Open Source Software in the Software Market ---
An Agent-based Modeling Approach
The prosperity of Open Source Software (OSS) has generated
many interests and debates. Surrounding the Open Source phenomenon, there
are four major issues that academic inquire gets involved in: motivation,
collaboration, market effects and business models. The research presented
here proposes an agent-based artificial market simulation model to investigate
the performance of OSS in the software market and its possible consequences.
While the fact that some OSS projects, notably Linux and
Apache, have gained significant market share in competition with their commercial
counterparts shows that OSS has good chances to succeed in the market; it
does not necessarily lead to the dominance of OSS in software industry promised
by proponents of OSS. The very future of OSS is still an open question, since
uncertainty and competition in the marketplace pose real challenges. Our
analysis on the nature of the software market suggests that coexistence of
OSS and commercial software could be the most possible scenario and it works
better for harvesting more innovation and productivity. For a single OSS
project, furthermore, its market performance depends on its initially incremental
development processes. An agent-based model is built to investigate such
threshold effects and to verify above hypotheses about dynamics and consequence
of software market after the entry of OSS. The basic scenario is a software
market for similar software that seeks to accomplish the same function/functions
and compete on price, features and attention of developers, to gain the same
user target. The model explicitly considers the heterogeneous market and
interaction between multiple agents: users, developers, commercial software
packages and OSS projects. Finally, we conclude the implication of OSS for
public policy and business operation.
Jing Deng
University of Michigan – Ann Arbor
School of Information
http://www-personal.si.umich.edu/~dengj/
dengj@umich.edu