Pacific Connection(英語)

Caldera Tries to Light a Fire with SCO Purchase

In the universe of Linux distributors, there's Red Hat, and then there's everyone else. At least that's how it might seem if you're a competitor. Take Caldera Systems, a bona fide Linux company since 1994, headed by co-founder Ransom Love. Caldera's OpenLinux operating system and eDesktop client software get good reviews. The company claims more than a thousand resellers worldwide and it offers the usual array of support and educational programs. And Caldera scored something of a coup by hiring away John Terpstra, former TurboLinux vice president and SAMBA project founder, as its new chief technology officer.

On the other hand, Caldera's revenues are but a small fraction of Red Hat's. Red Hat is widely perceived to have more than 50 percent of the market share for Linux distributions, with Red Hat Linux 7 capturing an estimated 40% of Linux sales in Japan. Red Hat's investors include Intel, Compaq, Novell, and IBM. And Red Hat's products are the most highly visible in server bundling agreements, including IBM, Compaq and Dell. Moreover, while Red Hat's stock has dropped some since the company first went public, Caldera's has plummeted, opening at around 30 last spring and hovering around 2.5 in December.

If you're behind in the race, you've got to think differently. And that's what Caldera did last August when it agreed to acquire the Santa Cruz Operation's Server Software and Professional Services divisions. The deal, which covers virtually all SCO assets except SCO OpenServer, will result in a new company: Caldera International, Inc. (The remainder of SCO will go by the name of its subsidiary, Tarantella, Inc.)

The move is controversial, especially among those who believe freeware is destined to triumph. For the true believers, SCO is a dinosaur, stuck in the old economy, using the old way to develop software, and gravely threatened by Linux. Why, they ask, would a Linux distributor dirty its hands with a company that seemed to be running from Linux, rather than embracing it? "While other software companies had devised Linux strategies months or years ago, SCO sat it out," wrote Lisa DiCarlo in Forbes magazine. "It contributed some intellectual property to Linux developers and sold services for Caldera and TurboLinux, but it never became a distributor. Additionally, the company wasn't exactly a public cheerleader for Linux, advising people to go with Unix if they wanted a stable, scalable server platform. Naturally, this hasn't sat well with the Linux community." So why the acquisition? DiCarlo says that SCO gives Caldera "the technology to build a more scalable, enterprise-class Linux that should help it gain business among big customers. It can also take advantage of SCO's distribution channel, which includes resellers and big computer makers."

Analysts have also been skeptical. "This acquisition marries two financially unstable companies that are urgently seeking a viable combination of new and old technologies in an effort to create a successful new business model," writes George Weiss, an analyst for the research firm Gartner Group. Weiss credits Caldera as the Linux distributor "most willing to raise Linux to the next level: from versions that businesses must customize entirely themselves to ready-to-run but customizable solutions." But he argues that Caldera will have its hands full showing a unified strategy for SCO's OpenServer and Caldera's OpenLinux. And Caldera will need to gives special care to SCO partners who have shied away from open source and still consider Linux to be alien turf.

And then there's the matter of brand awareness. In the shadow of Red Hat, Caldera didn't exactly pick up a household name with SCO. As the acquisition takes hold, it's unclear whether two minor brands will result in a better known identity. "As we've pointed out, both Caldera's Linux and SCO's UnixWare are excellent products in their class-but both are suffering from a popularity deficit," writes columnist Andrew Orlowski. "And that marrying the two cultures, while satisfying Caldera's basic business requirement to produce open source software, is not going to be easy. Especially when the crown jewels Caldera acquired really belong to Compaq and Veritas, among others, and are essentially out of the open source game."

ZDNet columnist John ??who? Taschek?writes that Caldera has been "despised" by Linux true believers, while remaining "one of the few trusted distributors of Linux software for business users. So what does Caldera do? It scores SCO, which is also almost universally despised by Linux enthusiasts. SCO, however, still has distribution power, although lately, of course, the company hasn't distributed a lot of anything." Taschek says that "Caldera clearly believes that to get its technology into play, it has to have a distribution channel and a support staff. The stock/cash agreement is an inexpensive way to accomplish both goals."

