|
The WEL will be a new system, not carrying the burden of legacy code. We are free to choose
any technology for the implementation, to achieve the goal of building an information system
which is...
- an advanced system meeting the users' needs.
- a reliable production system.
- open and thus able to interoperate with other systems.
The analysis of the problem has shown that an information system to support art historians will
have much in common with systems like hypermedia authoring tools and document management systems.
The possible use of such systems for the WEL has yet to be evaluated.
Currently the WEL data is stored in a relational database. On top of this several implementations of
the application are tested. The next alternatives to be pursued are:
- Tycoon: The current version of Tycoon is the outcome of a long systems development
tradition at the STS institute and its predecessors. It contains the purely object oriented Tycoon
Language, implemented in a compiler and a virtual machine. It offers orthogonal persistence and mobility
of data, code, and threads.
- Java: Java is a language which has achieved remarkable acceptance in the past
years.
It builds upon concepts of research prototypes like Tycoon, and is now being extended in many
directions.
Though it still lacks some features, like persistence, it has advantages because of its wide-spread
use.
The use of Java enables one to make use of many pre-fabricated components.
- Off-The-Shelf components: The mentioned systems like document archives seem to offer
much of the functionality needed by the WEL. There is hope that they can be integrated in
part, and that
ideas will be exchanged with the document processing community. Possible languages for additional programming
might be constrained by a concrete system.
It is planned to develop tools for different purposes using the different
technologies. This way all
technologies will be used on an every-day basis, and thus become comparable.
|