Computer scientists managed to put their money where their mouth is in 1981 only: since that year ETH Zurich has been offering a computer science curriculum. Before, computer science was but a minor presence. Mathematicians, physicists and electrical engineers, in particular, related computer science to early successes in computer engineering and forced it into existing curricula and disciplinary concepts.

Much too late did the industry vocalize its growing need for soundly educated computer scientists. Some students, on the other hand, feared to no longer be able to sell their programming skills which they had gained along the way. In this inhospitable environment computer science needed to be reinvented as a discipline in its own right.

This panel tries to answer the following questions:

• How did Computer Science turn into a science and how did this result in a computer science curriculum at ETH Zurich?
• Why is there no "Sihlikon Valley"?
• Should a university anticipate industrial needs or does it suffice to react swiftly?

Computer science - a recent and dynamic discipline, whose past future looks hardly less complex than its present future.

Faced with dwindling enrollments, many computer science departments are going through a soul-searching exercise as to which direction the discipline should be headed in, both in terms of teaching and research.

Two rather different views have emerged from this exercise. One view, which I will call the "fundamentalist" view, argues that we are on the verge of developing the core theories underlying our field, and that the study and application of these core theories will bring untold benefits in terms of reliability and performance. Teaching should therefore focus on mathematics and core computer science, both in theory and practice. It is further argued that "computational thinking" will cause breakthroughs in other fields, and that therefore our interaction with other disciplines should be driven by notions from computer science, rather than the other way around.

The other view, which I will call the "pragmatic" view, argues that the useful subset of the core of computer science has been developed, and that hence we should focus on applications. Teaching of core computer science should therefore be limited, and instead attention should be paid to knowledge of application areas and soft skills. The interaction with other fields should be driven by the needs of those other fields, with computer science viewed as providing a service to those other disciplines.

The pragmatic view sits well with politicians (inside or outside academia), funding agencies, students, and the general public. The fundamentalists tend to perform better in the classic academic endeavor of publishing, and hence they tend to do better in terms of academic job hunting or promotions.