The Open Source Movement has gotten a lot of attention recently, and it has moved into the realm of geometry kernels and closer to home for us. There is a classic paper on the Open Source Movement written by Eric Raymond. It can be found on a number of websites by searching for "The Cathedral and the Bazaar". The article traces the history of Raymond’s development of Unix fetchmail and provides an interesting socio-economic view of the open source philosophy. Unix and Linux are, of course, the classic examples of the advantages (and disadvantages) of open source.

The advantages are clearly that a community of skilled and motivated people can evolve a far better software product than can a highly bureaucratic and stifled "software team" despite all the efforts of modern software engineering. While it may by argued that during the "creative phase", this philosophy works very well, it can also be argued that during the "maintenance and enhancement phase", the argument suffers. Ultimately, some established organization with a solid business objective must take over the daunting task of maintaining the software and providing stability with the code, the documentation, and the configuration control of new versions.

The keys to retaining the sense of "community" to further the enhancements and development of the algorithms and code are commitment, communication, and follow-through. Raymond’s article gives a number of examples of successful open source projects focused primarily on Linux. Note that Red Hat became the entity that provided the solid business and financial foundation leading to a stable and reliable Linux product. It was essential that a business entity take on that role in order for Linux to thrive as a stable and viable product. Unix suffers from the fact there are so many versions of Unix in the industry that portability is seriously compromised.

Les Piegl and Wayne Tiller (a partner of Solid Modeling Solutions) wrote the definitive "The NURBS Book" on non-uniform rational B-splines (NURBS) with aids to designing geometry for computer-aided environment applications. The fundamental mathematics is well defined in this book, and the most faithful manifestation in software is in GeomWare’s Nlib. SMLib has Nlib as its NURBS kernel. Nlib is "open" in the sense that the design is fully documented, and the extensions to the basic kernel are available in source code. The NURBS history is documented in papers available on the SMS website at www.smlib.com.

The SMS team has preserved the dynamics of an open source philosophy. We continue to provide source (not just object code), we have a strong commitment to the evolution of the technology, and we are extremely focused on communication with our community of users. A key difference in our approach is the level of commitment we ask of our users.

Since our users are not doing rocket science (they are busy doing things that are much more difficult!), we try to free them from the burdens of reinventing the NURBS geometry utilizing an adaptation of the open source philosophy. Rather than freely giving out the source, we ask our users to commit with us to a financial relationship for support and enhancements. There is no startup payment, and users can cancel their participation at any time. After two years, they are free to retain the source without further financial obligation, but to date, all of our customers have continued to subscribe since they have seen the clear benefits of sustaining a robust relationship with the SMLib team.

We ask for a financial commitment for two basic reasons. First, we need it to provide the very best support and regular upgrades. Second, we will remain independent of proprietary bias in order to better serve our users. We are not owned or financed by other companies that could compromise our commitment to SMLib, and we have truly committed users with complex applications. The fees amount to less than a quarter of the cost of a fully-qualified expert on your staff. The commitment we ask of our users provides us the financial stability to continue to fix bugs, enhance the technology, and release new versions of the software with a frequency that fits our users needs.

Since our users are developing serious applications and have made a substantial commitment to SMLib, every user receives extraordinary support. We are not bogged done with casual users. Industrial strength application developers need and expect high levels of support to do their job efficiently. The SMLib business model allows us to look at bugs and often fix them within a week if not within a couple of days.

SMS prides itself on both its technical knowledge (the original creators and leading experts of NURBS technology are the core development team for SMLib) as well as responsiveness to our users. User development teams are extensions to the SMLib enterprise, and our staff feel that they are extensions of our user’s enterprise. That sense of teaming has led to a first-rate product (not just free code, but code with real industrial-strength quality), first-rate support, and the most cost-effective solution for our users. Our users have considered "Total Cost"; a function of license fees, as well as costs associated with the user’s productivity. Quality code and responsive support reduce costs associated with development time, software maintenance, ease of use, reliability of software in performing operations, relative performance of the basic operations, and turn around time for bug fixes. In the final analysis, those are the key ingredients that differentiate us from our competition.

Copyright © SMLib™, TSNLib™, GSNLib™, NLib™, and SDLib™ are trademarks of Solid Modeling Solutions.
All rights reserved.