Sucess of Open Source - Steven Weber (Book Review)
The Success of Open Source By Steven Weber Review
Open Source Software <ul><li>Open Source (OS) software as an experiment in social organisation </li></ul><ul><li>around a distinctive notion of property. </li></ul><ul><li>OS software is one which is governed by the following </li></ul><ul><li>Source code must be distributed / made available with the software at no more than the cost of distribution </li></ul><ul><li>Anyone may redistribute the software for free </li></ul><ul><li>Anyone may modify the software or derive from it and then distribute the software under the same terms. </li></ul><ul><li>This is in contrast to the intellectual property regime under which </li></ul><ul><li>proprietary source code functions. </li></ul>
The process of OS is explained in the political context. OS poses three questions for the political economy Motivation of Individuals - Why do software developers devote time and effort to open source projects without compensation? Coordination - How and why are open source projects coordinated? Complexity - And finally, how does the open source community get around "Brooks's Law," which says that adding programmers to a job increases the time to complete it? Open Source Software
History of Open Source <ul><li>1969 - UNICS (Uniplexed information and computing services) - UNIX </li></ul><ul><li>Till 1971 there was no official documentation </li></ul><ul><li>1972 – 73 - A total of 16 installations all at AT&T </li></ul><ul><li>Unix Philosophy </li></ul><ul><li>Write programs that do one thing and do it well, </li></ul><ul><li>Write programs that work well together, and </li></ul><ul><li>Write programs that handle text streams because that is a universal interface. </li></ul><ul><li>1973 a paper was presented to a larger audience which resulted in </li></ul><ul><li>a larger user base </li></ul>
<ul><li>Was introduced largely through universities and academic institutions between 1974 – 76 </li></ul><ul><li>UC Berkeley was one of the first clients </li></ul><ul><li>1976 – Unix too Unix copy programme to facilitate the use of the internet </li></ul><ul><li>1978 – Berkeley Software Distribution (BSD) a package of tools and utilities was created at UCB </li></ul><ul><li>1979 – 2 BSD on Unix version 7 and subsequently 3 BSD (1980) and 4 BSD (primarily for the network) </li></ul><ul><li>Unix had the advantage as </li></ul><ul><li>Availability of source code </li></ul><ul><li>Unix’s ability to move easily to other systems </li></ul>History of Open Source
<ul><li>1980 – Unix though declined in capturing markets, evolved technically </li></ul><ul><li>Deregulation </li></ul><ul><li>Change in the licensing fees where AT&T created Unix Systems laboratory and the fee was revised to $100,000 in 1988 and $ 250,000 later. </li></ul><ul><li>1989 – The Berkley group released its own TCP/IP code and a set </li></ul><ul><li>of utilities – Network Release 1. </li></ul><ul><li>1984 – The Free Software Foundation – Richard Stallman </li></ul><ul><li>Emergence of the General Purpose license (GPL) </li></ul><ul><li>January 1994 - Linux Version 1.0 by a computer science graduate student </li></ul><ul><li>at Helsinki. </li></ul><ul><li>By the end of the decade Linux was a success both technologically and in </li></ul><ul><li>the markets. </li></ul>
Elements of OS Voluntary participation and voluntary selection of tasks Distribution of labour Source Code is available freely General user base can and does propose ‘check ins’ to the code.
OSS and the Process The essence of OS is not the software but the process by which it is created. To explain the process the author raises the following questions: Who are the people who write OS code? What do they do? How do they collaborate? How do they deal with disagreements and deal with conflict?
Who Participates ? Through various surveys and this has its caveats Who ? Geographically far flung, extremely large and international. The US had the largest absolute number though Europe is more active on a per capita basis Largely the commercial organisations rather than the academia There is a relatively small inner circle contributing a majority of the code and a much larger outer circle who are less active Why ? Largely by commercial organisations to meet their in-house requirements and then contribute it back to the community for larger use in different settings.
What do they do? <ul><li>Eight general principles that the author outlines </li></ul><ul><li>Developers choose to participate and do so on tasks which are interesting. </li></ul><ul><li>The need to resolve a problem - an itch to scratch. </li></ul><ul><li>Rarely reinvent the wheel, work on the principle of maximum efficiency </li></ul><ul><li>Resolve problems through parallel work processes - OS rely on evolution archetype rather than on engineering archetype (traditional software) </li></ul><ul><li>OS developers take advantage of the large numbers </li></ul><ul><li>Document what is being done </li></ul><ul><li>Release early and release often </li></ul><ul><li>Talk a lot </li></ul>
How do they collaborate? Drop the notion of the ‘Invisible Hand’ and the ‘Self Organising System’ TECHNOLOGY IS AN ENABLER – Networking is an essential part of OS The advent of the internet is the main facilitator as it wiped away networking incompatibilities Sharing of code over the net became a seamless process Internet enabled communications were used to coordinate behaviour. LICENSING SCHEMES AS SOCIAL STRUCTURE - OS relies on IP regime though based on right and responsibility to distribute and not to exclude ARCHITECTURE TRACKS ORGANISATION
How do they Resolve Conflicts <ul><li>Conflicts in OS centre on three issues </li></ul><ul><li>Who makes the final decision </li></ul><ul><li>Who receives credit for specific contributions </li></ul><ul><li>Possibility of forking - the issue of legitimacy </li></ul><ul><li>Much of the conflict management takes place through behavioral </li></ul><ul><li>patterns and norms </li></ul><ul><li>Two elements of these norms </li></ul><ul><li>Visible nature of leadership </li></ul><ul><li>Structures of decision making </li></ul>
Explaining OS The author focuses on the Social – Collective phenomenon Political – formal and informal structures function to allocate scarce resources, manage conflicts and promote certain practices and values Technical – Core Product is a software code Economic - not in the traditional sense but to understand the reasoning behind choices.
