In 1999, a new team at MARKET MAKER Software AG began to develop a software product line for managing and displaying stock market data and financial market news. The basic idea was to use web technology in all applications for delivering services to customers. It soon turned out that the company had to change both the processes and the organization. This report summarizes the changes made and the lessons learned over the past five years, when the product line idea was introduced into a small company which faced the pressure to quickly market the first product line instances.

This article reports on the experience of monitoring the evolution of a product family of financial Web service software system with OO-metrics over a period of three years. Every four to six months, a snapshot of the system was taken and a set of OO metrics was computed. The classes with extreme metric values or important value changes became the focus of analysis and discussions among the developers, designers, and managers. This way, potential problems were identified and corrected early.

The software development process and its related activities are described, implemented, analyzed, and changed by so-called Process Engineers. Process Engineers provide descriptions of software development processes to Process Performers. Because the processes usually are complex, support is needed for both Process Engineers and Process Performers. This paper reports the development and application of the process modeling environment Spearmint1. The architecture of Spearmint allows for a flexible definition and addition of views which are used for retrieving filtered and tailored presentations of the process models. One distinct view, namely the Electronic Process Guide used for dissemination of process information and guidance of Process Performers, is discussed in more detail. The Spearmint environment has been validated in industrial process engineering cases.

Representations of activities dealing with the development or maintenance of software are called software process models. Process models allow for communication, reasoning, guidance, improvement, and automation. Two approaches for modeling processes and instantiating and managing the process models, namely CoMo-Kit and MVP-E, are combined to build a more powerful one. CoMo-Kit is based on AI/KE technology; it is a support tool system for general complex design processes, and was not been developed specifically with software development processes in mind. MVP-E is a process-sensitive software engineering environment for modeling and analyzing software development processes, and guides software developers. Additionally, it provides services to establish and run measurement programmes in software organizations. Because both approaches were developed independently from one another, major integration efforts had to be made to combine both their advantages. This article concentrates on the resulting language concepts, and their operationalization necessary for building automated process support.