Individual reservations should be guaranteed by check or major credit card for the first night's room and tax. Be sure to request a written confirmation.

Use this hotel registration form for all the co-located IEEE Computer Society sponsored conferences -- These special conference rates are available November 13 - 23, 1998.

Hotel Registration Form

Tutorial 1: Measurements for Managing Software Maintenance
George E. Stark, IBM, Austin, Texas

This half-day tutorial is divided into five sections: Background, which describes principles of software measurement along with definitions important to software measurement and a high-level discussion of how metrics information is used; Defining a Software Metric Set for Maintenance discusses two methods for defining a project-specific metrics set and presents an approach to evaluate the probability of success of a metrics program; Organizing for Software Maintenance compares and contrasts three common software maintenance organizations using measurement data and presents an example maintenance process checklist that includes activities, responsible group, and relative schedule; Common Maintenance Metrics presents metrics data organized for making decisions by software developers, project managers, and quality professionals responsible for system maintenance; Standards and the Future, describes industry standards that can be used to support a maintenance metrics program, and current research topics in the field of software maintenance metrics.

Target Audience: Practitioners chartered with starting or improving a software maintenance measurement program. Quality assurance professionals and managers who are responsible more managing software maintenance upgrades or releases.

Mr. Stark is a programming consultant with IBM in Austin, Texas. Previously, he was a principal scientist and project manager with the MITRE Corporation in Colorado Springs. George has supported the software maintenance efforts of the Missile Warning and Space Surveillance Sensors Program Management Office at Peterson AFB. This work has included supporting the maintenance process definition as well as measuring attributes of the more than 40 releases delivered to the field since that time. He received his bachelors degree in statistics from Colorado State University in 1983 and his masters degree in mathematics from the University of Houston in 1988. George has been involved in software measurement for 15 years and was the vice chairman of the AIAA blue-ribbon panel on software reliability. He has published more than 40 articles on software measurement and has been the manager of software testing and reliability for a local loop fiber optic telephone system. He was awarded NASA's Quality Partnership award for the definition and implementation of a software quality metric set for NASA's Mission Operations Directorate and the MITRE General Managers award for contributions to software measurement.

This half-day tutorial presents an overview of legacy systems evolution and comprehension. It reports both the research and practical experience. Past emphasis of software engineering community has been on development of new software. However, with the progress of the time, the emphasis is shifting towards the processes dealing with the legacy software. Legacy systems are characterized by one or more of the following properties: they were implemented many years ago, their technology became obsolete, their structure deteriorated, they represent a large investment, they contain business rules not available elsewhere, they cannot be easily replaced, and original authors are not available. These properties are the reason why dealing with the legacy systems is difficult. The most common processes of legacy software are maintenance, evolution, reuse, and reengineering. Of them, software evolution is a change in requirements (most often adding new functionality), and it is the most common legacy software process. It is widely reported that comprehension consumes the largest portion of the resources when dealing with the legacy systems. Hence together, they represent large part of the software legacy expenses.

Target Audience: The tutorial is aimed at immediate level. It is intended for software engineers, managers, researchers, and students. The prerequisite is familiarity with software engineering issues and terminology. Experience with software maintenance may be helpful but is not required.

Vaclav Rajlich received a Ph.D. from Case Western Reserve University. He published numerous papers on software engineering and computer science theory. His current research interests are program comprehension and legacy systems evolution. He is involved in projects involving impact of software architectures on software evolution, documentation of programs by hypertext, and modeling of evolution processes by graph rewriting. He is a professor and former chair of the Department of Computer Science at Wayne State University, Detroit, Michigan. Before that he was an associate professor of Computer and Communication Science at University of Michigan in Ann Arbor, and a research scientist and manager at the Research Institute for Mathematical Machines in Prague, Czech Republic. He was a founder and general chair of the IEEE International Workshop on Program Comprehension (1992, 1993, 1994, 1996). He has been a general chair (1992, 1997), program chair (1991), and steering committee chair (1994, 1997) of the IEEE International Conference on Software Maintenance. He is a vice-president for operations of the Conference and Tutorial Board of the IEEE Computer Society, and executive vice-president of the Technical Council on Software Engineering (TCSE) of the IEEE Computer Society. He is listed in Marquis' Who's Who in Science and Engineering, Who's Who in Midwest, Who's Who in America, Who's Who in the World, and in several other reference publications.

Imagine the following scenario: `Pat' the programmer, who could be a maintainer or developer, is contemplating a complicated piece of software. A change comes to mind, based on an incomplete understanding of the code. Pat asks the following --

1. Is there a way to determine analytically the effects of this change before it is implemented?
2. Is there a way to guarantee that this proposed change that will not have adverse impacts in the parts of the system that are not to change?
3. Can I make a reasonable estimate of the total effort required for this change before I make it?

Software Surgery techniques may be used in maintenance or development and are applicable to a wide variety of programming languages. They certainly apply to the year 2000 problem.

Dr. Keith Gallagher is an Associate Professor of Computer Science at Loyola College in Maryland and a Faculty Research Associate at the National Institute of Standards and Technology. He has just completed a sabbatical at the Commonwealth Scientific and Industrial Research Organization (CSIRO) in Australia, where he continued investigating techniques of software surgery. His research interests are best described by the question "What can I do to help software engineers `in the trenches'"?

