by Scott Tilley

The 23rd International Conference on Software Engineering (ICSE 2001) took place May 12-19 in Toronto, Canada. As the event in the academic calendar for software engineering researchers, ICSE  is a must-attend conference. This year's ICSE was a tremendous success. I participated in the regular ICSE tracks, as well as a co-located workshop (IWPC 2001), an ICSE-sponsored workshop (NCC 2001), a new track (FoSP), and a new symposium (NSEFS). Although the packed schedule made for a very busy and tiring week, I think ICSE 2001 was by far the best ICSE yet.

Last month it was unofficially "Software Engineering Week" in Toronto, Canada. The occasion was the 23rd International Conference on Software Engineering (ICSE 2001). ICSE is the flagship event in the academic calendar focused on all aspects of software engineering. It is sponsored by both the Association for Computing Machinery and the IEEE Computer Society. The conference itself was a huge success, due in large part to the efforts of the conference committee, which was lead by the General Chair, Hausi Müller of the University of Victoria. Having organized several smaller events myself, I know only too well how much time and effort is required in such an undertaking. Congratulations Hausi!

The ICSE program is packed full of events, including co-located workshops, ICSE-sponsored workshops, tutorials, a rich paper track, and several very enjoyable social events. I had a paper in ICSE, "Evaluating the Reverse Engineering Capabilities of Web Tools for Understanding Site Content and Structure: A Case Study," which was co-authored with and presented by Shihong Huang. This paper describes an evaluation of the reverse engineering capabilities of three Web tools for understanding site content and structure. The evaluation is based on partitioning Web sites into three classes (static, interactive, and dynamic), and is structured using an existing reverse engineering environment framework (REEF). This case study also represents an initial evaluation of the applicability of the REEF in the related but qualitatively different domain of Web sites.

In addition to the regular paper track, I was also fortunate to be able to participate in four other events at ICSE: IWPC, NCC, FoSP, and NSEFS. For the acronym-challenged, IWPC is the International Workshop on Program Comprehension, NCC is Net-Centric Computing, FoSP is Frontiers of Software Practice, and NSEFS is the New Software Engineering Faculty Symposium. Believe it or not, I didn't make up these acronyms. Well, not all of them anyway.

[back to the top]

IWPC 2001

The first event for the ICSE 2001 week that I was involved with was the 9th International Workshop on Program Comprehension (IWPC 2001), which took place on May 12-13. IWPC was one of four co-located events at ICSE, and attracted nearly 80 practitioners and researchers from academia, industry, and government. As stated by the workshop organizers,

"Comprehending programs is one of the core software engineering activities. Program comprehension is needed when one maintains, reuses, migrates, reengineers, inspects, or enhances software systems."

The workshop included dual-tracked technical papers, research tool demonstrations, and special "working sessions" that are like structured "birds of a feather" sessions found at other conferences. I presented a paper entitled "Spreading Knowledge About Gnutella: A Case Study in Understanding Net-Centric Applications," co-authored with Mohan DeSouza. This paper describes our experiences in attempting to understand the functional nature, high-level design, and implementation details of Gnut, a program that implements the Gnutella peer-to-peer Internet file system protocol. Gnutella is representative of a new breed of net-centric applications that is both qualitatively and quantitatively different than the typical legacy systems that are usually the focus of program understanding exercises.

I found this year's IWPC to be very good. It is always interesting to hear about research and recent results in a field that is critically important in the overall software engineering context. If French food suits your taste, then prepare for IWPC 2002: it will be in Paris, June 27-29, 2002. On y va!

[back to the top]

NCC 2001

I was one of the organizers of the 3rd International Workshop on Net-Centric Computing (NCC 2001), which took place on May 14. Previous NCC workshops were London in 1997 and Toronto in 1998. The underlying principle of NCC is a distributed environment where applications and data are downloaded from servers and exchanged with peers across a network on as as-needed basis. In some respects, NCC resembles an earlier computing era of mainframes and dumb terminals. However, there are significant differences. For example, NCC relies on portable applications running on multiple platforms, mobile data accessed via high-speed network connections, and low-cost appliances for local processing.

In keeping with the theme of "Migrating to the Web," NCC 2001 focused on issues related to reengineering legacy systems for use in an NCC environment. Of particular interest were holistic techniques for Web-enabling existing applications that integrate various reengineering aspects (e.g., code, data, and user interface reengineering) into a "whole system" modernization process. The workshop was structured around three central issues: decomposing legacy systems to identify logical components representing essential functionality, developing a new Web-enabled system using these components, and deploying the new system in an NCC environment.

I presented the paper "The State of Net-Centric Computing in Early 2001," which was co-authored with Shihong Huang and Kenny Wong. The paper comments on changes in NCC since the first workshop in 1997. For example, backbone bandwidth and computing power have increased tremendously. However, some things remain the same. For example, consumer-oriented network connections remain relatively slow. Two new developments that stand out are infrastructure support, such as Microsoft’s .NET initiative, and applications, such as peer-to-peer (P2P) networking.

