| June LJ Supplement |
Open Source is software which has been freed. Freed to allow the bits to be copied and reused endlessly. Freed to allow inspection of the source code. Freed to allow new innovations to be built upon old, without having to duplicate past efforts. It is free software, with an emphasis on freedom.
This past year has seen an explosive rise in visibility for Linux and Open Source software. The computer world at large has gained at least limited understanding and respect for the workings of free software. Much of this attention would have been unimaginable even a year or two ago.
During this time, Open Source has been put under heavy scrutiny. Significant benefits are undeniable. Yet every analysis invariably confronts two simple, critical questions: "How does one make a living on free software?" and "Who is motivated to innovate?"
The strength of the answers to these questions will largely determine if Open Source will achieve its full potential for the greatest possible audience. Economic viability is essential to continued growth of this market.
"Money rests on the axiom that every man is the owner of his mind and his effort ... Money permits you to obtain for your goods and your labor that which they are worth, but not more ... Money is your means of survival."
-- Ayn Rand, Atlas Shrugged
Business Models
The obvious challenge of Open Source is that it may be copied freely, even if purchased initially. A $10 Linux disc may legally be used to install one machine or a thousand machines. It would seem there is no incentive to put effort into improving a product like this.
In negative critiques, Open Source is often equated with a kind of communism. A system that offers something for nothing. A system that exploits the labor of others without rewarding them. In short, a system that is unsustainable because it puts people's self interest in conflict with the greater good.
These concerns should not be dismissed out of hand, nor taken as factual. The truth is more complicated. Copyright and patent laws are not inherently part of the free market. They create limited monopolies for the companies that own the rights. This is intentionally and artificially done by government in order to reward research and encourage innovation.
Open Source is a voluntary system that waives exclusive ownership of software, in exchange for other benefits. These benefits include wider adoption, faster collective innovation, and a level competitive playing field. This makes for a frictionless, dynamic, and highly competitive market without the very profitable "vendor lock-in" that is facilitated by traditional copyrighted software.
[Sidebar: Open Source: Consumer and Developer Views]
>From a consumer point of view, Open Source is cheap. Uncustomized packages are free for the downloading. Support, service, add-ons, and customization can be purchased from a highly competitive market of suppliers who all have equal access to the code. Open Source is widely used, and so it has a strong ability to the de-facto market standard. If it has been available for some time, it has been inspected and tested with great scrutiny by early adopters.
>From a developer point of view, Open Source puts you on equal footing with any other person or company. If something doesn't work, you have the code and can fix it. If some feature isn't provided, you have the code and can add it. And every project is a world-wide community that values and celebrates their contributors. You put your reputation and your efforts on display for the world to see.
[end sidebar]
Several kinds of companies have shown that Open Source can provide a foundation for profitability. We detail some of them below, and speculate on further opportunities. These are companies that leverage the unique new possibilities afforded by Open Source, in return for their sacrifice of copyright.
The open question and challenge to these companies is whether they can create as much innovation and value as traditional software companies, given the handicap that their work can benefit their competition as much as it benefits themselves.
The Research Economy
Open Source corporations are important growth engines, but to date they have mostly built upon the efforts of others. The bedrock of the market is the thousands of individual students and moonlighting professionals who make small and large contributions every day.
These developers are not paid for their efforts. They begin a project with no promises or commitments. They work at their own pace, use their own judgement, and set their own priorities. They are the university researchers and basement scientists of software, working together to make their contribution to the world.
Often these developers are only learning or honing their craft. So many projects fail. Yet out of this soup of individual and group efforts rises some of the best software available today.
The power of Open Source and the Internet is these successful efforts can instantly be copied and put into use world-wide. They can be enhanced and customized by thousands of others. They can continue to evolve like an organism, adopting to new software and hardware architectures as the years go on.
The first and most unshakable answer to "Who will innovate?" are these students and moonlighters, motivated by their desire to learn, and inspired by the energy and clarity of curiosity and youth. Whatever happens to the "commercial" Open Source market, these software research activities will go on. They will continue to add daily to the useful body of free software available to the world.
Yet despite the best efforts of the students and moonlighters, this free software has common flaws. Because development goals are driven by the author's own needs, it results in software "by developers, for developers." The threshold at which a developer is satisfied with ease of use is much lower than for typical users. In many ways these are research projects, with all the beauties and warts that implies.
The Support Economy
As this free software attained critical mass and started becoming widely useful, needs arose for many support services built around the software itself. These services add polish and value to the base provided by the initial project.
Distribution
The Internet is great for downloading small software, but for larger products it is often too slow. And finding the software you want among the jungle of projects on the web is difficult.
>From this need rose companies like RedHat, Caldera, Suse, and Walnut Creek CDROM. On one convienient disc, you have an organized collection of the best available packages.
Opportuntizes exist for further specialized services. LinuxPPC has made a solid business out of focusing on the PowerPC market. A small company could take this further. Take one of the most popular (Dell, Linux, eMachines) PCs and produce a distribution that is tailored to that hardware, guaranteed to recognize all devices install flawlessly. Obviously, the target machine would be updated every few months.
The potential for mass specialization is one of the key Open Source advantages that can be harnessed to provide a valuable and marketable product.
Technical Support
When a single company owns exclusive rights to a software product, it is obvious where technical support has to come from.
The fact that Open Source does not have an exclusive support provider has repeatedly been portrayed as a weakness. This is a mistaken notion. Rather, what it allows is a whole market of support providers, all competing on a level playing field of equal access to the code.
