First, Spring makes it possible to directly test the projects' object oriented methodology for application development in a commercial system. Indeed, Spring follows an object oriented data model that provides a representation for real world geographical entities, distinguishes among their various kinds and links geometrical and non-geometrical data, uniformly accounting for vector, grid and raster data geometries.
Second, Spring involves research and development in the extension of database technology, which is also part of domus' goals. Thanks to this characteristic, spring has been developed using a modular and layered architecture based on DBMS. Therefore, it is possible to use different database systems underneath it, by changing the interfaces between the upper-level modules and the supporting database. Thus, one of the experiments to be conducted is the coupling of the Spring system to an object oriented DBMS. The experiment will evaluate, for example, how to incorporate object oriented technology on the image processing subsystem, which typically handles objects of about 250M and implements methods such as histogram calculations, image filters, etc. The size of image objects makes it impractical to directly use page-to-page client-server architecture. In other words, one cannot simply transform images into persistent objects of such systems without running the risk of unacceptable performance degradation.
Third, the code has been made available to the DOMUS and GEOTEC projects. Therefore, it will be possible to make different experiments on alternative data structures, as well as on the extensibility of the system itself. For example, one will be able to introduce new storage methods for image objects to circumvent the above-mentioned problems and implement new access methods for vector data. Experiments with new user interface paradigms are also planned which, in particular, expand the merging of hypermedia and GIS concepts.
For more information on Spring, contact
Gilberto Camara