****************REQUIREMENTS ENGINEERING NEWSLETTER******************** No. 31. Contents 1. Message from the Moderator! (Anthony Finkelstein) 2. New Mailing Lists (Andrew Blyth) 3. REFSQ '94 Proceedings and Workshop Summaries (Klaus Pohl) 4. Software Process Automation on the Web (Alan Christie) 5. 1994 Monterey Workshop (James Brockett) 6. DIS-95 Designing Interactive Systems (Stephen Morris) 7. Fifth International Symposium of NCOSE (NCOSE) 8. CFP Special Issue Software Engineering Journal (Ian Sommerville) If you have questions about particular items appearing in the newsletter - send them to the originators. If you wish to contribute send your material to: re-list@doc.ic.ac.uk (will be moderated). Subscription or Removal requests should be sent to: re-request@doc.ic.ac.uk Back issues can be obtained via anonymous ftp from ftp-host: dse.doc.ic.ac.uk (IP number: 146.169.2.20). Directory: requirements. Files are called renl1, renl2, etc. If you cannot use ftp then you can get any back issues using email. Send email containing the following to ftpmail@doc.ic.ac.uk open dse.doc.ic.ac.uk cd requirements get quit The Requirements Engineering Newsletter and its archive is also accessible through WWW. The URL is: http://web.doc.ic.ac.uk/req-eng/index.html You may wish to link any Internet software engineering information resource you maintain to this and/or notify the manager of your local Web server by passing this message on to them. If you wish your requirements or software engineering ftp archive to be linked to the RE Newsletter archive please inform me. If you are unfamiliar with WWW you may wish to obtain a copy of the Mosaic public domain internet browser which is available for X-Windows, Macintosh or Microsoft Windows. The RE Newsletter can be conveniently accessed through the Imperial College, Department of Computing, United Kingdom, WWW Home Page (http://www.doc.ic.ac.uk/). Requirements Engineering Newsletter is published solely as an educational service. Copyright (c) 1994, Anthony Finkelstein; All Rights Reserved. ********************************************************************** From: acwf@doc.ic.ac.uk (Anthony Finkelstein) Subject: Message from the Moderator! The mailing of these newsletters has been somewhat erratic, partly because I have been busy and partly because of a shortage of interesting news over the summer. I have just returned from the 1994 IFIP World Computer Congress. It was a very mixed meeting, both in terms of topics and quality, with relatively little of interest to an RE audience. Two papers that are worth a look are: Morten Kyng; Experience with Participative Application Development; (In) Proc. 13th World Computer Congress 94, Vol. 2, K. Brunnstein & E. Raubold (Eds.); Elsevier Science BV, North Holland; pp 107-114. Abstract: Current trends in application development place increasing emphasis on early involvement of end-users. The paper presents an approach that takes user centered methods some steps further: cooperative design. Cooperative design views both users and developers as active participants who co-construct the design. Following an overview of the key elements of cooperative design the paper presents a set of strategies for the introduction of new applications based on the same notion of co-construction. Gerhard Fischer; Domain-Oriented Design Environments; (In) Proc. 13th World Computer Congress 94, Vol. 2, K. Brunnstein & E. Raubold (Eds.); Elsevier Science BV, North Holland; pp 115-122. Abstract: Domain-oriented design environments (DODEs) support the generation and evolution of applications within a particular domain. They reduce the large conceptual distance between problem-domain semantics and software artifacts and they emphasise a human-centred approach facilitatingcommunication about evolving systems among all stakeholders. The paper discusses the major problems and challenges for future software systems, develops a conceptual framework to address these problems, and describes the major architectural models required for the development of DODEs. Just to help interested parties. Morten Kyng's email is mkyng@daimi.aau.dk and Gerhard Fischer's email is gerhard@cs.colorado.edu ********************************************************************** From: A.J.C.Blyth@ncl.ac.uk (Andrew Blyth) Subject: New Mailing Lists I saw the following while reading news and thought that the RE Newsletter might find some of it interesting. You are invited to join the new software and systems engineering email list located at majordomo@world.std.com. The available lists are as follows: software-engineering software-design software-methods software-process software-tools systems-engineering systems-analysis systems-capability-maturity systems-design systems-management systems-methods systems-process systems-products systems-requirements [* Note from Moderator. The systems-requirements list appears to be inactive.] systems-synthesis systems-thought systems-tools autopoiesis (self-organizing systems) To subscribe to any of these send the