Topics related to database management software have been major research areas. Functional Data Language (FDL) tries to define entities and functions between entities. The integration of networks and relational data structures is possible using FDL. Object-oriented data models are becoming popular, and FDL may help in defining object-oriented database systems.

Poulovassilis’s paper is an attempt to improve FDL. Most of the FDL concepts use the relational model, and only functions are stored in the database. This paper overcomes this limitation.

FDL is based on &lgr;-calculus with recessive functions and partial definition of functions. All the functions are stored in a common repository. Data types are nested lists, sums, and products. They can be arbitrary and persistent.

The FDL architecture consists of a parser, a type checker, a printer/evaluator, an assembler, and an integrity constraint enforcer. The author describes these modules in detail. Meta-data can be handled like any other data. Matching is implemented without backtracking from the root to its leaves.

The paper is a definite contribution to the current research on FDL. Researchers and students in this field may find the details relevant. We will have to wait for a commercial implementation of these ideas. Recent papers cited by the author will help the reader to access additional FDL literature.