Permalink
Switch branches/tags
Commits on Jul 11, 2012
  1. release v0.08198

    frioux committed Jul 11, 2012
Commits on Jul 10, 2012
  1. fix silent Oracle connection failures

    ::Storage::DBI::_server_info was eating connection exceptions by
    ignoring exceptions from $dbh->get_info, but this information is
    important for _rebless when e.g. determining which class to use based on
    the DB server version, so rethrow the exceptions when
    $self->{_in_determine_driver} is true.
    rkitover committed Jun 24, 2012
  2. better type check for mssql+ado binary null strip

    The data_type in the t/lib result class for the blob for the tests
    changed to 'blob', while the binary type null stripper only expected
    'binary' or 'image' data types, which is wrong for possibly
    cross-deployed schemas.
    
    Change _strip_trailing_binary_nulls in ::ADO::CursorUtils to use the
    newly added and more generic ::Storage::DBI::_is_binary_type instead of
    just checking for 'binary' or 'image'.
    rkitover committed Jun 24, 2012
  3. mark some tests in admin/10script.t on Win32 TODO

    The test_dbicadmin() tests are failing on Win32 due to some sort of
    quoting issue with the temp directory. I'm not sure how the temp
    directory is being used here, so it will have to wait for someone else,
    in the meanwhile I marked the tests TODO.
    
    Also changed the " backwhacking to use Win32::ShellQuote instead and
    added it to optdeps, unfortunately that didn't fix the problem, since
    the problem seems to be some sort of DBICTest::RunMode interaction with
    dbicadmin.
    rkitover committed Jun 24, 2012
  4. preload threads.pm in storage/txn.t for fork error

    SineSwiper reported this error in t/storage/txn.t on Strawberry Perl
    5.12 32bit:
    
    You must load the 'threads' module before Moo::_Utils at t/storage/txn.t
    line 130.
    
    Preload threads in a BEGIN block at the beginning of the .t file for
    Win32. This fixes the error.
    rkitover committed Jun 18, 2012
  5. skip 51threadnodb.t on Win32 on < 5.14

    SineSwiper reported a crash with a dialog from "make test" in
    t/51threadnodb.t on Strawberry Perl 5.12 32bit.
    
    The test passes on my ActivePerl 64bit 5.14.2 and there is a passing
    test report for 5.14 here:
    http://www.cpantesters.org/cpan/report/239d2ae4-87c3-1015-ad08-a09405085451
    
    Disable test on Win32 for perls below 5.14 unless the environment
    variable TEST_AUTHOR is set to be safe.
    rkitover committed Jun 18, 2012
  6. release 0.08197

    frioux committed Jul 10, 2012
  7. fix pod for unicode name

    frioux committed Jul 10, 2012
Commits on Jun 20, 2012
Commits on Jun 15, 2012
  1. fix broken syntax

    arcanez committed Jun 15, 2012
  2. Add authority for new moules

    ribasushi committed Jun 15, 2012
  3. fix tests on solaris

    frioux committed Jun 15, 2012
Commits on Jun 5, 2012
Commits on May 30, 2012
  1. release 0.08197_01

    frioux committed May 30, 2012
Commits on May 23, 2012
  1. fix for RT#77241

    arcanez committed May 23, 2012
Commits on May 15, 2012
  1. fix typo

    arcanez committed May 15, 2012
  2. add more docs on software_limit resultset attr

    Add a paragraph to Storage/DBI/Sybase/ASE and a section under SEARCHING
    to Cookbook.
    rkitover committed May 15, 2012
Commits on May 8, 2012
Commits on May 6, 2012
  1. Remove mention of obsolete DBIx::Class::WebForm.

    Mention HTML::FormHandler::Model::DBIC instead, which is actively
    maintained.
    ilmari committed May 6, 2012
  2. Remove/fix bogus defaults values in example

    is_auto_increment already provides a default, and for the other columns
    we either don't want a default value or it needs to be an integer.
    Explicit is_auto_increment => 0 is also pointless and distracting.
    ilmari committed May 6, 2012
Commits on Apr 27, 2012
  1. Adjust tests for new Moo

    ribasushi committed Apr 27, 2012
Commits on Apr 21, 2012
  1. Remove realiasing overengineering introduced in 86bb5a2

    mst is right - there is no viable use case for this, cleanse with fire
    Also I managed to misspell subUery... twice
    ribasushi committed Apr 21, 2012
  2. More limit torture

    ribasushi committed Apr 21, 2012
  3. Do not artificially order the internals of a has_many prefetch subquery

    Thre is no point using the artificial order which is only needed for efficient
    collapse. The user didn't request an order - this means any rows will do
    ribasushi committed Apr 21, 2012
Commits on Apr 20, 2012
  1. Fix test failure on MSWin32

    ribasushi committed Apr 20, 2012
  2. Make sure order realiasing remains in proper sequence on sorting

    We now support 999 realiased order criteria, this is beyond sufficient
    ribasushi committed Apr 20, 2012
Commits on Apr 18, 2012