]> granicus.if.org Git - postgresql/log
postgresql
20 years agoUse gettext_noop() to mark strings. errmsg() perform the
Dennis Bjorklund [Wed, 4 Aug 2004 21:55:27 +0000 (21:55 +0000)]
Use gettext_noop() to mark strings. errmsg() perform the
real gettext() later on, so it was called twice before.

20 years agoLabel CVS tip as 8.0devel instead of 7.5devel. Adjust various comments
Tom Lane [Wed, 4 Aug 2004 21:34:35 +0000 (21:34 +0000)]
Label CVS tip as 8.0devel instead of 7.5devel.  Adjust various comments
and documentation to reference 8.0 instead of 7.5.

20 years agoAdd LOG_NOWAIT flag to openlog() call, per my note of 2004-06-24.
Tom Lane [Wed, 4 Aug 2004 20:58:46 +0000 (20:58 +0000)]
Add LOG_NOWAIT flag to openlog() call, per my note of 2004-06-24.

20 years agoFix silly thinko in ALTER COLUMN TYPE. Check for finding expected
Tom Lane [Wed, 4 Aug 2004 20:53:53 +0000 (20:53 +0000)]
Fix silly thinko in ALTER COLUMN TYPE.  Check for finding expected
dependency was looking at wrong columns and so would always fail.
Was not exposed by regression tests because we are only testing cases
involving built-in (pinned) types and so no actual dependency entry
exists to be removed.

20 years agoUpdate oidjoins regression test to match current catalog structure.
Tom Lane [Wed, 4 Aug 2004 20:33:49 +0000 (20:33 +0000)]
Update oidjoins regression test to match current catalog structure.

20 years agoTweak postmaster code to avoid double reporting when bgwriter crashes.
Tom Lane [Wed, 4 Aug 2004 20:09:47 +0000 (20:09 +0000)]
Tweak postmaster code to avoid double reporting when bgwriter crashes.

20 years agorecord_out and friends need to cope with dropped columns in the row
Tom Lane [Wed, 4 Aug 2004 19:31:15 +0000 (19:31 +0000)]
record_out and friends need to cope with dropped columns in the row
datatype.  Per example from Gaetano Mendola, 2004-07-25.

20 years agoAdd some notes about unimplemented aspects of PITR backup/recovery.
Tom Lane [Wed, 4 Aug 2004 17:37:09 +0000 (17:37 +0000)]
Add some notes about unimplemented aspects of PITR backup/recovery.

20 years agoDump comments on columns of composite types.
Tom Lane [Wed, 4 Aug 2004 17:13:03 +0000 (17:13 +0000)]
Dump comments on columns of composite types.

Instead of putting all the OWNER TO commands at the end, it dumps then
after each object.  This is WAY more readable and nice.  ACLs are still
at the end.

Christopher Kings-Lynne

20 years agoTranslation updates
Dennis Bjorklund [Wed, 4 Aug 2004 16:40:11 +0000 (16:40 +0000)]
Translation updates

20 years agoMake use of backup label/history files to control recovery properly.
Tom Lane [Wed, 4 Aug 2004 16:25:02 +0000 (16:25 +0000)]
Make use of backup label/history files to control recovery properly.

20 years agoFix typo in comment.
Tom Lane [Wed, 4 Aug 2004 16:24:26 +0000 (16:24 +0000)]
Fix typo in comment.

20 years agoAdd some strings for translation and remove some cut'n'paste
Dennis Bjorklund [Wed, 4 Aug 2004 16:05:13 +0000 (16:05 +0000)]
Add some strings for translation and remove some cut'n'paste
that makes it impossible to translate to other languages.

20 years agoFix syntax error by adding a space.
Bruce Momjian [Wed, 4 Aug 2004 16:02:08 +0000 (16:02 +0000)]
Fix syntax error by adding a space.

20 years agoPreliminary documentation for PITR.
Tom Lane [Tue, 3 Aug 2004 23:42:59 +0000 (23:42 +0000)]
Preliminary documentation for PITR.

