Guiding principles of natural ecosystems and their applicability to software ecosystems

摘要

Nature has developed optimal solutions to numerous problems in the course of the evolutionary process (Begon et al 1990). Scientists in engineering contexts and engineers often try to learn from their natural counterparts and solve problems by following approaches in nature. Software engineering (SE) has already exploited many natural phenomena to improve the efficiency of algorithms, tools, models, and processes. For example, evolution and genes serve as the role model for genetic algorithms, where natural selection is applied to computer programs and data (Goldberg 1989). Ant colony optimization has been inspired by ants and their behavior of finding shortest paths from their nest to sources of food (Bilchev and Parmee 1995). Other examples are neural networks, swarm robotics, and bee algorithms. Recent developments and trends in software product line engineering have made us ponder about the term software ecosystems. It is evident that the term was coined to reflect the organization of software vendors, third-party developers, suppliers, and users (Bosch 2009). The name is obviously derived from the archetype natural ecosystem. However, not much work has been reported which compares characteristics of natural and software ecosystems. Several research reports have emphasized the importance of considering software ecosystems from a business perspective (Jansen et al 2009). However, the key questions are how the two kinds of ecosystems can be mapped to each other, and how software ecosystems researchers and practitioners can benefit from insights taken from natural ecosystems.