Caldera has responded that nurturing channel partners, not open source, is the company's main focus. If purchasing SCO has put Caldera in the enemy's camp---so be it. And like all companies involved with mergers, it has argued that the whole will be greater than sum of its parts. "Caldera, Inc. will incorporate a worldwide network of sales and support offices, a strong commercial UNIX system business, and a rapidly growing open source company," said David McCrabb, president of the SCO Server Software Division, in a written statement. "This combination will be a force to contend with in the worldwide market for Internet solutions on high volume platforms." Caldera says that the resulting combined platform "will scale from the thinnest of clients to the clustering needs of the largest data center" by combining Linux at the low-end and Unix at the high-end." But how these two operating systems will be combined in a single platform remains to be seen.

From SCO and Back

Caldera is an established player in the Linux business. The company's genesis dates back to around 1992, when a group of engineers within Novell Corporation pulled Torvald's code off the Internet, and took it to then president and CEO Ray Noorda. The result was an R&D project code-named Corsair. Noorda and company later acquired the rights to Word Perfect, then a leading word processing program, as well as to DR-DOS, a competitor to MS-DOS. It also bought the rights to the Unix brand-name from AT&T. The aim of all these acquisitions was to compete with Microsoft, and the Gates-Noorda rivalry became something of an industry legend. When the Microsoft machine continued unphased, Noorda stepped down and Caldera later sold the Unix name to SCO. But Noorda didn't give up on Linux. His venture fund company, Canopy Group, provided the backing to turn the Corsair project into a company: Caldera. And now, Caldera has in effect repurchased the Unix brand.

But if you're finding this story difficult to track, you're not alone. Sam Williams of Upside magazine notes that things got even more complicated last year when Canopy, Novell, Sun Microsystems and Citrix became late-round investors in Caldera before it went public. "With so much money changing hands, it's hard to keep track of who owns whom nowadays. Maybe that's why both investors and analysts seemed a little under whelmed when news of the semi-merger finally became official."

Even before the acquisition, SCO had been thinking---and worrying---about Linux. David McCrabb told CNET that SCO was already working on its own Linux strategy-called Linux Kernel Personality, that would enable UnixWare to run Linux programs, thereby tapping into that operating systems' growing application base. He said that while the company also considered developing its own version of the Linux kernel, they realized that effort would be difficult to accomplish.

Caldera and SCO first began talking in the spring of last year [2000]. "We approached Doug Michaels, the president and CEO of SCO, back in April," said Benoy Tamang, Caldera vice president of strategic development, who oversees mergers and acquisitions. "It started with a phone conversation and then a quick follow-up meeting face to face. Thereafter, once there was an understanding of what we could do together as a company, then we had to have confidential meetings off-site, in hotels in Utah, San Francisco and elsewhere to fully flesh out the possibilities and to negotiate a price. Speed was important to both of us, and we signed a deal in August."

One arena that Caldera hopes to win is in wooing SCO integrators. Caldera focuses on its channel partners, including Tech Data, Ingram Micro, and Computer 2000. "Caldera is a great group to work with," said of Ron Herman, CEO of Blue Chip Computer Systems, in a interview with Ben Elgin of Sm@rt Partner. "More than SCO, they know what a channel is worth."

While Elgin says that there's no guarantee that SCO's integrators will hook up with Caldera, the acquisition certainly puts the Linux company up front. And of course, SCO integrators were clearly one of the chief assets Caldera is paying for. One of the biggest challenges Caldera faces is in proving it has the financial clout to drive UnixWare's future. As Elgin notes, "the company is tiny compared with most of its operating-system rivals. Caldera has never been profitable and isn't growing as quickly as Red Hat. During its most recent fiscal financial year, Red Hat's revenue topped $42 million, compared with $3 million for Caldera."