20 years agoAdd functions pg_start_backup, pg_stop_backup to create backup label
Tom Lane [Tue, 3 Aug 2004 20:32:36 +0000 (20:32 +0000)]
Add functions pg_start_backup, pg_stop_backup to create backup label
and history files as per recent discussion.  While at it, remove
pg_terminate_backend, since we have decided we do not have time during
this release cycle to address the reliability concerns it creates.
Split the 'Miscellaneous Functions' documentation section into
'System Information Functions' and 'System Administration Functions',
which hopefully will draw the eyes of those looking for such things.

20 years agoFix misplacement of savepointLevel test, per report from Chris K-L.
Tom Lane [Tue, 3 Aug 2004 15:57:26 +0000 (15:57 +0000)]
Fix misplacement of savepointLevel test, per report from Chris K-L.

20 years agoTranslation update
Peter Eisentraut [Tue, 3 Aug 2004 08:11:37 +0000 (08:11 +0000)]
Translation update

20 years agoMove dbsize/oid2name to open items list.
Bruce Momjian [Tue, 3 Aug 2004 01:14:57 +0000 (01:14 +0000)]
Move dbsize/oid2name to open items list.

20 years agoFix not-quite-right Assertion. Did not work at all in extended-query
Tom Lane [Mon, 2 Aug 2004 21:42:18 +0000 (21:42 +0000)]
Fix not-quite-right Assertion.  Did not work at all in extended-query
mode (per complaint from Kris Jurka) and it was only by chance that it
didn't fail in simple-query mode.  A COMMIT or ROLLBACK has to be
executed by a portal, therefore it's wrong to suppose that there aren't
any live portals at CleanupTransaction time.

20 years agoActually, there's no need to use pg_strcasecmp for checking exception
Tom Lane [Mon, 2 Aug 2004 17:03:48 +0000 (17:03 +0000)]
Actually, there's no need to use pg_strcasecmp for checking exception
names, because the name we got from the lexer is already downcased.
Just store the table in lower case and use strcmp ...

20 years agoWhile perusing SQL92 I realized that we are delivering the wrong SQLSTATE
Tom Lane [Mon, 2 Aug 2004 16:51:10 +0000 (16:51 +0000)]
While perusing SQL92 I realized that we are delivering the wrong SQLSTATE
error code for string-too-long errors.  It should be STRING_DATA_RIGHT_TRUNCATION
not STRING_DATA_LENGTH_MISMATCH.  The latter probably should only be
applied to cases where a string must be exactly so many bits --- there are
no cases at all where it applies to character strings, only bit strings.

20 years agoTranslation updates
Peter Eisentraut [Mon, 2 Aug 2004 15:17:21 +0000 (15:17 +0000)]
Translation updates

20 years agoTranslation update
Peter Eisentraut [Mon, 2 Aug 2004 15:11:19 +0000 (15:11 +0000)]
Translation update

20 years agoMark savepoints as supported.
Bruce Momjian [Mon, 2 Aug 2004 12:46:49 +0000 (12:46 +0000)]
Mark savepoints as supported.

20 years agoReadd pg_config --pgxs code.
Bruce Momjian [Mon, 2 Aug 2004 12:34:14 +0000 (12:34 +0000)]
Readd pg_config --pgxs code.

20 years agoSupport USING INDEX TABLESPACE clause for PRIMARY KEY and UNIQUE
Tom Lane [Mon, 2 Aug 2004 04:28:29 +0000 (04:28 +0000)]
Support USING INDEX TABLESPACE clause for PRIMARY KEY and UNIQUE
constraints.  Christopher Kings-Lynne.

20 years agoAllow DECLARE CURSOR to take parameters from the portal in which it is
Tom Lane [Mon, 2 Aug 2004 01:30:51 +0000 (01:30 +0000)]
Allow DECLARE CURSOR to take parameters from the portal in which it is
executed.  Previously, the DECLARE would succeed but subsequent FETCHes
would fail since the parameter values supplied to DECLARE were not
propagated to the portal created for the cursor.
In support of this, add type Oids to ParamListInfo entries, which seems
like a good idea anyway since code that extracts a value can double-check
that it got the type of value it was expecting.
Oliver Jowett, with minor editorialization by Tom Lane.

20 years agoUpdate the in-code documentation about the transaction system. Move it
Tom Lane [Sun, 1 Aug 2004 20:57:59 +0000 (20:57 +0000)]
Update the in-code documentation about the transaction system.  Move it
into a README file instead of being in xact.c's header comment.
Alvaro Herrera.

