Picking Up the Gauntlet

Computer science and software engineering are in the middle of a very critical phase. The impressive development of new technologies such as high-speed telecommunication networks, multimedia, and the Web is going to change the way computers and software are developed and used. We are moving towards a global computing environment (Internet) that is much more than a simple interconnection device. It will be the "operating environment" of the future, that will provide the infrastructure, the information, and the software that are presently supposed to reside on each user's machine.

This is a real revolution that is challenging the computer and software industry, and also the research community. Marketing strategies, product lines, and alliances have to be rethought. Researchers must take into account new factors, issues, and problems, and must also accept the existence of new de-facto standards and trends. It is a very critical but still creative and promising period, with a variety of opportunities and risks that we must be able to properly identify and address.

More specifically, software engineering is facing many problems and risks, perhaps more that other disciplines. This is due to several reasons. First, software engineering is an indirect technology. Its "products" are not directly used by end-users, and it is quite difficult to evaluate them in terms of economic benefits for the company that uses them. For instance, it is not easy to evaluate the impact that a CASE tool or a specification technique might have on the overall business performance of a software development company or, even more difficult, on the performances of the end-user that purchased the products developed with these tools. Clearly there are qualitative and subjective indications that can be instrumental in developing such an evaluation, but still we are often left with the doubt about the real cost-effectiveness and usefulness of software engineering products and results. Moreover, software development is a creative activity where humans play a key role. Again, this implies that it is difficult to distinguish the contribution of the human from that of technology and methodology. Second, most of the methods and technologies we have built so far are not so effective to be really felt as necessary and important by software developers. Software development tools such as most configuration management tools or visual development environments (e.g., Visual Basic) have been successful because they have become absolutely crucial and extremely effective to accomplish any non-toy development activity. Would it be ever possible to manage a large product without a configuration management tool? At the same time, other software engineering technologies have not reached the same level of maturity yet. For instance, we all know that software should be specified and designed before coding, possibly using some advanced upper CASE tool. But in many cases this activity is perceived as cumbersome and not productive. Developers are more inclined to "build" the product rather than designing and documenting it. Most CASE tools are still perceived as high-level word processors that do not contribute directly to the construction of the solution, i.e., the code. Their adoption does not appear as a means to radically change and improve the way developers operate. Third, software engineering research is more and more oriented to short term results. This means that universities and academic institutions are often competing with industry. At the same time, we often miss the breath and perspective of medium and long term research. What are the long term goals of research? What are the problems that we as researchers should try to anticipate and address?

Organizing a software engineering conference in such a turmoil could easily become a nightmare. Still the organizers of ICSE 97 have decided to "pick up the gauntlet" and play the game. We are organizing the conference having in mind specific solutions for the above problems:

1. The program and the proceedings will include industrial experiences, longer-term research works, and even opinions and suggestions for new research directions. This wide spectrum of contributions will allow us to collect and compare different results, experiences, and viewpoints. Nonetheless, they will be kept clearly distinct, in order to better appreciate their relevance, exploitability, and maturity. This is instrumental to the evolution and consolidation of the discipline.
2. We will give a special emphasis to industrial experiences and empirical studies. This is definitely needed to provide a solid background of facts and consolidated results that can guide future research activities, and facilitate the wide dissemination of the state-of-the-art in the field.
3. We will promote the evaluation and diffusion of new ideas. In particular, we will provide opportunities for young researchers (specifically, Ph.D. students) to present and discuss their research activities.
4. In the selection of the program committee and in the definition of the programme we are paying much attention to the disciplines which are related to software engineering, such as multimedia, Computer Supported Cooperative Work (CSCW), the Web, and human computer interaction.

As a sneak preview of the conference program, we can announce that Ed Yourdon has accepted to be invited speaker. Also we are aligning an interesting set of workshops and co-located events, that will be held before and after the conference:

• Symposium on Software Reuse
• Workshop on Living With Inconsistency
• Second Cleanroom Workshop
• Second IFIP International Workshop on Software Engineering for Parallel and Distributed Systems
• Workshop on Process Modeling and Empirical Studies of Software Evolution
• 7th International Workshop on Software Configuration Management (SCM7)
• Workshop on Software Engineering and the World Wide Web

Finally, we have received several interesting proposals for tutorials. These are some of titles that we are considering:

• A survey of object-oriented analysis and design methods
• Integration of CORBA and Java
• Distributed software architecture
• Methods for broadband and multimedia systems
• Evaluating software technology
• Making requirements measurable
• Framework patterns
• Workflow Management
• Software reliability & engineering testing

We are confident that ICSE 97 will be able to address most of the issues that have been discussed above. We hope that you will join us in this adventure and will "pick up the gauntlet", by attending the conference and providing your contribution to the success of ICSE 97.