|
As technology planners review their new kernel-related reviews about SMP, file system, and ARCnet support, practice leaders at Booz Allen & Hamilton carry on with a bias toward the science of business, as astute watchdogs for sustainable sources of enabling technologies beyond the gee-whiz features of technologies themselves. Barry Jaruzelski, lead partner of Booz Allen & Hamilton's Global Computers & Electronics practice, considers the future of industrial-strength Linux, placing that future on the bottom line: Will Linux survive its business models?
Scenario 1 · Player consolidations
Q We're asking you to give us a Booz Allen assessment on the future of Linux.
JARUZELSKI Well, we could easily make this a checklist discussion of technical features and how advancements push Linux's standing in the light of Windows NT, Solaris, and other platforms. But beyond the technology itself, the real question is: What's the Linux end game? Do the business models that we are seeing in the Open Source community need to change? What will it take for Linux to be viable in the enterprise space?
Q How seriously are these questions being posed and debated?
JARUZELSKI I'm hearing some strong views outside Booz Allen that there's no way for Linux to achieve its potential in that space unless we see some changes. Right now, the two most important players are IBM and Red Hat. IBM has placed high commitment and resources, and has had two of the biggest wins. On the commercial side alone, 15,000 servers for a convenience-store chain in Japan; putting applications such as DB2 out there for Linux as well as WebSphere. On the technical side, IBM has had a series of clustered supercomputer wins. [IBM last year announced that a Japanese convenience-store chain will install more than 15,000 IBM servers running Linux, making it one of the largest IBM Linux-based installations. IBM also announced DB2 Universal Database for Linux on Intel-based clusters and WebSphere for Linux.]
Q So what changes do you see and what's your estimated timeline?
JARUZELSKI The main move to expect in the near future, no matter what kind of end game we see later on, will be consolidation. Right now, there are so many distributors out there for bundling and services, and one can expect to see significant consolidation on the distributor side. Many players that presently stand alone are under financial pressures, in a climate where stock valuations have dampened. Consolidation waves generally are such that one is never sure when they will happen. But once they start, they happen fast. There's a flurry of actions in a compressed window. Once we see the consolidation wave pass, the clock will start in the direction of some alternative end games that will begin to present themselves.
Scenario 2 · Happy order-entry clerks
Q Before we get to those end games, I want to check to see if I'm right about your view of the future of Linux. I sense that you're not seeing the kernel as a signal for a Brand New Day. I sense you're taking a harder look than that.
JARUZELSKILinux has seen good traction in the technical-computing space. It benefited from the inherent infrastructure, nature of the community, and its preference for bare-bones technologies. But in order to penetrate the enterprise, a lot of stuff is still required.
Q Such as?
JARUZELSKI A number of enterprise applications are not yet available on Linux. We've seen some good signs in such commercial applications as from Oracle, for example, but commercial application availability is still running comparatively thin. If you talk about enterprise-strength platforms, you need to see a plethora of applications. And while attention surrounds Linux on servers for the enterprise, the important question is what to expect in Linux on desktops. The big problem in my view is we need to see stuff out there that won't just make the Ph.Ds from Stanford happy, but stuff that can make the order-entry clerks happy.
Q You're getting into the real world of business, apparently, where we have CEOs who do not want their IT workforce spending their time looking under the hood. They want an IT workforce that will drive their product lines and lines of business. What moves could push Linux forward dramatically as a desktop solution?
JARUZELSKI Well, Microsoft Office is not available on Linux, but think about how much more key the migration to Linux would be. Who knows what will happen with remedies in Microsoft? A new administration has at least hinted it might have a different perspective than the Clinton administration. But if the draconian remedy does happen in the form of a Microsoft breakup, and if there will be an independent applications company as a result, then the company might think how to exploit the Linux desktop market. This would dramatically enhance the visibility of Linux. You know, getting over hurdles in general are so much higher in the business space. Technologies are enablers, but business needs to invest in applications and people resources to implement those enablers. And we now get into the other area where Linux lags, which is service and support.
Scenario 3 · Beast-strength support
Q But don't many Open Source vendors stress service and support, and don't they count on service and support as their bread and butter?
JARUZELSKI How much support is actually out there? Let's say I am a customer who wants to do SAP on Linux. Who is going to install it? Where's my documentation? Where's the 24 x 7 support line? Who's going to deliver in a truck? Traditional computer vendors such as Compaq, HP, and IBM took years to build that level of support infrastructures to feed the large corporate beasts.
Q What about Linux players such as Red Hat?
JARUZELSKI There is not a level of service and support that is comparable to the service and support levels coming from older major vendors. After all, it's not that Red Hat or others are doing something wrong, but the question is, what would be reasonable to expect from Red Hat? It is a startup. They have half the market and have acquired other players and continue to grow. They are doing well, but enough to support the enterprise space? As of yet, the answer is "no." You also face a chicken-and-egg situation, where Linux was gaining the most traction in universities and research labs where there was the least willingness to pay for anything. So investments in enterprise-strength services and applications have been slow in coming.
Q We sense that you're looking at a future where hard realities drive home the point where, even if Linux makes considerable advancements technologically, it may not be able to sustain itself in the enterprise? It could be a hard pill to swallow for those who are so committed to an ever-upward spirit of Open Source.
JARUZELSKI We're talking about the future of Linux in business. This is not a fairy tale. The Linux story right now is one of evolution, where things are still embryonic. The situation is not yet sustainable. The excitement about Linux is about what Linux can do, not about what Linux can already do. The question is, can you have your cake and eat it too? You must not enter into the trap and confuse a sustainable economic model with religious fervor around free software and a posture of "anything but Microsoft."
END GAME
Q The emotional appeal of Linux has been an undeniable force…
JARUZELSKI Well, if there has been an inherent saintliness from some adherents, let me note that saints usually got crucified. The business world is nasty. If participants do not have a sustainable business model, the benefits of Linux will not happen. We're talking about the need to see proper levels of support, installation, and administration. There may be a point where it hits the wall, with not enough people and not enough money. If you come to a fork in the road, and I do not think it is unreasonable to imagine, where the system no longer sustains itself, and if Linux players say 'No, we cannot let it happen,' then two things must occur.
Q The alternative end games that you mentioned earlier on?
JARUZELSKI Two options: Either a very robust consortium or somebody with very deep pockets. The consortium could guide and focus the Linux investment, providing structure for Linux players and driving the growth of a bigger service arm to feed the enterprise beast and creating a vehicle for money to be injected into it. IBM and others appear to be starting down that path, as we saw with last year's announcement of a cooperative testing lab. The alternative end game might be a big player with enough market power to take control of the operating system and provide the continued level of investment for the evolution required.
| 