message "subscribe " to majordomo@world.std.com These lists are meant to provied a forum for more focused discussion than usenet groups can provide. Everyone with an interest in software and systems engineering is invited to participate in these new lists. For more information write thinknet@world.std.com ********************************************************************** From: Klaus Pohl Subject: REFSQ '94 Proceedings and Workshop Summaries Proceedings and Workshop Summaries R E F S Q ' 9 4 Klaus Pohl, Gernot Starke, Peter Peters The abstracts of the First International Workshop on "Requirements Engineering: Foundation of Software Quality" (REFSQ) were contained in a previous newsletter. The workshop was organized by Klaus Pohl, Gernot Starke, and Peter Peters and was held in conjunction with the CAiSE conference in Utrecht, Netherlands, June 6-7th. The proceedings of the workshop were published by Augustinus Verlag and can be ordered by sending an email to {peter,pohl}@informatik.rwth-aachen.de (they will cost 40,-- German Marks (17 eng. pound or 27 US$)) A workshop summary, including an evaluation of the talks, can be gained via ftp as compressed postscript file ftp.informatik.rwth-achen.de or harry.informtik.rwth-aachen.de (login: anonymous passwd: guest) directory: /pub/NATURE/REFSQ/WS-summary.ps.Z ********************************************************************** From: Alan Christie Subject: Software Process Automation on the Web We have recently started up a Software Process Automation topic area under the SEI home page on the World Wide Web. To access this area, the Universal Resource Locator (URL) address is: http://www.sei.cmu.edu/SEI/topics/process_auto/ProcAuto.html The objective of this topic area is principally for the dissemination of information to help promote the practical application of software process automation. Sections within the topic area cover the following subjects: * Guest list -- indentifies persons who are interested in process automation, * Related home pages -- identifies other home pages on the Web which are related to this one, * Commercial products -- identifies and describes relevant commercial products, * Research projects and products -- identifies research projects and products, * Experience reports -- describes projects undertaken, lessons learned etc., * Executable examples -- provides examples of processes which can be executed in commercial or research products. Alan Christie Software Engineering Institute ********************************************************************** From: brockett@cs.nps.navy.mil (James Brockett) Subject: 1994 Monterey Workshop 1994 MONTEREY WORKSHOP 7-9 September, 1994 Naval Postgraduate School Hyatt Regency Hotel Monterey, California 1. Description of the Problem DoD and the computer industry urgently need software systems that can meet user needs effectively and reliably. Formal methods that can be partially or completely automated provide a fundamental approach to this problem, because they influence the design of languages for programming-in-the- large and because of their application to specification and analysis tools. To enhance software quality, formal methods should play a more prominent role in the software production cycle. However, most developers are not using formal methods, and researchers have not addressed many of the issues that arise in large scale applications of formal methods. To remedy this situation, researchers in formal methods need to better understand where the developers of large software systems need help. Also software developers need to understand the benefits of formal methods, and of software tools that apply these methods to practical problems. The workshop will assess the practical impact of formal methods and tools, identify gaps between the capabilities of formal methods and the practical needs of software development, and define appropriate research directions. The workshop will also provide an opportunity to share recent advances in for- mal methods and their integration in software development environments. The workshop will help to clarify what formal methods are and what they are good for. Formal means definite, ord- erly, and methodical, and does not necessarily entail logic or proofs of correctness. Everything that computers do is formal in the sense that syntactic structures are manipu- lated according to definite rules. Formal methods are syn- tactic in essence and semantic in purpose. Given the motivations for the workshop, we believe this is the most appropriate sense for the word "formal" in the phrase "for- mal methods". The usual example of a formal system is first order logic. This system encodes first order model theory with certain formal rules of deduction that are provably sound and complete. A practical problem is that this system and associated theorem provers can be difficult to work with. Formal systems can also capture higher levels of meaning, e.g. languages for expressing requirements, but these sys- tems are harder to work