Emerging Database course Declarative programming Declarative programming Associated
Emerging Database course: Declarative programming
Declarative programming Associated with the deductive database struc ture is the declarative programming approach, in which data are grouped by properties. Controversially, the object-oriented database architecture uses im perative programming, and data are grouped by properties. Controversially, the object-oriented database architecture uses imperative programming, and data are grouped by objects. Kifer and co-workers [10] presented a criteria for classifying future database prototypes. In this classification, the prototypes Pascal-Rand GLUE represent imperative and relational approaches. The deduc tive object-oriented database architecture, such as F-logic, 0 -logic, and C-logic, stands between object-oriented and deductive database architectures. It utilizes declarative programming, grouping data by objects. Based on Kifer's point of view, "pure object-oriented relational languages are not flexible enough; some knowledge is naturally imperative whereas some is naturally declarative. "
Therefore, even though some problems exist, the DOOD will become important if those problems can be solved, because DOOD can combine the advantages of a deductive databases structure, such as explicit specification, high-level abstraction, extendibility, modifiability, and easy com prehension, with the merits of an objectoriented structure, such as the rich ness of object structure and the potential of integration. However, Bancilhon insisted that it is too early to combine object-oriented and deductive database concepts since there are still some difficulties in implementing object oriented concepts into an object-oriented database itself. He proclaimed that the object-oriented database and deductive database systems need to be developed separately rather than abruptly combining them. Moreover, in the long term, research efforts should focus on a deductive database approach, since it trans fers more work to machines. Ullman agrees with this view point. Furthermore, Ullman predicted that if it is necessary to combine these two architectures, some objectoriented features should be added to the deductive database but not vice versa. In conclusion, as Ullman addressed earlier, the full integration between objectoriented database and deductive database is impossible. However, if some concepts in these two approaches can be removed or relaxed, it is possible to combine the merits of the two approaches. Still to be solved are the questions: should we add deductive functions into object-oriented databases or add object-oriented functions into deductive databases? and which approach will provide the greatest promise for the next database generation?
REFERENCES • Timon C. Du. , Emerging Database System Architectures • Bochmann, G. Concepts for Distributed Systems Design • Capron, H. L. Computers: Tools for an Information Age
- Slides: 4