Introduction To The QPW Case Study

Jim — Thanks again for speaking at BIC '93. I'm also glad you could stop by Borland and experience what we call Borland Software Craftsmanship. We are a young company, started by a Frenchman, with young bright and excited developers. In my 8 years at Borland I have been in the center of it all and can't imagine another place to be. — David Intersimone, Director of Developer Relations, Borland

In 1993, Borland invited me to speak at the Fourth Annual Borland International conference in San Diego, California, and to visit their location in Scotts Valley, California. I made arrangements with David Intersimone, Director of Borland Developer Relations, to speak at the conference in exchange for access to one of their development organizations. Interviews with such development organizations have helped the process community better understand the high-level characteristics of software development organizations. We can use this understanding to help projects assess their development methods against those used in other development cultures. I was enthusiastically received and graciously hosted, a harbinger of other positive signs of the Borland culture that I would observe that day. I was treated to insights into one the most stunning development efforts I have had the pleasure to study.

In this chapter, I relate what I learned while meeting with the development team for Borland's Quattro Pro for Windows 1.0 (QPW) on May 20, 1993, in Scotts Valley. I feel there is much to be learned about their process, technology, and organization that we can apply to projects across the industry, including large projects and perhaps even embedded and real-time system developments such as we have in AT&T. This chapter is distilled from a paper published in Dr. Dobb's Journal in 1994 [BibRef-Coplien1994] — before the organizational patterns were first published!

It is important to understand that this is a retrospective on the development of the software for the initial offering of QPW. There was little or no embedded base, and the project didn't face the constraints one finds in the legacy code projects common in large, traditional telecommunication projects. Even so, the phenomenal productivity of this group and the factors contributing to that productivity are thought-provoking. Most organizations should be able to take a page from Borland's book as a basis for their own process improvements.

This chapter starts with a high-level description of the project and describes the personalities in the development effort. Analyses of the data from our process analysis technique follow in the next section. Subsequent sections of the paper describe aspects of the QPW development that stood out as contributing to its success.