Over the past year, I have watched and written about various Linux technical issues and gained invaluable experience within my own company, Command Prompt Inc. My experience and the revelations that I have had at the last two LinuxWorld expositions have given me some insight into where Linux is going.
The August 2000 LinuxWorld event was crazy. The show had outgrown the San Jose, California Convention Center, and some exhibitors were relegated to the halls. The floor was loud, with multiple magicians and presentations. You couldn't walk past three booths without somebody trying to get your attention or give you some swag. Many vendors hired women to wear costumes specifically designed to please the male attendees. The stock market had not yet reared up and quashed the IPO frenzy it started, and people still felt alive with Linux euphoria.
The feelings of euphoria disappeared in the third and fourth quarters of 2000. Corel, which was once considered by many to be one the larger forces in the commercial Linux arena, laid off at least 320 employees. Profit warnings from companies like VA Linux abounded, and stocks that once rode in the utopian clouds sank to US$8 per share.
There was a perception in the investor community that Linux had hit its ceiling, that some companies would merge and others would disappear. Some fellow consultants (the Microsoft kind) let the constant barrage of bold headlines and slightly skewed press reinforce their feelings that Linux was a fad. Within Command Prompt, we saw potential customers close their purses.
Many firms' cash flows were drying up, forcing them to adapt. Some Linux companies like Maxspeed adjusted their product lines to be more Microsoft-centric. The widespread hysteria of credit, VC funding, and mass spending was over and some companies found out the hard way that the traditional rules of business still applied.
As 2001 approached, many people wondered what would happen within the Linux market. There were rumors of a merger between Turbolinux and Linuxcare, and the distribution company Stormix went bankrupt.
Then IBM announced that it would invest $1 billion into Linux for the year 2001. The announcement itself was not a shock but it sent waves of gratification through the Linux community. IBM has been a proponent of Linux for several years and has continued to mount stronger support for our operating system every year. The shock of the announcement came from the number of dollars IBM was willing to invest into something that is "free."
But the investment wasn't the largest part of IBM's legacy for 2000. Unlike most Linux companies and many other technology giants, IBM proved profitable beyond expectations. IBM not only generated over $6 billion in operational cash for the company's continued growth but also recorded a backlog demand of over $85 billion in services. Apparently, IBM figured out how to apply old economy stability to new economy profits.
Granted, the majority of IBM's revenue didn't come from Linux. However, the fact that Linux now runs on every IBM platform except the AS/400 shows a serious commitment by one of the largest companies in the world. A commitment of that magnitude shows that Linux will be around for the long run and is no longer just a fad technology causing grief for traditional companies. Linux is becoming a true, marketable, business-quality technology.
The New York expo
At the end of January we had the LinuxWorld Conference & Expo in New York. Earlier, I had predicted that we would see a calmer and more mature audience, that the attendees would be people looking for Linux answers. I believed the expo wouldn't be as loud as the previous one and that the vendors would be more interested in proving their wares than in dressing up young women in grass skirts.
I think that I was correct. As I spoke with vendors in New York they were more interested in honest delivery of product. I didn't receive the same sales fervor and promises of vaporware. Instead, vendors made thoughtful requests for feedback and product review. They wanted to know what I thought of their product ideas. I believe that the expo in New York was more about showing real product. Some of the vendors I spoke to can be found on my product roundups from the show.
However, some of the old Linux hype persisted. Ximian's booth was made up like a Congo hut, the GNOME folks seemed more interested in playing foosball than in advocating their software, and Chilliware's booth teemed with voluptuous ladies. I will say that it appears Ximian has cornered the desktop market for Unix and Linux, though.
With the advent of the GNOME foundation, Ximian and Sun announced that Sun would replace CDE with GNOME as the default desktop for its Solaris systems. Ximian reciprocated with the CDs it distributed for free at the show: The platforms supported by Ximian GNOME were all Linux, save one -- Solaris/Sparc.
