Commit Graph

12158 Commits

Author SHA1 Message Date
Bruce Momjian
170c6760e8 Marc, you get this one. :-) You are the expert on the item:
> * Allow ability to easily build only certain interfaces (Marc)
2002-02-21 05:10:49 +00:00
Tom Lane
a40c17e17d Fix comment. 2002-02-20 22:47:12 +00:00
Bruce Momjian
8b9f293e7d Add:
> 	o Allow cursors outside transactions [cursor]
2002-02-20 03:50:33 +00:00
Bruce Momjian
9cfb55c55a Add cursors outside transactions thread. 2002-02-20 03:49:06 +00:00
Bruce Momjian
fe3045f5a9 Add savepoints discussion. 2002-02-20 00:44:30 +00:00
Bruce Momjian
6e73984d52 Add schema discussion. 2002-02-20 00:42:01 +00:00
Bruce Momjian
02f3a69571 Add:
> * Clarify use of 'application' and 'command' tags in SGML docs
2002-02-19 22:41:29 +00:00
Bruce Momjian
8a58c7f5bc Add:
> * Improve concurrency of hash indexes
> * Test hash index performance and recommend or discourage usage
2002-02-19 22:34:27 +00:00
Tom Lane
d9b01c13a6 Avoid failures in cash_out and cash_words for INT_MIN.
Also, 'fourty' -> 'forty'.
2002-02-19 22:19:34 +00:00
Bruce Momjian
a2b4dbd4a7 Clearify:
> * Make single-user local access permissions the default by limiting
>   permissions on the socket file (Peter E)
2002-02-19 21:38:29 +00:00
Bruce Momjian
0e8c26eeca Added:
---
> * Improve caching of attribute offsets when NULLs exist in the row
2002-02-19 21:36:58 +00:00
Tom Lane
2220a2b5d7 Add casts to suppress gcc warnings on Solaris (where apparently pid_t
is different from int).
2002-02-19 20:45:04 +00:00
Bruce Momjian
0519a1a1f7 Update as done:
> * -Cache system catalog information in per-database files (Tom)
2002-02-19 20:31:31 +00:00
Tom Lane
7863404417 A bunch of changes aimed at reducing backend startup time...
Improve 'pg_internal.init' relcache entry preload mechanism so that it is
safe to use for all system catalogs, and arrange to preload a realistic
set of system-catalog entries instead of only the three nailed-in-cache
indexes that were formerly loaded this way.  Fix mechanism for deleting
out-of-date pg_internal.init files: this must be synchronized with transaction
commit, not just done at random times within transactions.  Drive it off
relcache invalidation mechanism so that no special-case tests are needed.

Cache additional information in relcache entries for indexes (their pg_index
tuples and index-operator OIDs) to eliminate repeated lookups.  Also cache
index opclass info at the per-opclass level to avoid repeated lookups during
relcache load.

Generalize 'systable scan' utilities originally developed by Hiroshi,
move them into genam.c, use in a number of places where there was formerly
ugly code for choosing either heap or index scan.  In particular this allows
simplification of the logic that prevents infinite recursion between syscache
and relcache during startup: we can easily switch to heapscans in relcache.c
when and where needed to avoid recursion, so IndexScanOK becomes simpler and
does not need any expensive initialization.

