•  
  •  
 

Abstract

The prevailing approach to free/open source software and licenses has been that each system is developed, distributed, and used under the terms of a single license. But it is increasingly common for information systems and other software to be composed with components from a variety of sources, and with a diversity of licenses. This may result in possible license conflicts and organizational liability for failure to fulfill license obligations. Research and practice to date have not kept up with this sea-change in software licensing arising from free/open source software development. System consumers and users consequently rely on ad hoc heuristics (or costly legal advice) to determine which license rights and obligations are in effect, often with less than optimal results; consulting services are offered to identify unknowing unauthorized use of licensed software in information systems; and researchers have shown how the choice of a (single) specific license for a product affects project success and system adoption. Legal scholars have examined how pairs of software licenses conflict but only in simple contexts. We present an approach for understanding and modeling software licenses, as well as for analyzing conflicts among groups of licenses in realistic system contexts, and for guiding the acquisition, integration, or development of systems with free/open source components in such an environment. This work is based on an empirical analysis of representative software licenses and of heterogeneously-licensed systems. Our approach provides guidance for achieving a “best-of-breed” component strategy while obtaining desired license rights in exchange for acceptable obligations.

Share

COinS