dcsimg

In a World Without Taxes, Who Builds the Road Ahead?

One of the most powerful advantages that the open source development model offers is the way it handles growth. Each sale of an open source system creates a larger network of people who become contributors to the project. Open source project maintainers are typically major code contributors, but they must also be good managers. Successful projects make optimum use of the community to find problems, pro-vide bug fixes, and accept innovations and improvements to the code. With open source, a small number of developers can maintain a very large software project.








Trenches Art
ANDREA BARUFFI

One of the most powerful advantages that the open source development model offers is the way it handles growth. Each sale of an open source system creates a larger network of people who become contributors to the project. Open source project maintainers are typically major code contributors, but they must also be good managers. Successful projects make optimum use of the community to find problems, pro-vide bug fixes, and accept innovations and improvements to the code. With open source, a small number of developers can maintain a very large software project.

In the proprietary world, each software sale actually becomes a liability to the seller. When people buy commercial software, they expect someone to support it. This costs money. And often proprietary software vendors find themselves spending even more money to maintain their software as it becomes bloated, unstable, and increasingly more difficult to improve.

This is one of the reasons it is now common for vendors to charge for commercial software, and then charge again for technical support. In a healthy market, support to end users would be part of the purchase price of the product, but proprietary software companies need to find ways to offset their spiraling development and support costs. Here, as always, it’s the customer who pays the price.

As commercial interest in open source software increases, pressure is being placed on the development community to produce competitive products, provide predictable support, and be accountable for delivering according to predetermined schedules. Even the most enthusiastic open source contributor is reluctant to commit to anything other than “it will be done when it’s done.” A model has evolved where those companies who benefit from sales into the open source community are funding projects that will satisfy their business requirements. But, those requirements are often narrowly focused and don’t cover the costs of producing the increasingly complex software infrastructure that is necessary to support today’s computers. We have a lot of companies paying to build cars, but few will agree to fund the cost of the roads.

It’s not difficult to build the cost of infrastructure development into the cost of proprietary software. Commercial Unix vendors have traditionally factored infrastructure costs into the price of their computer systems. When you buy a Sun workstation, you’re contributing to the development of the entire Solaris OS. The advantage of this approach is that the software infrastructure can take full advantage of the hardware being sold. The downside is price.

The Microsoft Tax

One of the reasons that Windows has been so successful is that it allows hardware vendors to standardize on a consistent infrastructure. Consequently, Microsoft has been able to make things significantly less expensive. Unfortunately, this was done by imposing licensing and certification fees. The incentive to pay the “Microsoft tax” and not risk damaging a business relationship with Microsoft is firmly established in the PC industry. Many vendors are reluctant to publicize their involvement with alternate OS companies for fear of reprisals by Microsoft. Some complain they are forced to pay royalties for Windows on all systems shipped, even when that OS is not included with all units sold. These factors present a serious disincentive for vendors to stray from Windows on their systems.

The real issue we need to face in the open source world is how to fund a fully malleable infrastructure, so that we can all be free from the control of a single vendor. Linux represents freedom from single-source, proprietary-software dependency. It also poses an important question: In a world where there is no “Microsoft tax,” who pays for the infrastructure? To date, a variety of Linux vendors have contributed to the task. Red Hat and SuSE have traditionally provided significant funding, and VA Linux is now heavily contributing to this effort. Companies such as Matrox, ATI, 3dfx, SGI, and Intel have absorbed a lot of the costs of producing an open source graphics infrastructure, but there is much more work to be done. Open source development is being significantly outspent by proprietary interests. Fortunately, we don’t need to spend nearly as much money to get a much better result. Paraphrasing Albert Einstein, we should put as little money into it as possible…but not too little! If the open source paradigm is to succeed, all vendors in that space need to contribute, either by funding third parties or by their own internal efforts, to improve the infrastructure. Buying the infrastructure today will save a lot of money in rent tomorrow.

Linux has proven itself in the server market and is posing a serious challenge to the “powers that be” in that area of computing. Some vendors, however, are dismissing the desktop as unreachable with Linux. Their attitude suggests they believe Microsoft has already won the business and commercial desktop war. I disagree.

One Part of the Puzzle: The DRI

Thanks to industry funding, the company I work for, Precision Insight, has created a fully open source Direct Rendering Infrastructure (DRI) for Linux. This has enabled workstation-level performance for graphics on a Linux PC.

The DRI is a collection of functions that are integrated into the OS within various software layers that enable graphical output from applications to make optimum use of the capabilities present in graphics hardware. The DRI makes use of several standard graphical layers, including the X Window System. It uses portions of each layer, and connects them via the DRI control structures to provide the fastest possible path between a drawing request from an application and the realization of that request on the display. The DRI must continue to evolve as technology improves, and that is the main focus of my company’s business.

Linux itself was created by volunteer developers who believed in the aesthetic purpose of their effort, but they’ve created something that is far more important than they had first imagined. Commercial vendors are beginning to understand the benefits of Linux to their respective companies. Although Linux is free to use, it will take serious investment by commercial companies in order for this OS to achieve its full potential. Wall Street’s reaction to Linux has clearly shown that the world is watching. It will be up to all of us to make it happen.





Frank LaMonica is president and CEO of Precision Insight Inc. He can be reached at frankl@precisioninsight. com.

Comments are closed.