At the show HP and Ximian also announced that HP would eliminate CDE as the preferred windowing platform for HP-UX and replace it with Ximian GNOME. A trend toward GNOME as the preferred desktop on Linux is becoming evident. I am aware of the arguments for a choice of Linux desktops -- and agree with them -- but if a series of large companies take steps toward a given product, it will become dominant within the marketplace. That product is Ximian GNOME.
However, beyond a GNOME versus KDE argument, I believe that the announcements by Sun and HP suggest a more important event. Solaris and HP-UX are losing traction and their parent companies know it. Sun and HP are going after GNOME so they can influence the development through the GNOME foundation, and so they can make their respective Unixes act like Linux.
Yes, GNOME runs on *BSD. Heck, it even runs on Microsoft Windows but Linux remains the primary platform for GNOME. If HP and Sun can make Solaris or HP-UX "Linux compatible," they will have a stronger foothold for bringing market share back to their own operating systems. To be clear, I know that Solaris has more of the server market share than does Linux. Solaris, however, trails Linux on the desktop and in small business. If Solaris were made to run like Linux, Solaris just might get some of the desktop and small-business market. Sun has been making aggressive moves to enter that market, such as its recent purchase of Cobalt. Cobalt produces small-business Internet and file servers that run Linux. The same plan could potentially work for HP and HP-UX.
San Francisco here we come
As we travel the six-month road to the August 2001 LinuxWorld Conference & Expo in San Francisco, we will see more mergers. They will affect Red Hat the most. The more Linux companies that merge, the harder it will be for Red Hat to keep its market share. I believe that in the United States, Red Hat will start to become the 800-pound gorilla of the Linux world. Remembering the uproar over Red Hat 7.0, with its incompatible GCC and numerous bugs, I predict that if Red Hat continues to provide incompatible products and bug-filled releases, fewer and fewer people within the community will run Red Hat. As the faithful abandon Red Hat, so will new users. Unless Red Hat moves toward compatibility with other distributions you will begin to see the erosion this year.
I believe that SuSE will pick up steam in the United States and that Mandrake will begin to fall by the wayside. Caldera will remain a niche vendor but will offer unique technologies like the Volution distributed management product, and the integration of Tarantella into Linux. Such technologies will cement Caldera in the IT market in a position that is similar to Novell's. Caldera will not grow as big as Red Hat, but may become profitable sooner and offer a strong set of services.
Turbolinux and Linuxcare have announced their merger, and as a result you will see a new power behind Linux. Linuxcare is known in the industry as the support company for Linux. Turbolinux, although not yet predominant in the United States, has a stronghold of users within Asia and strong business allies like IBM. The combined company may have the opportunity to oust more solidified vendors within the community. The previously mentioned 800-pound gorilla comes to mind. That is, if Turbolinux and Linuxcare can merge successfully.
Linux is growing, and as many have suggested it is a disruptive technology. You cannot kill Linux because it proliferates, mutates, and adapts like a virus (a good one). It already provides new avenues for companies like PocketLinux, TIVO, and Linux Wizardry to make money in embedded devices and specialty product vending. Non-Linux companies that compete with Linux will have to adapt. Recently, even HP is considering the possibility of having Linux on its Jornada line of Pocket PCs. I would not be surprised to see a Microsoft Office for Linux.
Over the next year, Linux will continue to mature. Some companies will become profitable with it, while others will die. The ugly reality of business, with its profit requirements and "you can't spend forever" attitude, has started to appear. The Linux companies that have practiced reasonable business and worked toward profitability will survive. Some, like mine, will continue to focus on Linux but will start providing our services around other platforms as well. OS X comes to mind. I would suggest that by February 2002 Linux will have captured enough of the server and desktop market that Microsoft's revenue will start to erode, and that Redmond will have to continue refocusing its profit centers to alternative industries such as media and entertainment. In the end, the year 2001 will be the year in which Linux cements its future.