JDBC 3: Java Database Connectivity

ODBC versus Specific Drivers as a Subprotocol

ODBC is available on a wide variety of platforms, including the Windows, UNIX, and Macintosh environments. This wide use isn’t enough to make a Java-based project totally portable, but it is sufficient in many cases.

Using an ODBC leads to the same problem of deploying a non-Java, native underlying layer. It is very difficult to use such solutions in uncontrolled environments because the support for ODBC may be limited or not available at all, but you may consider using an ODBC bridge or a native driver during the development phase, for example.

The reasons such solutions are difficult to use are as follows:

• ODBC is a native component dependent on the platform.

• ODBC uses a local Registry or configuration file to look up data source names.

• ODBC uses native drivers that are dependent on the platform.

• ODBC is software that must be administered on each client platform separately.

On the other hand, a Java native-protocol driver, eventually using a third tier that does most of the job, offers these benefits:

• It is independent of the platform it is running on.

• It uses a universal mechanism to name data sources — JDBC URLs.

• It may be upgraded automatically (in the case of applets using Type 4 drivers).