TORONTO -- Day 2 of the KMDI Open Source conference started with Robert F. Young, cofounder of Red Hat, Inc. presenting a positive view about open source business models. At the end of a conference day lasting nearly 13 hours, Young returned to deliver the conference's keynote address. In between, Jason Matusow of Microsoft's Shared Source Initiative gave attendees a very different point of view about the value of open source in the enterprise. No real surprises there.
A company can have the world's best business plan, Young said, but still not be completely successful. It isn't simply the business plan that makes a business work; it's talking to customers and running the business that makes the plan produce results.
Young's basic business plan when he started Red Hat was to find a way to pay his rent after being laid off from a failing company.
The key: Always listen to customers
Companies, he said, that start with a perfect business plan don't necessarily succeed. The business plan can cause missed opportunities. The key is not the business plan or model that the company is planning to use but simply listening to what customers want. What your customers want is more important and more relevant than what the venture capital investors want.
His initial idea was to start an all-Unix and related systems book store. He went out and asked potential customers what it was they wanted, and their answer was categorically this: to find out what this free software thing they were hearing about was.
Red Hat created a business based around distributing Linux, and pretty soon it was making seven figures a year.
As the company grew and there was money to be invested, he started attending conferences of Unix vendors. At one such conference in 1995, Scott McNealy, CEO of Sun Microsystems said in a keynote address free software could have no P&L (profit and loss) because it had no P. That made Young happy, he said, because free software was already getting attention at that early stage.
Their customers, Young said, told them not to go proprietary with their software. The great strength of Red Hat to many customers was that the customers could fix the software if it was broken, because source is included, unlike existing proprietary software models.
Red Hat does not sell a product, he told the audience. Red Hat sells control of a product. People are willing to pay for Red Hat even though they know they can get the same product for free because of that.
Like so many analogies in the computer industry, Young compared it to the car industry. Proprietary software, he said, is like an auto manufacturer selling a car with a locked hood that only the manufacturer can open again. Open source allows people to see what's under the hood, even though not everyone will ever actually fix anything.
Red Hat's loss leader
A member of the audience asked if Red Hat loses any money from the $2 Red Hat CD sales that take place around the world.
A lot, said Young, but gracefully. It isn't a problem to the company, because in the long term it's still good for the company, and it's good for Linux as a whole. It allows Red Hat to get to markets that it would otherwise have trouble reaching, he said, by allowing people to sell it in places where the company doesn't market. Those markets would otherwise be essentially cut off from Red Hat if the practice was not allowed.
Besides, as people who buy the $2 version of Red Hat move on, many may get to a point where they can afford the official boxed set, so really they don't lose. It becomes something of a loss leader.
The attendee with the view most contrary to the majority of conference attendees was Microsoft's Matusow.
Matusow said that there is no correct way to distribute software. In fact, he told us, the GPL is itself a proprietary license as are all software licenses, because at the root is the assertion that someone owns it and can give someone else permission to use it.
Linux, he said, is growing largely because of the endorsement of large corporations such as IBM and Novell, which have at their heart their own corporate interests which happen, for the moment, to coincide with Linux. With about 60,000 software companies around the world, it is not just a matter of Linux versus Microsoft. There are a lot of open source and a lot of proprietary software companies.
He addressed Young's assertion that listening to customers is the most important part of a business plan. Of course it is, he said, that's why 80 percent to 90 percent of the features in Microsoft Office are customer-requested features.
Frequent upgrades impossible?
The open source philosophy of "release early, release often" does not work in large corporate settings, he said. Getting a release cycle every two years or 18 months is hard enough; frequent upgrades are simply impossible, he added.
Matusow went on to make a point about Red Hat's corporate Linux licensing, saying that Red Hat has a per-CPU licensing scheme with an auditing clause in the contract, and that client companies could not modify the (GPL'd) code for risk mitigation reasons on Red Hat's part.
He commented that some Microsoft executives' early comments about open source were unfortunate, considering Microsoft itself uses open source in some situations. BSD code, he said, is in Windows, and BSD is used at Microsoft's hotmail web mail service.
