Changes

Revision history for DBD::Oracle

{{$NEXT}}

  Test updates (GH#193, Victor)
  Fix up unit tests (Dean Hamstead)
  Remove outdated ifdefs (GH#186, H.Merijn Brand)
  Attempts to resolve SEGV issue (GH#187, Victor)
  Various clean ups and typos (GH#187, Victor)
  Makefile.PL fix for missing Test::NoWarnings build requirement (GH#175, Wesley Hinds)
  don't care about ExtUtils::MakeMaker version number anymore (GH#184, Wesley Hinds)
  Makefile.PL: help to find linking libraries on Gentoo (GH#184, Marco Genasci)

1.90    2024-04-16

  Specify resources metadata explicitly in dist.ini (GH#162, Graham Knop)
  Ensure dbivport.h is installed- (GH#150, Wesley Hinds)
  Add DBI to ConfigureRequires, BuildRequires, and Test Requires - (GH#155)
  Move from TravisCI to Github Actions (GHA) - (PR#152, Wesley Hinds)
  Update dbdcnx.c - (PR#152, Andrei Voropaev)
  Rewrite of login6. This is an enormous rewrite which should fix many problems - (PR#150, Andrei Voropaev)
  Improvements to Makefile.PL - (PR#150, Andrei Voropaev)
  Have find_headers() also search for header files based off the major version number - (GH#142, Wesley Hinds)
  Check for appropriate permissions before running 56embbeded.t tests - (#143, Wesley Hinds)
  Updated links in doc in Oracle.pm - (Gh#145, kjetillll)
  Update 25plsql.t and add 21.5 to @bad_oci_vers - (GH#144, kjetillll)
  Check permissions for tests in 28array_bind.t - (GH#141, Wesley Hinds)
  Exclude 19.9 and 19.13 from 25plsql.t - (GH#140, Wesley Hinds)

1.83

  Build improvements on Debian-ish systems (GH#112, Alex Muntada)
  Add rpath to ORACLE_HOME to DBD/Oracle/Oracle.bundle (GH#129, hackonhagland).

1.82      2021-12-29

  Second try at github #130.  Still not fully cooperating but better.

1.81      2021-12-24

  [BUG FIXES]

  Run nonassigned pl/sql tests based off versions known to have the issue. - (GH#70, Wesley Hinds)
  Oracle Instant Client 21 support - (Tux)
  Add rpath to ORACLE_HOME to DBD/Oracle/Oracle.bundle  - (GH#128, hakonhagland)
  fix materialised views being misclassified as tables in ->table_info - (GH#132, John Smith)
  Fix bugtracker ref in dist metadata github - (GH#130, John Smith)

  [MISCELLANEOUS]

  Updates to POD and README file
  Moved old README file to t/TESTING.md
  Updated links

1.80      2019-07-25

  [BUG FIXES]

  orphaned test code in t/28_array_bind.t - (GH#64, dzort)
  Made the code run so it is nolonger orphaned

  [MISCELLANEOUS]

  Updated bugtracker to git
  updated home page to metacpan

1.791     2019-07-22

  [BUG FIXES]

  Fix Avoid GCC-ism so that HP-UX can compile (GH#92, Dean Hamstead)

  Destroy envhp with last dbh (GH#93, GH#89, Dean Hamstead, CarstenGrohmann)


1.76      2018-11-22

  No Changes from 1.75, we just botched up publishing to CPAN

1.75   2018-11-22

  No Changes from 1.75_42

1.75_42   2018-08-22

  [BUG FIXES]

  Fix potential buffer overflow in dbdimp.c - (GH#57, Various)

  Fix truncation error on ROWIDs from an Index Organized table as they are
  not a fixed length. Code now allows up to a size of 2000.
  (GH#31, Martin J. Evans)

  Various fixes for compiler warnings, OCI handle leaks, and OCI programming
  errors. - (PR#38, Dag Lem)

  Corrections to t/25plsql.t - (GH#56, kjetillll)

  Fix: Invalid binding call for large undef arrays. - (GH#36, GH ghost)

  Fix: compile warnings about int vs long unsigned. - (PR#62, Dean Hamstead)

  Fix: Spelling errors in pod. - (PR#63, Jochen Hayek)

  Fix: Various changes in pod. - (PR#54, Mike O'Regan)

  [ENHANCEMENT]

  Connection informational messages like "ORA-28002: the password will
  expire" were lost. Thanks to J.D. Laub.

  Add new path to find 64 bit Oracle client on MAC OSX - (GH#20, Martin J. Evans)

  Ignore constraints which are not enabled in primary/foreign key_info
  (GH#23, Martin J. Evans)

  dist.init overhaul. - (PR#62, Dean Hamstead)

  Mailmap and TODO changes. - (PR#62, Dean Hamstead)

  Travis CI testing. - (PR#62, Dean Hamstead)
  Note: this uses Oracle XE which doesn't provide enough features to test
  the entire suite. Nor does it test anything other than Linux on Linux.
  Release tests are also NOT yet run in Travis.

  Rewrite of DRCP session pooling to make it work as intended - (PR#38, Dag Lem)

  Support for Oracle Fast Application Notification (FAN). - (PR#38, Dag Lem)

  Work by Dag Lem was sponsored by EVRY Information Services. Thank you!

1.75_2    2014-11-19

  [ENHANCEMENT]

  Try and set -l when the build would have failed (H.Merijn Brand)

1.75_1    2014-11-17

  [DOCUMENTATION]

  Change mentions of READMEs in Makefile.PL to troubleshooting guides.
  (GH#17, reported by Ken Williams)

  [BUG FIXES]

  Fix GH#15 and GH#21 (the same problem). Previous change for
  RT91698 broke other things in output parameters.

1.74      2014-04-24
  - Promote to stable.

1.73_01   2014-04-23
  - Tweak fix for RT-88185. (GH#14, Martin J. Evans)

1.73_00   2014-04-23
  - Reverts current fix for RT-88185, as it causes breakage. (GH#14)

1.72      2014-04-14
  - promote 1.71_00 to stable.

1.71_00   2014-03-31
    - Recognizes __CYGWIN64__. (RT88709, reported by Witold Petriczek)
    - CHOOSE hint is deprecated. (RT91217, reported by Andy Bucksch,
        fix by Martin J Evans)
    - Set UTF8 flag per-connection. (RT88185, reported by Heinrich Mislik, patch by Martin
      J. Evans)
    - Add a CONTRIBUTING.mkd file. (GH#2)
    - Add SELinux trick. (RT#87003, patch submitted by Mike Doherty)

1.70   2014-02-12
  - promote 1.69_02 to stable.

1.69_02 2014-01-19

    [IMPROVEMENTS]

    - The DSN 'dbi:Oracle:sid=foo' is now an alias for 'dbi:Oracle:foo'.
        (RT#91775, Yanick Champoux, requested by David Wheeler)

    - Support for ORA_SYSBACKUP, ORA_SYSDG and ORA_SYSKM. (RT#91473,
        Kris Lemaire)

    [BUG FIXES]

   - OCI_THREADED setting had been accidentally removed, causing potential
    crashes when using threads. (RT#92229, Martin J. Evans, reported
    by Detlef Lütticke)

  - When using fetch*_hashref the values are decoded but
    not the keys so if you have unicode column names they were not
    returned correctly.  (RT#92134, Martin J. Evans, reported by
    Marcel Montes)


1.69_01 2014-01-14

  [BUG FIXES]

  - Fix RT91698. If you bound an output parameter to a scalar and
    repeatedly called execute the memory allocated in your bound
    scalar could increase each time. (Martin J. Evans)

1.68      2013-11-25
  - promote 1.67_00 to stable.

1.67_00   2013-11-05

  [BUG FIXES]
  - Fix RT88135. Add statistics_info support (patch by Steffen Goeldner)
  - Fix RT89491. Add RULE hint (patch by Steffen Goeldner)

  [DOCUMENTATION]
  - POD typos (RT#88285, RT#88284, Gregor Herrman).
  - Grooming of Hpux troubleshooting pod (GH#7, Martin J. Evans,
    Yanick Champoux)

1.66      2013-08-23
  - promote 1.65_00 to stable.

1.65_00   2013-07-29

  [BUG FIXES]

  - Fix RT85886. The TYPE passed to bind_col is supposed to be sticky
  and it wasn't. Attributes passed to bind_col could be lost later if
  bind_col is called again without attributes. Both of these occur
  when fetchall_arrayref is called with a slice (Martin J. Evans).

  [DOCUMENTATION]

  - Fix a bunch of typos. [GH#5, David Steinbrunner]

1.64      2013-05-22
  - promote 1.63_00 to stable.

1.63_00   2013-05-03

  [ENHANCEMENTS]
  - DBD-Oracle: Use of uninitialized value $user_only in uc [RT#84657]
    (Steffen Goeldner)

  [BUG FIXES]
  - Make 50cursor.t Oracle8-friendly. (RT#84660, patch by Steffen Goeldner)
  - Makefile.PL's use of ACL tweaked for Suse Enterprise 11 SP2
        (RT#84530, patch by Alfred Nathaniel)

  [DOCUMENTATION]
  - Bogus 227 directory no longer required for MacOS. (GH#1, patch
    by theory)

1.62      2013-04-30
  - promote 1.61_00 to official release

1.61_00   2013-04-15

  [BUG FIXES]
  - Adjust the privs needed for the DROP/CREATE table test. [GH#35]
    (Joe Crotty)

  - Fixed RT84170 - when using scrollable cursors and you've done a
    positioned fetch and then keep fetching until the end of the
    result-set calls to fetch never return undef and you keep getting
    the last row forever. Also added test case to the 51scroll.t test
    (Martin J. Evans).

1.60      2013-04-01
 - Move github repository to github.com/pythian/DBD-Oracle.

1.58      2013-03-05
  - promote 1.57_00 to official release

1.57_00   2013-02-07
  [BUG FIXES]

  - fix RT46628 - bind_param_inout ORA_RSET causes MSWin32 access
    violation and RT82663 - Errors if a returned SYS_REFCURSOR is not
    opened (Martin J. Evans)

  - Fix RT82663. If a procedure/function returns a SYS_REFCURSOR which
    was never opened DBD::Oracle magics a DBI statement handle into
    existence and attempts to describe it (which fails). This change
    examines a returned SYS_REFCURSOR and if it it is initialised but
    not executed does not create a DBI statement handle and returns
    undef instead. So now if you have a procedure/function which
    returns a SYS_REFCURSOR and never open it you'll get undef back
    instead of a useless statement handle.  Extended 50cursor.t test
    to check the above fix. (Martin J. Evans)

  [DOCUMENTATION]
  - Update Lion instructions for 64-bit Instant Client. (GH#37, patch by
    theory)

1.56      2013-01-08
  - fix t/26exe_array.t in the case of no db connection (RT82506,
    reported by Peter Rabbitson)

1.54      2013-01-03
 - promote 1.53_00 to official release

1.53_00      2012-12-18

  [BUG FIXES]
  - Fix RT69350 - 31lob.t was using $lob after destroying its parent $sth
        (Rob Davies)

  - Fix memory leak in execute_array (John Scoles, Pierre-Alain Blanc)

  - Fix RT80349 - The error message in execute_for_fetch when a row fails
    can contain the wrong error count. Thanks to Steffen Goeldner for
    RT and patch.

  - Fix RT80375 - no exception when execute_for_fetch fails and
    ArrayTupleStatus is not specified. Also the tuple count calculation
    resulted in an undefined warning. Thanks to Steffen Goeldner for
    RT and patch.

  - Fix RT80487. Skip XMLType tests if Oracle less than V9. Thanks to
    Steffen Goeldner for RT and patch.

