on whether
a schema is our own temp schema or another backend's temp schema, and use
these in place of some former kluges in information_schema.  Per my
proposal of yesterday.

17 years agoSeems no one wants this:
Bruce Momjian [Thu, 14 Sep 2006 21:44:08 +0000 (21:44 +0000)]
Seems no one wants this:

< * -Allow PREPARE to automatically determine parameter types based on the SQL
<   statement (Neil)

17 years agoDone:
Bruce Momjian [Thu, 14 Sep 2006 21:41:29 +0000 (21:41 +0000)]
Done:

< * -Allow protocol-level BIND parameter values to be logged, if text mode
> * -Allow protocol-level BIND parameter values to be logged

17 years agoNot done yet:
Bruce Momjian [Thu, 14 Sep 2006 21:37:04 +0000 (21:37 +0000)]
Not done yet:

<  o -Allow commenting of variables in postgresql.conf to restore them
>  o Allow commenting of variables in postgresql.conf to restore them

17 years agoFix missing markup.
Tom Lane [Thu, 14 Sep 2006 21:15:07 +0000 (21:15 +0000)]
Fix missing markup.

17 years agoRemove sslinfo copyright with author permission, keep author attribution.
Bruce Momjian [Thu, 14 Sep 2006 20:50:51 +0000 (20:50 +0000)]
Remove sslinfo copyright with author permission, keep author attribution.

Victor Wagner

17 years agoRemove:
Bruce Momjian [Thu, 14 Sep 2006 20:46:10 +0000 (20:46 +0000)]
Remove:

< * Research storing only active XIDs in subtransaction cache

17 years agoAdd:
Bruce Momjian [Thu, 14 Sep 2006 20:30:35 +0000 (20:30 +0000)]
Add:

> * Research storing only active XIDs in subtransaction cache

17 years agoAdd XML item.
Bruce Momjian [Thu, 14 Sep 2006 18:10:38 +0000 (18:10 +0000)]
Add XML item.
>
> * Improve XML support
>
>   http://developer.postgresql.org/index.php/XML_Support

17 years agoAdd description for variable-length header:
Bruce Momjian [Thu, 14 Sep 2006 17:34:14 +0000 (17:34 +0000)]
Add description for variable-length header:

>    One idea is to store the header in network byte order (high bits
>    first), and read the high bits to determine the header length.

17 years agoAdd recent ICU URL:
Bruce Momjian [Thu, 14 Sep 2006 17:14:53 +0000 (17:14 +0000)]
Add recent ICU URL:

>   http://archives.postgresql.org/pgsql-hackers/2006-09/msg00662.php

17 years agoUpdate TODO items:
Bruce Momjian [Thu, 14 Sep 2006 17:11:51 +0000 (17:11 +0000)]
Update TODO items:

<  o Store disk pages with no alignment/padding?
>
>    This involves having the user-specified order of columns
>    be different from the physical order.  SELECT * would
>    need to reorder the physical values to match the
>    user-specified ordering.
>
>  o Store disk pages with no alignment/padding?
>
>    This necessitates adding CPU-required padding when moving
>    rows from disk to memory.
>
>
>    http://archives.postgresql.org/pgsql-hackers/2006-09/msg00848.php
>

17 years agoFor release notes checklist, add UTF8 URL.
Bruce Momjian [Thu, 14 Sep 2006 16:55:43 +0000 (16:55 +0000)]
For release notes checklist, add UTF8 URL.

17 years agoIn release notes checklist, add URL for HTML non-ASCII escapes, mention
Bruce Momjian [Thu, 14 Sep 2006 16:51:39 +0000 (16:51 +0000)]
In release notes checklist, add URL for HTML non-ASCII escapes, mention
UTF8 escapes.

17 years agoFix SGML markup
Teodor Sigaev [Thu, 14 Sep 2006 13:40:28 +0000 (13:40 +0000)]
Fix SGML markup

17 years agoImprove error message. Per discussion
Teodor Sigaev [Thu, 14 Sep 2006 11:26:49 +0000 (11:26 +0000)]
Improve error message. Per discussion
http://archives.postgresql.org/pgsql-general/2006-09/msg00186.php

17 years agoGIN documentation and slightly improving GiST docs.
Teodor Sigaev [Thu, 14 Sep 2006 11:16:27 +0000 (11:16 +0000)]
GIN documentation and slightly improving GiST docs.
Thanks to  Christopher Kings-Lynne <chris.kingslynne@gmail.com> for
initial version and Jeff Davis <pgsql@j-davis.com> for inspection

17 years agoCygwin systems seem to have a different error message when trying to connect to port...
Michael Meskes [Thu, 14 Sep 2006 08:46:58 +0000 (08:46 +0000)]
Cygwin systems seem to have a different error message when trying to connect to port 0. Let's try port 20 instead.

17 years agoAdded missing constuctor/destructor for interval and date.
Michael Meskes [Thu, 14 Sep 2006 08:02:38 +0000 (08:02 +0000)]
Added missing constuctor/destructor for interval and date.

17 years agoSome small editorialization on the description of CREATE INDEX
Tom Lane [Wed, 13 Sep 2006 23:42:26 +0000 (23:42 +0000)]
Some small editorialization on the description of CREATE INDEX
CONCURRENTLY.  Greg Stark, some further tweaks by me.

17 years agoMake logging of extended-protocol commands a bit more consistent, per
Tom Lane [Wed, 13 Sep 2006 21:59:04 +0000 (21:59 +0000)]
Make logging of extended-protocol commands a bit more consistent, per
discussion with Guillaume Smet.

