Surveying the Linux Landscape

Just as nature abhors a vacuum, the marketplace cannot seem to abide a steady state, especially one that stifles innovation. Not long ago, IT users were doomed to performing superhuman programming feats just to connect one departmental system to another. Then along came the Internet, complete with standards that allowed us to begin connecting everything and everyone -- in the process launching an era of what economist Joseph Schumpeter called "creative destruction."

Trenches Linux/IBM2

Just as nature abhors a vacuum, the marketplace cannot seem to abide a steady state, especially one that stifles innovation. Not long ago, IT users were doomed to performing superhuman programming feats just to connect one departmental system to another. Then along came the Internet, complete with standards that allowed us to begin connecting everything and everyone — in the process launching an era of what economist Joseph Schumpeter called “creative destruction.”

Not content with that, the marketplace — seeing an IT industry resigned to living with a Babel of different application interfaces and competing over proprietary operating systems — discovered yet another disruptive technology in the form of Linux. And in the two-or-so years since Linux began to enter the mainstream, its momentum has accelerated beyond anyone’s dreams.

Gaining Speed

In fact, IDC estimated that Linux shipments grew in 1999 by more than 93 percent over those of the previous year. Certainly, it was growing from a relatively small base. Nevertheless, Linux was number two in unit shipments with almost a 25 percent share, behind only Windows NT and ahead of the combined shipments of AIX, Solaris, and HP-UX.

Looking to the future, IDC projects that Linux unit shipments will grow at more than 28 percent, outpacing all other operating systems and becoming the highest volume operating system sometime in the middle of the decade. This projection is impressive when you consider that it counts only server pre-loads and retail sales — it doesn’t include the millions of copies of Linux that are downloaded from the Internet every year.

It is pretty clear that Linux is off to a fast start, especially considering the fact that operating systems tend to evolve over seemingly geologic time spans. The Linux phenomenon stems from a number of factors, but the most prominent of these are: its origins in an international community of programmers who are interested more in creative code than proprietary rights, an incredible flexibility that permits Linux to run anywhere, and a growing perception that Linux is the ideal base for future innovations around emerging technologies.

The Open Source Community: Strength in Numbers

Eric Raymond once said that “…the really great hacks come from harnessing the attention and brainpower of entire communities. The developer who uses only his or her own brain in a closed project is going to fall behind the developer who knows how to create an open, evolutionary context in which feedback, exploring the design space, code contributions, bug-spotting, and other improvements come back from hundreds (perhaps thousands) of people.”

With the growing open source soft-ware movement effectively crossing corporate, industry, and geographic boundaries, Linux reflects the collective, worldwide wisdom and experience to which Raymond alluded, and the community is truly open to any qualified developer wishing to contribute to its maturity.

Since the Linux operating system is not the property of any one company, its future is determined by the aggregate interests of its value-net, from distributors to value-add providers to customers. This neutral playing field makes it attractive for vendors to give customers a high-volume, standards-based foundation that easily accommodates leading-edge technologies, while competing with each other by layering unique value around that open-source base.

What’s more, the industrialized world will continue to grow profoundly dependent on huge, interconnected software programs. In fact, for precisely that reason, software was one of the critical research areas singled out for increased investment by the President’s Information Technology Advisory Committee. (I was a member of this committee —

the full recommendations are available at http://www.itrd.gov/ac/report/.)

Certainly, the role of government and industry is and will continue to be critical in this area. But with neither of them commanding the skills that one finds in such abundance in the open software community, it is reasonable to expect that the community will grow in influence and importance in the coming years and support the vast, interconnected software complexes that will undoubtedly characterize our society.

Application Flexibility: Freedom of Choice

While the Information Technology industry has seen operating systems emerge with great fanfare in the past — OS/360, VMS, Windows, Solaris, and AIX come to mind — never has one developed as quickly and spread as broadly as Linux. In its relatively short life, Linux has already become the most ubiquitous of server operating systems.

Unlike Linux, almost all other operating systems are proprietary, with their interfaces and capabilities controlled by the individual companies that own them. As we have discussed earlier, Linux is open; its interfaces and capabilities reflect the consensus of the professional community of developers that oversees it. In addition, Linux runs on the widest range of platforms in the industry: all of the Intel server platforms, all of the Unix RISC platforms, new information appliances, even mainframes run Linux.

For decades, the Holy Grail of interoperability has been sought by the industry; with Linux, it is within reach for the first time.

Never before have customers had this much flexibility in selecting server platforms to run their applications. And never before have customers been able to connect these servers as seamlessly as with Linux.

This flexibility can provide a colossal advantage in an e-business environment with its unpredictable transaction volumes. Imagine being able to build and deploy your application on a PC server and then, as business increases, being able to simply move it to a midrange server or ultimately to a mainframe.

Being able to select from the long list of server options available for any single Linux application lets e-business designers optimize just about any solution. And if the application eventually needs extensions or alterations of any sort, they will be available for a Linux application if they exist anywhere in the industry.

Supporting Technological Innovation

Because of their steadily declining cost and increasing power, high-volume technologies — which were originally developed for PCs and consumer products — are becoming the building blocks of choice for a wide variety of systems. As a result, these low-cost technologies — memory, microprocessors, I/O, storage, etc. — are increasingly finding their way into innovative, new configurations.

In fact, the prices of these technologies are reaching the point where it is possible to incorporate full-fledged computers into consumer electronics and billions and billions of other everyday devices. Thus we are seeing the emergence of special-purpose, information appliances.

These same technologies are being aggregated into much larger “super-dense” servers aimed at a whole variety of Internet applications and capable of achieving huge processing capacities at very reasonable prices. Even larger scientific and technical computers will be built by aggregating these technologies. IBM’s “Blue Gene” system, for example, will be capable of a quadrillion calculations per second as it performs very sophisticated analysis of proteins.

What does all of this have to do with Linux? Well, it all starts with those information appliances. As these appliances increasingly use more powerful, low-cost hardware (thus becoming capable of supporting ever more sophisticated software and applications), they will certainly require a more general-purpose operating system to manage it all.

Its small footprint makes Linux an excellent choice for deployment in information appliances and the emerging wave of pervasive devices. But even as Linux becomes richer in features, it remains very modular. Consequently, as a whole new class of servers and supercomputers eventually comes to market, leveraging many of these inexpensive high-volume technologies but aggregating them to achieve very large capacities, we should see Linux as the ideal base on which to develop innovative software and applications.

A World of Possibilities

The Linux community will continue to improve the basic functionality of Linux. Vendors will leverage Linux to bring to market many new products. And the marketplace, in turn, will surely leverage these open, scalable, and attractively-priced platforms to develop not just more new applications, but also entirely new classes of applications heretofore neither feasible nor affordable.

Dr. Irving Wladawsky-Berger is IBM’s Vice President of Technology and Strategy, IBM Enterprise Systems, where he serves as IBM’s chief visionary for Linux and e-business.

Comments are closed.