  - Fix RT80486 for 31lob_extended.t. In old old Oracle8,
    SYS_REFCURSOR is not defined. Instead of CREATE/DROP PROCEDURE,
    use anonymous block. Thanks to Steffen Goeldner for RT and patch.

  - Fix bug in 39attr.t which could fail if using an Oracle Client >
    11 but not >= 11.2 (Martin J. Evans)

  - ora_server_version was not documented.

  - Fix RT80566. 70meta.t test fails with Oracle 8 because
    ALL_TAB_COLUMNS.CHAR_LENGTH is new in Oracle 9. Use DATA_LENGTH
    instead on pre-9 versions. Thanks to Steffen Goeldner for RT and
    patch.

  - Fix RT80704. 51scroll.t test checks scrollable cursors but assumes
    all Oracles support them (only 9 and above). Thanks to Steffen
    Goeldner for RT and patch.

  - Fix RT81067. 58object.t has some subtype tests and subtypes were
    introduced in Oracle 9. Skip if < Oracle 9. Thanks to Steffen
    Goeldner for RT and patch.

  - Fix RT81317. 34pres_lobs.t uses the Data Interface for Persistent
    LOBs which is new in Oracle 9. Skip if < Oracle 9. Thanks to
    Steffen Goeldner for RT and patch.

  [MISCELLANEOUS]

  - The original 26exe_array test was replaced some time ago with a
    copy of the one from DBD::ODBC. Since then I've fixed issues in
    the DBD::ODBC one and added more tests (like tests for some RTs
    above). To make keeping them in synch easier I've modularised the
    tests. Hence new ExecuteArray.pm. (Martin J. Evans)

  - simple code clean up, replacing 3 uses of safemalloc with Newz
    (John Scoles)

  - Add DBI as a configure prereq for the META* files (thanks to Joe Crotty)

  - New FAQ entry on Solaris and setting linker library path
    (Martin J. Evans)

  - Removed ineffective commit in 34pres_lobs.t (Martin J. Evans)

  - Remove dead README link in Win32 documentation. (pointed out by Alexandr
    Ciornii, RT#82040)

  - Changed any use of if SvUPGRADE to remove the if test as per email
    from Dave Mitchell and posting at
    http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2012-12/msg00424.html
    (Martin J. Evans)

1.52      2012-10-19

 - promote 1.51_00 to official release

1.51_00   2012-09-28

  [BUG FIXES]

  - fix serious memory corruption in TAF support (Martin J. Evans)

  - fix finding client in situation where client and server both
    installed but different architectures (patch by H.Merijn Brand)

  - fix memory leak in TAF handling - the TAF function was leaked
    (Martin J. Evans)

  - fix issue with taf_function being set to a scalar which goes
    out of scope before the callback is made (Martin J. Evans)

  - fix RT46739 if a connection breaks the environment handle is
    not thrown away (Martin J. Evans)

  - ora_driver_name was not defaulted to the correct DBD::Oracle
    version (Martin J. Evans)

  - ora_driver_name, ora_client_info, ora_client_identifier,
    ora_action and ora_oci_success_warn were set twice (if specified)
    on connect as they were not deleted from the connect attributes
    once handled. Code now leaves the setting to the later STORE DBI
    calls (Martin J. Evans)

  - fixed some compiler warnings for %lf (Martin J. Evans)

  - fixed RT78700 - column_info reports wrong size for char semantic
    char type columns (Douglas Wilson).

  [CHANGE IN BEHAVIOUR]

  - ora_taf and ora_taf_sleep were redundant and have been removed.
    To enable/disable TAF simply set ora_taf_function and if you
    want to sleep do it in your callback (Martin J. Evans)

  - ora_taf_function can now be a code reference as well as a string
    (Martin J. Evans)

  [ENHANCEMENTS]

  - the ora_can_taf method was virtually useless since you can only
    call it after connecting and to enable TAF you had to do it in the
    connect call. Now you can enable and disable TAF at any time by
    simply setting or clearing the ora_taf_function (see RT78811)
    (Martin J. Evans)

  - the ora_taf_function is now passed a third argument of the
    connection handle (Martin J. Evans)

  - RT78987 - removed Oraperl.pm and oraperl.ph; these files will be
    available in a separate distribution named "Oraperl" (David Perry)

  [MISCELLANEOUS]

  - hide dr, db and st packages from PAUSE (Martin J. Evans)

  - added a few more simple TAF tests (Martin J. Evans)

1.50      2012-08-15
 - RT78965 - Remove Oraperl tests (which were forcing a require on Oraperl)

1.48      2012-08-09
 - promote 1.47_00 to official release

1.47_00   2012-07-11

  [BUG FIXES]

  - fixed redeclaration of $len in 31lob.t - (Martin J, Evans)

  - RT55028 - stop segfaulting when attempting to read empty lobs
    (Martin J. Evans)

  - RT69059 - Despite OCIPing being documented as added in 10.2 AIX
    does not seem to have it in 10.2 leading to undefined symbol -
    Martin J. Evans

  [DOCUMENTATION]
  - Promoted the troubleshooting for the different architectures to
    POD documents, for easier/prettier access.

  - Added a troubleshooting entry for RT71819 - bound output
    parameters may be returned in the wrong order (Martin J. Evans)

1.46      2012-07-11
 - promote 1.45_00 to official release

1.45_00   2012-06-21

 [CHANGE IN BEHAVIOUR]

 - In future versions of DBD::Oracle ora_verbose will be changed
   so that it is simply a switch to turn DBI's DBD tracing on or off.
   A true value will turn it on and a false value will turn it off.
   DBI's "DBD" tracing was not available when ora_verbose was created
   and ora_verbose adds an additional test to every trace test.

 [BUG FIXES]

 - Fixed RT76695 - offset passed to ora_fetch_scroll should not affect
   normal fetches (Martin J. Evans)

 - Fixed RT76410 - fetch after fetch absolute always returns
   the same row (Martin J. Evans);

 - Fixed RT75721 - does not build with Oracle 9.2 (Martin J. Evans)

 - Fixed RT71343 - Oracle 9i does not have OCI_ATTR_TAF_ENABLED
   or OCI_ATTR_RESERVED_15/16 so cannot build (Martin J. Evans)

 - skip 24implicit_utf8.t if chr set is not UTF-8 (Martin J. Evans)

 - Fixed RT76269 - ora_taf_sleep was documented as taf_sleep by
   accident. There was no way to stop the TAF reconnect attempts.
   If you want to try another connect attempt in your taf handler you
   now need to return OCI_FO_RETRY from it. (Martin J. Evans)

 [MISCELLANEOUS]

 - minor change to confusing debug output for input parameters
   (Martin J. Evans)

 - RT72989 - add note to trouble shooting guide re this RT and
   Module::Runtime (Martin J. Evans)

1.44      2012-04-23
 - promote 1.43_00 to official release

1.43_00   2012-03-30

 [BUG FIXES]
 - Applied patch from Rafael Kitover (Caelum) to column_info to handle
   DEFAULT columns greater in length than the DBI default of 80. The
   DEFAULT column is a long and it is a PITA to have to set
   LongReadLen which you can only do on a connection handle in
   DBD::Oracle. The default maximum size is now 1Mb; above that you
   will still have to set LongReadLen (Martin J. Evans)

 - Fixed 70meta and rt74753-utf8-encoded to not die if you cannot
   connect to Oracle or you cannot install from CPAN if you have not
   set up a valid Oracle connection.

 - Fixed 75163. Bfile lobs were not being opened before fetching if
   ora_auto_lobs was disabled (Martin J. Evans).

   Note: this has a minor impact on non bfile lobs when ora_auto_lobs
   is not in force as an additional call to OCILobFileIsOpen will be
   made.

 - Removed all DBIS usage fixing and speeding up threaded
   Perls (Martin J. Evans).

 - Minor fix to avoid use of uninitialised variable in 31lob.t (Martin J. Evans)

 [DOCUMENTATION]
 - clarification of when StrictlyTyped/DiscardString can be used and
   LongReadLen (Martin J. Evans)

 - Documented the 3rd type of placeholder and rewrote the existing
   pod for placeholders (Martin J. Evans).

1.42      2012-03-13
 - skip rt74753-utf8-encoded.t if db is not unicode

1.40      2012-03-08
 - promote 1.39_00 to official release

1.39_00   2012-02-24

 [BUG FIXES]
 - TAF supports now conditional to presence of OCI_ATTR_TAF_ENABLED
    [RT73798]
 - detect broken Win32::TieRegistry (patch by Rafael Kitover (Caelum))
    [RT74544]
 - PL/SQL out values were not utf8 encoded [RT74753]
     (Steve Baldwin + Martin J. Evans)

 [DOCUMENTATION]
 - Mention the release of Oracle Instant Client 64 bit which does not work
    on Lion. (Martin J. Evans)
 - fix DBD::Oracle::GetInfo blurb (patch by Julián Moreno Patiño) [rt74000]
 - fix typos. (patch by Julián Moreno Patiño) [rt73999]
 - add troubleshoot doc and diag for error with bequeather. [rt75263]

 [OTHERS]
 - change the shebang line of examples to the more modern '/usr/bin/env perl'
    [RT74001]

1.38   2012-01-13

 - promote 1.37_00 to official release

1.37_00   2011-12-30

 [ENHANCEMENTS]
  - added SYSASM session mode. [RT651211] (patch from
  Anthony DeRobertis, reported by Julián Moreno Patiño)

 [BUG FIXES]

 - applied patch from Charles Jardine avoiding undefined values
   warnings in ora_server_version when the database is not open
   [RT72623] (Martin J. Evans)
 - TNS_ADMIN was ignored [RT73456]

 [DOCUMENTATION]

 - Document possible problem with
   ora_connect_with_default_signals and connect_cached
   [RT72716] (Martin J. Evans)
 - Fix documentation for 'ora_fetch_scroll()'

Changes in DBD-Oracle 1.36 (6-12-2011)

 - promote 1.35_00 to official release

Changes in DBD-Oracle 1.35_00 (18-11-2011)

 [BUG FIXES]
 - if bind_col is called with a TYPE but no bind attributes like
   StrictlyTyped or DiscardString are set DBD::Oracle still attempts
   to call sql_type_cast which is pointless (Martin J. Evans)
 - if bind_col is called with a TYPE other than SQL_NUMERIC,
   SQL_INTEGER or SQL_DOUBLE and bind attributes like StrictlyTyped or
   DiscardString a warning was not issued that the type is unsupported
   and no data was returned (Martin J. Evans)
 - Fix test so it works with perl compiled with -Duselongdouble [RT71852]
 - Apply patch from Charles Jardine for better building against a full
   Oracle 11 install [RT72463] (Martin J. Evans)

 [DOCUMENTATION]
 - Added notes to bind_col documenting the fact that setting a TYPE
   does not affect how the column is bound in Oracle, only what
   happens after the column data is retrieved (Martin J. Evans)
 - fix typo (thanks to Julián Moreno Patiño) [RT72038]
 - shuffle POD around to improve documentation flow [RT72252]
 - major tidying up of the connect() documentation. (by Gwen Shapira)
 - Moved LONG examples out of POD and into examples/

 [OTHER]
 - Commented out some functions in oci8.c which were not used to
   reduce the size of the driver a little (Martin J. Evans)

Changes in DBD-Oracle 1.34 (31-10-2011)

 - promote 1.33_00 to official release

Changes in DBD-Oracle 1.33_00 (18-10-2011)

 [BUG FIXES]
 - COLUMN_SIZE of VARCHAR2 returns size in chars, not bytes. [RT#13865]
    (reported by Stefano and Laurent Dami)

 [DOCUMENTATION]
 - add mention of the github mirror of the subversion repository
 - add 'resources' info to META.yml
 - fixed broken link to Oracle DRCP doc in POD (John Scoles)

