Greetings, I think I have my very first patch to submit. "Contributing Patches" <http://wiki.services.openoffice.org/wiki/Contributing_Patches> suggests that I take it to d...@openoffice.org first, but I am bringing it here instead. Here, with some rationales and questions, is what I contemplate ...
component : Database access subcomponent : none ( because ODBC is not offered ) found in version: DEV300_m75 platform : other os : Linux priority : P5 ( unless some platform is less forgiving than mine, perhaps? ) issue type : patch initial state : new assigned to : oj ( because that is what <http://qa.openoffice.org/issue_handling/submission_gateway.html> does when you click "connectivity" ) summary : operator delete mismatch in ODatabaseMetaDataResultSet description ... In file connectivity/source/drivers/odbcbase/ODatabaseMetaDataResultSet.cxx, line 86 assigns m_pRowStatusArray from array new but line 100 does a scalar delete. In a non-production build, this causes numerous assertion failures at sal/cpprt/operators_new_delete.cxx line 95 when, for example, you open a query to an ODBC database. And then the patch ... --- ODatabaseMetaDataResultSet.cxx_m75 2010-03-21 23:37:01.000000000 -0400 +++ ODatabaseMetaDataResultSet.cxx_tje 2010-03-21 23:38:03.000000000 -0400 @@ -97,7 +97,7 @@ osl_incrementInterlockedCount( &m_refCount ); dispose(); } - delete m_pRowStatusArray; + delete [] m_pRowStatusArray; } // ------------------------------------------------------------------------- void ODatabaseMetaDataResultSet::disposing(void) I have tested this only to the extent of observing fewer assertions raised as I open my favourite query against an ODBC database. Should I say more? Less? Something else? I welcome your comments. Cheers, Terry. --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@dba.openoffice.org For additional commands, e-mail: dev-h...@dba.openoffice.org