13.2. SQL language in commercial implementations

Currently, SQL is implemented in almost all commercial relational DBMSs, all firms declare their implementation to the SQL standard, and in fact the implemented SQL dialects are very close. This did not happen immediately or simply.

The closest to System R were two IBM systems - SQL / DS and DB2. As it seems (the author does not have any documentary evidence of this), both of these systems directly used the System R implementation. Hence, the close proximity of the implemented SQL dialects to SQL System R. Only parts that were insufficiently developed (for example, savepoints) were deleted from SQL System R. ) or the implementation of which caused too many technical difficulties (for example, integrity constraints and triggers). You can call this path to the commercial implementation of SQL from top to bottom.

Another approach was used in systems such as Oracle and Informix. Despite the difference in the way these systems were developed, the implementation of SQL was "bottom-up." In the first SQL implementations released to the market, these systems used a minimal and very limited subset of SQL System R. In particular, the first known implementation of SQL in the Oracle DBMS did not allow the use of nested subqueries in the sample statements.

Nevertheless, despite these limitations and very poor at first efficiency, the orientation of firms to support different hardware platforms and user interest in the transition to relational systems allowed firms to achieve commercial success and begin to improve their implementations. In current versions of Oracle and Informix, quite powerful SQL dialects are supported, although the implementation is sometimes in doubt.

A feature of most modern commercial DBMSs that makes it difficult to analyze existing SQL dialects is the lack of a complete description of the language. Usually, the description is scattered in various manuals and mixed with a description of language-specific tools that are not related to SQL. Nevertheless, it can be said that the basic set of SQL statements, including database schema definition statements, data sampling and manipulation, data access authorization, support for embedding SQL in programming languages ​​and dynamic SQL statements, is relatively stable in commercial implementations and more or less corresponds to the standard .