Changes in DBD-Oracle 1.32  (16-10-2011)

 - promote 1.31_00 to official release

Changes in DBD-Oracle 1.31_00

 [ENHANCEMENTS]
 - Makefile.PL's options are now documented
 - move 'explain' to '/examples' directory

 [BUG FIXES]
 - support development release versions in GetInfo (Martin J. Evans)
 - don't gag diag() on the tests by default
 - SKIP condition in 10general.t was reversed (reported by Alois) [RT#46761]
 - Check for LD_LIBRARY_PATH_(32|64) as well for solaris [RT#46761]
 - convert a symbolically linked ORACLE_HOME to an absolute path
    (patch by H.Merijn Brand, applied by Martin J. Evans) [rt70785]

 [DOCUMENTATION]
 - announce that Oraperl will be removed from the dist by v1.38.
 - add Mac OS instructions for Lion (patch by David Wheeler) [rt71109]
 - fixed minor typo in POD (John Scoles)

Changes in DBD-Oracle 1.30
  [DOCUMENTATION]
 - add warning about RT#69350 in documentation

Changes in DBD-Oracle 1.29_1
 [ENHANCEMENTS]
 - added support for TAF callback (John Scoles)
 - now trap OCIServerAttach errors (patch by Marc Fielding, applied
    by Martin J. Evans) [rt68958]
 - added installation notes for MAC Snow Leopard (Martin J. Evans)
 - added '/etc' to the search paths for tnsnames.ora (Martin J. Evans, Jay Senseman)
    [rt67942]
 - removed support for ProcC connections (John Scoles)
 - added ora_db_shutdown and ora_db_startup private functions (Steffen Goeldner)
 - added Test::Simple 0.90 to build_requires as we use note etc (Martin J.  Evans)

 [BUG FIXES]
 - fixed bug in bind_col which was broken in 1.27 and stopped anyone using
    a string as a column number. e.g., '1' (Martin J. Evans, Alexander
    Foken)
 - removed obsolete oparse_lang (John Scoles)
 - fixed some compiler warnings in dbdimp.c and oci8.c (Martin J. Evans)
 - added missing OCIServerRelease to oci.def (Martin J. Evans) [rt68172]
 - fixed up the POD based on DBD::Pg (John Scoles)
 - POD review rephrasing, fixing typos and spelling mistakes up to
    "Placeholder Binding Attributes" (Martin J. Evans)
 - rt 56824 - add META_MERGE mentioning DBI required version (Martin J. Evans)

* Changes in DBD-Oracle 1.28 (svn rev 14765)

  Changed 26exe_array.t so it is compatible with older version of Test::More by John Scoles
  A fix from Charles Jardine for 58object.t to stop an error on some 64 bit systems
  A few quick changes from H. Merijn Brand to fix some compiler warings and a fix for Oracle 9 clients
  Added connection attribute 'ora_connect_with_default_signals' that will localize Perl's $SIG{INT} handler from Brian Phillips and T. Bunce
  Fix in execute_array to stop possible endless loop when using a fetch sub by Martin J. Evans
  Adapted Martin J. Evans' ODBC 70execute_array.t into t/26exe_array.t by John Scoles
  Fix for execute_array to bring it up to spec. by Martin J. Evans and John Scoles
  Marked ProC, Oraperl.pm, ora_explain.pl, ora_context, ora_use_proc_connection and ora_parse_lang as deprecated  to be removed in 1.29
  Added in 4 new server side debug/trace attributes, ora_driver_name, ora_client_info, ora_session_user and ora_action on the connection handle from John Scoles
  Cleaned up the pod a little by John Scoles
  Fix for function name length, Some function names are over 31char long which may cause problems for some OS/Compilers (VMS IA64 box.) from Jakob Snoer
  Fix for OCIPing in case where a 10 client tries to ping a <10 DB  from Tim Oertel
  Fix for DBD-Oracle stored proc with array bug where second call array size is unchanged from Tim Oertel
  Fix for rt.cpan.org Ticket #=63332: Spelling error in POD from jonasbn
  Fix for rt.cpan.org Ticket #=62152: t/28array_bind.t and t/31lob.t may call plan() twice and others do not fail on not connect from John Scoles
  Fix for rt.cpan.org Ticket #=61511 ORA-00942 when inserting into a table with a LOB column over a synonym on HP-UX from Kris Lemaire
  Fix for rt.cpan.org Ticket #=42842 Test 31lob fails with 64-bit Instant Client by John Scoles
  Fix for support for objects on big endian platforms from Charles Jardine, John R Pierce
  Fix for rt.cpan.org Ticket #=61225 Windows install (Stawberry Perl) fails on long path names from David Tulloh
  Fix for rt.cpan.org Ticket #=rt64524 Memory Leak when Oracle connection fails by Martin J. Evans
  Added all the missing ora_drcp values to dbh private_attribute_info by Martin J. Evans
  Removed a load of attributes from sth private_attribute_info which are not handle attributes but attributes to bind_param/prepare by Martin J. Evans
  Fix for rt.cpan.org Ticket #=64244 - don't bail out, skip tests when we cannot connect by Martin J. Evans and John Scoles
  Added DBI to PREREQ_PM in Makefile.PL by Martin J. Evans
  Added build_requires in Makefile.PL by Martin J. Evans
  Added workaround for ExtUtils::MakeMaker problems by Martin J. Evans
  Added LICENSE to Makefile.PL by Martin J. Evans


* Changes in DBD-Oracle 1.27 (svn rev 14583)

  This version removes 'PERL_POLLUTE' and adds in PL_ where required so it will be fully compatible with Perl 5.13

* Changes in DBD-Oracle 1.26 (svn rev 14411)

  Manual re-release of 1.25 with changes to file permissions.

* Changes in DBD-Oracle 1.25(svn rev 14411)

  Added support for the OCIPing by John Scoles
  Spell checked the pod (the first time in a while me thinks) updated the todo By John Scoles
  Added support for DCRP (Database Resident Connection Pooling) by John Scoles with Luben Karavelov
  Fix for odd error with Ping from Tom Payerle
  Removed the NEW_OCI_INIT compile directive and the deprecated OCIInitialize calls
  Fix for rt.cpan.org Ticket #=57256 :  Double free problem in dbdimp.c by John Scoles
  Fix for invalid format in trace of OCILobLocatorIsInit_log_stat reported by Martin Evans Fixed by John Scoles
  Fix for very odd UNKNOWN OCI STATUS 1041 (OCILobFreeTemporary) on disconnect reported by John Parker and Bob Mcgowan fixed by John Scoles
  Fix for rt.cpan.org Ticket #=55445: get_info(28) SQL_IDENTIFIER_CASE seems to return the wrong value from Martin J Evans and a bunch of re jigging from John Scoles
  Patch for PL/SQL: numeric or value error: character string buffer too small from Scott T. Hildreth
  Fix for rt.cpan.org Ticket #=51594 type_info and type_info_all miss vital information from John Scoles
  Added ora_lob_is_init function by John Scoles
  Fix for rt.cpan.org Ticket #=55031 Ubuntu Server  Building with Oracle XE under 32-bit from Brian Candler
  Fix for rt.cpan.org Ticket #=56810 bug with multiple nested cursor from John Scoles
  Fix for bug found only on Big-Endian hardware reported by Timothy Everett and others from Charles Jardine
  Fix for memory leak when using prepared_cached and lobs reported by Mark Bobak and Martin Evans found and fixed by John Scoles and a test from Martin Evans
  Added more entries to the Readmes from John Scoles

* Changes in DBD-Oracle 1.24_01(svn rev 14060)

  This release has been prepared specifically for the 'Debian' http://www.debian.org project.  It contains no changes
  to functionality or usage.  The following has been changed

  Fixed some formatting and typos in Pod from Julián Patiño
  The Copyright terms for ora_explain have changed and now read as follows:
    You may distribute under the terms of either the GNU General Public
    License or the Artistic License, as specified in the Perl README file.

* Changes in DBD-Oracle 1.24(svn rev 13793)

  Extended precision for OCIDateTimeToText to 6 instead of 0 for embedded types from John Scoles
  Extended support of Oracle Embedded objects from Charles Jardine
  Added support for RowsInCache as RO and RowCacheSize as a set-able value on the Statement Handle. So it would comply with DBI spec By John Scoles with thanks to Martin J. Evans
  Added extended support for 64 bit clients in Makefile.PL from Ralph Doncaster
  Added extended nvarchar support from Jan Mach
  Added support for the TYPE attribute on bind_col and the new DBI bind_col attributes StrictlyTyped and DiscardString from Martin J. Evans
  Added ora_ncs_buff_mtpl and environment var ORA_DBD_NCS_BUFFER so we can control the size of the buffer when doing nclob reads
  Fix for rt.cpan.org Ticket #=49741 Oracle.h has commented out params in OCIXMLTypeCreateFromSrc from Kartik Thakore
  Added from rt.cpan.org Ticket #=49436 Patch to add support for a few Oracle data types to type_info_all from David Hull
  Added from rt.cpan.org Ticket #=49435 Patch to add support for a few Oracle data types to dbd_describe from David Hull
  Fix for rt.cpan.org Ticket #=49331 Bad code example in POD from John Scoles
  Added support for looking up OCI_DTYPE_PARAM Attributes
  Added support for looking up csform values
  Fix for rt.cpan.org Ticket #=46763,46998 enhancement -Rowcache size is now being properly implemented with row fetch buffer from John Scoles
  Fix for rt.cpan.org Ticket #=46438 enhancement -Errors returned by procedures are now unicode strings from Martin Evans, John Scoles and Tim Bunce
  Fix for rt.cpan.org Ticket #=47503 bugfix - using more than 1 LOB in insert broken from APLA
  Fix for rt.cpan.org Ticket #=46613 bugfix - sig-abort on nested objects with ora_objects=1 from TomasP
  Fix for rt.cpan.org Ticket #=46661 DBD::Oracle hangs when insert/update with LOB and quoted table name from APLA
  Fix for rt.cpan.org Ticket #=46246 fetching from nested cursor (returned from procedure) leads to application crash (abort) from John Scoles
  Fix for rt.cpan.org Ticket #=46016  LOBs bound with ora_field broken from RKITOVER
  Fix for bug in 58object.t when test run as externally identified user from Charles Jardine

* Changes in DBD-Oracle 1.23(svn rev 12724)

  Fix from rt.cpan.org ticket #=44788 bool in_lite should be char in_literal
  Fix for UTF8 and blobs by John Scoles with Milo van der Leij
  Fix for some warnings and one bug in ocitrace.h from Charles Jardine
  Fix in case there may be a bug in 11 where the OCI_ATTR_DATA_SIZE my return 0 which should never happen, John Scoles
  Fix on the Makefile.PL for a possible sql bug in IC from Martin Evans
  Added a change from a suggestion from Martin Evans for making ref cursors faster.
  Added rt.cpan.org Ticket #=42328 ora_objects attribute for extended embedded objects support from Tomas Pokorny
  Fix for rt.cpan.org Ticket #=42328 user defined types from different schema in describe_obj from Tomas Pokorny
  Added a README for sun suggested by Jim McCullars
  Clean up of white space and formatting to 4 tabs  from John Scoles
  Fix for GCC 4.3 warnings from Eric Simon
  Standardized the dbd_verbose levels so they are all 3 and over.  from John Scoles
  Added private statement functions ora_stmt_type_name and ora_stmt_type from John Scoles
  Update to POD from Chris Underhill
  Added README.win64.txt with content from Alex Buttery
  Fix for rt.cpan.org Ticket #=21920 Bug with Oracle DBD for Mac OS X Instant Client From boingolover
  Added a few more constants to get rid of magic numbers from John Scoles
  Fix for rt.cpan.org Ticket #=38267 Inserts/Updates to BLOB's via synonyms fails from John Scoles
  Fix for rt.cpan.org Ticket #=39603 build problem and fix missing functions in oci.def from Zoltán Sebestyén
  Fix for rt.cpan.org Ticket #=39374 Makefile.PL: error when reducing echo messages from make from Tippa
  Fix for rt.cpan.org Ticket #=39232 binding large XMLTYPE fails on 64-bit perl from Jeff Klein
  Fix for rt.cpan.org Ticket #=38749 Warning of a NULL column in an aggregate function also added ora_oci_success_warn to display silent OCI warnings from John Scoles
  Patch for UTF8 check on execute_array from David Mansfield and a little by John Scoles