20 years agoCause ALTER OWNER commands to update the object's ACL, replacing references
Tom Lane [Sun, 1 Aug 2004 20:30:49 +0000 (20:30 +0000)]
Cause ALTER OWNER commands to update the object's ACL, replacing references
to the old owner with the new owner.  This is not necessarily right, but
it's sure a lot more likely to be what the user wants than doing nothing.
Christopher Kings-Lynne, some rework by Tom Lane.

20 years agoAdd libpgport to postgres.def for Windows build. Per Magnus Hagander.
Tom Lane [Sun, 1 Aug 2004 18:07:42 +0000 (18:07 +0000)]
Add libpgport to postgres.def for Windows build.  Per Magnus Hagander.

20 years agoTranslation updates (zh_TW converted to UTF-8)
Peter Eisentraut [Sun, 1 Aug 2004 17:54:22 +0000 (17:54 +0000)]
Translation updates (zh_TW converted to UTF-8)

20 years agoError message style adjustments, per Alvaro Herrera.
Tom Lane [Sun, 1 Aug 2004 17:45:43 +0000 (17:45 +0000)]
Error message style adjustments, per Alvaro Herrera.

20 years agoSome mop-up work for savepoints (nested transactions). Store a small
Tom Lane [Sun, 1 Aug 2004 17:32:22 +0000 (17:32 +0000)]
Some mop-up work for savepoints (nested transactions).  Store a small
number of active subtransaction XIDs in each backend's PGPROC entry,
and use this to avoid expensive probes into pg_subtrans during
TransactionIdIsInProgress.  Extend EOXactCallback API to allow add-on
modules to get control at subxact start/end.  (This is deliberately
not compatible with the former API, since any uses of that API probably
need manual review anyway.)  Add basic reference documentation for
SAVEPOINT and related commands.  Minor other cleanups to check off some
of the open issues for subtransactions.
Alvaro Herrera and Tom Lane.

20 years agoSome name cleanup.
Bruce Momjian [Sun, 1 Aug 2004 14:08:08 +0000 (14:08 +0000)]
Some name cleanup.

20 years agoSmall pg_config.c cleanup.
Bruce Momjian [Sun, 1 Aug 2004 14:01:36 +0000 (14:01 +0000)]
Small pg_config.c cleanup.

20 years agoFix :
Bruce Momjian [Sun, 1 Aug 2004 13:55:12 +0000 (13:55 +0000)]
Fix :

< * Magnus is Magnus Haglander (?)
> * Magnus is Magnus Hagander <mha@sollentuna.net>

20 years agoAdd missing file.
Bruce Momjian [Sun, 1 Aug 2004 13:54:05 +0000 (13:54 +0000)]
Add missing file.

20 years agoCreate a C version of pg_config.
Bruce Momjian [Sun, 1 Aug 2004 06:56:39 +0000 (06:56 +0000)]
Create a C version of pg_config.

Andrew Dunstan

20 years agoMinor adjustments for dirmod.c.
Bruce Momjian [Sun, 1 Aug 2004 06:22:38 +0000 (06:22 +0000)]
Minor adjustments for dirmod.c.

20 years agoAdd docs for initdb --auth.
Bruce Momjian [Sun, 1 Aug 2004 06:19:26 +0000 (06:19 +0000)]
Add docs for initdb --auth.

20 years ago>I got a new idea on this. I think we should add an initdb option that
Bruce Momjian [Sun, 1 Aug 2004 05:59:13 +0000 (05:59 +0000)]
>I got a new idea on this.  I think we should add an initdb option that
>takes a string to specify the local authentication method:
>
>       initdb --auth 'ident'
>
>or whatever the user wants.  I think this is more flexible and more
>compact.  It would default to 'trust', and the packagers could
>set it to
>whatever they want.  If their OS supports local ident, they can use
>that.
>
>Also keep in mind you might want some ident map file:
>
>       initdb --auth 'ident mymap'
>
>so you would need to allow multiple words in the string.

Magnus Hagander

20 years agoMinor release wording improvments.
Bruce Momjian [Sun, 1 Aug 2004 05:22:26 +0000 (05:22 +0000)]
Minor release wording improvments.