With SCO, Caldera's overall revenues will rise and its marketing position will shift. In the history of technology mergers, all started optimistically, but many fizzled out. Caldera now faces twin challenges: nurturing a user base that has not yet seen fit to migrate to Linux, while the same time, bringing the best of SCO's technology to its Linux distribution. For a company often lost in the shadow of its larger competitor, the move sure beats standing still.

An interview with Benoy Tamang: Caldera VP of Strategic Development

As vice president of strategic development, Benoy Tamang was Caldera's point man in orchestrating the SCO acquisition. Tamang was born in Nepal, but his British military family moved around the world. He came to Caldera two years ago as vice president of marketing, back when the company employed about 40 people---now it's almost 200, and with the SCO acquisition, about 800. Before that, he had short stints with Novell and Pacific Bell. Tamang holds a bachelor's degree in computer information systems from Brigham Young University - Hawaii and a master's of business administration from the Marriott School of Management at Brigham Young University in Utah.

Why are you acquiring SCO?
The SCO acquisition really perpetuates the principles upon which Caldera is based: make a rock solid product, deploy it through a channel, with the right business metrics and internal discipline to deliver. SCO is a channel company, as are we. It has a business product and technology that can augment Linux. UnixWare, with its non-stop clustering technology, is a natural fit. SCO has a very good support infrastructure with the ability serve customers worldwide. Many of our customers said we did a good job but were U.S.-centric. SCO gives us 30 offices, and with 22,000 resellers worldwide. SCO also has a talented professional services organization with a revenue stream of millions of dollars. And it has $100 million-plus in annual revenues, from companies to that are using UNIX and looking forward to Linux.
I'm sure you're aware there's been some adverse reaction. Forbes, for example, called it a temporary life preserver for SCO. Is that a fair assessment?
It's too harsh. SCO had already announced its own Linux product to combat Linux's taking away some of its Open Server marketshare. And their investment in a technology called Tarantella was booming, so much so, that it warranted its own division, along with SCO's Professional Services and Server divisions. In effect, they had already prepared themselves to diversify. We just accelerated what they would have done on their own, but now there is even a better chance for success.
But with two minor brands, what are your chances for synergy? How do you make one plus one add up to three?
Without a doubt, both companies have not marketed themselves well. So yes, we agree that we have not done that job well. The foremost agenda item will be the branding issue. Caldera will be known as the Linux/UNIX company. That is, we will take the very large installed base of UNIX customers, allow them to continue with UNIX, and give them the future of Linux when and if it is appropriate to do so. In marrying Linux and Unix, we will go beyond the Linux companies who just rely on companies---using Linux for web-centric deployment. Caldera will have capability to have backend database application services that UNIX currently provides, as well as the web serving needs that Linux provides.
So this is more than just a transition for UnixWare customers to Linux?
No. We see value in UnixWare continuing for a long time. UnixWare has strengths in areas where Linux is only now starting to catch up. But Linux also has some unique attributes that UnixWare does not. So where they fulfill their respective callings, we want to make sure that they are able to address those issues of compatibility and integration that only we can provide---since we hold the IP to both. So UnixWare customers will have a UnixWare roadmap that will continue for a long time. We will enable those customers, when appropriate, to use Linux at the firewall stage for now, or later on as the kernel develops to more enterprise class services.
Linux enthusiasts believe there is no room for anything that is not open source. Have you diluted your position by not just being a Linux company anymore?
We diluted our position from the day we were created. We've always straddled the fence between closed and open solutions. For Netscape, we ported their client Navigator on Linux. We helped Corel with their Word Perfect port. We helped Novell with their Netware port. In the Linux world, the most important thing is to communicate what you are doing. We've said from day one, with our Novell background, that we're here to make Linux technology into Linux products for business deployment. That in itself was heresy back then, but we have been very consistent. We have dialog with Linus (Torvalds) and he knows who we are, what we do.