* Changes in DBD-Oracle 1.22(svn rev 11618)  1st Aug 2008

  Patch to remove compiler warnings from H.Merijn Brand
  Patch to Makefile for 64bit boxes from Alex Laslavic
  Added OCILobGetLength to lob functions from Milo van der Leij
  Updated readmes to state the test user has to have create, call and drop a procedure privileges by John Scoles suggested by Gisle Aas
  Patch to Makfile to prevent the installation of the lib/DBD/mkta.pl fil from Gisle Aas
  Added new Test 31lob_extended.t for use of LOBs when returned via stored procedures with bind_param_inout from Martin Evans
  Update to connection part of POD from  John Scoles
  Fix to test suite to bring it up to standard from Martin Evans
  Fix for memory hemorrhage in bind_param_inout_array found by Ricky Egeland, Fix by John Scoles
  Fix for a typo in oracle.xs from Milo van der Leij
  Fix for bugs on SPs with Lobs reported by Martin Evans, Fix by J Scoles
  Changed the way Ping works rather than using prepare and execute it now makes a single round trip call to DB by John Scoles
  Fix for rt.cpan.org Ticket #=37501 fail HP-UX Itanium 11.31 makefile also added the OS and version to the output of the Makefile.PL for easier debugging. from John Scoles and Rich Roemer
  Added a number of internal functions for decoding OCI debug values from John Scoles
  Fix for  hpux 11.23 linker error unrecognized argument on the Makefile from someone on CPAN forum
  Added fetch by piece for lobs, fixed persistent lobs and expansed thier usage for LONG and LONG RAW and changed to pod to reflect the changes from John Scoles
  Added comment to POD on case sensitivity of ORACLE environment variables suggested by Gerhard Lausser
  Added patch to fix a number of harmless, but annoying, GCC warnings from Eric Simon
  Added (finally) ora_verbose for DBD only tracking from John Scoles and thanks to H.Merijn Brand
  Fix for rt.cpan.org Ticket #=32396 from John Scoles
  Fix for memory leak that snuck into 1.21 from John Scoles
  Fix for rt.cpan.org Ticket #=36069: Problem with synonym from John Scoles
  Fix for rt.cpan.org Ticket #=28811 ORA_CHAR(s) not returning correct length in functions and procedures from John Scoles
  Makefile.PL now working without flags for Linux 11.1.0.6 instant client and regular client from John Scoles, Andy Sautins, H.Merijn Brand, Nathan Vonnahme and Karun Dutt
  Fixed how persistent lob fetch works now uses callback correctly, from John Scoles & Darren Kipp


* Changes in DBD-Oracle 1.21(svn rev 11067) 11th April 2008

  Added Notes to README.win32.txt on installing Instant Client 11.1.0.6.0 from John Scoles
  Added the oci_typecode_name method to get the name rather than just the number of an OCI_TYPECODE from John Scoles
  Fixed a unreported bug with Embedded Objects from John Scoles
  Fixes for #34621 & 33791 from RT cpan
  Added patch to allow faster fetch from REF CURSORs from Biswadeep Chowdhury
  Updated the Todo file for next version from John Scoles
  Added support for the 10.2 Data Interface for Persistent LOBs by John Scoles
  Changed the way pre-fetching is done by John Scoles
  Added support for Scrollable cursors from John Scoles
  Changed the max size of cache_rows to a sb4 rather than a int and or a ub4 from John Scoles
  Added support for Lobs in select of OCI Embedded Objects from John Scoles with a big thankyou to  Paul Weiss
  Fixed for embedded object in object from Paul Weiss
  Added support for direct insert of large XML data into XMLType fields from Hendrik Fuss & John Scoles
  Fixed memory leak (not releasing Temp Lob with OCILobFreeTemporary) when created for a bind from John Scoles
  Added support for bind_param_inout_array for use with execute_array from John Scoles
  Added enhancement for Embedded Objects handling from Paul G. Weiss
  Fixed to Makefile.PL let it read makefiles from other makes from Alexander V Alekseev
  Updated POD to tell users to Avoid Using "SQL Call" from Charles Jardine
  Updated POD to account for rt.cpan.org #30910: "DBD-Oracle crashes when trying to read empty LOB" from John Scoles
  Added DBD::Oracle impdata/threads patch from Jeffrey Klein

* Changes in DBD-Oracle 1.20(svn rev 10517) 11th January 2008

  Fixed lob test so it skips the one test that relies on it if v$ session. from Rafael Kitover
  Fixed // with /* */ in dbdimp.c from John Scoles
  Fixed for execute_for_fetch in Oracle.pm returning 0 instead of 0E0.  from Martin J. Evans
  Added README.64bit.txt that contains help for compiling on 64 bit boxes from John Scoles
  Fixed typo in Oracle.pm from Tom R.
  Added support for ora_charset, ora_ncharset from Stephen J. Smith
  Fixed Makefile.PL for better handling of empty array in File::Find::find from Slaven Rezic
  Fixed references to README.clients.txt in Makefile.PL from  John Scoles
  Added PERL_NO_GET_CONTEXT for better multi-threaded support from John Scoles
  Changed required version of DBI to be 1.51 from John Scoles
  Fixed bug in 31lob.t from John Scoles
  Added notes on installing Instantclient .rpm to README.Lunix.txt
  Added support for OCI array bind from Alexander V Alekseev
  Added support for select of OCI Embedded Objects from John Scoles
  Added a tip in README.64bit.txt from cartmanltd
  Added fix to Makefile.PL for finding SQLplus for Ubuntu Server (but should work for others) from Martin J. Evans
  Added fix to Makefile.PL for Gentoo AMD64 from Tom R.
  Added fix to dbdimp.c for speed up of Null-Operations from Andreas Behal
  Added fix to dbdimp.c for SQLCS_NCHAR index use on varchar2s from  Peter J. Holzer

* Changes in DBD-Oracle 1.19 (svn rev 8002) 3rd November 2006

  Fixed execute_array to comply with DBI standard from Martin J. Evans, Xho Jingleheimerschmidt and others
  Fixed execute_array so it will not throw a Perl warning on undef values in Tuples from John Scoles
  Fixed execute_array so it will take the ora_array_chunk_size DB handle attribute
  Fixed some typos in code and READMEs from John Scoles
  Fixed a few other little bugs dealing with  compatibility with Oracle 8
  Changes to README from Karl Auer
  Suppress warning in 26exe_array.t from Philip Garrett
  Added support for array context aware execute_for_fetch from Martin J. Evans
  Fixed Makefile.PL for an incompatibility with ExtUtils::MM_Unix v1.50 (invoked byExtUtils::MakeMaker) from Dennis McRitchie
  Updated POD to reflect that OCI after 9.2 no longer strips trialing spaces


* Changes in DBD-Oracle 1.18 (svn rev 6697)

  Added support for native Oracle Array interface thanks Kristian Nielsen
  Added suppot for LOB Locators from Jeffrey Klein.
  Updated README.win32.txt for Oracle 10xe and new Visual C++ version
  Updated README.lunix.txt for work-araound for UTF8 bug
  Fixed a number of compile warings

please enjoy.

* Changes in DBD-Oracle 1.17 (svn rev 3726)

  Updated README.win32.txt fixed some typos
  Fixed expanded support for Lobs support from Jeffrey Klein
  Added notes on expanded support for Lobs to Oracle.pm

* Changes in DBD-Oracle 1.17 (svn rev 3725)

  Added expanded support for Lobs from Jeffrey Klein

* Changes in DBD-Oracle 1.17 (svn rev 2487)   7th February 2006

  NOTE: With this release of DBD::Oracle pythian.com (http://www.pythian.com)
  are taking on the role of lead maintainer - with my support and gratitude.
  John Scoles at pythian.com is now the man in the hot seat for support and
  maintenance!

  Fixed automatic csform setting for some UTF8 cases and for Oracle 8.0
  Fixed truncation error on fetch into UTF8 charset thanks to Honza Pazdziora.
  Fixed INTERVAL DAY TO SECOND thanks to Honza Pazdziora.
  Fixed unicode tests for cygwin thanks to Andy Hassall.
  Fixed undef warnings when connecting with undef $user.
  Fixed undef warnings from $dbh->get_info(18);
  Fixed LOB streaming example thanks to Pablo Zorzoli.

  Added support for nested cursors in select lists thanks to Charles Jardine.
  Added "Trailing Spaces" section to docs thanks to Michael A Chase.
  Added support for binary floats/doubles thanks to Dennis Box.
  Added INSTANCE_NAME, SERVER and SERVICE_NAME as valid connect keywords
    in the 'dbi:Oracle:x=y' short form of connecting without tnsnames.ora.
    For example 'dbi:Oracle:host=localhost;service_name=xe;server=dedicated'
  Added auto-detection of ORACLE_HOME in some configurations.

  Changed "Binding Cursors" docs, clarifying examples thanks to Charles Jardine.
  Changed tests to use ORACLE_DSN or DBI_DSN env vars if defined thanks to Jill Vogel.
  Updated README.vms re logical name tables thanks to Jakob Snoer.
  Removed README.utf8 since the topic was covered better in the main docs.
  Updated README.hpux.txt and restructured examples thanks to H.Merijn Brand.
  Updated README.aix thanks to Stephen de Vries and Nathan Vonnahme.
  Updated README.macosx thanks to Stephen de Vries.
  Renamed README.*'s to add .txt suffix to make life easier for some.

  Changes to Makefile.PL:
    Instant Client support thanks to Hilmar Lapp, John Scoles and others.
    Improved HP-UX support thanks to H.Merijn Brand.
    Avoid risk of sqlplus hanging thanks to Mark Dedlow.
    More reliably use correct sqlplus thanks to Honza Pazdziora.
    Improved build rule and Oracle header file detection.
    Improved cygwin build thanks to Andy Hassall.
    VMS logical name checks thanks to Jakob Snoer.

  The Copyright terms for DBD::Oracle have been simplified and now read:
    The DBD::Oracle module is free open source software; you can
    redistribute it and/or modify it under the same terms as Perl 5.

* Changes in DBD-Oracle 1.16 (svn rev 515)   22nd October 2004

  NOTE:
  This release has major changes to Unicode support. See below.
  It no longer supports the old Oracle 7 OCI interface.
  It requires DBI >= 1.38 for some of the tests if using Perl 5.6.
  It no longer supports Perl 5.5 or earlier.

  Fixed placeholder names to be case insensitive thanks to Charles Jardine.
  Fixed some LOB test problems with Oracle 8.1.7 by implementing ora_lob_append
    with OCILobGetLength() and OCILobWrite(), instead of buggy OCILobWriteAppend(),
    if the Oracle client version is < 9.0. Thanks to Jeff Urlwin.
  Fixed handling of temporary LOBs thanks to Chris Donnelly.
  Fixed memory leaks in auto LOB refetch code thanks to Dongqiang Bai.
  Fixed reporting of length truncated in error message thanks to Jeff Urlwin.
  Fixed column_info() to handle TIMESTAMP and INTERVAL datatypes
    for Oracle >= 8 thanks to Stephen Clouse.
  Fixed STORE to cache attribute value in handle cache.
  Fixed seg fault returning LOB Locators reported by Raj Chandran.
  Fixed binding to allow overloaded scalars (not for 'inout' params).
  Fixed setting of $DBI::err to 0 triggering PrintWarn in DBI >= 1.41.
  Fixed some edge cases in row cache sizing.
  Fixed truncation error fetching very small numbers (1 ^ -130).
  Fixed Oraperl to not enable PrintError or AutoCommit (broken since 1.13).

