by Jose Luis Borbinha
Following the global interest about the Web, INESC has carried out a few experiments using this technology for organisational issues.
Our experiences started in the spring of 1994, with the implementation of a concept we called the "Organisational Folder". This would be a public information system, based in an X.500 repository and a WWW interface. It would also provide information about people working for INESC, as well as information about each individual research group. The information would be maintained by both the administrative services and the owners themselves (workers and groups). However, the lack of flexibility of X.500 and the poor capabilities of HTML 1.0 soon unveiled the dificulty of achieving a system that could do more than just the usual and simple X.500/WWW interfaces.
Meanwhile, the use of the Internet in internal corporate services became a subject of growing interest. Much work has been done in this area, which has been marked as one of the more exciting opportunities in the near future. A new expression was even coined about it: "Intranet services".
At the end of 1994 a project was started from this new perspective called: InInWeb (INESC In the Web).
The opportunity arose when the administrative services felt the need for a management system that could help them keep track of the INESC participation in the European Community (EC) programs. The system should know all about the EC projects in which INESC participates, such as their contracts, researchers and groups involved and their mile stones (such as deliverables and cost statements to issue). As almost everyone at INESC works, or will work, for an EC project, it required information about almost everything and everyone.
It was decided to use the database developed for this purpose and, with the same effort, to develop the InInWeb system as an initiative to provide a uniform WWW interface for both an internal and a public view of the organisation (preserving the necessary security levels).
The first prototype was scheduled for January 1995, but it was ready in June. There were several reasons for this delay, but one of the most important was again the state-of-the-art Web technology. In fact, when we started the project, the only available clients supported HTML 1.0, with extensions for forms (there were no such things as tables and Java). We were able to produce a useful public information service, but not an administrative information service for users not familiar with the hypertext paradigm. This taught us a well known lesson: you can have the vision, but it will fail if you start it too soon.
The final version takes advantage of HTML 2.0, and has been in a test phase since November 1995. It has been running in a closed environment with a good acceptance (for now, it is available only inside the INESC domain). We are still improving the interface and tuning some administrative procedures related to the existence of the system, before making it fully accessible to the public.
The implementation of the InInWeb service consists of a CGI application and an Oracle database. The CGI application is a C++ program based in the GENESIS library. This is a library of C++ classes that was developed for this kind of system and using concepts like:
It has been possible to use the GENESIS concept in other similar systems. Using the accumulated experience it is now possible to reformulate the model. The purpose is to develop an Object-Oriented approach (and the technology to support it) for the coherent development of Web interactive services.