17 years agoCode review for patch to avoid second scan when vacuuming index-less
Tom Lane [Wed, 13 Sep 2006 17:47:08 +0000 (17:47 +0000)]
Code review for patch to avoid second scan when vacuuming index-less
table: avoid invoking LockBufferForCleanup without need, put out the
same log message we would have before, minor code beautification.

17 years agoChange "tps" to "scale" to avoid confusion
Tatsuo Ishii [Wed, 13 Sep 2006 00:39:19 +0000 (00:39 +0000)]
Change "tps" to "scale" to avoid confusion

Fix bug with handling default scaling factor in the default
scenarios

17 years agoAdd David Fetter:
Bruce Momjian [Tue, 12 Sep 2006 21:02:43 +0000 (21:02 +0000)]
Add David Fetter:

> * David is David Fetter <david@fetter.org>

17 years agoAdd uninstall script
Teodor Sigaev [Tue, 12 Sep 2006 16:11:44 +0000 (16:11 +0000)]
Add uninstall script

17 years agoSpell readme file's name correctly, per buildfarm reports.
Tom Lane [Tue, 12 Sep 2006 14:47:24 +0000 (14:47 +0000)]
Spell readme file's name correctly, per buildfarm reports.

17 years agoSmall fix for the README install for contrib/sslinfo.
Peter Eisentraut [Tue, 12 Sep 2006 11:46:11 +0000 (11:46 +0000)]
Small fix for the README install for contrib/sslinfo.

Dave Page

17 years agoSeems this wasn't committed originally
Michael Meskes [Tue, 12 Sep 2006 09:09:31 +0000 (09:09 +0000)]
Seems this wasn't committed originally

17 years agoMove set_pglocale_pgservice() from path.c to exec.c, so that pulling in
Tom Lane [Mon, 11 Sep 2006 20:10:30 +0000 (20:10 +0000)]
Move set_pglocale_pgservice() from path.c to exec.c, so that pulling in
path.c does not in itself force linking of both exec.c and libintl.
Should fix current ecpglib build failure on pickier platforms.

17 years agoMove xml2's PG_MODULE_MAGIC block to the right source file, per Michael Fuhr.
Tom Lane [Mon, 11 Sep 2006 15:30:32 +0000 (15:30 +0000)]
Move xml2's PG_MODULE_MAGIC block to the right source file, per Michael Fuhr.

17 years agoRename the uninstall scripts for contrib/lo and contrib/tsearch2 to
Tom Lane [Mon, 11 Sep 2006 15:14:46 +0000 (15:14 +0000)]
Rename the uninstall scripts for contrib/lo and contrib/tsearch2 to
match the convention that foo's uninstall script is uninstall_foo.sql.
Also, stop installing lo_test.sql, which really ought to be made into
a regression test anyway (though it's unclear how to avoid a dependency
on the current OID counter...)

17 years agoAdd LDAP_LIBS_FE to $(libpq) for static or AIX linking, per Albe Laurenz.
Tom Lane [Mon, 11 Sep 2006 13:35:08 +0000 (13:35 +0000)]
Add LDAP_LIBS_FE to $(libpq) for static or AIX linking, per Albe Laurenz.

17 years agoOoops, forgot to update ecpg comparison files to match
Tom Lane [Mon, 11 Sep 2006 02:50:07 +0000 (02:50 +0000)]
Ooops, forgot to update ecpg comparison files to match
source-code change.

17 years agoFix omissions in contrib uninstall scripts. Michael Fuhr
Tom Lane [Mon, 11 Sep 2006 02:10:26 +0000 (02:10 +0000)]
Fix omissions in contrib uninstall scripts.  Michael Fuhr

17 years agoFix unportable usage complained of by icc. Per buildfarm results.
Tom Lane [Mon, 11 Sep 2006 01:41:35 +0000 (01:41 +0000)]
Fix unportable usage complained of by icc.  Per buildfarm results.

17 years agoRemove unnecessary brace pair.
Bruce Momjian [Sun, 10 Sep 2006 23:33:22 +0000 (23:33 +0000)]
Remove unnecessary brace pair.

17 years agoBack out patch added during 8.2.X development:
Bruce Momjian [Sun, 10 Sep 2006 22:54:47 +0000 (22:54 +0000)]
Back out patch added during 8.2.X development:

Allow to_char() "D" format specifiers for interval/time.

It doesn't work, and I doubt it is useful enough to fix ("D" = day of
week).

17 years agoAdd items:
Bruce Momjian [Sun, 10 Sep 2006 22:20:01 +0000 (22:20 +0000)]
Add items:

< * Research storing disk pages with no alignment/padding
> * Consider ways of storing rows more compactly on disk
>
>  o Store disk pages with no alignment/padding?
>  o Reorder physical storage order to reduce padding?
>  o Support a smaller header for short variable-length fields?
>  o Reduce the row header size?

17 years agoInstall a cleaner solution to the AIX libpq linking problem, as per
Tom Lane [Sun, 10 Sep 2006 22:07:02 +0000 (22:07 +0000)]
Install a cleaner solution to the AIX libpq linking problem, as per
an earlier discussion.  Centralize assumptions about what libpq depends
on in one place in Makefile.global.  I am unconvinced that this list
is complete, but since ecpg seems to have gotten along with just these
entries, we'll try it this way and see what happens.

17 years agoFix dblink build for --enable-nls or --enable-openssl on AIX.
Tom Lane [Sun, 10 Sep 2006 21:38:26 +0000 (21:38 +0000)]
Fix dblink build for --enable-nls or --enable-openssl on AIX.
Per Chris Browne.

17 years agoRemove unhelpful/misleading advice about how to use SPI_saveplan().
Tom Lane [Sun, 10 Sep 2006 20:56:42 +0000 (20:56 +0000)]
Remove unhelpful/misleading advice about how to use SPI_saveplan().
Per gripe from Jack Orenstein.