  Changed some utf8 internals for LOBs.
  Changed ORA_OCI constant from being just 7 or 8 to being a dualvar:
    in numeric context returns the major.minor version number (8.1, 9.2 etc)
    in string context it returns the full "major.minor.foo.bar" version string.
  Changed some SUCCESS_WITH_INFO situations to be treated as a "warning"
    by setting $DBI::err to "0" (and so trigger PrintWarn in DBI >= 1.43)
    eg "ORA-28011: the account will expire soon; change your password now"
    and package compilation errors.

  Added automatic support for UTF-8 for both NLS_LANG and NLS_NCHAR
    Many thanks to Lincoln Baxter who did most of the hard work and testing
    and to Jeff Urlwin and others who also helped out.
    Perl 5.8.x and Oracle 9+ are highly recommended if you want to use Unicode.
    See POD for more information and documentation.
  Added support for "... RETURNING lob_locator_column INTO ?"
    using $sth->bind_param_inout(2, \$loc, 0, {ora_type => ORA_BLOB});
  Added bind_param() ora_csform, ora_csid, and ora_maxdata_size attributes.
  Added bind_param() support for SQL_BLOB & SQL_CLOB thanks to Stephen Clouse.
  Added $dbh->ora_can_unicode and $dbh->ora_nls_parameters thanks to Andy Hassall.

  Documentation changes:
    Corrected typo in ora_lob_read() example thanks to Johannes Wierny.
    Corrected LOB example thanks to Sascha Pfalz and Thomas Upton.
    Updated README.macosx thanks to Hilmar Lapp.
    Updated README.hpux thanks to Gram Ludlow and Lincoln Baxter.
    Added $dbh->reauthenticate($user,$pass) docs thanks to Andy Hassall.
    Added $dbh->{ora_parse_error_offset} docs thanks to Andy Hassall.
    Added gcc example to README.aix thanks to Adrian Terranova.
    Added INSERT ... RETURNING ... example prompted by Steven Lembark.

  Build Changes:
    Fixed build32/build64 Makefile.PL bug spotted by Sean Kelly.
    Fixed Oracle version detection on clients thanks to Marko Asplund.
    Fixed Oracle 10 version detection on clients thanks to Andy Hassall.
    Fixed HP-UX 64bit build issues thanks to Weiguo Sun.
    Fixed VMS build issues thanks to Jakob Snoer.
    Fixed suprious warning from Makefile.PL thanks to Marko Asplund.
    Fixed compiler warnings thanks to Robert Baron.
    Fixed oci.def for Cygwin by adding OCILobWriteAppend thanks to Gert-Jan Paulissen.
    Changed to use ANSI C prototypes thanks to Steffen Goeldner.
    Changed to not warn if ORACLE_HOME env var is not defined
      as it's not used with Oracle's new 'Instant Client'.
    Added Test::More as a prerequisite module.
    Added many Unicode related tests thanks to Lincoln Baxter.

* Changes in DBD-Oracle 1.15   27th January 2004

  NOTE: DBD::Oracle now requires DBI version 1.28 (June 2002) or later.

  NOTE: This is probably the last release that will support being
  built with the old Oracle 7 OCI interface.

  NOTE: This release will build for perl 5.5 only after some manual edits
  which are detailed in README.help.  Future releases may not build for 5.5.
  Perl 5.5.3 is very old and and upgrading to at least 5.6.1 is recommended.
  The DBI itself has required perl >= 5.6.0 since DBI 1.38, August 2003.

  Fixed Makefile.PL for HPUX thanks to H.Merijn Brand & Sweth Chandramouli.
  Fixed Makefile.PL for handling 32/64bit libs thanks to Lincoln Baxter,
    Sean Kelly, Joel Van Boeckel and others. Oh the fun we've had.
  Fixed Makefile.PL error typo thanks to Martijn Koster.
  Fixed to build okay for Oracle 8.0.x and for older C compilers.
  Fixed minor omission in t/ph_test.t thanks to Tom Malaher.
  Fixed for extproc_perl (http://search.cpan.org/author/JHORWITZ/extproc_perl/)
  Fixed Oraperl handling of ora_errno and ora_errstr thanks to Martin Busik.
  Fixed PRECISION for "NUMBER" to be 126 not 0 thanks to Steffen Goeldner.
  Fixed bind_param_inout() for placeholders not assigned to in PL/SQL.

  Changed bind_param_inout for CHAR types to no longer use a minimum
    length of 28 characters. Warning: this change may break code that
    doesn't pass bind_param_inout() a length value large enough for the
    returned string. (The minimum length was not documented and should
    not have been relied upon. This change currently only applies to the
    CHAR type but may extended to all string types in a later release.)
  Changed type_info_all() to return the same type info as Oracle's own
    ODBC driver does, thanks to Andy Hassall for the data.
    The types include LOBs but not the new TIMESTAMP and INTERVAL types.

  Added direct access to LOB Locators and major LOB Locator functions
    such as $sth->ora_lob_read(...), $sth->ora_lob_write(...) etc.
    This work was sponsored by Geospiza Inc.
  Added LOB Locator example docs thanks to Mark Dilger at Geospiza.
  Added TIMESTAMP [WITH [LOCAL] TIME ZONE]] support
    thanks to Stephen Clouse and Robert Wyrick.
  Added INTERVAL YEAR TO MONTH, INTERVAL DAY TO SECOND support.
  Added warning for bad DSN string format thanks to Michael A Chase.
  Added /*+RULE*/ hint to metadata method SQL thanks to Andy Hassall.
  Added README.linux thanks to William Fishburne, Stephen Clouse and Brent LaVelle.
  Added README.aix thanks to KC, Mike Paladino and Rafael Caceres.
  Added connect example using OS authentication thanks to Bob Thomson.
  Added prepare("...", { ora_placeholders => 0 }) to disable placeholders.
  Added docs for returning a recordset (table/array) using Oracle >=9.0.1
    via "FUNCTION foo RETURN type PIPELINED" thanks to Steve Baldwin.
  Added docs on ora_check_sql=>0 in prepare() to avoid server-side parses.
  Added support for sharing database connections with ProC/SQLLIB code
    via ora_use_proc_connection attribute, thanks to Kristian Nielsen
    needs build time option thanks to Steffen Goeldner
  Added (restored) the error "possibly near <*> indicator" marker for
    syntax errors thanks to Jason Hitt and Andy Hassall.
  Added $dbh->{ora_parse_error_offset} attribute thanks to Andy Hassall.
  Added auto setting of $dbh->{Username} if not given to connect
    i.e. using Oracle OS authentication and connecting as "/"
    by selecting SYS_CONTEXT('userenv','session_user') from the db
    thanks to Eric Lenio and Andy Hassall.

  Updated README.macosx thanks to Danial Pearce and William Goedicke.
  Updated README.help for UnixWare thanks to Earle Nietzel.
  Updated PL/SQL cursor examples, plus Oracle.ex/README and curref.pl
    examples thanks to Michael A Chase.
  Updated PL/SQL example setup thanks to Bob Kline.

* Changes in DBD-Oracle 1.14    27th March 2003

  NOTE: OCI 7 and Oraperl will not be supported in future releases.

  Implemented inserting NULL values into LOB fields using placeholders,
    (sponsored by bayerwald-fenster.de thanks to Bernhard Donaubauer).
  Fixed broken get_info() due to missing GetInfo.pm #2274.
  Fixed compiler warnings/errors thanks to H.Merijn Brand.
  Fixed bind_param with array ref to croak not warn, reported by Steven Roels.
  Fixed "Bad hash" error using $sth->{ParamValues} if there were no placeholders.
  Fixed Makefile.PL for more HPUX/Oracle configurations thanks to Lincoln Baxter.
  Simplified MacOS X build thanks to Brook Schofield, see README.macosx.
  Updated README.hpux thanks to H.Merijn Brand, Jay Strauss, and Lincoln Baxter.
  Updated README and README.utf8 with minor changes thanks to Alexey Mahotkin.

* Changes in DBD-Oracle 1.13    14th March 2003

  Fixed null user issue when using OS Authentication thanks to Christopher R. Baker
  Fixed precision for Raw types in oci7.c thanks to J.D. Laub.
  Fixed LOB code for table names containing $ thanks to Wayne Volkmuth.
  Fixed LOB synonym handling thanks to John Milton.
  Fixed dbms_output_get to be immune to ora_ph_type changes thanks to Michael Fox.
  Fixed to now not treat SUCCESS_WITH_INFO (eg ORA-28011) on connect as an error.

  Updated table_info to decode SYS and SYSTEM to prepend 'SYSTEM ' thanks to Olga Voronina
  Updated table_info and added get_info, foreign_key_info thanks to Steffen Goeldner.

  Added that ShowErrorStatement is now enabled by default.
  Added ParamValues attribute so ShowErrorStatement includes bind_param values.
  Added experimental utf8 support thanks to Stefan Eissing.  See README.utf8
  Added BFILE support thanks to David Hull.
  Added :ora_session_modes export tag for ORA_SYSDBA ORA_SYSOPER constants,
   added ORA_STRING and ORA_CHARZ to :ora_types export tag and added docs
   for ora_ph_type attribute and other doc fixes, thanks to Michael A Chase.
  Added t/cursor.t tests thanks to Jeffrey Horn.
  Added ability to share connections between threads via ora_dbh_share
    attribute, thanks to Gerald Richter.
  Added ability to connect() via 'external process context', for embedding
    Perl DBI into Oracle server, thanks to Jeff Horwitz.

  Documentation changes:
    Updated README.hpux thanks to Lincoln Baxter and Jay Strauss.
    Documented ora_parse_lang, ora_auto_lob, and ora_check_sql prepare()
      method attributes thanks to Michael A Chase.

  Configuration and build related changes:
    Added Win32 support on systems with multiple Oracle Homes thanks To Jeff Urlwin.
    Fixed prototype for constant() thanks to Sreeji K Das.
    Fixed warning in Oraperl.pm with recent perl versions thanks to Ilya V.Rachkov.
    Fixed handling of shell backticks in make output during build rule discovery.
    Fixed Win32 build using gcc thanks to Chris R. Donnelly and Michael A Chase.
    Help HPUX configs by removing -l:libcl.a from liblist, thanks to Frans Postma.
    Assorted improvements to Makefile.PL thanks to Michael A Chase and Stephen Clouse.
    Updated tests and Oraperl.pm to not use old style connect (deprecated in DBI 1.33).


* Changes in DBD::Oracle 1.12   31th August 2001

  Improved LD_RUN_PATH code in Makefile.PL thanks to John Groenveld.
  Improved HPUX code in Makefile.PL thanks to H.Merijn Brand.
  Fixed handle/descriptor name code to be less Oracle version sensitive.
  Updated $dbh->primary_key_info() to new DBI spec thanks to Steffen Goeldner.
  Version 1.11 was skipped (uploaded to PAUSE with an error).

* Changes in DBD::Oracle 1.10   30th August 2001

  Explicitly require DBI version 1.20 (or later).
  Support multiple oracle home's in Win32 registry thanks to Jeff Urlwin.
  Improved build support for Oracle >=8.1.7 (I_SYM) thanks to Stephen Close.
  Fixed LOB bug when writing non-string values thanks to Stephen Close.
  Fixed t/ph_type.t to be better behaved thanks to Ville Skytta.
  Added OCI8 handle/descriptor name to trace output for some API calls.

