- 23 Jul, 2004 6 commits
-
-
unknown authored
-
unknown authored
merge sql/ha_ndbcluster.cc: merge
-
unknown authored
sql/ha_ndbcluster.h: Auto merged sql/ha_ndbcluster.cc: merge
-
unknown authored
ndb_supported_type, include all types in switch sql/ha_ndbcluster.cc: ndb_supported_type, include all types in switch
-
unknown authored
-
unknown authored
-
- 22 Jul, 2004 4 commits
- 21 Jul, 2004 2 commits
- 20 Jul, 2004 3 commits
- 19 Jul, 2004 3 commits
- 16 Jul, 2004 8 commits
-
-
unknown authored
into sanja.is.com.ua:/home/bell/mysql/bk/work-subs-4.1 sql/item_subselect.cc: Auto merged
-
unknown authored
-
unknown authored
into sanja.is.com.ua:/home/bell/mysql/bk/work-4.1
-
unknown authored
into sanja.is.com.ua:/home/bell/mysql/bk/work-subs-4.1
-
unknown authored
into sanja.is.com.ua:/home/bell/mysql/bk/work-4.1
-
unknown authored
into sanja.is.com.ua:/home/bell/mysql/bk/work-subs-4.1
-
unknown authored
into serg.mylan:/usr/home/serg/Abk/mysql-4.1
-
unknown authored
-
- 15 Jul, 2004 4 commits
-
-
unknown authored
into mysql.com:/space/my/mysql-4.1
-
unknown authored
into mysql.com:/space/my/mysql-4.1-build
-
unknown authored
when copying a small index file because the value returned for $length is < 1024. This can happen if the filehandle was open()ed as an UTF-8 encoded file with Unicode characters (In this case read() returns characters not bytes) (Thanks to Mike Bethune) for this hint)
-
unknown authored
-
- 14 Jul, 2004 3 commits
- 13 Jul, 2004 7 commits
-
-
unknown authored
Mac OS X: the name of the startup script itself must match the name of the subdirectory it's located in. Changed MySQL->MySQLCOM in the Do-pkg script and renamed the file in BK. (Thanks to Bryan McCormack for reporting this) Build-tools/Do-pkg: - renamed script from MySQL to MySQLCOM to match the name of the directory it's installed in support-files/MacOSX/MySQLCOM: - Fixed the file and directory name to match the actual location
-
unknown authored
into mysql.com:/home/mysql_src/mysql-4.1-874
-
unknown authored
by KILL or shutdown, do not mark the table as corrupted". It is indeed more logical to leave the corruption flag unchanged. This cannot be extended to REPAIR/OPTIMIZE as they make no backup copy of the MYI. This patch was tested with KILL and mysqladmin shutdown while a CHECK TABLE was running. Without the patch, the table becomes unusable (can't INSERT to it, error 145). With the patch, no. sql/ha_myisam.cc: When CHECK TABLE or ANALYZE TABLE is killed by KILL, do not mark the table corrupted; leave its corruption flag unchanged.
-
unknown authored
linking with PHP5) (thanks to Georg Richter for the patch)
-
unknown authored
into serg.mylan:/usr/home/serg/Abk/mysql-4.1
-
unknown authored
forced X/Open mode on HPUX removed. The correct way (that affects not only C but also C++ and linking) would be CFLAGS="-D_XOPEN_SOURCE -D_XOPEN_SOURCE_EXTENDED -D_HPUX_SOURCE" CXXFLAGS="-D_XOPEN_SOURCE -D_XOPEN_SOURCE_EXTENDED -D_HPUX_SOURCE" LDFLAGS="-lxnet" but apparently MySQL works without too.
-
unknown authored
bug#4527
-