SIGSOFT Annual Report
July, 1997 - June, 1998
Submitted by: David Notkin, SIGSOFT Chair
SIGSOFT has had an excellent year, the first with a new set of officers.
On the awards front, we continued to make our annual service and research awards. This year's ACM SIGSOFT Distinguished Service award was presented to Bruce Barnes, who was largely responsible for putting an effective software engineering research program into place at NSF. This year's ACM SIGSOFT Outstanding Research award was presented to David Parnas, for his many contributions to the area of software design. Joint with IEEE TCSE, we also continued to award the Most Influential Paper from ICSE-10, which this year went to the ICSE-10 paper "On Modeling and Analyzing System Behavior: Myths, Facts and Challenges", authored by David Harel and colleagues. We made several awards to students for travel support to SIGSOFT-sponsored conferences, under our CAPS program. Several of the new ACM Fellows are members of SIGSOFT, a fact that is very rewarding. As a final note, several years ago the SIGSOFT Executive Committee agreed to have Program Committees name up to 10% of the papers from conferences as Distinguished Papers; we have, however, never put this decision into action. We plan to do so this year, also understanding that we need to clear this through the ACM awards committee.
We have been involved this year in a number of significant and innovative programs.
- We have agreed, with the support of ACM HQ, to continue to take a leadership role in issues surrounding software engineering as a profession. By cementing decisions and participation in the past, we are ensuring that we will be co-equals in this effort with the IEEE Computer Society. Specifically, we have agreed to continue our participation in the curricular committees related to software engineering professionalism (at-large SIGSOFT members Tony Wasserman and Martin Griss are serving on this committee) and on the Body of Knowledge effort (Michal Young has agreed to serve on this committee). In addition, we have committed up to $65K in SIGSOFT funds, to be at least matched by ACM HQ funding, to support this effort over the next several years.
- At our request, ACM has revised its policy about SIGs being in-cooperation with commercial ventures. This is now permitted, which is clearly in the interest of SIG members.
- Our major conferences (Foundations of Software Engineering, the International Conference on Software Testing and Analysis, and the International Conference on Software Engineering) continue to be strong. (See some comments on ICSE near the end of the report.)
- To work to (re)strengthen our ties with SIGPLAN, we were in-cooperation with the new PASTE workshop Program Analysis for Software Tools and Engineering. To strength them even further, we will co-sponsor the next PASTE, which will be held collocated with the ACM SIGSOFT 2000 Symposium on the Foundations of Software Engineering (which itself is joint with the European Software Engineering Conference).
- SIGSOFT, along with three other SIGs (CHI, GROUP, and MOD), is co-sponsoring WACC '99, the International Joint Conference on Work Activities and Coordination and Collaboration.
- Will Tracz, with a growing cadre of volunteers, has continued to make our newsletter, SEN, stronger and stronger. Among the specific improvements this year are: a web-based questionnaire about SEN to help us gain feedback (from SEN subscribers as well as those who don't subscribe) about how to improve SEN (we're now gathering the results); a new, on-going column containing interviews with SIGSOFT members who are ACM Fellows, where our intent is to recognize and promote role models; a semi-improved web site for SEN; and our largest total page count in history (1224 pages, including proceedings).
There are several other key issues that SIGSOFT is aggressively addressing.
- We need to continue to increase interactions with other SIGs; PASTE and WACC are two examples of this, but we need to press this model further.
- We need to work to reduce overlapping conferences and workshops; the problem here, of course, is that when there are multiple events that have very similar agendas, the quality suffers and confusion among the membership increases. In addition, a cluttered calendar makes it more difficult for us to develop new events over time.
One example of an existing, unfortunate overlap is SIGSOFT's Symposium on Software Reuse (SSR) and IEEE CS's International Conference on Software Reuse (ICSR). Will Tracz (who is involved in both conferences) and I pushed hard to merge these conferences; although initially agreed upon by ICSR, they later backed off when we asserted that the steering committee for the new, merged conference should also change.
- We need to fix our relationship with IEEE CS, in particular IEEE TCSE (Technical Committee on Software Engineering). The ICSR example is one example, but our joint sponsorship of ICSE (International Conference on Software Engineering) is an increasingly bad situation. This is due to bad will between SIGSOFT and TCSE, and because of the difficulties that volunteers face in dealing with two societies: two TMRFs, with different rules, different financial policies, a multitude of people to deal with, and fundamentally different philosophies and goals. Ideally, we could negotiate a situation in which IEEE CS would give up its sponsorship of ICSE; another possibility would be for SIGSOFT to become a "managing partner" for the conference.
There is one more crucial issue facing SIGSOFT (in this specific regard, SIGSOFT is a clear microcosm of ACM). We claim that we are a professional society geared towards the needs of researchers, practitioners, and educators. We do an adequate job for educators, and a good job for researchers. But it's clear that we are not doing a satisfactory job with respect to the needs of practitioners. (Our strongest voice in industry is to industrial researchers, not active practitioners.) This is because SIGSOFT (and ACM in general) has not made clear to itself (to say nothing of to the membership and potential membership) what we can and should provide to them. Our involvement in software engineering as a profession issues is a great example of a plus, but we need a clearer overall vision. Without this, SIGSOFT membership will at best be stable and will more likely continue to decline slowly but surely.