* Changes in DBD::Oracle 1.09   27th August 2001

  Changed behaviour when OCIStmtExecute() returns OCI_SUCCESS_WITH_INFO:
    used to be treated as OCI_SUCCESS, now also sets $DBI::err/$DBI::errstr
    and will therefore trigger RaiseError/PrintError if they are set.
  Fixed ins/upd of LOBs in views thanks to Tom Moore.
  Fixed reading of CLOBs in wide characters sets.
  Fixed 'inherited AUTOLOAD for non-method DBD::Oracle::ORA_OCI' problem.
  Workaround very long column names causing failure thanks to Martin Busik.
  Added $dbh->primary_key_info() thanks to Steffen Goeldner.
  Added $sth->cancel method thanks to Fredrik Sjoholm.
  Added mention to registry to login failed hint thanks to Tim Callaghan.
  Changed type_info_all VARCHAR2 max length to 4000 if using OCI8.
  Removed use of perl global PL_no_modify to avoid linkage problems.
  Old OCI7 API:
    Fixed OCI7 memory leak from statement handles thanks to Bret Haeler.
    Fixed OCI7 PRECISION for RAW data types thanks to J.D. Laub.
    Fixed old OCI7 LONG buffer overwrite bug thanks to Bruce Templeton.
      (I thought I'd done this before but perhaps now it's done right.)
  Build:
    Fixed default build to set LD_RUN_PATH at build time. Yeah.
    Fixed parsing of Oracle's .mk files to handle += (appending values).
    Fixed detection of 'OCI' dire for Win32 Oracle >=8.1 thanks to many.
    Improved test for gcc in Makefile.PL thanks to Ian Kallen.
    Improved build rule detection thanks to Jonathan Segal.
    Don't merge text from Oracle's .mk into generated Makefile (use -f to reenable).
    Fixed bad test in plsql.t so shouldn't give false negatives.
    Added t/ph_type.t to test placeholder type issues thanks to James Jurach.
    Retested for perl 5.4.4 for possibly the last time. You have been warned.
      In future I'll only be testing on perl >= 5.5.3 and >=5.6.1.
  Documentation:
    Documented workaround for hang in "repetitive connect" test thanks to Alexi Lookin.
      Short version: add bequeath_detach=yes to sqlnet.ora file.
    Documented ora_oratab_orahome connect attribute (Oracle OCI 7 only).
    Documented ora_module_name connect attribute thanks to Renzo Toma.
    Added IRIX linking advice into README.help thanks to Dennis Box.
    Added AIX core dump on disconnect advice into README.help.
    Added docs on how Oracle determines the timezone thanks to Brad Howerter.
    Added 'return cursor function' example to docs thanks to Michael A. Chase
    Added more notes to README.java thanks to Peter Ludemann and Dave C.
    Added more notes to README.login thanks to Geoff Young.

* Changes in DBD::Oracle 1.08 7th August 2001

  Fix Oracle 7 build broken in previous release.
  Fix for 64bit builds thanks to Alan Burlison.
  Fix for 9i Makefile.PL thanks to Dong Wang.
  Added Oracle.ex/curref.pl thanks to Geoffery Young.
  REF types are now returned in a form usable with Oracle::OCI module
    when using Oracle >= 8.1.x (untested at the moment)
  Updated Makefile.PL and README.hpux thanks to Lincoln Baxter.

* Changes in DBD::Oracle 1.07 5th June 2001

  This release is mostly targeted at the needs of the new Oracle::OCI module.
    It's not a general release and not recommended for production use.

  Fixed SCALE/PRECISION/TYPE values for numeric types, including FLOAT,
    to be more correct thanks to Steffen Goeldner.
    NOTE: this may break code that depends on the old values!
  Fixed Win32::TieRegistry and Tie::Registry usage thanks to David Bradford.
  Fixed insert/update of LOBs while AutoCommit enabled.
  Fixed error in 'LOB value truncated' error thanks to Paul Walmsley.
  Fixed old oci7 LONG buffer overwrite bug thanks to Bruce Templeton.
  Fixed ocitrace.h not to trip up some compilers.
  Added table_info() search attribs thanks to Steffen Goeldner and Olga Voronova.
  Implemented some reworking of the placeholder handling, especially
    in relation to the start of some basic array support.
  Updated email and web page details to dbi-users@perl.org and dbi.perl.org.

* Changes in DBD::Oracle 1.06 14th July 2000

  Fixed ora_ph_type attribute to allow 96 or 97.
  Fixed compile failure with Oracle 7.

* Changes in DBD::Oracle 1.05 13th July 2000

  Added $dbh->{ora_ph_type} attribute to define default bind type:
     1=> VARCHAR2,  does strip trailing spaces, embedded \0 bytes okay
     5=> STRING, doesn't strip trailing spaces, embedded \0 ends string
    96=> CHAR,   doesn't strip trailing spaces, embedded \0 okay
    97=> CHARZ,  doesn't strip trailing spaces, embedded \0 ends string?
    The two CHAR types force 'blank-padded comparison semantics'.
    Needs testing and documenting, volunteers most welcome!
  Added (many) typecasts to avoid compiler warnings thanks to Denis Goddard.
  Added ability to pass existing DBD::Oracle select statement handle
    (cursor) back _into_ Oracle as a ref cursor type thanks to Mike Moretti.
    Note that this enables a workaround for closing ref cursors:
    $dbh->do("BEGIN CLOSE :cursor; END;", undef, $sth_ref_csr_to_close);
  Improved Win32 support in Makefile.PL thanks to Michael A. Chase.
  Workaround Win32::TieRegistry FETCH error during global destruction.
  Re-enable row cache by default for OCI8 (can give big speed increase).
  Fixed bug in OCI8 row cache calculation thanks to Greg Stark.

* Changes in DBD::Oracle 1.04 11th July 2000

  Added info on workarounds for Java thread related linker errors
    on Solaris with Oracle 8.1.6. Thanks to Andi Lamprecht.
  Fixed memory leak on connect. Thanks to Jeffrey W. Baker.
  Fixed memory and ref cursor leaks. Thanks to Mike Moretti and John Tobey.
  Fixed SEGV on insert/update or many LOBS. Thanks to Honza Pazdziora.
  Fixed treating SUCCESS_WITH_INFO from finish() as an error.
  Fixed passing of attribs to connect thanks to K.Takaoka.
  Fixed cursor binding example in docs thanks to Michael Chase.
  Build using otrace/demo/atmoci.mk as last resort thanks to Chris Sylvain.
  Improved reliability of "perl Makefile.PL -b" configure option.
  Improved reliability of t/plsql.t cursor tests. Hopefully.
  Improved reliability of ping method.
  Fixed broken SQL type warning in bind_param.

* Changes in DBD::Oracle 1.03 12th July 1999

  Added "perl Makefile.PL -b" configure option. Links DBD::Oracle
    using same linker args as Oracle's own OCI demo applications.
  Added connect("dbi:Oracle:host=foo;sid=bar", ...) syntax.
  Added OCI8 function trace at level 6, useful for Oracle support.
  Added ora_session_mode attribute to connect (eg SYSDBA/SYSOPER).
  Added test for Oracle security problem. See README.sec.
  Updated ora_explain tool to v1.1 thanks to Alan Burlison.
  Contributions from Michael Chase (plus docs from Andy Duncan):
  Documented plsql_errstr & dbms_output_enable/_put/_get.
  Enhanced $dbh->func('plsql_errstr') output.
  Replaced old Oraperl examples with DBI/DBD::Oracle ones.

* Changes in DBD::Oracle 1.02 14th June 1999

  LongReadLen no longer limited to 64KB for OCI 7 & 8!
    But beware of OCI 7 bug when fetching >64KB.
    See t/long.t for more details.
  Fixed OCI7 ref cursor missing data row.
  Fixed OCI8 LOB statement handle leak & improved trace.

* Changes in DBD::Oracle 1.01 8th June 1999

  Enhanced diagnostics in t/long.t test suite.
  Removed byte with high-bit set from t/long.t test data.
  Disable finish if database disconnected or perl is terminating.
  Made t/general.t work with other NLS settings.
  Added Cygwin support thanks to Alexander Smishlajev.
  Fixed 'undeclared identifier' error building with old Oracle's.

* Changes in DBD::Oracle 1.00 4th June 1999

  Increased default row cache size for improved performance.
  Added OCI8 binding of cursors! Sponsored by cp.net.
  Added OCI7 binding of cursors! (Was easier after OCI8 work :-)
  Added OCI8 blob_read (only for LOBs not LONGs) thanks to Jim Lawson.
  Added OCI8 re-authentication thanks to John Tobey.
  Added t/long.t test script for LONG/LONG RAW/CLOB/BLOB handling.
  Length of fetched LONG RAW string now 2 * LongReadLen for OCI7 & 8.
  Fixed LONG fetches being one byte longer than LongReadLen.
  Fixed OCI8 non-reporting of LobWrite error.
  Fixed OCI8 "LongReadLen too small and/or LongTruncOk not set" hint.
  Fixed OCI7 probable cursor leak.
  Fixed ping method to be more robust. This should fix
    the "morning bug" problem with Apache::DBI.
  Fixed t/general.t core dump thanks to Donald Buczek.
  Fixed ora_check_sql prepare attribute (for selects).
  Fetch errors (non-row level) now turn off the Active attribute.

* Changes in DBD::Oracle 0.61, 9th April 1999

  Fixed execute() always returning 0 rows! (with OCI 7).
  Fixed $sth->bind_param(..., SQL_CHAR);
  Assorted minor Makefile.PL improvements.
  Added ora_check_sql attribute to prepare() for OCI8.

* Changes in DBD::Oracle 0.60, 10th March 1999

  Fetching all records now resets Active flag as it should.
  Finally fixed "Can't bind unknown placeholder" (hopefully :-).
  Fixed placeholder parser to ignore comments, thanks to JD Laub.
  Fixed placeholder parser to ignore double-quotes strings.
  No longer changes ORACLE_HOME to match oratab, by default.
  Fixes for OCI 8:
    Fixed selection of very large numbers (132 digits).
    Fixed small memory leak in parameter binding.
    Fixed small memory leak in prepare of statements using LOBs.
    Fixed binding empty strings.
    Fixed OCILobRead error when fetching of zero length LOBs.
    Fixed OCILobWrite of zero length to use OCILobTrim instead.
    Treat SUCCESS_WITH_INFO from OCIStmtExecute as SUCCESS.
    Enabled OCI_CRED_EXT login thanks to Alan Burlison & Jeremy Brinkley.
  Modified hints/svr4.pl - SVR4 users please test.
  Updated Alan Burlison's ora_explain Tk tool to version 1.0.

* Changes in DBD::Oracle 0.59 (Oraperl 1.37), 27th December 1998

  Fixed detection of ambiguous LOB-param-to-table-field assignment
  (previous fix had typo). Only affects multiple LOBs in same table.
  Added support for Oracle 8's "... RETURNING foo,bar INTO :foo,:bar"
  for simple scalar types (not LOBs or arrays) via bind_param_inout.
  Cleaned up the code.
  Improved Oracle 8 library selection code for Win32 (untested).

* Changes in DBD::Oracle 0.58 (Oraperl 1.37), 22nd December 1998

  Fixed detection of ambiguous LOB-param-to-table-field assignment.
  Added bind_param ora_field attribute to disambiguate if needed.