20 years agoAdd descriptions to TODO items and make adjustments based on 7.5.
Bruce Momjian [Sun, 1 Aug 2004 05:15:58 +0000 (05:15 +0000)]
Add descriptions to TODO items and make adjustments based on 7.5.

20 years agoRemove pool discussion. All old stuff.
Bruce Momjian [Sun, 1 Aug 2004 04:53:38 +0000 (04:53 +0000)]
Remove pool discussion. All old stuff.

20 years agoSupport "OR condition ..." in plpgsql EXCEPTION clauses to make the syntax
Tom Lane [Sat, 31 Jul 2004 23:04:58 +0000 (23:04 +0000)]
Support "OR condition ..." in plpgsql EXCEPTION clauses to make the syntax
more nearly Oracle-equivalent.  Allow matching by category as well as
specific error code.  Document the set of available condition names
(or more accurately, synchronize it with the existing documentation).  In
passing, update errcodes.sgml to include codes added during 7.5 development.

20 years agoBe more consistent about reporting SPI errors in the various PLs.
Tom Lane [Sat, 31 Jul 2004 20:55:45 +0000 (20:55 +0000)]
Be more consistent about reporting SPI errors in the various PLs.
Create a shared function to convert a SPI error code into a string
(replacing near-duplicate code in several PLs), and use it anywhere
that a SPI function call error is reported.

20 years agoEscape backslashes from translation message rather than removing them.
Bruce Momjian [Sat, 31 Jul 2004 20:00:26 +0000 (20:00 +0000)]
Escape backslashes from translation message rather than removing them.

20 years agoStill another try at matching system timezone nicely. On non-Windows
Tom Lane [Sat, 31 Jul 2004 19:12:15 +0000 (19:12 +0000)]
Still another try at matching system timezone nicely.  On non-Windows
machines, break tie scores by preferring shorter zone names over longer;
for names of equal length, prefer the alphabetically first name.  This
yields for example 'EST5EDT' not 'America/New_York' for US eastern time.
On Windows, abandon the whole concept of inspecting the detailed behavior
of the system TZ library, because it doesn't bear inspection :-(.  Instead
use a hardwired mapping table to select our zone name based on the
result of strftime %Z output.  Windows code from Magnus Hagander.

20 years agoFor EXEC_BACKEND case, BootstrapMain needs to recompute paths, same as
Tom Lane [Sat, 31 Jul 2004 17:57:11 +0000 (17:57 +0000)]
For EXEC_BACKEND case, BootstrapMain needs to recompute paths, same as
PostgresMain.  Per Magnus Hagander.

20 years agoRemove obsolete files
Peter Eisentraut [Sat, 31 Jul 2004 16:35:17 +0000 (16:35 +0000)]
Remove obsolete files

20 years agoTranslation update
Peter Eisentraut [Sat, 31 Jul 2004 07:57:41 +0000 (07:57 +0000)]
Translation update

20 years agoplpgsql does exceptions.
Tom Lane [Sat, 31 Jul 2004 07:39:21 +0000 (07:39 +0000)]
plpgsql does exceptions.
There are still some things that need refinement; in particular I fear
that the recognized set of error condition names probably has little in
common with what Oracle recognizes.  But it's a start.

20 years agoFix broken markup, release date.
Tom Lane [Sat, 31 Jul 2004 07:11:07 +0000 (07:11 +0000)]
Fix broken markup, release date.

20 years ago>OK, if you look at the last few commits to interfaces/libpq/win32.mak,
Bruce Momjian [Sat, 31 Jul 2004 06:19:23 +0000 (06:19 +0000)]
>OK, if you look at the last few commits to interfaces/libpq/win32.mak,
>you will see changes for pg_config_paths.h and for thread safety.  Can
>you look at those changes, merge them into bcc.mak.  There might be
>other changes in there you want to add to bcc.mak too.
>
>
>
OK, having no M$ diff, I attach files src\interfaces\libpq\bcc32.mak and
src\interfaces\libpq\win32.h

Steve Holdoway

20 years agoAdd:
Bruce Momjian [Sat, 31 Jul 2004 06:13:52 +0000 (06:13 +0000)]
Add:

> * Allow point-in-time recovery to archive partially filled logs

20 years agoFix zh_TW.po initdb translation.
Bruce Momjian [Sat, 31 Jul 2004 04:41:54 +0000 (04:41 +0000)]
Fix zh_TW.po initdb translation.