Eliminate useless opening of a heapscan data structure while doing an indexscan
(this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 20:11:20 +00:00
Tom Lane
8e2998d8a6 Remove some unnecessary pqsignal() calls to shave a few cycles off
backend startup.
2002-02-19 19:54:43 +00:00
Tom Lane
6e546c286c Arrange to call localtime() during postmaster startup. On most Unixen,
the first call of localtime() in a process will read /usr/lib/tztab or
local equivalent.  Better to do this once in the postmaster and inherit
the data by fork() than to have to do it during every backend start.
2002-02-19 19:53:35 +00:00
Tom Lane
9103372f52 Avoid calling pq_flush just after sending AUTH_REQ_OK; this saves one
send() and probable process context swap during backend startup.
2002-02-19 19:49:09 +00:00
Peter Eisentraut
8adf56f77a Privileges on functions and procedural languages 2002-02-18 23:11:58 +00:00
Peter Eisentraut
5e03503126 Set version to 7.3devel. 2002-02-18 21:46:17 +00:00
Bruce Momjian
04802852b6 Add:
> 	o Generate failure on short COPY lines rather than pad NULLs
2002-02-18 16:54:38 +00:00
Tom Lane
c89e0d9666 Minor copy-editing. 2002-02-18 16:13:10 +00:00
Tom Lane
e4dd067398 Replace number-of-distinct-values estimator equation, per recent
pghackers discussion.
2002-02-18 16:04:14 +00:00
Bruce Momjian
b4a5fa4518 Remove MAX/MIN() macros, use c.h Max/Min() instead. 2002-02-18 14:25:40 +00:00
Bruce Momjian
d622cf867b Disable LIMIT #,# syntax, as agreed to months ago. Print message saying
to use separate LIMIT/OFFSET clauses.
2002-02-18 06:49:20 +00:00
Bruce Momjian
890a012d0a Reorder postgresql.conf WAL section to be clearer. 2002-02-18 06:42:54 +00:00
Bruce Momjian
5651a1665f Don't mention TIOGA in Makefile and move strdup.c rule into proper
'ifdef' in Makefile.
2002-02-18 06:03:23 +00:00
Bruce Momjian
fb8e3f15c0 Add code to preserve paren level display after \g, as submitted months ago. 2002-02-18 05:57:41 +00:00
Hiroshi Inoue
c03e7fbadc Remove compile warnings in multibute mode. 2002-02-18 05:54:50 +00:00
Bruce Momjian
fbcc0d69e8 Fix param handling of create* admin scripts as described months ago.
Properly handles default values.
2002-02-18 05:48:45 +00:00
Bruce Momjian
c9361a7a9c Remove CHECKPOINT call from pgbench -- Not needed and affects performance. 2002-02-18 05:46:41 +00:00
Bruce Momjian
c448847378 Add better error text:
elog(LOG, "XLogWrite: new log file created - "
                     "consider increasing 'wal_files' in postgresql.conf.");
2002-02-18 05:44:45 +00:00
Bruce Momjian
b0bf66dfbe Add HTML output option to pgcvslog. 2002-02-18 05:33:13 +00:00
Bruce Momjian
865d26fdd6 Update FAQ. 2002-02-18 05:09:01 +00:00
Tatsuo Ishii
ab604e8ac3 Fix kanji-coversion key binding. This has been broken since 7.1
Per Yoshinori Ariie's report.
2002-02-18 04:01:27 +00:00
Bruce Momjian
14c1c33f3b Add Neil's Name:
> * Have pg_dump use ADD PRIMARY KEY after COPY, for performance (Neil Conway)
2002-02-18 03:40:13 +00:00
Hiroshi Inoue
3c16d095b5 The version is now 7.01.0010.
1) Handle parameter array.
2) Allow re-use of the connection handle after SQLDisconnect.
3) Reject NULL if no indicator specified.
4) Improve the handling of '_' in table name.
5) Unify internal begin/commit/abort operations.
6) Change SQLTables() to return null not "" for the
   table_owner.
7) Fix a bug about parameter handling reported by Benoit Menendez.
8) Add cast in handling ODBC date/time escape sequences.
9) Fix a bug about cache_size handing in declare/fetch mode.

[ODBC3.0 related]
10) Improve the handling of descriptor handles(ODBC3.0).
11) Improve the type handling of some types for ODBC3.0.

[Thanks to Marcelo Aceto for his useful patches]
12) Allow nested ODBC escape.
13) Allow changing autocommit on/off inside the transaction
    block.
14) Improve the handling of ODBC scalar functions.
2002-02-18 03:16:11 +00:00
Hiroshi Inoue
b6db89a02d Separate info30.c from info.c. 2002-02-18 02:15:04 +00:00
Bruce Momjian
aedc8a99d5 Add:
> * Have pg_dump use ADD PRIMARY KEY after COPY, for performance
2002-02-18 02:00:43 +00:00
Hiroshi Inoue
22d7fa7de4 Add files for ODBC3.0 support. 2002-02-18 01:43:24 +00:00
Bruce Momjian
26d6e199d9 Added:
> * ODBC
> 	o ODBC 3.0 support
> 	o Unicode(UCS-2) support
> 	o Updatable cursors support
2002-02-18 01:17:16 +00:00
Bruce Momjian
4f5991552c Remove ODBC todo, add to main TODO. 2002-02-18 01:15:46 +00:00
Bruce Momjian
62feef6c95 Fix SGML typo in previous patch. 2002-02-17 13:29:00 +00:00
Bruce Momjian
bf4507e3af I think it's important that it's actually documented that they can add
primary keys after the fact!

Also, we need to add regression tests for alter table / add primary key
and alter table / drop constraint.  These shouldn't be added until 7.3
tho
methinks...

Chris
2002-02-17 11:50:09 +00:00
Bruce Momjian
a966ac1df0 Clarify params to ALTER TABLE to clearly show single parameters.
e.g. table contraint definition -> table_constraint_definition.
2002-02-16 23:45:48 +00:00
Peter Eisentraut
673b48becb Remove warning about automatic inclusion of sqlca. 2002-02-15 17:46:57 +00:00
Bruce Momjian
ceec779ab5 Update FAQ. 2002-02-14 17:15:00 +00:00
Tom Lane
3576820e78 Ensure that a cursor is scanned under the same scanCommandId it was
originally created with, so that the set of visible tuples does not
change as a result of other activity.  This essentially makes PG cursors
INSENSITIVE per the SQL92 definition.  See bug report of 13-Feb-02.
2002-02-14 15:24:10 +00:00
Tom Lane
1392042346 Point out that --adduser actually makes the new user a superuser. This
was mentioned on the man page for the underlying CREATE USER command,
but it should be explained here too.
2002-02-13 19:32:17 +00:00
Bruce Momjian
a164407b64 Add DROP COLUMN status from Hiroshi. 2002-02-13 14:36:43 +00:00
Tom Lane
fa046b6a92 Use RTLD_NOW, not RTLD_LAZY, as binding mode for dlopen() on all platforms.
This restores the Linux behavior to what it was in PG 7.0 and 7.1, and
causes other platforms to agree.  (Other well-tested platforms like HPUX
were doing it this way already.)  Per pghackers discussion over the past
month or so.
2002-02-12 23:41:25 +00:00