Now
MAIN commitmail json YAML
pkgsrc/databases/libzdb/Makefile@1.30
/
diff
pkgsrc/databases/libzdb/distinfo@1.11 / diff
pkgsrc/databases/libzdb/patches/patch-src_system_Time.c@1.2 / diff
pkgsrc/databases/libzdb/distinfo@1.11 / diff
pkgsrc/databases/libzdb/patches/patch-src_system_Time.c@1.2 / diff
Update databases/libzdb to version 3.2.3.
Pkgsrc changes:
* More patching of isxxxx() uses to u_char.
Upstream changes:
Version 3.2.3
-------------
* New: Provide better error reporting if a Connection cannot be
obtained from the Connection Pool by introducing the method
ConnectionPool_getConnectionOrException(). In C++ the exception
thrown now contains the actual cause instead of a generic error.
* Fix: In C++ guard against stopping the Connection Pool with active
Connections. I.e. all Connections should be returned to the pool
before explicitly attempting to stop the pool. This is due to how
C++ destructors are called at scope end; if there is a reference
to a live Connection object after the pool is stopped it might cause
a dealloc sequence problem.
* Fix: Issue #50 where unit test 5 could fail for MariaDB because
mysql_stmt_affected_rows() only returns actual affected rows, not
matched rows which MySQL does.
* Fix: plus other minor improvements and fixes
Pkgsrc changes:
* More patching of isxxxx() uses to u_char.
Upstream changes:
Version 3.2.3
-------------
* New: Provide better error reporting if a Connection cannot be
obtained from the Connection Pool by introducing the method
ConnectionPool_getConnectionOrException(). In C++ the exception
thrown now contains the actual cause instead of a generic error.
* Fix: In C++ guard against stopping the Connection Pool with active
Connections. I.e. all Connections should be returned to the pool
before explicitly attempting to stop the pool. This is due to how
C++ destructors are called at scope end; if there is a reference
to a live Connection object after the pool is stopped it might cause
a dealloc sequence problem.
* Fix: Issue #50 where unit test 5 could fail for MariaDB because
mysql_stmt_affected_rows() only returns actual affected rows, not
matched rows which MySQL does.
* Fix: plus other minor improvements and fixes