Towards a Theory for Understanding the Open Source Phenomenon

1
Towards a Theory for Understanding the Open
Source Phenomenon

• Kasper Edwards
• Technical University of Denmark
• Department of Manufacturing Engineering and
  Management

2
Agenda

• My perspective
• Introducing open source software
• Open source software as economic goods
• A community-based institutional framework
• A capitalistic, institutional framework

3
My perspective

• Background
• Engineer
• Tainted with economics
• Economics of technology
• The technology must be taken into account
• Data
• Hours of interviews with open source developers
• Personal interest / involvement

4
Introducing open source software

• What is open source software?
• Software like any other software
• Access to the source code, which may be modified
• Create derived works
• Create and distribute copies
• How is open source software developed?
• Organised in individual projects
• The central figure is the maintainer
• Communications are done mostly using web-based
  media
• Mailing lists are the central means of
  communication

5

• The open source software development cycle
• Maintainer releases software and source code
• User downloads software and source code
• User identifies problems or needed features
• User implements corrections
• Contributor returns corrections to the Maintainer
  for inclusion
• Corrections are discussed
• Maintainer includes changes and release a new
  version

6
The problem of understanding Open Source Software

• Observations
• Open source software is being developed and
  exchanged
• Some open source products have market dominance
• Private individuals contribute to the development
• Commercial enterprises contribute to the
  development
• Open source software development is not without
  cost
• Time and/or money
• People and especially enterprises must make a
  living

7
Economic theory of goods

• There is a long tradition that economists try to
  understand goods
• How benefits can be appropriated

8
Open source software as a good

• Technical properties
• A digital being - unlimited copies at
  insignificant cost
• Instantaneous mass-distribution
• Open source software is non-rival in consumption
• License properties
• Free redistribution
• The source code must be available
• Open source software is non-excludable

A pure public good
9
Theoretical consequences of being pure public good

• Under-provision
• What is underprovided has not been developed
• Massive free riding
• Free-riding in open source Development not
  returned to the project
• There is a penalty from not returning
  developments to the maintainer
• The question
• Why is open source software being developed?

10
Different approaches to the question

• Why is open source software is being developed?
• A research object
• Altruism
• It is intrinsically rewarding
• To gain reputation
• To build a CV
• To develop open source software instead of buying
  software

11
An analytical problem

• Two groups of actors
• Unpaid voluntary developers
• Commercial enterprises
• Could we understand both within the same theory?
• I believe not
• Prima facie it must be assumed that they have
  different incentives
• They might even adhere to different inner logic
• Two different institutional frameworks
• The community-based institutional framework
• The capitalistic, institutional framework

12
The community based institutional framework

• Observation
• Many are developing open source software
• OSS development is time consuming
• There is no monetary reward
• Properties of the open source organisation
• Very loosely coupled network
• Limited communications bandwidth
• Characterising members of epistemic communities
• A shared set of normative and principled beliefs
• Shared causal beliefs
• Shared notions of validity
• A common policy enterprise

13
OSS projects as an epistemic community

• Shared normative and principled beliefs
• Strong belief in empowerment of users
• A counter culture
• Shared causal beliefs
• Contributors have programming experience (or are
  gaining)
• Provision of a common understanding of how to
  solve a problem
• Shared notions of validity
• Important when choosing between solutions
• Provision of a common understanding of why a
  solution was chosen
• Two criteria 1) Performance, and 2) Beauty
• Common policy enterprise
• Freedom of choice
• Freedom to expand and change software to fit
  personal needs

14
Theoretical consequences of epistemic communities

• Possible to collaborate with minimal
  communication
• A shared mindset
• The code say more than a thousand words
• Little or no need for co-ordination
• Implicit understanding of the direction of the
  project
• Problems of epistemic communities
• A static analysis to a dynamic phenomenon
• Epistemic communities does not explain entry into
  projects

15
Legitimate peripheral participation

• Becoming part of a project is a learning process
• Every project has its own idiosyncrasies
• Learning is situated
• Knowledge cannot be de-coupled from situation
• Learning can only be done through participation
• Learners are trying to become insiders
• Consequences of legitimate peripheral
  participation
• Learners are not able to participate in core
  activities
• Learners can contribute to peripheral activities
• Learners must be allowed to participate
• Learners must be allowed to be part of the
  community practice
• By participating learners become part of the
  community

16

• Summing up the community based institutional
  framework
• Possible to collaborate with minimal
  communication
• Development is a learning process
• Situated learning describes the learning process

17
The Capitalistic Institutional Framework

• Observations
• Commercial enterprises contribute to open source
  development
• A market divided
• The Windows platform
• The other platforms
• Perspective
• Understanding at the level of the industry
• Understanding motivation
• Applications matter to the user platforms are
  just an enabler

18

• Theory
• Computing platforms and applications are
  compatibility regimes
• Network effects
• Increasing returns on several levels
• Application developers - Retail outlet - Users
• Competing technologies and lock-in
• Theoretical Consequences
• The others can only survive as niche players as
  the number of applications diminish
• To compete they need to establish a credible
  alternative

19

• Open Source Software is one such alternative
• Difficult to hijack development
• Combined effort makes for fast development
• Commercial enterprises have incentives
• Create an alternative platform and attract
  applications
• Develop applications to attract users/costumers
• Influence on platform development
• Freedom to develop new hardware for the platform
• Concluding the capitalistic institutional
  framework
• It makes sense to make a perspective of competing
  technologies
• OSS as a platform provides a singular opportunity
  to create one credible alternative to the Windows
  platform

20
Conclusion

• Are we on the way towards a theory?
• Yes, but only parts of the phenomenon
• Different institutional frameworks seem
  appropriate
• Community-based institutional framework
• The code says more than a thousand words
• Describes and helps to understand the process
• Capitalistic institutional framework
• Platform competition show an incentive to
  contribute to development
• An open source platform might be the credible
  alternative