No one in this business is free from what we call the Linux zealots. We want to give back to the roots upon which we got our business established. And so we contribute heavily, but at the same time, we have a business to run and we run it the way we need to. We've always realized that open source is a fabulous way to engineer and develop. But the conservative business world needs transition time. We believe in the long run, Open Source will be widely received by even the most conservative business institutions.
Do you plan to make UnixWare open source as well?
We are planning on some elements of it. At this stage, we are just beginning to understand the UnixWare components. Remember, we're not one company yet so there's a bit of professional distance we have to keep. But both companies are mutually agreed that there are some elements that would be good to open source.
What do you say to SCO resellers who are concerned that you are going to just take the best of SCO's technology and then dump the rest.
Don't worry. From a channel focus, the worst thing we could do is dilute the very company that we purchased. It's not just SCO's technology, but the 20,000 plus resellers who are going to be the feet on the street deploying UNIX and in the future, hopefully, Linux. We've seen enough studies to show that these channel partners are looking for new ways to make margins. We want to make sure that we continue to grow their current Unix investment, while, at the same time show them appropriate levels of Linux implementation. With SCO, we took a relatively bold, and to some, a weird, conservative step. Why would a Linux company go buy an old UNIX company and try to make something happen out of it? We did that because we knew that the future of business required a predictable revenue stream, and predictable revenue stream and Internet ecommerce [the main revenue source for Linux] are not synonymous. SCO has thousands of customers. Every McDonald's has SCO UNIX in it. K-Mart is a customer. All these brick and mortar companies are currently paying for value that SCO is providing. But they are also looking at what else is going to be happening, how do they get to the Internet. Linux represents that.
Linux represents a better Internet play than UnixWare?
Correct, without a doubt. It was born and bred in there. Starting out the chute, Linux is much more suited to that than UnixWare, and apart from Solaris, Linux is the main Internet play.
What kinds of businesses will you be addressing?
Two kinds. On one hand, we'll continue selling to dot.coms and ISPs---those companies who are looking for the lowest cost Web implementation. We've already done well here with small to medium businesses because of our channel partners. The newer areas for us are the large Wall Street financial houses, as well as other companies who rely on proprietary technologies. These companies generate more IT sales, but are more conservative, and much slower to adopt. But they, too, are opening up. I've gone to some of the largest SCO accounts, and they've all opened their doors, and expressed interest in our helping them with their Linux strategy.
Besides the SCO acquisition, how else are you competing with Red Hat?
One way is by not going in the same market or the same measuring stick as they have traditionally been leaders in. In principle, we've seen them have a tremendous lead in the retail space with their branding and boxes ever present. We've gone from day one with a channel focus on having our Linux product, OpenLinux, and we've targeted towards channel partners who can provide solutions for businesses. In that space there's a clear benchmark that is different from the retail barometer. We've also differed in the physical creation of our product. Other distributors have created multiple CDs for developers and enthusiasts. But because our target is business and we use channel partners, we provide a single CD---a compiled, well-tested, well- documented product that could be used with business.
What have you eliminated to pare it down?
Who needs 53 different types of clocks? Who needs 400 Tetris-like games? Let's make sure we have a very strong implementation of Samba for file and print service. Let's make sure that we do have one rock-solid clock. Let's make sure that we have an environment where if people want to create their own web-serving engine from Linux, we have SSH [Secure Shell] in there. We also, from the beginning, delayed the release cycles so that we didn't have a new release coming on every two weeks. Now everybody's doing it.
How are you doing in Japan?
Japan has always been strategic for Caldera, receiving our Linux for Business products and quality of service well. Our Japanese OEM customers demand 24 X 7 support in the local language, a high level of technical expertise and 100 percent commitment. The acquisition of SCO will help us deliver with an in-country office and team. We are also experiencing high levels of business in education and training. We have seen this in the US too, where businesses, no longer worried about whether Linux is a viable technology, have sought a Linux business partner who can help them with their internal Linux strategy and technical implementation. With the strong technical competence we have seen in Japan, this naturally leads to a thirst for solid education, products and training services, and Caldera is experiencing solid business in this category.