|
| 1 | +2001-04-06 05:23 inoue |
| 2 | + |
| 3 | + * src/interfaces/odbc/convert.c: 1) Fit the precision of floating |
| 4 | + point to that of server side (fix by Hiroki kataoka). 2) Ensure |
| 5 | + the definition of atof()(#include stdlib.h). (suggestion by |
| 6 | + Masaaki Sakaida). |
| 7 | + |
| 8 | +2001-04-06 01:52 thomas |
| 9 | + |
| 10 | + * doc/src/sgml/installation.sgml: Final (?) platform update for the |
| 11 | + 7.1 release. Includes new platform NetBSD/PPC from Henry Hotz, |
| 12 | + several other platform updates, and move QNX to the unsupported |
| 13 | + list for now. |
| 14 | + |
| 15 | +2001-04-06 01:50 thomas |
| 16 | + |
| 17 | + * src/test/regress/: expected/horology-no-DST-before-1970.out, |
| 18 | + expected/horology-solaris-1947.out, expected/horology.out, |
| 19 | + sql/horology.sql: Alter the previous test for "time with time zone" |
| 20 | + implicit time zone (which failed miserably during DST) to just |
| 21 | + compare against a time derived from a timestamp value. Certainly |
| 22 | + not a direct check for a correct result, but should work at any |
| 23 | + time of year. |
| 24 | + |
| 25 | +2001-04-05 22:06 tgl |
| 26 | + |
| 27 | + * src/pl/plpgsql/src/pl_comp.c: Since plpgsql keeps its compiled |
| 28 | + function parsetrees until backend exit, it needs to ensure that |
| 29 | + data structures attached to fmgr info records in the trees will |
| 30 | + stick around that long, too. Current code was crashing on cases |
| 31 | + like datatypes with old-style I/O functions. |
| 32 | + |
| 33 | +2001-04-05 12:55 tgl |
| 34 | + |
| 35 | + * src/backend/access/transam/xlog.c: Fix comment that Vadim found |
| 36 | + confusing. |
| 37 | + |
| 38 | +2001-04-05 05:34 vadim |
| 39 | + |
| 40 | + * src/backend/access/transam/xlog.c: StartupXLOG(): initialize |
| 41 | + XLogCtl->Insert to new page if there is no room for a record on |
| 42 | + last log page. |
| 43 | + |
| 44 | +2001-04-05 04:21 meskes |
| 45 | + |
| 46 | + * src/interfaces/ecpg/: ChangeLog, lib/typename.c, |
| 47 | + preproc/preproc.y: Hopefully fixed the long long problem. |
| 48 | + |
| 49 | +2001-04-04 22:50 tgl |
| 50 | + |
| 51 | + * src/bin/pg_dump/pg_dump.c: pg_dump needs to use ONLY when |
| 52 | + selecting data from user tables, in case they have children. As it |
| 53 | + stood, -d dump mode was badly broken. |
| 54 | + |
| 55 | +2001-04-04 18:57 tgl |
| 56 | + |
| 57 | + * src/makefiles/Makefile.aix: Clean up AIX makefile to recognize |
| 58 | + subversions of AIX 4.1 correctly. This patch is Peter E's, but I'm |
| 59 | + applying it on his behalf so it'll get into 7.1RC3 tonight (I'm |
| 60 | + guessing Peter went to bed already). |
| 61 | + |
| 62 | +2001-04-04 17:15 tgl |
| 63 | + |
| 64 | + * contrib/spi/Makefile, src/backend/Makefile, |
| 65 | + src/makefiles/Makefile.win, src/pl/plpgsql/src/Makefile, |
| 66 | + src/test/regress/GNUmakefile: Fix/cleanup DLLLIBS handling for |
| 67 | + Cygwin port. If it works it's to Jason Tishler's credit, if it's |
| 68 | + broken it's my fault ... |
| 69 | + |
| 70 | +2001-04-04 16:03 tgl |
| 71 | + |
| 72 | + * src/test/regress/: expected/horology-no-DST-before-1970.out, |
| 73 | + expected/horology-solaris-1947.out, expected/horology.out, |
| 74 | + sql/horology.sql: Remove daylight-savings-time-dependent tests of |
| 75 | + 'time with time zone' datatype. Perhaps someday we can figure out |
| 76 | + a way of getting reproducible results from testing this type, but |
| 77 | + for now ... |
| 78 | + |
| 79 | +2001-04-04 16:02 petere |
| 80 | + |
| 81 | + * doc/FAQ_Solaris: update |
| 82 | + |
| 83 | +2001-04-04 11:43 vadim |
| 84 | + |
| 85 | + * src/backend/commands/sequence.c: Have to lock buffer while |
| 86 | + changing page' LSN. Thanks to Tom. |
| 87 | + |
| 88 | +2001-04-04 02:47 pjw |
| 89 | + |
| 90 | + * src/bin/pg_dump/: pg_backup_archiver.h, pg_dump.c: Don't dump |
| 91 | + CHECK constraints with same source and names both starting with |
| 92 | + '$'. |
| 93 | + |
| 94 | +2001-04-03 17:58 vadim |
| 95 | + |
| 96 | + * src/backend/commands/sequence.c: Log sequence creation (to |
| 97 | + initialize magic number on recovery). |
| 98 | + |
| 99 | +2001-04-03 15:06 tgl |
| 100 | + |
| 101 | + * contrib/pg_controldata/Makefile: Provide snprintf() if system |
