Fresh from Cisco Systems's buyout of their last startup, IP multicast play Precept Software, in 1998 for $84 million, Ms. Estrin and Mr. Carrico have the clout to develop their latest venture in whatever way they choose. That means Packet Design is structured to keep its cofounders free and clear of the marketing, production, and financial hassles that go along with a more conventional startup.

But even though they've landed in the rose bed of startup situations, the couple is faced with an obvious question: why work at all? With enough money in the bank to make Gordon Gekko jealous, why aren't Ms. Estrin and Mr. Carrico walking hand-in-hand on some sun-drenched beach, on a remote island purchased with their cashed-in stock options? Why, instead, are they camped out over a single laptop in a freshly painted office, waiting for the cubicles to arrive? And why are they happy about it?

For Ms. Estrin and Mr. Carrico, making millions of dollars was never the goal. "I like starting up businesses," says Ms. Estrin. "I gravitate toward it." And neither she nor her husband has any interest in early retirement. They say that friends of theirs who have tried it end up feeling bored or out of touch.

With a life as venture capitalists out of the question ("You just don't have enough input from a board member perspective," says Ms. Estrin), the only thing left for the technically oriented pair was to design a startup that would remain forever in the problem-solving mode, a geek's self-employment dream (see "Packet Watch").

But Ms. Estrin, most recently the chief technology officer at Cisco, says that current market realities don't allow technology startups the luxury of indulging in long-term contemplation. Instead, companies are forced to develop products that either support a fast IPO or can be sold to the highest bidder.

To ensure that Packet Design doesn't miss out on any brilliant ideas, the founders have signed up an impressive list of tech-savvy individual investors -- including Jim Barksdale and Bill Joy -- not to mention a star-studded advisory board that includes prominent Net-heads like Vint Cerf, a senior vice president at MCI WorldCom and one of the Internet's founding fathers.

The company's list of friends and investors is a testament to its cofounders' industry connections, which go back to the earliest days of the Internet. Ms. Estrin, perhaps the better known of the duo, can trace her networking history back to the '70s, when she was a master's degree candidate at Stanford University, helping Mr. Cerf perfect TCP/IP, the basic language of the Internet.

While working at chip maker ZiLog a few years later, Ms. Estrin found a soul mate in Mr. Carrico, who shares her passion for technology and business. Since then, the two have combined their interests and talents at several other career stops. The enterprises have varied, but the role of each partner has remained the same. Mr. Carrico prefers to stay in the background ("Bill's the antisocial one," his wife jokes), while Ms. Estrin plays the public spokesperson.

Cisco's purchase of Precept Software had the side effect of making Ms. Estrin a bit of a media star, since few women hold chief technology roles in companies of Cisco's size. But she never seemed to warm to the big-company spokesperson role; she seems much happier in a nearly empty office, talking about working on the cutting edge of technology with a new plan that lets her enjoy all the fun of a startup without ever having to progress to the messy parts -- marketing and selling products.

For a serial entrepreneur, that sounds like quite the adrenaline rush.

PACKET DESIGN AT A GLANCECEO: Judy EstrinChairman: Bill CarricoLocation: Menlo Park, CAPhone: 650/617-6102URL: www.packetdesign.comOwnership: PrivateFounded: 2000Employees: 8Product: IP infrastructure technologiesCompetitors: Bell Labs, IBM Labs, Microsoft LabsProfitable?: NoFinancing: $24MInvestors: Foundation Capital, Jim Barksdale, Bill Joy, Judy Estrin, Bill Carrico

According to its founders, husband-and-wife entrepreneurial team Judy Estrin and Bill Carrico, Packet Design is "chartered with developing IP technologies," but will not sell branded products. Instead, it will either license its technology to existing companies or spin off separate companies to market the actual wares.

Packet Design's initial areas of research include Internet scalability, wireless infrastructure, and network security. While the company's founders haven't announced specific plans, they are confident they'll find plenty of problems that require marketable solutions.

Foundation Capital was the lead investor in Packet Design's $24 million seed round, with the balance coming from an impressive mix of individual investors, including ex-Netscape boss Jim Barksdale and Sun Microsystems chief scientist Bill Joy.

