By Sam Jaffe Imagine that you run a computer company that oversaw the development of seven different operating systems. It's not pretty. Microsoft (MSFT) struggled mightily to get the latest version of its OS, Windows XP, out the door. Meanwhile, Apple had to delay the launch of its OS X several times. What if either Apple or Microsoft had to deal with shepherding seven of these puppies, and do so all at once? Microsoft CEO Bill Gates might tear his hair out. Apple CEO Steve Jobs would be a candidate for an ulcer.
So what about Carly Fiorina, CEO of the to-be-merged Compaq and Hewlett-Packard (HWP)? We'll soon find out. When it comes to operating systems, "they've got Himalaya, VMS, Tru64, HPUX, and others," notes Laurie McCabe, an analyst with Summit Strategies. "It makes my head swim trying to keep track of each one." That could create a huge opportunity for Linux, the so-called open-source OS that's becoming an increasingly prominent player in the field.
If the new HP adopts Linux as its primary platform outside the Windows arena, it would break down the obstacles that have been standing in the way of Linux' growth and further market penetration. As for HP, by winnowing the OS plethora and unifying software-development efforts around two systems, Windows and Linux, the new company could create desperately needed economies of scale -- and perhaps save hundreds of millions of dollars in wasted development costs. More important, it could ease the integration of the two companies by giving them a common goal. Smart investors will watch what Fiorina does very carefully.
One company that could win big if HP were to make such a switch would be Linux vendor Red Hat (RHAT). Although it probably wouldn't win any license revenue from such a move, the adoption of Linux would be a big boon to Red Hat's overall goal of making Linux a top choice for corporate customers.
BIG DOG AND UNDERDOG. Not surprisingly, executives haven't had a chance yet to talk publicly about future software development at the combined company. HP's main Linux evangelist, Bruce Perens, expects see its Linux development efforts continue as before. "In the long run," says Perens, "I expect this merger to be good for Linux."
The dominant OS, of course, is Windows. It runs all the PCs that Compaq and HP produce. Sales of those PCs account for 38% of combined revenue. Fiorina has said she would gladly go with another OS for her desktop line. But for the most part, corporate buyers have little interest in switching, which ensures that Windows remains virtually the only game in town.
The consolidation of software offerings could really help the merged company in the server market. Windows 2000 is making great strides in penetrating the low and middle regions of that business. That's where Linux is also strongest right now.
In higher-end servers, companies offer many flavors of proprietary OS based on Unix. Those include HPUX (Hewlett-Packard's version) and Tru64 (Compaq's entry), Sun's Solaris, and IBM's AIX. Except for Solaris, all are slowly losing market share. No one thinks proprietary Unix systems have much of a future, save the folks at Sun Microsystems (SUNW).
STRATEGY. Compaq, HP, and IBM have made big noises about supporting Linux development. So far, IBM has taken the lead. It says it will spend more than $1 billion developing new Linux products in 2001. But its Linux-related revenues are still paltry, and most of its R&D resources go to developing and supporting AIX.
Most OS experts say Linux could handle anything that any other Unix variant does. But corporate buyers have been hesitant to adopt Linux for frontline operations without some major hand-holding by the recognized leaders in the IT-services industry. And IBM, the corporate leader in Linux cheerleading, hasn't yet advanced the ball to any great degree.
If Fiorina's new company were to adopt Linux wholeheartedly and set a goal of winning that market, HP could ride a rising tide of Linux adoption to sell Linux servers. It would be a win-win situation. HP could plant a flag as the go-to guys for Linux services. That would fulfill Fiorina's long-held goal of turning HP into a service powerhouse.
HP would have plenty of work to do to build Linux beyond entry-level servers. Linux has to be adapted by corporations for mission-critical operations. Right now, no bank will run its wire-transfer network on Linux, and no catalog company will power its call center with Linux. "Business customers aren't ready to entrust many of their databases and applications to Linux," says Jonathan Eunice, principal analyst at Illuminata. "IT managers tend to be a little more risk-averse than 'Sure, if anything breaks...my factories grind to a halt. But what the heck! Let's risk it!'"
SIMPLE SOLUTION. Linux has made progress on the corporate front. According to IDC, it powers more than 10% of network servers. True, Microsoft's Windows NT dominates with its 55% market share, but Linux' share is growing. And no one knows the exact penetration of Linux in the enterprise market, since it's available for free and in so many flavors. But tales of how Linux is popping up all over are now a staple around IT department watercoolers everywhere.
Moving talk of Linux' spread into the boardroom would be a boon to HP. Moreover, it would be competing against Linux startups that lack the global heft and infrastructure to support major corporate-enterprise installations.
This doesn't mean HP needs to abandon its other operating systems. As long as a computer runs Tru64 or HPUX, HP should service it. But HP shouldn't waste development dollars on software that is becoming marginal. "Over the next few months, they will be deciding how to focus their company and how to spend their software-research budget," says George Weiss, an analyst with Gartner Group. "They can either emphasize Linux and force change, or they can cultivate the OS they support and continue to send a muddled message."
UNIFY AND PREVAIL. If HP continues to place equal emphasis on the various operating systems, it will likely survive. But HP could have difficulty growing in the high-end server market and in services. Fiorina's job could become one of managing decline rather than leading growth.
Should HP not step up to the plate, the biggest loser would be Linux. Without HP's staunch sponsorship -- prior to the merger announcement, the company matched IBM's $1 billion Linux-development pledge -- the only corporate backer will be IBM. And if Big Blue's efforts to incorporate Linux into its line of core products are unsuccessful, it could ultimately abandon Linux. That would leave Linux to revert to its original role: a revolutionary solution for serving up Web pages -- but little else.
It could also, potentially, leave HP-Compaq as a great memory in the history of technology and deny it a place in tech's future. So, if you want to know how the merger will turn out, don't focus on hardware announcements from the combined company. Watch to see where the merged entity steers its software strategy. If HP-Compaq tries to unify its platforms, it means that integration of the two companies has a chance and might be worthwhile for investors with a taste for risk. But if it continues with seven primary operating systems -- well, that's not a good sign. Jaffe writes about the markets for BusinessWeek Online