Rationale Management and Agile Development
Hauptseminar für Überfachliche Grundlagen

Prof. Bernd Bruegge, Ph.D.
Allen Dutoit, Ph.D.

Registration To register for this seminar, send mail to Allen Dutoit with a topic from the list below no later than February 11, 2005. Topics are awarded on a first come, first serve basis. Requirements High motivation for the topic A presentation of about 45 minutes Independent literature research Attendence to all presentations and active discussion in the seminar Attendence to the rationale exercise Notes This seminar is conducted in English This seminar is an "Überfachliches Grundlagenseminar".

Intended Audience Students with a Vordiplom and some system development experience (e.g., a Software Engineering Praktikum). Time and Location Room 01.07.014 (Lehrstuhl Bruegge Seminar Room) Wednesdays, 16:15-17:45 Preparation meeting on April 13, 2005. First student presentation on May 4, 2005. Supervisor Allen Dutoit Room 01.07.053 Office hours: by appointment

Topics

Date	Presenter	Paper	Slides
13.4.2005	Allen Dutoit	Preparation Meeting Bibliographical Research	pdf
4.5.2005	Boris Borisov	Cockburn & Williams. The Costs and Benefits of Pair Programming	pdf
11.5.2005	Dimitar Marinov	Müller & Hagner. Experiment About Test First Testing	pdf
18.5.2005		No seminar (Pfingstferien)
25.5.2005		No seminar
1.6.2005	Mei E	Henninger. Tool Support for Iterative, Adaptable and Agile Software Development Methodologies	pdf
8.6.2005		No seminar
15.6.2005	Benjamin Becker	Burge & Brown. An integrated approach for software design checking using design rationale	pdf
22.6.2005	Hristo Dakev	Briggs & Grünbacher. EasyWinWin: Managing Complexity in Requirements Negotiation with GSS	pdf
29.6.2005	Simon Bierbaum	Sauer. Using Design Rationales for Agile Documentation	pdf
6.7.2005	Andreas Kauer	Conklin & Begeman. gIBIS: A Hypertext Tool for Exploratory Policy Discussion	pdf
13.7.2005	All	Rationale Exercise

Content

Motivation. The primary goal of a software development project is to build and deliver a software product with its accompanying support material (requirements specification, design documentation, test cases, and test results). Much of the focus is spent on the system under construction:

• What should it do (requirements)?
• How should it do it (system and detailed design)?
• Does the system under construction meet its requirements (testing)?
• Do the intended users find is usable (usability testing)?

Currently, little effort during development is spent describing the why of the system. For example:

• Why certain feature is included in the system?
• Why was a certain component selected?
• Why was a certain programming language chosen?

Unfortunately, this decision making information, called the design rationale of the system, is often lost when the project is completed. The focus of design rationale research is to find cost effective methods and tools for capturing, structuring, and accessing design rationale information.

Goal. The goal of this seminar is to survey current research, encourage critical thinking from its participants, and generate new ideas for design rationale. Each week, one or two participants will prepare an hour long presentation from a list of selected topics (see table below). The presenter will:

• summarize the topic,
• explain its significance in relation to similar work,
• evaluate the author's approach and results, and
• conclude with a list of issues that have been left open by the author.

The presentation will be followed by an active discussion conducted by the other participants.

Rationale Exercise. In parallel to the presentation cycle, the seminar participants will conduct a 2-3 session exercise on design rationale, making changes to an existing specification for a meeting scheduler system. Unlike typical specification exercises, the specification you are changing includes rationale provided by the original developers. You will use the rationale attached to the specification and provide additionale rationale knowledge for justifying the decisions you make. The goal of the exercise is to emphasize the importance (and complexity) of rationale during specific phases of software engineering.

Useful links

• Tips for the presentation
  • How to Present a Paper in Theoretical Computer Science: A Speaker's Guide for Students
• Paper about writing good papers
  • This is not an article
• Literature search
  • Citeseer (free access)
  • ACM Digital Library (access from TUM, configure your browser to use proxy.biblio.tu-muenchen.de:8080 as http proxy)