with and have fewer regularities and properties. It is an error to conflate mathematical models with the concrete realities they are supposed to represent. We can not prove the correctness of a real program running on a real computer, but we can prove the correctness of an algo- rithm, and we can expect the behavior of a program on a com- puter to conform to the algorithm if the intervening engineering has been done well. Formal methods can be very effective if the right methods are applied to the right problems. The excessive optimism of the attitude that everything important is prov- able helps to explain the excessive pessimism of the the attitude that nothing important is provable. Formal methods are often considered useful for proving that programs satisfy certain mathematical properties and are also often considered too expensive to be practical. This view is nar- row and limited because formal methods can reduce time to market, provide better documentation, improve communication, facilitate maintenance, and organize activities throughout the life cycle. We seek to clarify the conditions under which these benefits can best be realized. 2. Workshop on Formal Methods for Computer Aided Software Development The workshop will take place at NPS 7-9 September 1994, with a three day schedule. The purpose is to assess current efforts, to identify results and directions for increasing the degree of automation, to build a common understanding about the integration of methods and tools, and ultimately to help bring formal methods into practical use. The workshop should advance the state of Software Engineering by identifying applicable recent developments in formal methods and by focusing research on critical open problems that are barriers to the practical application of formal methods. Workshop papers will be in a proceedings distributed at the workshop. A volume of the IEEE technology series has been published as an outgrowth of last year's workshop. Sponsors for the workshop are AFOSR, ARO, ARPA, and ONR. 3. Justification for the Workshop Although formal methods are needed to achieve highly reliable software, they are not commonly used in practice. Fine-grain formal methods have a mathematical basis at the level of individual statements and small programs, but rapidly hit a complexity barrier when programs get large. Formal support for large-grain aspects of software develop- ment is necessary to extend research results to large prob- lems. The Monterey Workshop series on increasing the practi- cal impact of formal methods in computer-aided software development addresses different aspects of this problem each year. The first (1992) Monterey Workshop focused on real-time and concurrent systems. The second (1993) Monterey Workshop focused on software slicing and merging, and was considered a great success by the attendees. Researchers working on different aspects learned about each other's work for the first time, worked together to strengthen the connections between practical and formal aspects, and identified realis- tic research goals for this newly formed community. The 1994 Monterey Workshop will identify factors that split practice and theory in software evolution and develop- ment environments, and will suggest actions that can solve this problem. For example, many of the existing formal methods do not cover the entire software development pro- cess, and many of the existing tools are not sufficiently automated or do not provide user interfaces that are suit- able for engineers. Existing tools based on formal methods often require a high degree of skill to apply, so that efforts are needed to reduce the required skill levels and to enable use by engineers rather than researchers. Formal methods and tools do not yet adequately support the development of large and complex systems. Many research efforts on formal methods have focused on narrow parts of the problem, and are difficult to integrate. We have no accurate models of the software development process and its relation to internal software structures. Existing models have either focused on narrow aspects of the process or have tried to cover the entire software life cycle through infor- mal approaches that cannot support automation. Although there are formal models for several aspects of the software development process, each aspect has been considered in iso- lation, and effects that span the entire process are missing from the models. Such global dependencies are significant in actual software development, and the difficulty of maintain- ing these dependencies as software evolves is a limiting factor for large systems. We need to better understand the interactions among different parts of the process and to develop compatible models that enable solutions to different parts of the problem to be combined into systems that span the entire process. Progress in this direction should increase the practical impact of formal methods because the dependencies in practical size projects overwhelm unaided human understanding. 