An empirical study on selection of Open Source Software - Preliminary results

Growing attention on component-based development has inspired the development of several normative methods for selection of software components. Despite these efforts, empirical studies show only minor adoption of such methods. To understand how research can contribute to improving the selection of components we interviewed developers from 16 Norwegian software companies which integrate Open Source Software (OSS) components into their systems. We find that the selection of OSS components has a situational nature where project specific properties significantly constrain the selection's outcome, and that developers employ a dasiafirst fitpsila rather than dasiabest fitpsila approach when selecting OSS components. This could explain the limited adoption of normative selection approaches and general evaluation schemas. Moreover, it motivates a shift from developing such methods and schemas towards understanding the situational nature of software selection.