The trade press and industry pundits have put a spotlight on SOA, Web 2.0 and search engines. But enterprise architects and system architects understand the contribution of industrial-strength enabling technologies to the success of distributed applications and web sites.

There’s a steady stream of innovation in areas such as data visualization, user presentation, 3-D graphics, social networking, streaming video and business intelligence. But innovative applications often rely on mature core technologies, such as TCP/IP and SQL. Database software vendors, such as IBM, Microsoft and Oracle, continue to evolve their SQL platforms, spurred on in part by competition from open source products such as MySQL and PostgreSQL. Nonetheless SQL is mature technology for persistence and applications such as online transaction processing (OLTP).

The software architecture for multi-database access and the APIs represent a mature, enabling technology. Perhaps that’s why some system builders incorrectly assume database middleware is commodity software, with all drivers and providers being equal in quality, capabilities and compliance with standards. Database gurus and knowledgeable system architects don’t buy into the “all drivers are equal” argument. Nor do technologists at leading software vendors who select drivers and data providers for their products. More than 300 software companies license and distribute DataDirect middleware including:

• The top four vendors in the SQL database market
• The six leading vendors of application servers for J2EE and Java EE platforms
• 16 of the top 20 business intelligence (BI) vendors
• 12 of the largest vendors of content management, knowledge management and portal software
• 11 of the industry leaders with data integration and information integration products
• More than 80% of the leading application software, systems infrastructure and packaged software vendors.

SQL and dropping hardware prices helped launch the distributed computing wave when DBMS vendors moved to client-server architecture. The decreasing cost of computers was a driver for application partitioning, client-server computing and distributed processing. A distributed architecture offers benefits such as eliminating single points of failure and throwing more hardware at performance problems. The adoption of distributed computing has gone hand in hand with organic growth of high-performance and high-availability computing. Distributed processing and distributed data were a causative factor in the surge of interest in integration with enterprise applications, mainframe computers and legacy databases.

More computing capacity enables serving more users and improving application and database performance. It improves the response time for users and reduces the time lag for making information available. For example, processing during the mainframe era was batch-oriented because of limited processing capacity. During that period, business users would accept a window of five days for month-end processing on the computer. Today IT departments deal with totally different expectations, such as users wanting real-time business intelligence. Today’s software solutions must not only be robust, they must also be fast and scalable.

The adoption of client-server SQL technology brought major changes. Data access techniques evolved, dynamic SQL gained prominence and a new type of middleware emerged. In the early days of SQL application programming, developers used embedded SQL or a proprietary API. Embedding static SQL or dynamic SQL in source code requires re-compiling programs for adapting to heterogeneous databases having different types and features.

Embedded SQL, compile-time optimization and static SQL are a classic, performance performance-oriented solution for database developers. But programmers who required greater flexibility opted for dynamic SQL and run-time invocation of data access libraries. Developers who were writing multi-DBMS client software weren’t working with uniform data distribution, known queries and schemas. They wanted a solution that, unlike embedded SQL, didn’t require recompiling the client code for each targeted DBMS.

The desire for database platform independence and a vendor-neutral API led to the emergence of Open Database Connectivity (ODBC), and subsequently JDBC. Both APIs enable developers to write dynamic SQL, although DataDirect middleware uses bind packages when possible to provide better performance than dynamic SQL. The .NET Framework introduced another platform-neutral data access solution. ADO.NET is an API, optimized for the .NET platform, which complements the Common Language Runtime, ASP.NET and other .NET technologies.