4. Emphasis of the Workshop The 1994 Monterey Workshop will focus on software evo- lution control, because it is a practically significant large-scale problem that promises to be amenable to formali- zation. Two aspects of this problem are: formalizing the dependencies between parts of the requirements, design, pro- grams, etc.; and developing methods for maintaining and using these dependencies to solve other problems. These problems include maintaining the integrity of evolving designs, assessing compatibility of different versions, determining the impact of proposed changes, etc. Similar problems arise in integrating the tools in a software development environment and coping with the evolution of the tool set. The workshop will help researchers working on formal methods for different aspects of software development to understand recent progress on formalizing other, related aspects of the problem, and to identify issues from those other areas that have direct implications for their own work. Presentations given by speakers who are actively involved in developing formal techniques and tools for dif- ferent aspects of computer aided software development will be interleaved with periods for related discussion. Poten- tial topics include the following: I. Review the state of the art in formal methods related to the workshop focus and their use: What is currently feasible, and what is likely to be feasible in the near future without any major breakthroughs? (1) Formal models and tools for computer-aided software evo- lution, including the management of design dependencies, document versions, and work assignments. (2) Formal models and tools for assessing and reformulating requirements in response to problem reports, and data con- cerning both clients and the environment of the system. (3) Formal models and tools for representing and determining dependencies between requirements, specifications, designs, code, etc. (4) Formal models of interactions and dependencies between tools in software development environments, and methods for managing an evolving tool set and the associated data. II. Review the problems and barriers faced by software developers in using and integrating formal methods in their work: How can formal methods help in the software develop- ment cycle and what are the major problems that need to be solved? (1) The problem of changing a system without damaging it. (2) The problem of coordinating many interrelated develop- ment tasks whose magnitude is highly uncertain. (3) The problem of maintaining consistency in structures so large that no single person can know all of the constraints. (4) The problem of ensuring consistency in requirements specifications, and how formal methods and tools can help. (5) The problem of adapting designs to evolving require- ments. (6) The problem of combining systems that were designed in isolation. (7) The problem of developing user interfaces that can make formal methods more accessible to practicing software engineers. 5. Workshop Papers Participants in the workshop will write papers summar- izing their research results and views on the applicability of formal methods and the associated software tools to aspects of software development with which they are most concerned. Workshop papers are to: (1) Describe the application of formal methods and tools to the development problems of large software systems, with emphasis on aspects related to software evolution and integration of software development environments. (2) Assess the degree to which current formal methods and tools can be used to solve practical problems in software development, (3) Identify gaps between the needs of practical software development projects and the capabilities of current formal methods and tools, and (4) Define research problems whose solution will increase the practical impact of formal methods and tools by overcom- ing current barriers to their use. Workshop papers should be 3-5 single spaced camera- ready pages, including all text, figures, and references to the literature. Workshop papers will be due August 1, 1994 and will be collected in a proceedings to be distributed at the workshop. Papers will be grouped by the aspects of software development on which they focus, and will be used to choose specific subjects for discussion at the workshop. related discussion session. Submit papers to: Prof. Luqi / LT Jim Brockett Computer Science Department Naval Postgraduate School Monterey, CA 93943 e-mail : brockett@cs.nps.navy.mil phone : (408) 656-3394 fax : (408) 656-2814 Group rate rooms are available at the Hyatt Regency Hotel, Monterey. Reservations can be made by calling 1-408-372-1234 or 1-800-233-1234. When making reservations, identify your- self as a member of the Naval Postgraduate School Monterey Formal Methods Workshop for the group discount. ********************************************************************** From: Stephen J Morris Subject: DIS-95, Designing Interactive Systems ***** Call for Submissions ***** DIS'95 