* Changes in DBD::Oracle 0.57 (Oraperl 1.37), 21st December 1998

  Fixed bug preventing fetching LONGs when using OCI 8, better.
  (Oracle bug #641812 not involved. Should work for all 8.0.x)

* Changes in DBD::Oracle 0.56 (Oraperl 1.37), 19th December 1998

  Fixed bug preventing fetching LONGs when using OCI 8.
  (Oracle bug #641812 may still cause failure prior to v8.0.5.)
  Fixed LongTruncOk to work when using OCI 8.
  Fixed bug in table name detection code for OCI8 LOB refetch.
  SCALE & PRECISION work for OCI 7 & 8.

* Changes in DBD::Oracle 0.55 (Oraperl 1.37), 16th December 1998

  Major internal work to support Oracle 8 OCI.
  Oracle 8 LOBs are supported and treated as LONGs (DBD::Oracle works
  hard to hide the complexities of handling Lob Locators for you).
  See Oracle 8 section in the docs for more details.
  Added $sth->{TYPE}.
  SCALE & PRECISION are implemented but return 0 (Oracle bug?).
  DBI 1.02 or later is required.
  Assorted build time Makefile.PL improvements.
  Builds with 5.004_04, 5.005_02 and 5.005_54 (not _53).
  Added "use DBD::Oracle qw(:ora_types);"

* Changes in DBD::Oracle 0.54 (Oraperl 1.37), 14th August 1998

  Added $dbh->type_info_all.
  Fixed $dbh->tables (partly by renaming to new $dbh->table_info).
  data_sources no longer adds abbreviated versions of tnsnames.
  Alan Burlison's whizzo Tk based explain script now bundled.
  Revised workaround for "Can't bind unknown placeholder '3'" errors.
  Doubled default automatic row cache size (now approx 5KB).
  The resetting of Oracle's SIGCHD handler to SA_RESTART can
  now be disabled by setting the env var DBD_ORACLE_SIGCHLD=0.
  Fixed Makefile.PL -c to better avoid shared Oracle lib.

* Changes in DBD::Oracle 0.53 (Oraperl 1.37), 3rd August 1998

  Further build fixes (esp kpudc problem with Oracle 8).
  Now prefers oracle.mk over proc.mk again.
  Only $ENV{ORA_CLIENT_LIB} ||= 'shared' if shared lib exists.
  Builds okay with 5.005-thread (not tested).

* Changes in DBD::Oracle 0.52 (Oraperl 1.37), 28th July 1998

  Assorted build fixes (esp. Win32, HP-UX and AIX).
  More hints on error messages, especially long truncation and
  field-level errors when fetching.
  Compiles okay now for systems without SQLT_CUR defined.
  Only sets SA_RESTART on SIGCLD if connect was successful.

* Changes in DBD::Oracle 0.51 (Oraperl 1.37), 5rd July 1998

  Makefile.PL no longer tries to link with just -lclntsh directly :-(
  Improvements to some HP-UX builds (hopefully).
  DBI->data_sources ' dummy ' value removed (and list now sorted).
  connect failure now shows actual Oracle error message again.
  Initial (incomplete) support for binding cursor vars (see t/plsql.t).

* Changes in DBD::Oracle 0.50 (Oraperl 1.36), 3rd June 1998

  Makefile.PL changes: fixed -c option, now searches for .h files,
  tries alternate location for sysliblist, checks for executable
  orainst/inspdver before using it.
  Fixed cursor leak.
  Added first word of tnsnames.ora name as aliases if no clash.

* Changes in DBD::Oracle 0.49 (Oraperl 1.36), 1st June 1998

  Further improvements to build process over 0.48.
  Fixed broken truncation error in 0.48.
  Fetch ORACLE_SID from Win32 registry (thanks to Preston Bannister)
  Improved automatic row cache sizing (prompted by Jon Meek).
  Added $sth->{ora_cache_rows} and $sth->{ora_est_row_width}
  as read-only attributes to make cache size logic easier to test.

* Changes in DBD::Oracle 0.48 (Oraperl 1.36), 25th May 1998

  THIS IS AN EXPERIMENTAL RELEASE - USE WITH CAUTION!
  Now links to -lclntsh directly (Thanks to Bruce Nelson and others)
  Workaround for broken backticks after login (Thanks to Warren Jones)
  Now finds and reads tnsnames.ora to disambiguate dbnames in connect.
  Added basic support for bind_param(..., SQL_TYPE).
  Fixed bind_param_inout after execute.
  Added dbms_output_(enable|put|get) functions.
  Added $dbh->ping.
  Added DBI->data_sources('Oracle');
  $sth->rows now warns if called for select before rows fetched.
  Fixed RAW types to not truncate.
  Improved quality and clarity of trace information.
  Requires DBI 0.92

* Changes in DBD::Oracle 0.47 (Oraperl 1.35), 8th Sept 1997

  $h->{InactiveDestroy} = 1; now works reliably (with DBI 0.90).
  Makefile.PL changed for Oracle8. Thanks to Philippe Vanhaesendonck.
  Long params now work. Thanks to Michael Harvey.
  (Long params don't yet work for inout params.)
  AutoCommit flag now per-dbh. Thanks to Irving Reid.
  Fixed panic: _dbd_rebind_ph when binding an undef.
  Added $dbh->ping method (for Apache::DBI users).
  Some field-level fetch errors didn't cause the fetch to fail
  (the field was simply set undef).
  LongReadLen now works (if $Oraperl::ora_trunc unset or <= 0)
  LongTruncOk now works (for non oraperl mode handle).

* Changes in DBD::Oracle 0.46 (Oraperl 1.34), 20th June 1997

  Fixed Makefile.PL to work with 5.004_01.
  Some VMS support from Dan Sugalski <sugalsd@stargate.lbcc.cc.or.us>
  If ORACLE_HOME isn't set, Oracle.pm no longer tries to guess it.
  bind_param_inout now checks for read-only variables.
  Requires DBI 0.84.

* Changes in DBD::Oracle 0.45 (Oraperl 1.33), 16th June 1997

  A $dbh DESTROY without an explicit disconnect does a rollback.
  Note that this may 'break' existing 'lazy' code but is completely
  essential for robust applications. See comments in Oracle.xs.

  Added Makefile.PL changes from Eric Bartley and others.
  The changes should fix build problems for Oracle 7.3.x.

  Requires DBI 0.83.

  Oraperl now uses DBI->connect and thus works with DBI 0.81 to
  automatically support Apache without requiring script changes.

  Reworked parameter binding in preparation for future changes.
  - mutated placeholder values are now automatically rebound.
  - in/out vars that become undef/null after binding now work.
  - transparent support for longs should be easier to implement.

  Added Win32 support from Jeff Urlwin.

  Added some documentation to DBD::Oracle for 'perldoc DBD::Oracle'.

  Most tests now converted to standard t/*.t format.
  Added $sth->{NULLABLE}->[$field].
  Added private plsql_errstr method: $txt=$dbh->func('plsql_errstr')
  to fetch PL/SQL error messages. Thanks to Bob Menteer.
  Added $sth->{ora_pad_empty} and ORAPERL_PAD_EMPTY env var
  for better compatibility with old oraperl.
  Added $sth->{AutoCommit} FETCH.
  Added $sth->{ChopBlanks} (but not yet tested).
  No longer asks Oracle for text of login failure message since
  that can cause oracle's code to hang (sigh). We provide fake text
  for the most common errors and a useful default for the rest.
  You can set DBD_ORACLE_LOGIN_ERR env var to revert to old behaviour.

    The Copyright terms for DBD::Oracle have changed and now read as follows:
  You may distribute under the terms of either the GNU General Public
  License or the Artistic License, as specified in the Perl README file,
  with the exception that it cannot be placed on a CD-ROM or similar media
  for commercial distribution without the prior approval of the author.

* Changes in DBD::Oracle 0.44 (Oraperl 1.30), 14th Jan 1997

  Fixed leak in read_blob (thanks to Jurgen Botz for the patch).
  Improved automatic cache sizing (so better default caching).
  Negative cache size specifies desired cache/transfer size in bytes.
  Added $rowid = $csr->{ora_rowid} attribute (untested, please test).
  (Use via $csr->bind_param(1, $rowid, { ora_type => 11 });)
  Queries returning LONG's are no longer cached (so there's no
  need to set the cache to 1 explicitly to get read_blob to work).
  Added a test using string type with bind_param_inout in test.pl.
  Worked around the rather sad VMS linker case insensitivity.
  Worked around VMS linker length warning on XS...disconnect_all.
  Makefile.PL deletes non-existant files from $(COMPOBJS)
  (thanks to aburlison@cix.compulink.co.uk for the original patch)

* Changes in DBD::Oracle 0.43 (Oraperl 1.30), 29nd Oct 1996

  Fixed serious 'false ora_errno 1 after short select' bug.
  Worked around oracle bug that makes cda->ft unreliable.
  Do not use DBD::Oracle 0.41 or 0.42.
  Cursors are now 'describe'd at prepare time thus making
  NUM_OF_FIELDS always available. Describe does nothing for
  non-select operations. NUM_OF_FIELDS > 0 is now used to
  select between oexec() and oexfet() in execute().
  Added more internal debugging. Improved test.pl.

* Changes in DBD::Oracle 0.42 (Oraperl 1.30), 28nd Oct 1996

  Fixed serious 'cache empty after re-bind' bug.
  Do not use DBD::Oracle 0.41.
  Implemented oexfet (combined execute and cache fetch) for
  select operations. This is a further significant speed up.
  Many selects now make only one trip to Oracle (after prepare)
  which combines the execute and fetching multiple rows.

* Changes in DBD::Oracle 0.41 (Oraperl 1.30), 22nd Oct 1996

  Added the long overdue row cache to DBD::Oracle.
  (Thanks to Reetnem@aol.com for providing a patch that prompted
  me to complete the work.)
  Oraperl $ora_cache and cache parameter to ora_open now work.
  Default cache size is adjusted automatically for row width.
  Major reworking of field buffer memory management.
  Added a more internal debugging.
  Further updates to the README files.

* Changes in DBD::Oracle 0.40 (Oraperl 1.29), 14th Oct 1996

  WARNING - This release contains significant changes to the
  placeholding binding code. You should test it carefully
  before using in live systems.

  Implemented PL/SQL output values via $sth->bind_param_inout.
  See the code at the end of test.pl for example usage.

  Binding is now implemented using obndra rather than obndrv.
  This may have a subtle effect when matching char fields
  against placeholders with trailing spaces.

  Fixed bind_param ora_type attribute. Thanks to Stephen Zander
  for the patch.  Updated README's. Added README.longs.

* Changes in DBD::Oracle 0.39 (Oraperl 1.29), 23rd Sep 1996

  Fix for DEC "target := MACRO = string" Makefile syntax.
  Added README notes from Dave Moellenhoff, Lou Henefeld and others.
  Added README.login from James Taylor.
  Added README.client with various notes about building DBD::Oracle
  on minimaly configured client systems.
  Extra parameters to ora_do are now passed to DBI's do().
  (This is an extension to the original oraperl ora_do.)

* Changes in DBD::Oracle 0.38 (Oraperl 1.28), 22th Aug 1996

  Overhaul of Makefile variable parsing. Should now cope with
  complex variables which expand to nested shell escapes.
  Try perl Makefile.PL -v to watch the fun.
  Updated README, Makefile.PL and test.pl messages for clarity.
  Fixed possible memory corruption in dbd_bind_ph().

* Changes in DBD::Oracle 0.37 (Oraperl 1.28), 25th July 1996

  Fixed Makefile.PL for Oracle 7.3.2.
  Fixed $num_fields = ora_fetch($csr) before first fetch for
  	queries with bind vars.
  Fixed occasional core dump on global destruct.