The cash will be used to hire a staff, which may eventually grow to a total of 100 people, working on as many as ten projects at a time, according to the founders. Mr. Carrico says that Packet Design already has some ideas in development and could possibly produce licensable technology within six months. Officially, the company is giving itself two years to form its first product spin-off.

Mr. Hawkins gave several newsworthy answers to the audience, in fact. But the most interesting was in reply to a question about telephone connectivity for the Visor devices, which currently all support plug-in modules based on the company's so-called "Springboard" slot design. While several modem modules are already available, no company has yet formally announced a cell-phone module. But Mr. Hawkins says that situation will change soon.

"The Springboard slot was really designed for wireless communications," he said. "And you will see a cell phone module available before the end of the year."

LOOKING AROUND

Mr. Hawkins also said that the company's operating-system licensing agreement with Palm currently stretches for a five-year period, and doesn't see another OS for Visors anytime soon -- though he's not ruling it out, either.

"Our license [with Palm] extends for another three years with a two-year overhang," Mr. Hawkins said. "We have no plans to do another operating system today, but we might if it makes sense."

While keeping most future product specifics under wraps, Mr. Hawkins did offer some visions about Handspring's future to the chat audience, including a surprisingly open attitude toward the possibility of licensing operating-system software from handheld competitor Microsoft.

"They [Microsoft] are getting better, but we still don't see them being successful with their current products, but we continue to worry about them," Mr. Hawkins said. "In the future, if they had a really great offering, Handspring could build a product using Windows CE. But for today, we have no plans for that."

PROMISES HE CAN KEEP

While interested audience members tried to pin Mr. Hawkins down on delivery dates for thinner Visors, color-screen Visors, or any future Visors, Mr. Hawkins evaded all the queries by claiming that Handspring would deliver future Visors in "all shapes and sizes." And while saying that not all of the future products might contain a Springboard slot, Mr. Hawkins did pledge that any future versions of the Springboard technology would be backward-compatible with current designs.

Like Palm executives, Mr. Hawkins thinks that handhelds are heading toward more connectivity, and more attractive pricing. When asked to describe a far-reaching vision of handhelds, Mr. Hawkins opined, "In 20 years I would say that almost everyone will own a handheld device with high-speed wireless access for both voice and data communications. And the most surprising thing I would predict is that the wireless service will be essentially free."

For the full chat transcript, please see: Jeff Hawkins Q & A.

Discuss the handheld technology space in the Handheld Technology discussion forum, or check out forums, video, and events at the Discussions home page.

Whereas most of that cash is generated in the Internet and network server space -- where Linux has already claimed considerable market share -- Tuesday also saw the announcement of a concerted, multi-vendor attempt to address one of Linux's major shortcomings: its lack of a comprehensive desktop and application-development environment.

A BUNCH OF GNOMES

Backed by a group called the Gnome Foundation, the open-source application project (which counts Sun Microsystems, Compaq Computer, and IBM among its members) faces an incredibly tough opponent in Microsoft's Office software, which dominates the computing landscape today. But with the anything-can-happen nature of the Linux community (along with continued corporate support and venture capital funding), apparently no dream is too large.

Given that Mr. Dell has created his empire mostly by selling Intel-based PCs running Microsoft operating systems, there was some delicious irony (and plenty of conspiracy-theory subject matter) surrounding his address to the T-shirted, colored-haired Linux enthusiasts in the audience. Although the techies who nurtured and promoted the free operating system through its infancy are still a big part of the Linux market, they are increasingly joined by mainstream computer vendors like Dell Computer, IBM, Hewlett-Packard, Compaq, and even Sun Microsystems, who can all smell profits riding atop the open-source platform.

"We see Linux as a significant growth opportunity for Dell," Mr. Dell said at the beginning of his speech Tuesday. In his half-hour address, Mr. Dell cited numerous meaningless Linux-on-Dell sales figures, the sum of which could be simply stated as: Linux-based servers are selling like hotcakes, and we'll deliver as many as we can. During a question-and-answer session, Mr. Dell also deflected some attempts to criticize his company as being one that favors Microsoft products over Linux, a claim he repeatedly said was not true.