Microsoft's Shared Source iniative is so-named because Microsoft wanted to avoid using the term "open" for anything, lest it open too large a can of worms. It did anyway, he noted. Eric S. Raymond, for one, finds the term and concept offensive.
Microsoft's Shared Source is a framework, not a license, he told us. Governments, corporations, educational institutions and others get a read-only license for Microsoft code. Showing the source code is a source of trust, he said. Less than 5 percent of users with access to the source will ever actually modify anything, he said.
Open source, he said, is a pejorative term. The Linux Standard Base, he told us, is necessary for independant software vendors to build on top of Linux.
In the question-and-answer section following his presentation, a member of the audience asked when Microsoft would allow users to save as Open Office XML files, an official standard document format. He avoided the question rather expertly.
Microsoft needed the SCO license
He was next asked about Microsoft's relationship to the SCO lawsuit. He said that Microsoft needed the license from SCO and that the amount paid for that license would not cover the legal costs of the company's suits.
He finished with the statement: "[Microsoft] will continue to compete based on product merit."
After a reception and break, Young took the floor again, this time to deliver an energetic hourlong keynote address.
Young told of his educational background. He acquired a General Arts honors degree from the University of Toronto, the site of the conference, around 30 years ago. He said he was never a very good student, and following the university, he became a typewriter salesman after discovering that he needed several more years of education at institutes he couldn't get into with his grades. He wanted to become a lawyer.
Like many so-so students, he spent much of his homework time in the library reading about any topic possible not related to his homework. There, he learned about the thinker Adam Smith, whose view of the world he explained. Fundamentally, it is that people who go out to make their own lives better can make the world a better place in the process more easily than the world's most benevolent king.
Corporations, he told us, work well when they are small. They are generally responsive and innovative, but as they grow, they lose it. They become dysfunctional, stop innovating, and stop talking to their customers as they take over a larger and larger share of their market -- until they're a monopoly with no idea what's going on and increasingly high prices.
When the 'umbrella' effect kicks in
At that point, he said, the umbrella effect kicks in, where high prices and lack of attention allow smaller companies to enter the corporate lifecycle themselves.
After some discussion of Red Hat's corporate evolution and his own personal wealth, he spent a fiew minutes giving an overview of copyright law history -- primarily in the U.S.
In 1950, he said, copyright lasted for 20 years after the creation of a work -- the same as a patent. Now, 54 years later, the life of a copyright is 75 years after the death of the person or company that created the work.
Up until 1976, in order for a work to be copyrighted, it had to display the familiar "(c)" on the work. In 1976, that requirement was eliminated. All works were assumed to be copyrighted, and the public domain ended in the U.S. in a penstroke. Since 1976, he told us, it is assumed if that a work exists, someone owns it. And therefore that if you use that copyrighted work you could be sued at pretty much any time in the future.
The GPL, and Richard M. Stallman, who he called a true visionary, replaced the public domain. The license filled the void in copyright law created by the elimination of the public domain.
Patents, he said, stopped having to be specifically for inventions when the US PTO -- Patent and Trademark Office -- ran out of space to store all the inventions that had been delivered to them as part of the patenting process.
Got the money? Get a patent
Patents last for only 20 years. What can be patented has over the years devolved from a physical invention to an idea or concept. Anything, he said, can now be patented if you have the money.
He created the Center for the Public Domain with a $20 million budget and a mission to increase public awareness and debate on the issues surrounding copyright and copyright law.
Software patents are stupid, he said. When the USPTO surveyed software companies to find out if software patents should be implemented in the 1980s, the answer was a unanimous no. Lawyers and willing academics, he told the audience, managed to bring software patents into reality.
Intellectual property law, he said, is being written now -- not in actual, highly debated laws, but in hastily negotiated international treaties. Governments around the world are creating their IP law based on treaties signed among each other, which does not come up for public debate the way a regular law would.
The result is that many countries around the world are finding themselves with U.S. corporate-interest intellectual property laws.