* Changes in DBD::Oracle 0.36 (Oraperl 1.28), 10th July 1996

  Fixed bind_param ora_type attribute.
  Fixed preparse to allocate enough memory for worst case.
  Fixed broken HP-UX 10 check in Makefile.PL.
  Other assorted Makefile.PL improvements.
  (Many thanks to those who sent in fixes.)

* Changes in DBD::Oracle 0.35 (Oraperl 1.28), 21st June 1996

  Fixed broken Solaris 2.5 check in Makefile.PL.
  Added ld path to the log.

* Changes in DBD::Oracle 0.34 (Oraperl 1.28), 21st June 1996

  Workaround Solaris 2 bug #1224467 (_rmutex_unlock).
  With many thanks to James Taylor.

  Added 'Bad free()' warning suppression to ora_logon and
  ora_logoff(). Setting the DBD_DUMP environment variable
  will trigger a (handy for me) core dump if a Bad free
  warning is detected.

  Further additions to the README about Bad free()'s.

* Changes in DBD::Oracle 0.33 (Oraperl 1.27), 19th June 1996

  Added Makefile.PL -g option to enable debugging.
  Added Makefile.PL -s symbol_name option to search for symbols.
  Reorganised the way Makefile.PL uses MakeMaker liblist code.
  Oraperl defaults to the 'safe' (normal) mode of using the DBI.
  Oraperl uses sigtrap on SEGV & BUS to give a perl stack trace.
  Added README notes about -g option, core files and stack traces.
  Small change to $dbname/$user/$passwd logic in connect.

* Changes in DBD::Oracle 0.32 (Oraperl 1.25), 30th May 1996

  Fixed memory leak when FETCH'ing attributes.
  Fixed Makefile.PL FileHandle problem (forgot 'use FileHandle;').
  Enhanced Makefile.PL support for Oracle 7.3.

* Changes in DBD::Oracle 0.31 (Oraperl 1.25), 20th May 1996

  Makefile.PL for HP-UX now defaults to dynamic for hpux >= 10.

  execute (ora_bind/ora_do) now returns undef on error or the
  number of rows affected (0 is returned as 0E0, hence true, for
  okay but no rows affected or no row count available/applicable).
  This matches the oraperl version 2.4 behaviour.

  Made an attempt at supporting Oracle 7.3 (e.g. include's in proc.mk)
  Please let me know if it works (else supply patches to fix it :-)

  Fixed small memory leak in ora_titles etc functions.

* Changes in DBD::Oracle 0.30, 7th May 1996

  THE ORAPERL EMULATION LAYER IS NOW FORMALLY RELEASED (NO LONGER ALPHA).

  Note that the underlying DBI and DBD::Oracle interfaces remain alpha
  because they are still subject to (possibly significant) change.

  Oraperl v2 used to return the string 'OK' to indicate success
  with a zero numeric value. The Oraperl emulation now uses the
  string '0E0' to achieve the same effect since it does not cause
  any -w warnings when used in a numeric context.

  Fixed typecast warning (s/safefree/Safefree).
  Automatically sets/resets ORACLE_HOME from oratab value for sid.

  TO DO: Automatic configuration from Oracle 7.3 is not yet working
  (Oracle have reorganised the makefiles yet again!).

* Changes in DBD::Oracle 0.29, 2 March 1996

  Fixed Makefile.PL to get DBIXS.h from right spot.
  Changes to suit perl5.002 and site_lib directories.
  Detects old versions ahead of new in @INC.
  Random tidy-ups.

* Changes in DBD::Oracle 0.28, 29 Jan 1996:

  Minor release for Perl5.002 (beta2 or later).
  Requires Perl5.002 and DBI 0.66;

  Only functional change is to suppress (rare) 'handle not setup'
  warnings unless debugging.

* Changes in DBD::Oracle 0.27, 16 Nov 1995:

  Improved oraperl compatibility in assorted ways.
  Added the original oraperl manual to Oraperl.pm as pod.
  (perldoc Oraperl will display the manual).
  The manual highlights remaining differences in the emulation.
  Added original oraperl examples into oraperl.ex directory.

  Improved error reporting in test.pl.
  Removed some internal limits.
  Renamed readblob to blob_read and fixed return value.
  Implemented $sth = $dbh->tables method (see DBI changes file).
  Uses $Config{archlibexp} not $Config{archlib} in makefile.PL
  Requires DBI-0.65.

  I've still not got around to implementing a row cache (array
  fetch) but it's next on my list.

  This should be the last alpha release of the oraperl emulation.
  (The DBD::Oracle module itself will remain alpha for awhile yet.)


* Changes in DBD::Oracle 0.26, 23 Oct 1995:

  Removed error message (sql) size limits in parse and ora_error.
  Fixed FETCH NUM_OF_PARAMS so it doesn't trigger a describe.
  The ora_lengths and ora_types attributes work in non-oraperl mode.


* Changes in DBD::Oracle 0.25, 26 Aug 1995:

  User visible changes:

  Fixed bug in parsing '?' style placeholders.
  $sth->bind_param now checks for a too long LONG type string.
  $sth->prepare now takes attributes:
  	ora_parse_lang  => 0 | 1 | 2  (v6, auto v6/v7, v7)
  	ora_parse_defer => 0 | 1      (control defered parsing)

  Other changes to be propogated to other drivers:

  Added usage of new DBD_ATTRIB_* macros to XS and C code.
  dbd_bind_ph return value inverted to make it consistent.
  Added braces around XSRETURN_UNDEF in execute.
  Checked for describe failure in FETCH.
  Removed NumParams since it's now called NUM_OF_PARAMS in DBI.
  Updated NEED_DBIXS_VERSION in Oracle.h


* Changes in DBD::Oracle 0.24, 22 Aug 1995:

  User visible changes:

  $sth->bind_param method implemented (needs testing :-)
  and new DBI $sth->bind_col* methods work for DBD::Oracle.

  Other changes to be propogated to other drivers:

  - Slight change to STORE functions to allow them to return status
    if called as $h->STORE(...).

  - Revised signature of dbd_bind_ph function and added a bind_param
    method for it. Prototype moved from dbdimp.h to Oracle.h

  - Revised execute method to use modified dbd_bind_ph function and
    remove remaining oracle specifics.

  - imp_xxh_t structures first element now called 'com' not 'dbihcom'
    and new DBI macros used to access fields.


* Changes in DBD::Oracle 0.23, 18 Aug 1995:

  Fixed Oraperl.pm debug which was left on by default by accident.
  Added small patches from Davide.Migliavacca@inferentia.it for
  longs (dbtype_is_long, dbd_describe and dbd_st_readblob).
  Added svr4 hints from Alan Burlison <aburlison@cix.compulink.co.uk>
  Changed dbd_describe to return true for success (to be consistent).


* Changes in DBD::Oracle 0.22, 17 Aug 1995:

  Much more maturity, in line with the required DBI-0.60 release.
  The .xs file is now very 'clean'. It's an excellent base for
  developing other drivers. See notes in the DBI Changes file.

  The oraperl &ora_do should be working now (along with $sth->do()).
  It does not leak (as far as I can tell).

  This release is stable enough for general use again (like 0.20).
  Please test heavily.


* Changes in DBD::Oracle 0.21, 15 Aug 1995:

  NOTE: THIS IS AN UNSTABLE RELEASE!
  It requires the closely related DBI 0.59 release.
  See Changes file in that release.
  Major reworking of internal data management!

  Only execute and fetchrow have yet to move into dbdimp.c.
  These were kept back since I plan other associated changes.

  Known problems:
  Warning mode is on by default so certain oraperl coding
  styles cause warnings such as:
      Statement handle ... destroyed without finish() at ...
      disconnect(...) invalidates 1 associated cursor(s) at ...
  In future oraperl mode handles will have those warnings disabled.
  ...other changes to numerous/minor to mention

* Changes in DBD::Oracle 0.20, 1 Aug 1995:

  Fixed core dump when binding an undef (treated as a NULL)
  Binding a string longer that 2000 bytes will use LONG type.
  Workaround OSF makefile and oratype.h problems
  PL/SQL := construct will no longer confuse dbd_preparse()

* Changes in DBD::Oracle 0.19, 21 June 1995:

  Added $VERSION
  disconnect_all now gives error 'not implemented'.
  Reworked memory management, imp_dbh's are cached.
  imp_dbh_t now has an in_use flag and generation counter.
  st::DESTROY now checks its imp_dbh for validity.
  &ora_do() now returns "OK" for 0 rows (as per oraperl).
  Reworked logging to use DBILOGFP macro.
  Now test code added to loop through logon/prep/fin/logoff.

* Changes in DBD::Oracle 0.18, 17th June 1995

  Makefile.PL: HPUX now builds with LINKTYPE=static automatically.
  Fixed errors on logout/global destruction.
  Added logout/global destruction test to test.pl.
  Changed Oraperl.pm default mode from safe to fast (see func_ref()).
  Added platform who's-who list to README.
  Added $sth->readblob($field, $offset, $length [, \$dest) method.
  (You need to add the following after line 80 in DBI/DBI.pm:
     'readblob' =>   {'U'=>[4,5,'$field, $offset, $len [, \\$buf]']},
  in order to access the new readblob method.)

* Changes in DBD::Oracle 0.17, 9th June 1995

  Further enhancements to proc.mk parsing (this should fix
    the bug introduced in the last version which required the
    manual editing of -lora etc into the link line).
  Automatic use of proc16.mk if proc.mk not found.
  Automatic use of -Xa if compiler is 'clcc' (CenterLine).
  Further migration of code from Oracle.xs to dbdimp.c. This is
    not complete but may give you a flavor of my current direction.
  Fixed more warnings from pedantic compilers.
  [Released with DBI 0.56]

* Changes in DBD::Oracle 0.16, 25 May 1995

  Makefile.PL more robust and flexible.
  Code restructured - much more still to do.
  Changes to ora_do. It's probably still not right.
  [Released with DBI 0.55]

* Changes in DBD::Oracle 0.15, 25 Apr 1995

  Fixed all known build problems/core dumps/truncated data etc.
  Returns undef for NULLS.
  Bind vars still not supported (soon I hope).
  Bind vars produce errors rather than being ignored.
  Login usage of ORACLE_SID and TWO_TASK slightly revised.
  test.pl now uses ORACLE_SID, TWO_TASK and ORACLE_USER if defined.

* Changes in DBD::Oracle 0.14, 5 Apr 1995

  Fixed assorted build and core-dump problems.

* Changes in DBD::Oracle 0.13, 8th March 1995

  "Another alpha release but hey, guess what, this one actually allows
  you to fetch data :-)"

* Changes in DBD::Oracle 0.12, 1st March 1995

  "For DBD::Oracle I have implemented a fairly radical way to deal with
  the nightmare of Oracle OCI makefiles.  Rather than try to maintain a
  *large* and complicated set of Makefile macros the DBD::Oracle
  Makefile.PL now actually extracts the macros which oracle itself uses
  from $ORACLE_HOME/rdbms/lib/oracle.mk!  (The horses mouth, so to speak.)"
  [Released with DBI 0.53]

* Changes in DBD::Oracle 0.11, 24th Feb 1995

  "The first fruits of that work are now available for the brave (or
  foolhardy).  Take note: this very much *unsupported* *alpha* software
  (it does not even fetch records from oracle yet). Feel free to ftp it
  but don't expect to do anything very useful with it."
  Also first release of the Oraperl emulation module:
  "Yes, that's real oraperl code and it works! The only addition is
  the 'use DBD::Oraperl' line."
  [Released with DBI 0.51]

* ANCIENT HISTORY

12th Oct 1994: First public release of the DBI module.
               (for Perl 5.000-beta-3h)

19th Sep 1994: DBperl project renamed to DBI.

29th Sep 1992: DBperl project started.