Visit Report

Visit by Daniel Berry, Technion, to Erik Kamsties, IESE at Fraunhofer Institute in Kaiserslautern, 17-18 February 1999

The institute is the Institute for Experimental Software Engineering (IESE), which, investigates the most promising innovative SE techniques and methods available, to develop their applicability for industrial-stength environments, and finally, to transfer them into industrial practice. It carries this focus to its RE work, which has just begun and is headed by Dr. Barbara Paech with the assistance of Erik Kamsties.

The main objective of the meeting was to explore Ph.D. thesis topics for Kamsties and the possibility of Berry being the academic co-advisor. (The institute provides the institutional co-advisor, most likely Prof. Dieter Rombach.) Kamsties showed Berry a draft of a presentation of his thesis proposal. In it, Kamsties proposes to deal with problems concerning the creation and validation of system models in requirements engineering. Kamsties and Berry agreed that the proposed work was probably too ambitious for a Ph.D. done by one person, raising the risk of never finishing. One topic that excited Kamsties during the presentation was that of spotting ambiguities in early requirements documents before they caused inconstent specifications or code. Berry pounced on that issue as a perfect Ph.D. topic. It is a very hard problem; it has never been solved; it is a real problem faced by nearly all development projects; the topic is focused enough to be doable by one person; and the student is enthusiastic about the topic. Kamsties will prepare a new proposal and Berry will provide feedback.

Berry discussed the mechanics of being co-advisor with Rombach, the head of the institute.

Having achieved the major objective of the visit, Berry, Kamsties, and Paech used the rest of the visit to discuss a number of other RE issues.

1. Paech described the plans for the RE program at the institute.

2. Paech described a course offered by the institute teaching non-programmers software documentation. Bob Glass once observed that English majors write better maintenance documentation than do programmers. This observation was being exploited to solve multiple problems with one solution. The problems are:

1. a severe shortage in Germany of programmers
2. a severe dislike by programmers of documentation
3. a severe unemployment in many areas other than programming
The solution provides:
1. non-programmers to do documentation that programmers dislike doing
2. jobs for unemployed people doing a sorely needed task
3. Kamsties, Paech, and Berry discussed plans for an RE workshop to be held at the HCI conference in Munich in August 1999. The workshop is organized by Kamsties and Paech, and Berry has been invited to participate. Berry presented a draft of his presentation and Kamsties and Paech offered ideas for improving it.

4. Kamsties and Paech are the local organizers of a Dagstuhl workshop on requirements elicitation, documentation, and validation organized by Boerger, Hoerger, Parnas, and Rombach in June 1999. Berry and a number of RENOIR members are invited participants. Berry provided feedback on Kamsties's and Paech's ideas for organizing the workshop.

Last up-date: 10 March 1999