SYMPOSIUM ON DESIGNING INTERACTIVE SYSTEMS: PROCESSES, PRACTICES, METHODS, AND TECHNIQUES Sponsored by ACM SIGCHI in cooperation with IFIP WG 13.2 and ASD University of Michigan, Ann Arbor USA August 23-25, 1995 How designing is done has a strong impact on what is produced. There has been growing interest by researchers from a variety of disciplines in understanding the processes of designing and in finding ways to improve the result. Computer conferences are mostly concerned about the products of design. The design process sometimes appears, but its treatment is usually constrained by the viewpoint of particular disciplines, by particular approaches to design, or by particular technologies and techniques. The time is ripe to address designing as a coherent activity -- technical, cognitive, social, organizational, and cultural. The goal is to come to a better understanding of how designing works in practice and how we can improve it -- by broad-based empirical observations, by formulating theories and perspectives, by developing methods and techniques, and adopting effective practices. The purpose of this symposium is to bring together researchers who study design processes and practices, researchers who develop, analyze, and evaluate design methods and techniques, and professional designers who want to report and reflect on their design practices. The focus of the conference is on designing in the domain of interactive systems, which covers a broad range of artifacts, but perspectives and insights from other design domains are encouraged. The Symposium will have a single-track program in order to build a common ground among the variety of participants. TOPICS Contributions are encouraged on topics including, but not limited to, the following: * Empirical studies of design practices. * New design methods; evaluation and comparison of methods. * Design support tools and environments. * Requirements capture techniques. * Design rationale: capture, presentation, and use. * Software processes for interactive system design. * Design approaches: e.g. participatory or scenario-based design. * Formal notations and cognitive models for design. * New theoretical perspectives. * Critiques of existing approaches or perspectives. * Case study experiences in specific design situations. * Experiences, perspectives, and lessons from other design domains. * Distributed design practices and techniques. IMPORTANT DATES: Deadline for contributions: February 10, 1995 Notification to contributors: April 21, 1995 Final versions due: May 20, 1995 Conference: August 23-25, 1995 Sponsored by the Association for Computing Machinery Special Interest Group on Computer and Human Interaction (ACM SIGCHI) in cooperation with the International Federation for Information Processing (IFIP) Working Group 13.2 on Methodology for User Centered System Design, and the Association for Software Design (ASD). TYPES OF SUBMISSIONS: PAPERS. High quality papers which present original work relevant to the symposium are invited. Papers should be at most 12 pages (about 6000 words). The title page should include an abstract (no more than 200 words) and keywords. PANELS. The symposium will include a small number of panels. Proposals for panels that synthesize and orient research in the area, especially across disciplinary boundaries, are particularly encouraged. Panel proposals should define the issues the panel is addressing, list the panel members, their background, and their basic positions. Panel proposals should be two pages long. DESIGN CASES. These are descriptions of real design cases that serve the important role of grounding the discussions of design processes, practices, methods, and techniques. Design cases should focus on concrete detail and should describe the basic design problem, technical constraints, and the organizational settings of the designers and the target users. Case descriptions should be between 2 and 4 pages long. SUBMISSION INSTRUCTIONS Five copies of your contribution along with a cover letter indicating the primary contact person (including name, affiliation, address, phone number, fax number, and email address) should be sent to the address on the below. FAX OR E-MAIL SUBMISSIONS WILL NOT BE ACCEPTED. However, in order to assign submissions to reviewers quickly, we request that you e-mail a copy of your abstract only (along with the title and authors) to sues@umich.edu. SUBMISSIONS AND INQUIRIES: Send all submissions and address all inquiries to (email is preferred): Sue Schuon Collaboratory for Research on Electronic Work University of Michigan 701 Tappan Street Ann Arbor, Michigan 48109-1234 USA Email: sues@umich.edu Phone: +1 (313) 747-3110 Fax: +1 (313) 936-3168 ORGANIZING COMMITTEE: Allan MacLean Ruven Brooks John Carroll Gerhard Fischer Christiane Floyd Thomas Moran Gary Olson PROGRAM COMMITTEE: Bob Anderson Philip Barnard Susanne Boedker Joelle Coutaz Gillian Crampton Smith Bill Curtis Kate Ehrlich Michael Harrison Austin Henderson John Karat Reinhard Keil-Slawik Wendy Kellogg