About 20 years ago, all database system was used to face the problem of data integrity, data security, data recovery and back up of databases because of the enormity of data and concurrent changes. It was difficult to handle these databases along with the problem of sharing information among different applications and different servers (Batty, 1992).
Earlier GIS information was stored in a two-step way. Graphical information was stored separately in dedicated data structure, whereas alphanumeric data was stored in relational database. This was mostly done because of performance issues. This was known as non-database approach, which faces the biggest problem that graphical and non-graphical was kept in two different data structure.
At this point of time, IBM came into picture with an object oriented approach. According to this approach, the objects were managed by software and tool launched by IBM and all the attributes of an object such as alphanumeric, graphical representation, etc. are stored in relational database. The name of the tool was geoManager. This approach was used to store and manage GIS data, not for analysis and to manipulate data.