20 years agoRemove mention that we can't trap function errors.
Bruce Momjian [Sat, 31 Jul 2004 04:22:46 +0000 (04:22 +0000)]
Remove mention that we can't trap function errors.

20 years agoRestructure error handling as recently discussed. It is now really
Tom Lane [Sat, 31 Jul 2004 00:45:57 +0000 (00:45 +0000)]
Restructure error handling as recently discussed.  It is now really
possible to trap an error inside a function rather than letting it
propagate out to PostgresMain.  You still have to use AbortCurrentTransaction
to clean up, but at least the error handling itself will cooperate.

20 years agoMust guard against NULL return from localtime() when probing pre-1970
Tom Lane [Fri, 30 Jul 2004 17:31:24 +0000 (17:31 +0000)]
Must guard against NULL return from localtime() when probing pre-1970
dates.  Per Magnus Hagander.

20 years agoFix number for FAQ item.
Bruce Momjian [Fri, 30 Jul 2004 16:51:59 +0000 (16:51 +0000)]
Fix number for FAQ item.

20 years agoPostgreSQL extension makefile framework ("pgxs"), by Fabien Coelho, with
Peter Eisentraut [Fri, 30 Jul 2004 12:26:40 +0000 (12:26 +0000)]
PostgreSQL extension makefile framework ("pgxs"), by Fabien Coelho, with
some massaging by Peter Eisentraut.  This is basically a simple
generalization of the existing contrib makefiles.

20 years agoTranslation updates
Peter Eisentraut [Fri, 30 Jul 2004 05:28:53 +0000 (05:28 +0000)]
Translation updates

20 years agoAdd:
Bruce Momjian [Thu, 29 Jul 2004 16:47:07 +0000 (16:47 +0000)]
Add:

> * Allow administrators to safely terminate individual sessions

20 years agoUpdate Turkish FAQ.
Bruce Momjian [Thu, 29 Jul 2004 16:30:32 +0000 (16:30 +0000)]
Update Turkish FAQ.

Devrim GUNDUZ

20 years agoAdd:
Bruce Momjian [Thu, 29 Jul 2004 16:24:02 +0000 (16:24 +0000)]
Add:

> * Allow server logs to be read using SQL commands
> * Allow server configuration parameters to be modified remotetly

20 years agoFix for pg_ctl -o processing. Report from Sean Chittenden.
Bruce Momjian [Thu, 29 Jul 2004 16:11:11 +0000 (16:11 +0000)]
Fix for pg_ctl -o processing.  Report from  Sean Chittenden.

20 years agoTranslation update
Peter Eisentraut [Thu, 29 Jul 2004 14:54:09 +0000 (14:54 +0000)]
Translation update

20 years agoRevert ill-conceived patch that made elog(FATAL) the same as elog(ERROR)
Tom Lane [Wed, 28 Jul 2004 22:05:47 +0000 (22:05 +0000)]
Revert ill-conceived patch that made elog(FATAL) the same as elog(ERROR)
followed by seeing EOF from client.  If we want a safe session-kill
capability we will need to write one, not break our error handling
mechanism.

20 years agoUpdate release notes with better wording.
Bruce Momjian [Wed, 28 Jul 2004 14:28:57 +0000 (14:28 +0000)]
Update release notes with better wording.

Brian B.

20 years agoFix subtransaction behavior for large objects, temp namespace, files,
Tom Lane [Wed, 28 Jul 2004 14:23:31 +0000 (14:23 +0000)]
Fix subtransaction behavior for large objects, temp namespace, files,
password/group files.  Also allow read-only subtransactions of a read-write
parent, but not vice versa.  These are the reasonably noncontroversial
parts of Alvaro's recent mop-up patch, plus further work on large objects
to minimize use of the TopTransactionResourceOwner.

