Do you remember the old grade-school joke kids used tell about the Vindow viper? In the story, a woman gets a phone call from a man with a foreign accent. "Hello! I am the viper, and I vill be there soon!" Over and over, for the length of the joke, she answers the phone to hear the voice repeat, "I am the viper, and I vill be there soon!"

So, Linus Torvalds! Where are you now? Will you be viping my Vindows anytime soon?

Linux 2.6.0 Kernel Announced

Some developers think so, especially after last week's announcement by Torvalds that the new 2.6.0 version of the Linux kernel is now ready. "The beaver is out of detox," he proclaimed, mystically. According to Torvalds, the new kernel has been tested on servers with up to 64 processors, supporting up to 64 GB of memory on 32-bit systems. And on the low end, the new kernel includes support for low-cost, low-power processors with limited memory management capabilities, often used in embedded devices.

This new version is a significant advancement for Linux: It transforms the capabilities of Linux beyond the realm of a toy and places it firmly in technical competition with Windows 2003 Server. At the server level, this is the race to watch in 2004. With lower licensing fees and a growing number of applications, Linux may have a banner year in 2004. And, with SCO losing the PR war, Linux may overcome the FUD factor that Microsoft has been promoting.

But is this latest version of the Linux kernel really the Vindow Viper we've been waiting for? Probably not! And the reasons are, still, our users' expectations.

Windows Users Demand More

As long as an integrated desktop is the measure of a PC operating system's success in our users' minds, Linux will suffer the same misunderstandings that have historically plagued the OS/400 operating system. To these users, the GUI is everything. That is what Microsoft learned from its theft of the Apple Macintosh paradigm: The key to marketing a PC operating system is to integrate everything to a mouse click. That requires a consistent use of rules for all applications and services, with embedded APIs that reach into every nook and cranny of the software base. Users will suffer reboots, security flaws, and even file loss as long as they think they are in control of the mouse.

This is the fact of life that OS/400 developers in the 1990s bemoaned as they tried to convince CFOs that their applications were a better investment than Windows client-server applications. Now, 10 years later, OS/400--in the form of the iSeries--is making a comeback, but not as a favorite of the desktop user. It survives and thrives as a consolidation operating system, doing the work that users don't want to handle with their mouses.

Will Linux suffer the same marketing obstacles? Does the fact that Linux is not fully integrated to a GUI really keep it from being accepted?

Don't get me wrong. There are many significant Windows-like user environments for Linux today, including X Windows, Lindows, KDE, and others. And there is a growing list of office productivity tools, most importantly the StarOffice suite from Sun. But to the majority of day-to-day Windows users, Linux still looks like a more-cuddly incarnation of DOS: a command-level, terminal interface with cryptic keywords. In short, to a Windows user, Linux is still a Charles Shultz cartoon image of an operating system sucking its thumb and holding a security blanket.

That's a long way from the image of the powerful, fear-inspiring Vindow viper.

How Linux May Ascend

But perhaps, you'll counter, that's the wrong perspective anyway. Maybe the real value of Linux will be not its integration of the desktop, but its ability to become a pervasive part of the corporate infrastructure. After all, it's currently invading corporations by providing more stable, more secure file serving, Web serving, and application serving. And that's what corporations need most right now. Not a just another pretty GUI face.

In fact, if one looks at the trends in the industry--trends that show corporations investing more new money into application serving than desktop productivity suites--the need to satisfy GUI users seems to be less important than it once was. And it is here that Linux developers are following the path taken by past OS/400 developers: They are implementing Web-facing tools, screen scrapers, and client-server models that allow the users to retain their Microsoft-based client operating systems while accessing the power of the newer Linux offerings.

That's the first wave of Linux acceptance within corporate America. The second wave may arrive through the use of blade servers and thin-client architectures. These technologies remove the complexity of the desktop operating system to present the users with only those applications that are necessary to do their jobs. For most users, that's a word processing program, an email client, a spreadsheet program, and a browser. No more solitaire, no more screen savers, no more secretly loaded user applications. From the IT perspective, this second wave of Linux implementation consolidates licenses and IT support and better enables IT to control the environments where the users do their work. No need to support that fat PC client software. In this second wave, application suites like StarOffice make sense: They provide the functionality and compatibility of the Microsoft Office products without the constant IT tinkering that drives up costs.