There were 28 registered participants for NCC 2001. A report summarizing the workshop is currently being written and will be published as a CMU/SEI Technical Note in the near future. Plans for NCC 2002 are currently under discussion. If you have an idea about possible themes and co-location, please send me a note.

[back to the top]

FoSP 2001

A new addition to the ICSE program was the the Frontiers of Software Practice (FoSP) sessions. I helped organize FoSP with Hausi Müller. The FoSP track consisted of eight presentations by industry leaders, highlighting new developments in real-world software engineering. With the unprecedented amount of change in the computing industry, keeping abreast of new developments in both research and practice is extremely challenging. FoSP benefited software professionals of all stripes by introducing them to topics representing cutting-edge software engineering practice.

The eight FoSP sessions were:

1. "Enabling Technologies for the Future of Voice-Based Web Access," by Steve Woods and Jeromy Carriere of America Online Interactive Properties
2. "Jini™ Network Technology: Devices, Desires, and Designs," by Aleta Ricciardi of Valaran Corp.
3. "Inter-Language Object Sharing with the Common Language Runtime: Infrastructure for MS .NET," by Jennifer Hamilton of Microsoft Corp.
4. "Collaborative Software Engineering," by Alan Brown of Rational Software Corp. (presented by Scott Tilley)
5. "AntiPatterns in Software Architecture," by Tom Mobray of the World Wide Institute of Software Architects
6. "Dependability of Embedded Systems," by John Knight of the University of Virginia
7. "Simple Object Access Protocol (SOAP) and Web Services," by Arthur Ryman of IBM Canada, Ltd.
8. "Auditing Legacy Systems for Security and Survivability," by Tom Longstaff of Carnegie Mellon University's Software Engineering Institute

By all accounts, the first FoSP track was a success. It is planned to hold similar events at future ICSEs. Personally I feel that such presentations are an excellent addition to the ICSE program.

[back to the top]

NSEFS 2001

Last but not least, I participated in another new addition to the ICSE program: the New Software Engineering Faculty Symposium (NSEFS), which was organized by Leon Osterweil of the University of Massachusetts at Amherst, Gabriel Silberman of the IBM Center for Advanced Studies, and Kenny Wong of the University of Alberta. This unique event brought together seasons faculty, recent survivors of the tenure review process, and new faculty just beginning their academic careers.

As someone in between the latter two of these stages of my own career, I found this symposium extremely enlightening. Although there is a tremendous shortage of software engineering faculty in many countries around the world, it seems that new faculty are under continuous pressure to "teach courses for new software engineering programs, supervise many graduate students, collaborate with industry, raise research funds, be leaders in their field, and publish journal papers." Mea culpa.

I hope that there are NSEFS sessions at future ICSE conferences. It is an invaluable experience to receive guidance from veterans on what to do, and what not to do, to advance your academic career. Oh well, maybe I can succeed in spite of not following this advice J

[back to the top]

Final Comments and Photos

During the week I prowled the hallways (and environs) of the Toronto Westin Harbor Castle Hotel, snapping photos with my trusty Canon Powershot S100 Digital Elph As usual, the following images were converted to compressed JPEG files to reduce download time. I think the photos give a glimpse into the activities at ICSE 2001, both professional and otherwise.

Next year ICSE 2002 will be in Beunos Aires. If you're really planning ahead, ICSE 2003 will be in Portland, Oregon and ICSE 2004 will be in Edinburgh, Scotland. Finally, since I've been talking about so many conferences in this issue, the following two links may be of interest for events in Fall 2001:

• SIGDOC 2001: www.cs.ucr.edu/~stilley/sigdoc2001 -- this is the ACM SIG on Design of Communication. It takes place Oct. 21-24 in Santa Fe, New Mexico. The theme is communicating in the new millennium, which addresses topics such as multicultural audiences and multilingual documentation.
• WSE 2001: www.cs.ucr.edu/~stilley/wse2001 -- this is the 3rd International Workshop on Web Site Evolution. It takes place Nov. 10 in Florence, Italy and is co-located with ICSM 2001 (software maintenance). The theme is "access for all" and refers to making web site available to multiple users (e.g., those with impaired vision) and multiple platforms (e.g., WAP phones).

	Jochen and Shihong getting ready for the IWPC dinner cruise on Lake Ontario. (135KB)
	Leaving the inner harbor for the cruise. (130KB)
	Peggy onboard, waiting to sail. When does dinner start? (130KB)
	A view of the Toronto skyline near dusk. (65KB)
	Participants of NCC deep in discussion. (103KB)
	Ken and Kostas leading the discussion in the NCC workshop. (82KB)
	Lunch time at ICSE, tutorial day 1. (115KB)
	Hausi welcomes everyone to ICSE 2001. (70KB, picture slightly blurry)
	The banquet extravaganza finishes. What a show! (132KB)
	The ICSE General Chair (center) shaking a leg. (79KB)
	Steve talking at one of the FoSP sessions. (76KB)