| 102 | + hasn't got it. |
| 103 | + |
| 104 | +2001-04-03 15:01 tgl |
| 105 | + |
| 106 | + * contrib/pg_resetxlog/Makefile: Provide snprintf() if system |
| 107 | + hasn't got it. |
| 108 | + |
| 109 | +2001-04-03 14:14 momjian |
| 110 | + |
| 111 | + * doc/FAQ_MSWIN: See attached for a minor documentation patch that |
| 112 | + more accurately reflects the current state of and the issues with |
| 113 | + PostgreSQL on Windows. |
| 114 | + |
| 115 | + Thanks, Jason |
| 116 | + |
| 117 | + -- Jason Tishler |
| 118 | + |
| 119 | +2001-04-03 14:05 tgl |
| 120 | + |
| 121 | + * src/backend/utils/adt/: datetime.c, timestamp.c: Correct bogus |
| 122 | + mktime() calls. |
| 123 | + |
| 124 | +2001-04-03 09:38 momjian |
| 125 | + |
| 126 | + * doc/src/sgml/release.sgml: Remove last-updated date stamp for |
| 127 | + 7.1. |
| 128 | + |
| 129 | +2001-04-03 04:52 pjw |
| 130 | + |
| 131 | + * src/bin/pg_dump/: common.c, pg_backup_archiver.h, pg_dump.c, |
| 132 | + pg_dump.h: - Handle overridden attrs as per discussions 2-Apr-2001 |
| 133 | + - Dump CHECK constraints in OID order |
| 134 | + |
| 135 | +2001-04-03 00:07 tgl |
| 136 | + |
| 137 | + * src/backend/storage/file/fd.c: Check for failure of malloc() and |
| 138 | + realloc() when allocating space for VFD entries. On platforms |
| 139 | + where dereferencing a null pointer doesn't lead to coredump, it's |
| 140 | + possible that this omission could have led to unpleasant behavior |
| 141 | + like deleting the wrong file. |
| 142 | + |
| 143 | +2001-04-02 22:31 tgl |
| 144 | + |
| 145 | + * src/backend/storage/file/fd.c: open(2) flags saved for re-opening |
| 146 | + a virtual file should probably not include O_CREAT. |
| 147 | + |
| 148 | +2001-04-02 19:30 tgl |
| 149 | + |
| 150 | + * src/backend/utils/cache/relcache.c: Restore pre-7.1 behavior of |
| 151 | + allowing DROP of a table whose underlying physical file has |
| 152 | + disappeared. There is no really good reason why relcache should be |
| 153 | + opening the underlying file at all, AFAICS. In any case we needn't |
| 154 | + raise a hard error here. |
| 155 | + |
| 156 | +2001-04-02 19:20 tgl |
| 157 | + |
| 158 | + * src/backend/storage/smgr/md.c: _mdfd_getrelnfd() should include |
| 159 | + kernel error code in failure message. |
| 160 | + |
| 161 | +2001-04-02 14:30 tgl |
| 162 | + |
| 163 | + * src/backend/commands/creatinh.c: Complain if the same column name |
| 164 | + is inherited from multiple parents with different default values, |
| 165 | + unless the child table redeclares the column with an explicit |
| 166 | + default. This was judged to be the approach least likely to cause |
| 167 | + unpleasant surprises. |
| 168 | + |
| 169 | +2001-04-02 13:59 momjian |
| 170 | + |
| 171 | + * HISTORY, doc/src/sgml/release.sgml: Update HISTORY/release.sgml |
| 172 | + to current. |
| 173 | + |
| 174 | +2001-04-02 11:19 thomas |
| 175 | + |
| 176 | + * doc/src/sgml/installation.sgml: Update info for platforms BeOS, |
| 177 | + Tru64, Linux/MIPS, Linux/x86, NetBSD/Alpha, NetBSD/arm, |
| 178 | + NetBSD/VAX, OpenBSD/Sparc, Solaris, Ultrix. |
| 179 | + |
| 180 | +2001-04-02 10:34 momjian |
| 181 | + |
| 182 | + * src/backend/catalog/index.c: Disable creation of indexes on |
| 183 | + system tables. |
| 184 | + |
| 185 | +2001-04-02 04:17 meskes |
| 186 | + |
| 187 | + * src/interfaces/ecpg/: ChangeLog, preproc/pgc.l: Synced pgc.l with |
| 188 | + scan.l. |
| 189 | + |
| 190 | +2001-04-01 23:21 tgl |
| 191 | + |
| 192 | + * src/backend/Makefile: Remove duplicate definition of DLLLIBS, per |
| 193 | + gripe from Fred Yankowski. |
| 194 | + |
| 195 | +2001-04-01 23:08 tgl |
| 196 | + |
| 197 | + * src/backend/port/dynloader/: freebsd.c, netbsd.c, openbsd.c: Tom |
| 198 | + Ivar Helbekkmo <tih@kpnQwest.no> says that dynamic linking works |
| 199 | + now on VAX NetBSD. Hence, remove #ifdef'd disable tests. |
| 200 | + |
| 201 | +2001-04-01 18:37 tgl |
| 202 | + |
| 203 | + * src/backend/optimizer/plan/planner.c: Planner wasn't correctly |
| 204 | + handling adjustment of tuple_fraction for the case of LIMIT in a |
| 205 | + sub-select. |
| 206 | + |
| 207 | +2001-04-01 15:15 scrappy |
| 208 | + |
| 209 | + * configure, configure.in, ChangeLogs/ChangeLog-7.1RC1-to-7.1RC2: |
| 210 | + RC2 preparations ... |
| 211 | + |
0 commit comments