Beyond PC Productivity

The third wave of Linux will arrive in the form of device-specific systems in which the Linux OS will be the embedded operating system of intelligent mechanisms, like PDAs, cell phones, portable wireless devices, and a slew of other personal appliances. That's where the new Linux kernel may achieve a level of user acceptance that Microsoft CE has never claimed. Linux will be extended significantly toward inexpensive wireless devices that connect the mobile worker with the virtual workplace. Most users will never type in a Linux command: Linux will work behind the scenes through device-specific user interfaces.

IT at a PC Productivity Crossroad with Microsoft

At the same time, the industry trend of smaller, wireless devices has delivered IT to a crossroad: Is the PC really the wave of the future? Will IT be saddled with Microsoft Windows PC products forever? Is the client/server paradigm really the way to go? And if not, is there room for an upstart Linux operating system to provide the platform for solutions that corporations need? The whole issue of PC productivity--and Microsoft's support of it--will be a defining moment for both IT and Microsoft in 2004.

As noted in a previous article, Microsoft will end its support of Windows 98 on January 17. In June, Microsoft will also end support for Windows 2000. Microsoft offers Windows XP and Windows 2003 as its replacement products, but users will notice that--for many of them--these newer versions don't work exactly the same out of the box as the older versions.

For instance, Microsoft can no longer bundle the Sun Java Virtual Machine (JVM) with its latest versions of Internet Explorer. Users (or their IT technicians) are now required to directly download and install the Sun JVM just to sustain the level of compatibility that they once had with out-of-the-box Windows 98 and Windows 2000. While this will be, at most, an inconvenience to Windows user community, it is an indicator of the kinds of change that will impact loyal Windows users in 2004: increasing incompatibilities between the Microsoft proprietary products, and advances in other Internet technologies.

We can expect similar areas of what I call "proprietary incompatibilities" in XML, Microsoft .Net, Internet Web serving, and browser functionality. The DOJ lawsuit has put the fear of governmental regulation into the minds of Microsoft lawyers, and we can expect that Microsoft-proprietary technologies will be increasingly more controlled--and significantly different--than technologies and products offered by the larger global development community.

In addition, basic file formats for Office applications like Excel and Word will probably evolve yet again as Microsoft brings out new versions, with new APIs that make integrating these products more difficult for third-party Office tool providers. In the past, Microsoft could afford to fund these forward-incompatibility features as a part of its embrace-and-extend marketing scheme to inspire users to upgrade. But the intellectual property disputes and the legal wrangling of lawyers within the industry will make this market tactic much more expensive for Microsoft.

In the meantime, as Microsoft licenses become more expensive, IT may actually balk. After all, what is the value of these increased licensing fees if the result is merely an increased load on IT technical resources? Who, in this evolving world of Microsoft intellectual property, will really be in control of the desktop? Not the user! Not IT!

The Future Challenge of Linux

All of these issues frame the future challenges of Linux in 2004. Linux has made great strides in 2003, and the road that has led it to the doors of corporate acceptability has allowed it so far to skirt the issue of the user desktop. Linux has a great heritage of UNIX applications fueling its future, too. It has the tremendous resources of IBM's marketing support. And it has the goodwill of a worldwide community of developers. In IT, we tend to believe that its next big barrier to acceptance will be the desktop, but more than likely, that will be the least of its challenges because the desktop will be less important.

Indeed, Linux's greatest challenge may come if IT actually decides to embrace it as the real future IT platform for application development. With a developer community dispersed all over the world, a fragmented support structure delivered by a slew of OS re-packagers, and a complex array of nested licensing agreements between distributors, developers, and OEMs, Linux is a potential train wreck that could leave IT in a support mess that could make a proprietary system like Windows seem quite alluring.

That's what I believe is the true challenge of Linux in 2004. If it can better position itself as a supportable "solution"--and not as an upstart operating system--it might actually achieve corporate acceptance by both large and small organizations around the world. Until then, it will probably remain merely the Vindow viper joke told to scare the legions of Microsoft users who have settled in with their mouses for a good night's sleep.

Thomas M. Stockwell is Editor in Chief of MC Press.