WWW8 Notes
Open-Source Software and Software Patents
7/27/1999
by Edward Piou
Open-source software and software patents were two noticeable themes at the WWW8 conference in Toronto this past May. Both the opening and closing keynote speakers - Tim Berners-Lee and Bob Metcalfe, respectively - took care to mention the open source movement, though they had different takes on its future. Richard Stallman, founder of the Free Software Foundation, won a $10,000 award (partially funded by one of his frequent targets, Microsoft) at the conference, and used his acceptance speech to warn against the dangers of software patents. And attendees had a chance to meet and talk to various open source developers on the post-conference Developer's Day. This article will attempt to bring together the various open source and patent issues that came up at the conference.
(Note: a brief overview of the conference is also available.)
Tim Berners-Lee, the inventor of the World-Wide Web, gave a wide-ranging talk that focused mainly on the future of the web - how he hopes to see the network grow. One threat to his vision which he saw was the proliferation of patents on interesting, but non-innovative, Internet technology. "The bar for innovation, [for] a patent to be innovative, seems to be too low," he said. "When you look at them, it is very very hard to figure out which bit actually the patent filers thought was innovative." Taking a normal offline activity - for example, shopping - and putting it on the Web is not innovative, and should not be patentable, he said. But under the US Patent and Trademark Office's (USPTO's) current guidelines, that is exactly the type of thing which many companies try, and succeed in, patenting.
The result of frivolous patents, he thought, would be a great slowdown in the development of the Internet. Rather than rewarding innovation, and encouraging new and interesting ideas, he characterized the current patent process as rewarding those whose business plan is to make money by suing people and licensing obvious technology.
He held up open-source software, and open standards in general, as one way to make the Web a place where information and ideas can be easily exchanged. As examples of open-source software that help drive the web forward, he cited Amaya and Jigsaw, the open-source web browser and web server developed by the World Wide Web Consortium to serve as reference implemenations of the W3C's recommendations, as well as the libwww library, released freely by the W3C many years ago to demonstrate to programmers the technology underlying the World-Wide Web. He attributed the robustness and pervasiveness of the World-Wide Web to the free, open standards that underly it; standards which everyone is free to examine, both to determine their own implementation of it, and to point out those deficiencies which need to be eradicated for the medium to grow. This spirit of cooperation and peer review, prevalent in the academic world, is something he would like to see on the Internet in general.
He also mentioned that the W3C is looking into ways o get around the plethora of frivolous patents that are being issued, including patents on the technology underlying style sheets and P3P. "We have changed the world before," he said, "we can change it again. We change it every six months!"
Bob Metcalfe, the closing keynote speaker, was decidely less optimistic about the future of open-source software. In his speech, he gave seven predictions regarding the future of the Internet. His prediction regarding open source? "I predict it's going to fizzle." He gave several reasons for his prediction of the demise of the open-source movement, and took care to state "I'm not celebrating this, I'm predicting this."
This prediction was met by silence from the audience. When Metcalfe asked why the silence - no booing or clapping, as had accompanied his previous predictions - an audience member replied, "We're mourning."
As soon as his face appeared on the display screens, even before the announcer mentioned his name, the applause for Richard Stallman started.
From an office in Mountain View, California, Stallman appeared, via videocamera, to accept the Yuri Rubinsky Memorial Award. The $10,000 cash award is meant to recognize leaders in the development of the Internet's infrastructure. Rather than talk about technology, or thank a long list of people for helping him to get to this point, Stallman took the opportunity to talk to the audience about patents, and the danger software patents hold for the Internet.
"It would be very good if we could simply continue working on software," he said, "writing programs to carry out the ideas we have and nobody would try to stop us. But unfortunately that's not true. In the US, you're very likely to step on a software patent when you make a decision about writing a program."
While the United States does allow and enforce software patents, the status of software patents in European Union countries is unclear. He warned that multinational corporations are encouraging EU countries to allow US-style patents, specifically the ability to patent basic software processes. He predicted that should the European Union decide to enforce software patents (it will make a decision by the end of the year 2000), a number of lawsuits would crop up as a result of "stealth patents" that have been filed in Europe, but never enforced. He encouraged all citizens of European Union countries to take a look at the web site http://www.freepatents.org/ to become familiar with the situation.
It was fairly surreal when the announcer, without a trace of humor in his voice, thanked the companies which helped put together the $10,000 prize: Sun Microsystems, the Graphic Communications Association, OASIS, and Microsoft. The audience broke out in laughter at the mention of Microsoft giving Stallman its money. In parting, Stallman warned the audience against pursuing short-term gain at the cost of future freedoms: "Keep a sense of proportion. It's more important to be safe from the software patents than to get money to give an award to somebody, much as I appreciate it."
Aside from the major speeches, conference attendees had a chance to get more up-close and personal with various open-source software developers through both a BOF (Birds of a Feather) meeting with PHP and Apache developers, and a half-day session on Developer's Day.
Rasmus Lerdorf and Stig Baaken, two of the core group of PHP developers, spoke at the BOF meeting and the Developers' Day session. Two fairly interesting things they discussed were the upcoming PHP book to be published by O'Reilly, and the development of Zend, the new engine for PHP which was announced earlier this year.
The O'Reilly PHP book is being coauthored by the entire core group of PHP developers. While the content of the PHP book is essentially done, the integration of all the parts, to make it a cohesive whole, has not occurred yet. Unfortunately, the developers didn't have a good estimate of when the book would be out. (Nor did they know what animal would grace the cover.)
At the BOF meeting, the developers also revealed the story behind the development of Zend, a scripting engine developed by Zeev Suraski and Andi Gutmans, the core PHP developers currently residing in Israel. The development of Zend was the result of a decision to really modularize PHP - to seperate the core engine from the language functions and the rest of the language. Lerdorf mentioned that the developers hoped that Zend might be plugged into other products - MySQL was one example given - at some point. This would allow the Zend creators to work on creating the best scripting engine possible, rather than worrying about the external environment that it would sit in.