Through this heightened competition, the level and quality of Open Source support is capable of rising above the best standards of today's closed software market.
RedHat, LinuxCare, and many other companies and individual consultants have stepped up to serve this market. The opportunities are so great that even the largest companies have recognised the potential. In 1999, IBM announced Linux support and consulting services.
In coming years, many companies and consumers will be motivated to select Open Source software because of the quality and wide range of entry-level and premium support services.
[table 1. Existing Open Source Business Models] Model Example -------------- ----------------- Distribution RedHat Support LinuxCare Custom Development Cygnus Training and Education O'Reilly Open/Closed Mix Sendmail Loss Leader Netscape Hardware Systems VA Research Hardware Appliances Cobalt Networks Hardware Accessories Cyclades
Nearly all companies are in fact hybrids of the business models above. [end table: Existing Open Source Business Models]
Education
Tim O'Reilly has built a booming business publishing books about Open Source software products. O'Reilly books are the standard by which others are compared: consistently top quality, and often the standard reference for a given subject. In 1998, O'Reilly revenue topped $40 million dollars. More than half of this revenue is from Open Source related books.
Opportunities in this market still abound.
Demand exists for a company that leverages the fact that: 1) An accelerating stream of Open Source code and documentation is being created by authors around the world. 2) Most people like to read this material (a) in their native language (b) in print.
A service could be imagined to gather these materials, translate them, market them on the web, fill orders with just in time printing, and ship direct to the customer. Amazon and UPublish are filling some of these roles. But as the market grows, a targeted, vertically integrated (but highly automated) company could tap these unique opportunities.
Computer-book authors may also benefit from Open Source.
The time investment to produce a book or video is daunting. Unfortunately, because of the "planned obsolescence" of many proprietary products, books and educational materials are quickly out of date. That leaves little room for authors to make a good return on their time investment.
By contrast, Open Source products cannot be declared "obsolete" and abandoned by any one company. Valuable software is carried forward as long as it is useful. Beside my Linux system sits Exploring the Unix System, published in 1984. It is proof that a dusty old tome can still be a good reference.
With Open Source, a book written today is more likely to be providing good royalties years from now. If this market keeps growing, it could become an even greater publishing hotbed.
[table: Open Source Venture Funding] Company Size Date Who ----------- ------- --------- -------------------------------------------- RedHat unknown Sept 1998 Greylock, Benchmark Partners, Intel, Netscape Sendmail $7M Fall 1998 Silicon Valley Band of Angels Cygnus $6M Feb 1997 Greylock, August Capital VA Research unknown Fall 1998 Sequoia Capital ActiveState unknown 1997 Tim O'Reilly Scriptics $400K Jan 1998 Advance Sales [end table]
The Customization Economy
The most profitable new opportunities result from the unfettered ability to customize that open source provides.
Hardware Bundles
Many more businesses now have obvious potential: Network server appliances -- Web Servers, SQL Servers, Storage Servers, etc. with an Open Source embedded OS.
Consultants
[sidebar: Survey of the Open Source Consulting Market] In February 1999, the members of the Linux Consultants HOWTO were surveyed to determine the size of the Linux/Open Source consulting market.
They were asked: 1. How many Open Source consultants are on staff? 2. How much was earned in Open Source related work in 1998? 3. What is the expected growth in 1999? 4. Did they feel that a living could be made on Open Source consulting in 1999?
Of 20 respondents, The median number of consultants was 2. The median predicted 1999 growth was 50% The median earnings per employee was $15000 US.
This survey is not definitive, and certainly not scientifically rigorous (for example, it was unfortunately unclear about revenue vs. earnings). But it does give some flavor of the Open Source consulting market.
[end sidebar]
New Markets
Many other new businesses might be possible in this new environment.
For example, there is significant demand for software solutions which can servce a customer for more than a few years. This is difficult with traditional commercial software. Unless you are IBM or Microsoft, you have to rely on other vendors, who very well may have gone out of business or no longer support that product.
But with Open Source, even a small company has control over every detail from high-level applications down to low-level drivers. By having access to the whole software system, even a small company has the access to everything needed to keep the software running and updated, no matter what changes occur in the world around them.
Stengths and Weaknesses
Despite all these strengths and opportunities, key problems remain in the Open Source market.
* The financial motivations for innovation are still weak, because they are not kept from competitors. Most of today's business models do not have software development as their primary goal. Investments in Open Source research and development must be profitable, if progress is to accelerate. * Open Source is still largely "by developers, for developers." To achieve mass market success, it must become more customer driven and consumer friendly.
Traditional software products harness the free market to solve these issues. Consumers pay to buy a software product if it meets their needs. Successful products are profitable for the companies that create them. Unsuccessful products die off. Through these mechanisms, good developers make a living and consumers get good choices.
Open Source needs equivalent consumer checks and balances. One interesting but unproven idea to tackle some of these unresolved issues is "cosourcing."
It is a system where the consumers of free software can use the web to commit funds to pay for the development of specific applications, feature enhancements, or bug fixes which are critical to them. The system turns the software market on its head: Consumers pool their resources to pay developers for the enhancements and features they need, rather than relying on one company to produce a new version of a commercial product and charge for the upgrade. And with Open Source licensing, everyone gains the benefit of the enhancements once they are complete.
In part two of this series, we analyze this idea in detail and describe our attempt to create a web service which provides the necessary mechanisms.