isn't a stock-picking service, but I do try to flag upcoming tech IPOs that look hot. Sometime in the next few weeks, probably during the second week of August,
readers who scramble now to line up some shares with their favorite broker -- and getting these shares is going to take a very good relationship indeed -- are going to have a shot at what looks like one very hot IPO.
The company going public is
, a tiny Durham, N.C., shop that reported revenue last year of only $10.8 million, producing a loss of $91,000. So why all the buzz? How did a tiny company like this -- planning an IPO of just 6 million shares at a certain-to-be-increased offering price of $10 to $12 -- attract
as its lead underwriter, with
Thomas Weisel Partners
in the game, too?
Don't touch that dial: This two-part column will run a little longer than usual, because understanding the Red Hat opportunity means understanding the
phenomenon -- and the body of myth and legend that has grown up around Linux and its prospects in the market. But you just can't evaluate the Red Hat IPO without knowing something about the often-strange world of Linuxland. So buckle up, get out that pad and pencil, and let's dig in.
The buzz arises because Red Hat has hitched its star to the free operating system, Linux, which is in the white-hot zone in technology these days. So it's impossible to look at Red Hat's prospects in any meaningful way without first looking at Linux, and the paradoxes of making money off something that's, well, free.
The Roots of Linux
Linux competes head-on with
Windows NT, both NT's server and desktop/workstation versions. As a Unix-based system, it also competes with the proprietary, for-profit Unixes such as Solaris from
, Irix from
, HP-UX from
, AIX from
, UnixWare from
Santa Cruz Operation
and others, plus other "free" Unix versions from
, the famous
Berkeley Software Distribution for Unix
disks and other sources.
In its 30-year life, Unix has been through an amazing series of misadventures. Hatched in 1969 at
, Unix was designed from the ground up to do things then considered heretical, but now commonplace in business operating systems, such as delivering highly crash-proof performance, multitasking and supporting multiple processors. Unix was given to university computing environments for free by then-owner
, which I am long. It was an immediate hit. AT&T's idea was to produce almost overnight a generation of Unix-trained, Unix-loving computer-science and engineering graduates who would, presumably, pound on the table at their new employers, demanding the companies move to Unix.
The strategy worked. Unix became a cult. Hundreds of thousands of Unix-loving computer-science grads fanned out from the nation's institutions of higher education into the business world and began boosting Unix as both product and ideal. The early inclusion of TCP/IP capability -- the "Internet protocol" upon which we rely today for Web communications -- proved critically important to the ultimate success of Unix.
The actual ownership of Unix itself followed a twisted path. Never quite sure what to make of its baby, AT&T sold it to
in 1993, which in 1995 sold it to Santa Cruz Operation. Today, Unix has wound up almost in the public domain. (In the case of Linux, users can create their own variant, but must in turn contribute their additions, for free, to the world community of Linux users.)
Major computer vendors produced their own proprietary Unix versions, optimized for their products -- typically workstations and servers. Those proprietary versions have now been through many generations and are today finely tuned, highly evolved products. Certainly much of the success of Sun, for example, can be traced to its founders' Unix experience at Berkeley and their early decision to produce only Unix workstations -- period. What became Solaris, Sun's version of Unix, was key to selling standardized, high-performance machines at what were then unbelievably low prices. Now Sun's most important business is servers, but those servers still run optimized versions of the company's own flavor of Unix.
But the problem of so many "flavors" of Unix, as Unixoids like to say, became overwhelming. What was originally supposed to be a universal platform -- a single, standardized product upon which any "Unix application" could be run, no matter who produced the underlying Unix version -- fragmented into a welter of competing, incompatible Unixes. In theory, it was a simple matter to just recompile your application for any given flavor of Unix, using tools provided by the maker of that version. In practice, it has rarely been so simple.
Into this melange of competing and often very expensive products came
, a Finnish college student. Sitting at an IBM PC in his apartment in Helsinki, he created the bare bones of the operating system we know today as Linux. (Every creator and vendor of Unix feels compelled to create a cute variation on the "-unix" name; Linus fused his first name to that stub. The result is pronounced LINN-ux, please, not LYE-nux. FYI, it could have been worse. Originally, Torvalds wanted to call the product Freax, from "free" plus "freaks" plus "Unix." Wiser heads, etc.)
He shared his work with a few friends; they shared it with a few of
friends ... and today, there are at least 10 million people working on Linux-based computers around the world. Along the way, literally thousands of programmers have contributed their own little bits of Linux code, to increase the number of functions the operating system provides, or to make it faster and more stable. And a well-regarded Web-server software package, Apache (we won't go into the origins of that name here) emerged; today, Apache running atop Linux, powers more than half the Web servers in the world. Go back and read that sentence again:
Apache on Linux has a 50%-plus world market share for Web servers
And you were wondering why Linux is considered important?
