| Next: Licensing, Open Source, or Up: Giving It Away Previous: We Are in the |
Linux fills this role naturally and extremely well. The primary complaint about the market leader is the control that vendor has over the industry. A new OS must deliver control over the OS platform to its user and not become just another proprietary binary-only OS whose owner would then gain the same dominant market position that consumers are currently complaining about.
Consider that Linux is not really an OS. It has come to describe a whole collection of open-source components much like the term ``car'' describes an industry better than the thing we drive on the highway. We don't drive cars -- we drive Ford Tauruses or Honda Accords. Red Hat is the equivalent of an OS assembly plant of the Free Software operating system industry. Red Hat succeeds when customers perceive themselves not as purchasing an operating system, or even purchasing Linux, but purchasing Red Hat first and foremost.
Honda buys tires from Michelin, airbags from TRW, and paint from Dupont and assembles these diverse pieces into an Accord that comes with certification, warranties, and a network of Honda and independent repair shops.
Red Hat takes compilers from Cygnus, web servers from Apache, an X Window System from the X Consortium (who built it with support from Digital, HP, IBM, Sun, and others), and assembles these into a certifiable, warranted, and award-winning Red Hat Linux OS.
Much like the car industry, it is Red Hat's job to take what it considers the best of the available open-source components to build the best OS we can. But control over the OS is not held by Red Hat or anyone else. If a Red Hat customer disagrees with our choice of Sendmail and want to use Qmail or some other solution, they continue to have the control that enables them to do this. In much the same way, someone buying a Ford Taurus may want a higher performance manifold installed on the engine in place of the one that was shipped from the factory. Because the Taurus owner can open the hood of the car they have control over the car. Similarly, Red Hat users have control over the Linux OS they use, because they have license to open and modify the source code.
You can't compete with a monopoly by playing the game by the monopolist's rules. The monopoly has the resources, the distribution channels, the R&D resources; in short, they just have too many strengths. You compete with a monopoly by changing the rules of the game into a set that favors your strengths.
At the end of the 19th century, the big American monopoly concern was not operating systems, but railroads. The major railroads held effective monopolies on transportation between major cities. Indeed, major American cities, like Chicago, had grown up around the central railway terminals owned by the railroad companies.
These monopolies were not overcome by building new railroads and charging several fewer dollars. They were overcome with the building of the interstate highway system and the benefit of door-to-door delivery that the trucking companies could offer over the more limited point-to-point delivery that the railroad model previously offered.
Today the owners of the existing proprietary OSes own a technology that is much like owning the railway system. The APIs of a proprietary OS are much like the routes and timetables of a railroad. The OS vendors can charge whatever toll they like. They can also control and change the ``route'' the APIs take through the OS to suit the needs of the applications they sell, without regard to the needs of the applications that their competitors sell. These OS vendors' biggest competitive advantage is that they control access to the source code that both their applications and the Independent Software Vendors (ISVs) applications must run on.
To escape the confines of this model, ISVs need an OS model where the vendor of that OS (Linux) does not control the OS; where the supplier of the OS is responsible for the maintenance of the OS only; and where the ISV can sell his application secure in the knowledge that the OS vendor is not his biggest competitive threat. The appeal of this OS model has begun to take hold in the software world. This is a big part of the reasoning behind Corel's port of WordPerfect to Linux, behind Oracle's port of their database software to Linux, and behind IBM's support for Apache.
The benefit an open-source OS offers over the proprietary binary-only OSes is the control the users gain over the technology they are using. The proprietary OS vendors, with their huge investment in the proprietary software that their products consist of, would be crazy to try and match the benefit we are offering their customers, as we generate a fraction of the revenue per user that the current proprietary OS vendors rely on.
Of course if our technology model becomes accepted by a large enough group of computer users, the existing OS vendors are going to have to react somehow. But that's still several years in the future. If they do react by ``freeing'' their code the way Netscape ``freed'' the code to the Navigator browser, it would result in better products at dramatically lower cost. The industry at large will be well served if that were the only result of our efforts. Of course it is not Red Hat's goal to stop there.
As an illustration of the importance of the ``control'' benefit of the Linux OS, it is interesting to note Fermilab's experience. Fermilab is the big particle accelerator research laboratory outside Chicago. They employ over a thousand high-level physics engineers who need state-of-the-art technology that they can customize to the needs of the projects they are working on. An example of the benefit of Linux is its ability to be used in cluster farms to build massively parallel super-computers. Fermilab needs this feature, as they are proposing to increase the performance of their accelerator. As a result of this performance increase, they expect to need to analyze almost 10 times more data per second than they have been. Their budgets simply will not enable them to acquire the computing power they need from the existing super-computer suppliers.
For this and other reasons Fermilab wanted something Open Source. They recognized that Red Hat Linux was one of the more popular Open Source choices, so they called us. In fact they called us six times in the four months during the system selection phase of the project, and we did not respond even once to their inquiries. Nonetheless the result of their study was to select Red Hat Linux as an officially supported OS at Fermilab. The moral here is that (a) we need to learn how to answer our phones better (we have), and (b) that Fermilab was able to recognize that our business model was delivering them the control over the Red Hat Linux OS they were intending to use -- whether or not Red Hat Software, Inc. was in position to support them.
So whether it is the large computer consuming organizations, or the large computer technology suppliers (ISVs), the Linux OS provides benefits and is free from the major limitations of all the proprietary binary-only OSes available today. Careful brand management of Red Hat Linux among Linux distributions, and careful market position of Linux among OS alternatives, enables Red Hat to enjoy the growth and success we have today.