Object-Oriented Databases Management Systems (OODBMS) allow management of data which is described as objects and classes, instead of e.g. tables and columns as in relational database systems.
The object-oriented model for describing data (data model) has been found to have a number of advantages:
- More suited to manage complex data, which is not easily modelled as tables and columns, e.g. documents, drawings and product structures.
- Higher performance for certain types of applications, by not joining data at run-time. Data is linked using machine generated object identifiers.
- Not limited to built-in simple data types such as integer, float, string, bitmaps etc. New data types (classes) may easily be added. Examples are Video or Sound data types for multimedia type applications.
- Support for encapsulation and modularity, for managing the complexity of large models.
- Simpler to use in combination with object-oriented programming languages, since the data model of the programming language and the DBMS are identical.
- In some implementations, these database systems support storing and execution of code (methods) in the database. This is useful for developing distributed applications, e.g. by reducing network traffic.
There are many commercial products available which support this data model, ranging from simple class libraries providing persistent storage of objects for C++ or Smalltalk, to powerful multi-user systems, supporting dynamic schema evolution, transaction handling etc.
previous |
next |
start |
contents |
dictionary |
help |
evaluation
© Ericsson Telecom AB, 1995, Stockholm, Sweden