SIGSOFT Annual Report
July, 1996 - June, 1997

Submitted by: Lori A. Clarke, SIGSOFT Chair

SIGSOFT continues to be a vital organization that offers its members a first rate newsletter and an excellent conference program. The highlights of this last year are described below.

SIGSOFT sponsors two major conferences each year, the International Conference on Software Engineering (ICSE) and the Foundations of Software Engineering(FSE), plus some specialized conferences and workshops. FSE was held in San Francisco this year, with Mark Moriconi serving as general chair and David Garlan as program chair. The speakers and panels were selected to support the theme of the conference, software architecture. There was a successful tutorial program and attendance was up to 200 participants. Overall the technical program was excellent and participant satisfaction seemed extremely high.

ICSE was held in Boston this year, with Rick Adrion serving as general chair and Dick Taylor and Alfonso Fuggetta serving as program co-chairs. The organizing team experimented with the format and advertising of the conference. They hired a conference management company, organized a large trade show, a large tutorial program, a demo track, and a student poster track. The trade show did not have a major draw, but in contrast the demonstration track was extremely successful and very well attended. It was later decided by the steering committee that there are currently too many competing trade shows and it is more fruitful for ICSE to concentrate on relevant demonstrations; The tutorial program was also successful but the large number of additional tutorials did not significantly increase tutorial attendance. The steering committee and organizing committee agreed that it was worthwhile to use a conference management firm to help in organizing a conference of this magnitude. Overall the conference was a major success. Attendance was over 900 and the conference succeeded in drawing individuals from academia, industry, and government.

Both FSE and ICSE had workshops that co-located with these conference. The clustering of conferences and workshops seems to be very satisfactory in that it helps reduce overall travel costs and increase attendance. In addition, the Symposium on Software Reuse also co-located with ICSE. While both events were very successful, there is on-going discussion about whether events that might be viewed as competing for papers should co-locate. There was also some confusion about the financial relationships between ICSE and some of its co-located events. It was agreed that these relationships need to be more carefully explained since the volunteers running these events often do not understand ACM's procedures.

In terms of future plans, plans for ICSE'98 in Kyoto, Japan and ICSE'99 in Los Angeles are well underway. An organizer for ICSE'2000 has been selected and is putting together a team to propose to the ICSE steering committee at the next steering committee in September. FSE'97 is combining with the European Software Engineering Conference (ESEC) this Fall and plans for FSE'98 to be held in Orlando are proceeding. An outstanding issue is how to coordinate ESEC and FSE in the future since both are Fall events.

Another concern with regard to conferences is how to best interact with other organizations. Recently SIGPLAN has held or proposed workshops that overlap in scope with ISSTA, the International Symposium on Software Testing and Analysis. ISSTA is scheduled for this March and then will move to a biannual summer schedule. Instead of competing events, it would be preferable for SIGPLAN and SIGSOFT to cooperate. Although there is general agreement in favor of cooperation, neither organization has been able to move in that direction. More troublesome, SIGSOFT has had several serious conflicts with the IEEE Technical Committee on Software Engineering (IEEE/TCSE). IEEE/TCSE has introduced numerous conferences and workshops over the years to the point that they now have about two events a month, which negatively compete with each other. This was a major issue discussed at the ICSE steering committee, since this conference is co-sponsored by both IEEE and SIGSOFT. During this past year, SIGSOFT withdrew under duress and with much protest from co-sponsoring the Working Conference on Reverse Engineering, since it was being run without appropriate oversight and regard for SIGSOFT policies. With new IEEE/TCSE leadership it is hoped that many of these problems can be resolved in the future.

Under the oversight of Mary Lou Soffa, the student travel fund has been attracting applications and funding travel expenses for students to attend SIGSOFT sponsored conferences and events. Although less than a half dozenstudents have been funded so far, this is a very worthwhile endeavor that has strong support and should be continued.

This year SIGSOFT initiated an Outstanding Research Award. Lee Osterweil formed a committee that recommended that such an award be given annually and proposed a procedure for selecting the yearly award recipient. This recommendation was wholeheartedly endorsed and followed by the SIGSOFT executive committee.

Barry Boehm was the recipient of the first SIGSOFT award for Outstanding Research. Barry was one of the first individuals to recognize the impending software crisis. His seminal 1973 Datamation paper, "Software and Its Impact," was one of the first papers to quantitatively evaluate the software problem. He was one of the primary contributors to the evolution and adoption of the Waterfall Model and the developer of the Spiral Model. He also developed the Cocomo schedule and resource prediction model. He is a Fellow of the ACM and the IEEE and a member of the National Academy of Engineering. Barry is clearly one of the pioneer researchers in quantitative evaluation of software and software processes and a most appropriate recipient for our first research award.

This year the Distinguished Service award was given to Peter Neumann. Peter was the founding editor-in-chief of SEN. He served as editor in chief for 17 years before "retiring" to serve as the RISKS column's associate editor. He has served on numerous national boards and panels. He has dedicated his career to informing the world about the inherent risks associated with computers. It seems only fitting that SIGSOFT should honor Peter considering that so many of his contributions have served the software engineering community as well as all of society.

The executive committee selected a logo for SIGSOFT. This logo uses a mobius strip to represent the "S" in SOFT, signifying the often convoluted, ever evolving software engineering process. Sandy Wise proposed the original concept, which was then refined by a design artist. The logo will be used on our web pages and in our advertising. Along with many other issues, the logo was presented and discussed at the two yearly general SIGSOFT meetings, held in conjunction with ICSE and FSE.

A new slate of officers was elected this year. The nomination slate was extremely strong, indicating that a number of outstanding individuals are willing to serve SIGSOFT. It was particularly difficult to find a Chair candidate, however, since people were reticent to run against the current vice chair, David Notkin. Like most of the large SIGs, the next chair is usually someone who has been active in the organization and is well aware of the procedures and on-going issues. It is unfortunate that ACM does not openly recognize this and have a procedure for selecting the next chair from the existing executive committee. Not being able to find a chair candidate delayed the printing of the ballot and increased the cost considerably. At the last SIGCHAIRS meeting, a recommendation to change the election process was put forth to the SIGBOARD. It is hoped that ACM election procedures will be modified so that the election process will be more forthright in the future.

Finally, Software Engineering Notes, our newsletter, continues to improve. There are several volunteers who serve as column editors and book reviewers. The new length and format guidelines have decreased the backlog and improved the appearance of the newsletter. We are most fortunate to have Will Tracz as the newsletter editor!

Future plans include better cooperation with SIGPLAN and IEEE, as mentioned above. We also want to do more in the area of electronic publishing, making it easier for our members to access information. Moving to electronic advertising will also reduce mailing costs considerably, but needs to be considered carefully since internet overload seems to be reducing the effectiveness of this media for advertising. Although our retention figures are about average for SIGs, it seems that we should be doing much better with our strong conference program and newsletter. As noted in last year's report, ACM could do a better job of advertising SIGSOFT and the benefits that come bundled with membership, instead of focusing on benefits that require additional cost. Overall, SIGSOFT has a strong program and an active membership.