Although there is no "Linux Inc." corporation today that "owns" Linux, the core operating system is effectively controlled by an international cooperative of long-time Linux programmers, many chosen by Torvalds himself, that regularly releases new "standard" versions incorporating the latest round of tested-and-approved additions to the Linux kernel. (For more information, go to
Linux was neither the first student-created version of Unix nor the first "subversively free" one: Those credits belong to GNU (as in
nix ... sorry), a Unix variant created by
Massachusetts Institute of Technology
, who began creating his wonder 15 years ago, and has gone on to become the leading -- or at least the loudest -- exponent of the notion that
software should be free.
But it was Linux that won the hearts and minds of programmers everywhere, and the millions of hours invested by them -- all for free, with neither expectation nor even hope of compensation for their work -- has produced a remarkable product that today runs, in various fragments and forms, on everything from Palm Pilots to the experimental, massively parallel supercomputers being created at Los Alamos National Laboratories.
Torvalds subsequently emigrated to the U.S. and today works in Silicon Valley at the famously secretive chip-design firm
-- where, he says, a substantial part of his time, by contract, is spent on Linux. But true to company reputation and policy, he won't say more about what he or Transmeta is doing.
Despite his substantial fame -- Torvalds is a godlike figure to many in the computer community -- he remains a remarkably unaffected, direct and modest man, whose simple personal Web page, complete with snapshots of his 2-year-old daughter, Patricia, can be found at
www.cs.helsinki.fi/~torvalds/. He even provides on the page his address and --
-- telephone number. Just offhand, do you think
post their phone numbers somewhere on the Web?
Red Hat: Standards, Support
Today you can get copies of Linux on CD-ROM from a number of sources, for small fees, or you can get it completely free by downloading it on the Web -- if you're willing to sit through a few hundred megabytes worth of download. Almost no one wants to suffer through that, so outfits such as Red Hat that have put current versions of Linux on CD are popular among Linux buffs.
But distributing "free" software on CDs, even when you charge a nominal fee for the package -- the current list price for Red Hat Linux is about $80 -- is a tough way to make money. Moreover, for many Linux users, especially in business settings, the free-distribution model is only half an answer: What about readable, usable documentation? Even more important, what about postsale technical support? Enter the Red Hat business model.
Red Hat makes money four ways:
- It improves, commercially packages and sells Red Hat Linux in frequently upgraded versions (currently version 6.0).
It extends the product in ways commercial users of Linux request. It has just introduced, for example, its Commerce Server edition, offered at twice the base Red Hat Linux price, including transaction subsystems, credit-card-security systems and more.
It develops and sells coherent documentation -- coherent, at least, in terms of the complex and arcane world of Unix.
Most important for investors, Red Hat offers ongoing telephone and email support for Linux users. This "services" opportunity for Red Hat is large and growing daily. To be sure, there are a lot of Unix/Linux hackers out there for whom "support" means jumping on one of the many Web-based discussion boards to ask a question -- and, often, getting an answer back in minutes or hours. But for many more Linux users -- and for those not-yet-Linux users who must be drawn into the Linux web if this is to be an enduring success, not a short-lived phenomenon -- the arcana of
any variation of Unix means they need and want live postsale tech support through the channel with which they're most familiar: the telephone. And they're happy to pay for it.
Red Hat itself is also key to the larger-scale success of Linux by bringing a kind of commercial respectability to the product. Many Linux buffs with roots in the hacker community find it a hoot that anyone would actually
to buy and pay for an operating system, especially from such an evil and hated source as Microsoft. But stories of network managers called on the carpet by corporate executives who have just found to their horror that the company is running its servers on a software product produced by a bunch of crazed hippie hackers around the world, who don't even
for it, are commonplace.
For businesses, a commercial product that comes on a CD in a shrink-wrapped package, with real, printed documentation, from a company that demands to be paid for its work and offers ongoing technical support, is a far more appealing prospect than committing to a freeware product of largely unknown and shifting parentage, without an 800 number to call when things go dark.
Bob Young, recruited to Red Hat as CEO last year, understands that and has been pushing Red Hat in the right direction. With a customer list including names such as
Burlington Coat Factory
New York Life
Internal Revenue Service
and many more heavy hitters, he's off to a roaring start.
Tomorrow: Some of the advantages -- and some of the baggage -- Red Hat brings to market.
Jim Seymour is president of Seymour Group, an information-strategies consulting firm working with corporate clients in the U.S., Europe and Asia, and a longtime columnist for PC Magazine. Under no circumstances does the information in this column represent a recommendation to buy or sell stocks. At time of publication, Seymour was long AT&T, although positions can change at any time. Seymour does not write about companies that are consulting clients of Seymour Group, or have been in recent years. While Seymour cannot provide investment advice or recommendations, he invites your feedback at