Eiji Kuwana Judith Olson Harold Ossher Rolf Pfeifer Gustav Pomberger John Rheinfrank Tom Rodden Mary Beth Rosson Gitta Salomon Keiichi Sato Andrew Singer Alistair Sutcliffe Michael Tauber Terry Winograd Heinz Zuellighoven ********************************************************************** From: NCOSE@ROO.FIT.EDU Subject: Fifth International Symposium of NCOSE Fifth International Symposium of NCOSE July 24-26, 1995 Saint Louis, Missouri Call for papers: The symposium theme for 1995 is "Systems Engineering in the Global Market Place." Papers that describe SE needs, innovative practives, practical theory, application of tools and techniques, and excellence in training and education will be especially relevant. Topics include: Systems Engineering Value Added * Value added in management decisions * Risk Management * Best practices and their business results * SE and Quality * SE during concept exploration and before * SE and multicultural product needs Process Definition and Management * Processes * Case studies in change * Commercial vs. Government * SE outside the U.S. World Class Systems Engineering and How to Know if You Have It * Systems engineering process maturity * Systems engineering Metrics * Benchmarking processes, techniques, and results Enablers * Education and training * Automation, tools, and Techniques * Requirements management * Concurrent engineering * Role of standards Design and Systems Engineering * Tailoring SE for design * Designing for quality * Role of design in the SE process Systems Engineering in Emerging Applications * Resource and environmental management * Departments of Energy, Agriculture, and Commerce * Using SE to design processes and organizations Prospective authors are required to submit 2 copies of a preliminary paper. 4 pages preferred, but no more than 8 pages, with an abstract as the first section. Please include author's name, business affiliations, address, phone , fax, and email; contact name, contact's phone, fax and email. Upon acceptance, paper format requirements will be sent to the contact. Authors are eligible to be accepted as eith technical session presenters or poster session presenters. Technical session paper presentations are planned for 20 to 30 minutes, to include 5-10 minute question and answer periods. If interested in other than a single paper presentation, please submit your proposal. Send papers to: Ellen Barker Engineering Professional Programs Univ. of Washington, XD-51 3201 Fremont Avenue North Seattle, WA 98103 (206) 543-5539 Direct Technical Questions to: Bill Schoening 2 Mosley Acres St. Louis, MO 63141 (314) 234-9;651 Schedule: Draft Paper Submission: Sept 29, 1994 Acceptance Notification: Jan 15, 1995 Photo-ready Paper Due: Apr 1, 1995 What is NCOSE? The National Council on Systems Engineering (NCOSE) was created in 1990 by individuals to address the need to have a forum for discussion of system engineering issues, to improve the practice of systems engineering, and to create a national voice for systems engineering. It has grown to over 1900 members nationwide and has held four national symposia. ********************************************************************** From: Ian Sommerville Subject: CFP Special Issue IEE/BCS Software Engineering Journal CFP Special Issue of the IEE/BCS Software Engineering Journal Viewpoints for Software Engineering The notion of viewpoints in the area of requirements engineering and systems design recognises the need to adopt multiple perspectives when developing complex software and hardware systems. Various different models of viewpoints and viewpoint-oriented methods have been proposed but the dissemination of these approaches has been relatively limited. We believe that the time os now right to bring this important research to the attention of a wider community. The objective of this special issue is to publish recent research in this area and to disseminate this work to the general software engineering community. Papers are therefore invited which address one or more of the following areas: Models of viewpoints and viewpoint classes Viewpoint-oriented requirements and design methods Viewpoints in the requirements elicitation process Practical experiences in the use of a viewpoint-oriented approach The integration of viewpoints with existing design methods Viewpoints for domain analysis Authors should ensure that their work is accessible to software engineers working outside of this speciality and should demonstrate the practical application of their research. Papers on industrial experience with viewpoint-oriented techniques are particularly welcome. This special issue will be edited by Ian Sommerville and Anthony Finkelstein. 5 copies of papers should be submitted to Prof. Ian Sommerville, Computing Department Lancaster University, LANCASTER LA1 4YR UK is@comp.lancs.ac.uk The closing date for submissions is 30th December 1994. It is hoped to publish the special issue before the end of 1995. **********************************************************************