Prior to the fix for DBCP-333, DriverManager was always used to load drivers. As documented in the patch and updated javadoc, DriverManager is now used only when loading via configured driverClassLoader or the context classloader fails. Might this change cause problems for implementations that somehow rely on the fact that DBCP has up to now used DriverManager to load drivers? What brought it to mind for me was the test case submitted in the patch for DBCP-342, which did rely on this (undocumented) behavior.
Phil --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org