There was much debate about whether data access middleware should support a proprietary application programming interface (API) or a standard, multi-database API. The issue was settled by widespread adoption of multi-database APIs. Today tools, frameworks and developer languages make extensive use of SQL API standards, including JDBC™ and Open Database Connectivity (ODBC). Following the emergence of middleware based on those standards, the software industry raced to develop libraries, database engines and data-aware components that operate with standards-based middleware. It was here that many performance bottlenecks were found that were erroneously attributed to standard APIs and middleware.

When the performance myths were de-bunked, the major software vendors started using standards-based middleware for multi-database access by Oracle Open Gateway, IBM DataJoiner and other products. Eventually data access middleware and drivers were bundled with application servers, integration servers, business intelligence servers and many platform products.

Data access middleware enables connected and disconnected clients, including middle-tier servers, to communicate with remote or local SQL servers. But distributed application design requires attention to security, scalability and performance.