MariaDB Connector/ODBC 3.0.0 Alpha Changelog
Release date: 19 Jan 2017
For the highlights of this release, see the release notes.
The revision number links will take you to the revision's page on GitHub. On GitHub you can view more details of the revision and view diffs of the code modified in that revision.
- Revision #db72181
2017-01-17 23:46:20 +0100
- Fix of build problems on rhel5.
- Revision #b2f32ca
2017-01-17 00:46:27 +0100
- MADB_OPT_FLAG_PAD_SPACE was doing what MADB_OPT_FLAG_IGNORE_SPACE is supposed to do. Fixed that + in the setup dialog. Changed FindMariaDB to skip search if directory values are set (in command line parameters).
- Revision #766e375
2017-01-13 18:21:55 +0100
- Updated dll version to 3.0.0.
- Revision #1bf80e2
2017-01-08 23:47:41 +0100
- Added one more native error (specific to mariadb_stmt_execute_direct) to be treated as "lost connection" Changed CMakeLists to configure odbc*.ini only if tests directory is present.
- Revision #0542505
2016-12-02 14:11:54 +0100
- Slightly refined multi-statement query detection - it now allows having semicolons at the end of a single statement. Reset rs columns number before preparing query. Changed a few testcases in error.c to make output more informative
- Revision #02bef5f
2016-12-01 17:31:27 +0100
- mariadb_stmt_execute_direct is not used in case of params array, since it even doesn't make sense, not talking about problems it causes. Once again fixed code around "where current of" execution - index fields count now added to ParamCount, and deducted after execution (in case of execute_direct). Simpler and no problems with prepare+execute or older servers.
- Revision #d777bf8
2016-12-01 01:50:21 +0100
- Some more fixes around "WHERE COUNT OF". Fixed params allocation and works in conjunction with DAE
- Revision #81d5e44
2016-11-30 16:55:54 +0100
- Fixed "where current of" operation - there was error in total parameters count calculation. stmt is now closed and re-ignited before new Prepare on the same handle. Number of pre-bound parameters is set even if it is 0. ParamCount type is changed to SMALLINT.
- Revision #e4ac5ec
2016-11-24 00:56:00 +0100
- Removed couple of not really needed intermediate functions called by SQLFetch and SQLFetchScroll.
- Revision #4242e3c
2016-11-22 11:10:45 +0100
- Changed max->MAX, min->MIN. Added more data to generated DSN config in odbc.ini. Added cmake macro to easier populate values (for ini generations so far).
- Revision #7d73e86
2016-11-21 17:52:19 +0100
- Added ini files generation for testing with UnixODBC Set correct version.
- Revision #7cd56ca
2016-11-17 15:01:39 +0100
- As ma_errormsg.h has been renamed in C/C back to errmsg.h, had to change its inclusion.
- Revision #fbc65f3
2016-11-17 00:30:07 +0200
- Some more fixes for build on *nix. Mainly addint defines for TRUE and FALSE does that. Also had to add dl to tests linking list in dirty way. Fixed warnings.
- Revision #c5af166
2016-11-16 22:41:01 +0100
- Copied code for LIST, DYNAMIC_ARRAY, DYNAMIC_STRING from C/C to enable build on *nix, there my_global.h and my_sys.h are not installed any more, and to eventually make dynamic linking of C/C possible. All types and functions naming patterns have been changed. Also copied some macros, defined in aforementioned headers, and used in either c/odbc or in newly copied code.
- Revision #e58af25
2016-11-14 20:52:19 +0100
- Support of mariadb_stmt_execute_direct in first approach(ODBC-63).
- Revision #ded6696 2016-10-28 00:32:38 +0200
- Revision #572bdb1
2016-10-27 19:44:59 +0200
- Fix and testcase for ODBC-57. The problem was that MS Access adds parenthesis around each SELECT in the UNION. And the function determining query type wasn't ready for that. Now it skips query string characters till first alpha.
- Revision #3f1bb8b
2016-10-21 18:55:37 +0200
- Fix of problems with cursors and positional operations. Many things were relying on internal C/C stuff; that has been changed in C/C 3.0. That code had to be refactored. Removed unused fields from one of descriptor structures.
- Revision #db232db
2016-10-13 19:31:18 +0200
- Enabling build against C/C 3.0
- Revision #43c8d5a
2016-08-16 01:02:24 +0200
- Fixed some bugs in SQLSetPos with SQL_COLUMN_IGNORE - we had failing test in cursor.c Also fixed error it returns if there were errors on some of the updated rows.
- Revision #48bc3e1
2016-10-15 00:31:19 +0200
- mysql_stmt_bind_param instead of direct copying of BIND structures to stmt handle property. mysql_stmt_row_tell/row_seek instead of direc MYSQL_ROWS pointer manipulation. Removed couple of TODOs (simple or done). Fixed a bit my_dynamic_pos_cursor1 test.
- Revision #0c101a2
2016-10-14 13:22:03 +0200
- Changed MADB_CALLOC to follow old ma_malloc (MY_ZEROFILL)) behavior in case of 0 length. Solves many problems comped to simple calloc.
- Revision #609489d
2016-10-14 11:13:54 +0200
- Fix of crashes caused by relying on C/C internal features, changed in 3.0. In this case that was setting length_value of param bind structure, and not letting length buffer pointer. Older C/C does that, if length is NULL. Changed some helper functions parameters to "const" to avoid warnings, and just because that is right.
- Revision #9eab0b8
2016-10-13 19:31:18 +0200
- Enabling build against C/C 3.0.
- Revision #774a5a9
2016-04-29 14:00:01 +0200
- Build with C/C 3.0 enabler.