The last day of the conference was a "Developer's Day," in which attendees could choose to listen (and talk) to prominent developers in their field of interest. At the Open Source Developer's Half-Day (which had some subject overlap with the Scripting Developer's Half-Day), a number of developers talked about the goings-on in their various areas of interest. Aside from many of the usual suspects - including PHP, Python, and Apache - there were two notable talks.
One was given by a representative of the developers of UIML - User Interface Markup Language. The idea behind UIML is to develop an open-source language that people can use to build interfaces for on web applications, standalone computer applications, cell phones, and elsewhere. UIML separates the interface of a system from the backend, allowing programmers so inclined to focus on just the guts of a system, while multiple people could build the interfaces to connect users to the backend functionality. Similarly, a single interface designer would be able to spread a single interface across multiple applications, each of which might have been developed by a different backend designer.
Paul Everitt, from Digital Creations, also gave an interesting talk - but about the business side of open source, rather than development or technical issues.
Digital Creations makes and supports a product called Zope, a Python-based web application server. His story of how Digital Creations started out, how it got venture capital funding and released its application server as open source, and how it continues to grow is a useful lesson in open-source business.
As he told it, there were three basic things that he and the rest of the company had to deal with, in going completely open source:
Digital Creations was already familiar with the open source community, and supporting open-source software, before releasing Zope as open source. Their first product was Bobo, a free Python object publisher. Members of the six-person Digital Creations team were also active in the Python community, and, according to Everitt, had donated the first machine which served as the host python.org.
Their next product, an application server (originally named Principia), they hoped to sell; and they had dreams of taking on the world with this next product. As Everitt put it: "Netscape, watch out!"
But the impetus to give away Zope ended up coming from an investor. In late 1998, a venture capitalist whose Wall Street company had used open-source software - including Python and Bobo - to help them make millions of dollars, bought a minority investment in Digital Creations.
After making the investment, the VC suggested that they give away their product. Why? First, they expected that while they would not make money from sales, more people would use Zope if it were free, providing more consulting and support revenue than they would have made through selling.
Second, as the venture capitalist asked, "Is this the last good idea you're going to have?" In his eyes, any individual golden egg, like the product Zope, was expendable; the goose that laid them, Digital Creations, would not lose value by giving them away.
So, on the advice of their investor, they started giving Zope away.
Everitt pointed to three ways in which going completely open source, in terms of the company's products, changed the company's business plans.
The first, and most obvious, way in which the business changed was that they became completely service-oriented rather than product-oriented. A lot of technology businesses are moving towards models where most money comes from services and support. By going straight for a 100% services firm, Digital Creations skipped the intermediate step at which many companies are today, where a large, and less profitable, part of business is selling a product.
Second, Digital Creations has no problems in either hiring qualified programmers or finding consultants to subcontract with them. Because any programmer can look under the hood and see how Zope works, there are numerous people in the field who are training themselves to be the programmers that Digital Creations will need in the future. (On the other hand, there are also people training themselves to be Digital Creations' toughest competitors.)
Third, Digital Creations decided to put a lot of money into branding and expanding Zope, rather than advertising themselves. The existence of a large user community for Zope, they felt, would be more likely to inspire confidence and buy-in from their customers than just building up Digital Creations, the company. With a vibrant and well-developed user base, customers could be confident that even if Digital Creations fails, the software will live on.
The last issue with going open source was that of business value. Traditional software companies tend to count their intellectual property - the software products they sell - as a major part of their business value. When you give away your software, all that value disappears.
In the open-source business, the intellectual property is the people who are doing the work. Just as traditional software is valuable because it provides a revenue stream through sales, the people in a company are valuable because they provide a revenue stream through services. Having a number of good programmers working in your open-source business also makes you a valuable takeover target, Everitt said, if your people have a reputation for doing work that is both good and fast. Increasingly, takeovers in the computer industry are being done not for tangible assets, but for know-how and business processes. Having an open-source business just values the people, and what they know, to an even greater degree than traditional businesses.
When the World-Wide Web was first created (and Gopher before it), there was no "open-source movement," and the chances of software patents keeping programmers from building the tools they wanted were much lower. From the beginning, the World Wide Web Consortium has been releasing its browsers, servers, and associated software for no charge, and with access to the source code insured. Without that openness, the web wouldn't be as ubiquitous as it is today, with numerous commercial entities profiting from the freely available standards and software. What I saw at WWW8 gives me confidence that free software will continue to have a major impact on the development of the Internet, Bob Metcalfe's predictions notwithstanding.
The W3C, as well as being an arbiter and promoter of open web standards, looks like it will take up the fight against frivolous patents. It is already working on finding prior art to dispute a recently-revealed patent on the technology underlying the Platform for Privacy Preferences.
The talks given by the PHP, Apache, and other open source developers demonstrate that there is still a lot of energy among the folks who bring us the already-established free software we use; and the UIML developers show that the philosophy continues to infect those seeking widespread acceptance for new technology. Finally, the presentation by Paul Everitt of Digital Creations demonstrates that there is money to be made in open-source software - in contradiction with Bob Metcalfe's criticism of the movement as anti-capitalist.
Whether "open source" loses its buzzword status in the next year or not, the talks at WWW8 indicate it will remain a powerful force in shaping the Internet and Internet business.
To be notified when new conference reports, and other content, appear on ahref.com, join our mailing list.
Edward Piou is an ahref.com producer and runs ep Productions, Incorporated, a Washington, D.C.-area development company.
© 1998-1999 ep Productions, Inc. All rights reserved.