Process OS process poses two puzzles Micro level Individual Motivations Economic Logic of Collective good Macro level Coordination Complexity
Micro-foundations <ul><li>Individual Motivation </li></ul><ul><li>Art and beauty </li></ul><ul><li>Job as a vocation </li></ul><ul><li>The joint enemy </li></ul><ul><li>Ego Boosting </li></ul><ul><li>Reputation </li></ul><ul><li>Identity and belief systems </li></ul>
Micro-foundations <ul><li>Economic logic </li></ul><ul><li>Two frameworks based on argument of abundance and that of scarcity. </li></ul><ul><li>Negates the argument that open source is based on abundance rather than scarcity </li></ul><ul><li>OSS is a non excludable and a non rival good therefore making it a pure public good. </li></ul><ul><li>Free Rider Problem </li></ul><ul><li>Software in many circumstances is more than a non rival good – </li></ul><ul><li>subject to network externalities. In OS the increased number of </li></ul><ul><li>users help in maintenance and debugging. </li></ul><ul><li>The system as a whole benefits positively from the free riders, </li></ul><ul><li>however everybody cannot be a free rider – the larger the group </li></ul><ul><li>the more the contributions </li></ul>
Macro Foundations <ul><li>COORDINATION </li></ul><ul><li>How do individuals coordinate their contributions of specialised knowledge on a focal point? </li></ul><ul><li>What governance institutions manage the implications of complexity in the division of labour? </li></ul><ul><li>The issue of coordination is addressed through ‘code forking’ </li></ul><ul><li>In the commercial software world code forking is restricted by </li></ul><ul><li>authoritative decision making. </li></ul><ul><li>What holds the process together in the OS </li></ul><ul><li>A combination of individual incentives, cultural norms and </li></ul><ul><li>leadership practices </li></ul>
Coordination Individual incentives Economic – Forking reduces the size and liquidity of the market place as it reduces interoperability. Cultural - Reputation risk Political - Relationship between the leader and the followers. Forking would require a new community of followers to de developed. Cultural Norms Ownership customs (granting of rights to distribute rather than exclude). Initiating the project Inheriting the ownership Pick up an abandoned project.
Coordination Leadership Practices which when strong holds the process and disallows forking. Too much forking undermines the process in terms of scattered efforts, duplication of work, incompatibilities etc. Too little forking as it may inhibit innovation.
Macro foundation COMPLEXITY Source Code Modularisation – reduces the compexity of the system as it limits the reverberations??? Sanctioning License as a Social Structure
Business Models and the Law Conventional business models for proprietary software creates a market where the power lies with the supplier. Control of the source code is the foundation In OS the underlying structure of the market shifts where the user is more powerful than the supplier. Source code is free. It reduces the potential for supplier lock in. This poses a huge challenge for the business model – what can generate sustained economic returns? Brands and trademarks. Accumulation of tacit knowledge
Generic Business Models for OS Support Sellers – Technical support is the most obvious may to make money Loss leaders and creating a market for linked commercial product Sell it free it – Release the source code when the benefits of the OS are believed to outweigh the proprietary licensing revenues Accessorizing – selling accessories that make the use of OS easier Service enablers
Legal Structures Copyright - grants an exclusive bundle of rights to the creator of original work. OS licenses generally depend on copyright law for their claim to enforceability. Copyright and GPL. Patents cover inventions and new processes, not expressions. The OS community is experimenting with a range of partial defenses against patent restrictions.
<ul><li>With the development of OS market shifts are plausible and can be highly disruptive and revolutionary. </li></ul><ul><li>Inherent risks </li></ul><ul><li>Market dominance – Free code is a commitment against certain forms of market opportunism. Lock ins are not possible. </li></ul><ul><li>Remove the competitive advantage of a new entrant </li></ul><ul><li>Financial Risk </li></ul>Conclusion
Conclusion OS is a way of organizing production jointly based on an understanding of property rights that are configured around distribution. OS inverts the core notion of property It demonstrates large scale non hierarchical cooperation in a leading economic sector which disregards national boundaries – the implication of this for international politics is large. Is it possible to build a working economic system on a foundation of property rights configured on distribution and what would such a system look like?
A particular slide catching your eye?
Clipping is a handy way to collect important slides you want to go back to later.