20 years agoReplace nested-BEGIN syntax for subtransactions with spec-compliant
Tom Lane [Tue, 27 Jul 2004 05:11:48 +0000 (05:11 +0000)]
Replace nested-BEGIN syntax for subtransactions with spec-compliant
SAVEPOINT/RELEASE/ROLLBACK-TO syntax.  (Alvaro)
Cause COMMIT of a failed transaction to report ROLLBACK instead of
COMMIT in its command tag.  (Tom)
Fix a few loose ends in the nested-transactions stuff.

20 years agoUpdate tuple header size.
Bruce Momjian [Tue, 27 Jul 2004 03:27:48 +0000 (03:27 +0000)]
Update tuple header size.

20 years agoFix oversight: in case where SIGTERM is received while there are
Tom Lane [Tue, 27 Jul 2004 01:46:03 +0000 (01:46 +0000)]
Fix oversight: in case where SIGTERM is received while there are
live backends, the archiver and stats processes never got sent a
kill signal.  They'd eventually exit on their own, but not for awhile,
which is a bit annoying when you are trying to replace the executable
file on a platform that doesn't allow removal of busy executables.
Also, tweak main loop logic so that we will perform the background
tasks after select() returns EINTR.

20 years agoAdd some more logging to simplify diagnosis of problems in IDENT
Tom Lane [Mon, 26 Jul 2004 18:53:38 +0000 (18:53 +0000)]
Add some more logging to simplify diagnosis of problems in IDENT
authorization.  Inspired by problem report from Ian Pilcher.

20 years agoSQL defines are only used in SQL space in Informix mode.
Michael Meskes [Mon, 26 Jul 2004 10:28:28 +0000 (10:28 +0000)]
SQL defines are only used in SQL space in Informix mode.

20 years agoHere is a patch that fixes the pipes used in find_other_exec() when
Bruce Momjian [Mon, 26 Jul 2004 01:48:00 +0000 (01:48 +0000)]
Here is a patch that fixes the pipes used in find_other_exec() when
running as a service on windows <= 2000. Required to make the pg_ctl
service wrapper to work at all.

Magnus Hagander

20 years agoAdjust release notes based on community feedback.
Bruce Momjian [Mon, 26 Jul 2004 00:26:42 +0000 (00:26 +0000)]
Adjust release notes based on community feedback.

20 years agoTranslation updates
Peter Eisentraut [Sun, 25 Jul 2004 12:00:07 +0000 (12:00 +0000)]
Translation updates

20 years agoUpdate 7.5 release notes.
Bruce Momjian [Sun, 25 Jul 2004 04:18:05 +0000 (04:18 +0000)]
Update 7.5 release notes.

20 years agoImprove description of release note creation.
Bruce Momjian [Sun, 25 Jul 2004 03:30:31 +0000 (03:30 +0000)]
Improve description of release note creation.

20 years agoFix obsolete comment.
Tom Lane [Sat, 24 Jul 2004 20:01:42 +0000 (20:01 +0000)]
Fix obsolete comment.

20 years agolog_destination should only be PGC_SIGHUP, not PGC_POSTMASTER, as per
Tom Lane [Sat, 24 Jul 2004 19:51:23 +0000 (19:51 +0000)]
log_destination should only be PGC_SIGHUP, not PGC_POSTMASTER, as per
discussion yesterday.  Also a few improvements in the associated
documentation.

20 years agoFirst draft of 7.5 release notes. Still needs markup, additional major
Bruce Momjian [Sat, 24 Jul 2004 09:36:53 +0000 (09:36 +0000)]
First draft of 7.5 release notes.  Still needs markup, additional major
items, and detailed text descriptions.

20 years agoAdd steps for creating release notes.
Bruce Momjian [Sat, 24 Jul 2004 03:06:47 +0000 (03:06 +0000)]
Add steps for creating release notes.

20 years agoMove Win32 defines from port.h to win32.h.
Bruce Momjian [Fri, 23 Jul 2004 01:58:36 +0000 (01:58 +0000)]
Move Win32 defines from port.h to win32.h.

20 years agoAdd cross-check that current timeline of pg_control is an ancestor of
Tom Lane [Thu, 22 Jul 2004 21:09:37 +0000 (21:09 +0000)]
Add cross-check that current timeline of pg_control is an ancestor of
recovery_target_timeline --- otherwise there is no path from the backup
to the requested timeline.  This check was foreseen in the original
discussion but I forgot to implement it.