Many organizations have initiated software process improvement (SPI) programs, based upon the recognition that software processes are key determinants of the cost, quality, and schedule performance of software efforts. This full-day tutorial focuses on knowledge, techniques,and examples directly related to the activities of developing improved software maintenance processes and documenting them in such a way that they can be understood, communicated, and followed by the personnel expected to perform these new processes in the organization. The following specific topics will be covered:

• the importance of software process improvement in the practice of software maintenance, including its quantitative business value
• an overall method for developing, defining, and evolving software processes, including where the other components of this tutorial fit in
• the most important information content elements and interrelationships that should be included in process documentation for practitioners' use
• recommended layouts and structures for process documentation on paper and on the Web (taking full advantage of Web technologies and capabilities), including comprehensive examples of each
• fundamental process modeling techniques and applications, and quantitative simulations of processes using these models

This tutorial will focus on material that can be applied in practice immediately. It will also introduce important concepts, as well as summarize research and advanced applications, practical applications experience, and results achieved. The tutorial will be oriented toward software maintenance practitioners, but will be of interest to researchers as well. The recommended prerequisite is a basic familiarity with any approach to SPI.

Dr. Kellner is a senior scientist at the Software Engineering Institute (SEI), and has pioneered much of the work on software process modeling and definition conducted at the SEI. He has published more than 30 papers on software process issues, and has delivered approximately 100 technical presentations at numerous conferences world-wide. He has also taught tutorials on process modeling, definition, and related topics to more than 1,100 software professionals. Currently, Kellner leads a team developing exemplary process guides for paper and for the Web, as well as continuing his research and development work in other areas such as quantitative process model simulation. Prior to joining the SEI in 1986, Kellner was a professor at Carnegie Mellon University, and has also served on the faculty of The University of Texas (Austin). Kellner received his Ph.D. in Systems Sciences (specializing in MIS) from Carnegie Mellon University.

1. Organization of the maintenance function
2. Decision on whether to maintain, replace, re-engineer, etc.
3. Supervision of people in software maintenance
4. Control of the quality of software maintenance done
5. Selection, assignment, and retention of people in maintenance
6. Performance of the process of software maintenance
7. Making positive changes in the manageability of maintenance
8. Improvement of personnel motivation and morale

CONFERENCE LOCATION

TRANSPORTATION

Directions from Ronald Reagan National Airport (DCA):

• Via Metro/Subway: Get on the Yellow Line going toward Mt. Vernon Sq.-UDC. Change lines at Gallery Pl-Chinatown getting on the Red Line going toward Shady Grove. Get off at Bethesda Station and use elevator to access Hotel.
• Via Roadways (~18 miles): Take George Washington Parkway (west) to I-495 North. Go north on I-495, to Wisconsin Avenue (South). Go south on Wisconsin Avenue to the Hyatt Regency Bethesda (~2 miles on the left).

• Via Roadways (~18 miles): Take Dulles Access road to I-495 North. Go north on I-495, to Wisconsin Avenue (South). Go south on Wisconsin Avenue to the Hyatt Regency Bethesda (~2 miles on the left).

• Via Roadways (~18 miles): Take I-195 West to I-95 South. Go south to I-495 South (towards Northern Virginia) to Wisconsin Avenue (South). Go south on Wisconsin Avenue to the Hyatt Regency Bethesda (~2 miles on the left).

The Workshop on Empirical Studies of Software Maintenance (WESS)

• maintenance cost models (changes, releases)
• reliability of maintained systems
• assessment of system maintainability based on measurement
• models for impact analysis and their evaluation
• maintenance process assessment, modeling, and improvement

1. promote discussion and exchange of ideas among researchers and practitioners in this area
2. better identify practical problems and research issues
3. share and identify existing and potential solutions

Software maintenance requires significant effort, but we actually know little based on empirical studies. Much of what we know quantitatively is based on old studies, where software was developed and maintained (perhaps) differently from the way it is handled now. Tools, techniques and processes are very different today from those of even a decade ago. But we need a solid base of empirical support for making maintenance and life-cycle decisions. As a consequence, this year's Workshop on Empirical Studies of Software Maintenance (WESS) will be focused on two things: packaging what we already know about maintenance so that others can use it, and brainstorming about what we need to know so that we can proceed to investigate, understand and control our maintenance projects.

Metrics '98, the Fifth International Symposium on Software Metrics, sponsored by the IEEE Technical Council on Software Engineering, is being held Friday and Saturday, November 20-21, in conjunction with ICSM. We know that effective technology depends upon measurement and that effective measurement depends upon a three-pronged approach of science to develop new metrics, engineering to evaluate and test those metrics, and applications that use metrics in industry. The 1998 Symposium expands the reach of measurement into both the academic and industrial community with a full program of parallel sessions containing 22 full-length papers, 7 short presentations, and several state of the art reports.

Sessions are planned on architectural, object-oriented, error, testing, and program management and process improvement measures. Additional sessions on measures for multimedia systems and understanding experimental research results are also planned. Tom McCabe, whose work in 1976 was one of the earliest yet still widely used models for software complexity, has been invited as our keynote speaker.