20 years agoAdd a check on file size as an additional safety check that a WAL file
Tom Lane [Thu, 22 Jul 2004 20:18:40 +0000 (20:18 +0000)]
Add a check on file size as an additional safety check that a WAL file
recovered from archive is not corrupt.  It's not much but it will catch
one common problem, viz out-of-disk-space.
Also, force a WAL recovery scan when recovery.conf is present, even if
pg_control shows a clean shutdown.  This allows recovery with a tar backup
that was taken with the postmaster shut down, as per complaint from
Mark Kirkwood.

20 years agoUpdated wording:
Bruce Momjian [Thu, 22 Jul 2004 12:24:13 +0000 (12:24 +0000)]
Updated wording:

> * Have psql show current values for a sequence

20 years agoDone:
Bruce Momjian [Thu, 22 Jul 2004 11:49:19 +0000 (11:49 +0000)]
Done:

> * -Have psql show more information about sequences

20 years agoStill another try at automatically detecting the best match in the zic
Tom Lane [Thu, 22 Jul 2004 05:28:30 +0000 (05:28 +0000)]
Still another try at automatically detecting the best match in the zic
timezone database for the system behavior we find ourselves in.  Scan
backwards from current time and choose the zone that matches furthest
back.  As per discussion a week or so back.

20 years agoAvoid default: followed by no statement, per Dann Corbit.
Tom Lane [Thu, 22 Jul 2004 01:44:36 +0000 (01:44 +0000)]
Avoid default: followed by no statement, per Dann Corbit.

20 years agoAdd missing <getopt.h>, per Dann Corbit.
Tom Lane [Thu, 22 Jul 2004 01:41:24 +0000 (01:41 +0000)]
Add missing <getopt.h>, per Dann Corbit.

20 years agoInvent WAL timelines, as per recent discussion, to make point-in-time
Tom Lane [Wed, 21 Jul 2004 22:31:26 +0000 (22:31 +0000)]
Invent WAL timelines, as per recent discussion, to make point-in-time
recovery more manageable.  Also, undo recent change to add FILE_HEADER
and WASTED_SPACE records to XLOG; instead make the XLOG page header
variable-size with extra fields in the first page of an XLOG file.
This should fix the boundary-case bugs observed by Mark Kirkwood.
initdb forced due to change of XLOG representation.

20 years agoThe attached patch allows 'select foo()' as well as 'select * from
Bruce Momjian [Wed, 21 Jul 2004 20:45:54 +0000 (20:45 +0000)]
The attached patch allows 'select foo()' as well as 'select * from
foo()' where foo() is a plperl function that returns a single composite.

Andrew Dunstan

20 years agoPlease find enclosed a patch that matches the PL/Perl documentation
Bruce Momjian [Wed, 21 Jul 2004 20:44:52 +0000 (20:44 +0000)]
Please find enclosed a patch that matches the PL/Perl documentation
(fairly closely, I hope) to the current PL/Perl implementation.

David Fetter

20 years agoThe attached patch shows the new column "tablespace" in the mentioned
Bruce Momjian [Wed, 21 Jul 2004 20:43:53 +0000 (20:43 +0000)]
The attached patch shows the new column "tablespace" in the mentioned
views.

Klaus Naumann

20 years agoBack out pg_autovacuum commit after cvs clean failure causes commit.
Bruce Momjian [Wed, 21 Jul 2004 20:34:50 +0000 (20:34 +0000)]
Back out pg_autovacuum commit after cvs clean failure causes commit.

20 years agolease find enclosed a patch that matches the PL/Perl documentation
Bruce Momjian [Wed, 21 Jul 2004 20:23:05 +0000 (20:23 +0000)]
lease find enclosed a patch that matches the PL/Perl documentation
(fairly closely, I hope) to the current PL/Perl implementation.

David Fetter

20 years agoMention plpgsql:
Bruce Momjian [Wed, 21 Jul 2004 13:46:44 +0000 (13:46 +0000)]
Mention plpgsql:

<  o -Allow parameters to be specified by name and type during definition
>  o -Allow PL/pgSQL parameters to be specified by name and type during definition