"Those accusations were mostly based on old information," says Mr. Dell. "People think you can't buy a laptop with Linux, that you have to pay extra to get Linux on our machines -- it's just not true."

LET'S NOT CONTEMPLATE OUR NAVELS

Like the early days of the Internet, the formative years of the Linux market were marked more by a spirit of creation and collaboration, rather than dirty old profit seeking. But even Linus Torvalds, the original creator of Linux, says the addition of commercial concerns is one the Linux community should embrace.

Mr. Torvalds took the stage after Mr. Dell. Resplendent in his vacation-grade shorts-and-sandals ensemble, he told the crowd, "I used to worry about Linux becoming too commercial, but I don't anymore. It turns out that it's been very important to see the old-fashioned Linux community become augmented by the new people. I'd really hate to see Linux turn into some self-centered, navel-contemplating kind of thing."

At the day's top news event, the contemplation surrounded the idea of an open-source, freely available desktop and application framework environment, a project that could make Linux even more appealing to corporate developers and users. By combining one of the top Linux graphical interfaces with products like Sun's StarOffice (a suite of freely available desktop productivity applications), the Gnome project aims to build not just a nice interface for Linux, but to provide organizational, financial, and legal support for the creation of an entire desktop environment, one that could include a wide range of applications.

In addition to Sun, Compaq, and IBM, the Gnome Foundation includes pure-play Linux concerns like Red Hat, VA Linux Systems, and Eazel, and open-source development entities like Collabnet. Whether or not the Gnome Foundation will be able to deliver an environment that could dislodge Microsoft's Office is a time-will-tell story.

But it could be a long time. Corel, which refocused itself as a Linux applications vendor last year (although it isn't part of the Gnome Foundation), has seen its stock performance suffer, and Tuesday it announced the departure of its flamboyant founder and CEO, Michael Cowpland.

SLOW CLOCK

Despite trying times for Linux stocks generally, the Linux players believe that time is on their side.

"Given the choice between a free product that does all the same things as a proprietary product, I'm confident that people will choose the free, open one," says Bud Tribble, vice president of engineering for Eazel, a startup developing a Linux interface and related services. "But that's dependent on the free product having all the same functionality as the proprietary one."

Having been part of the original Macintosh development team, as well as having worked at Next Computer and Sun Microsystems, Mr. Tribble says the power of open-source development -- which theoretically can draw upon an infinite number of developers to improve the source code, rapidly -- is the main reason why Linux has succeeded where other operating systems have failed.

"The huge difference between Linux and, say, Macintosh or Next, is the ever-expanding base of developers," Dr. Tribble says. "Clearly, it's already made a difference in the server market, and there are signs now that it may do the same to the desktop market."

Mr. Dell, for one, is hesitant to endorse Linux's immediate desktop market chances, especially since the StarOffice package belongs to one of his biggest competitors in the server market.

"I'm a little bit skeptical because of where it [StarOffice] comes from," Mr. Dell says. "I'm always thinking it may just be a diabolical plot to sell more Sun servers."

But, like Sun, which has pledged to lend its support to the Gnome Foundation, Mr. Dell will take profits over political positions any day of the week -- no matter what operating system drives the sale. In fact, Mr. Dell says that his company has already had discussions with Eazel and other Linux desktop players, to explore the possibilities.

"If customers start asking for it [StarOffice], we'll sell it," Mr. Dell says.

Discuss Linux and Open-Source trends in the ongoing Linux Explosion discussion forum, or check out forums, video, and events at the Discussions home page.

The bigger challenge for the combined firm, as yet unnamed, will be to hold and keep the minds and pocketbooks of its customers: telephone companies. Phone.com sells software and services for wireless telephone Internet access, and Software.com sells carrier-side email and messaging infrastructure. Both target telephone companies.

The deal happened because courting carrier behemoths is perhaps better achieved by a larger, luminary-led entity than by two smaller startups, particularly when well-funded players like Microsoft and Nokia have signaled their intentions to compete in the infrastructure-services arena.

TWO THUMBS UP

Investors certainly gave a thumbs-up to the planned $6.4 billion merger, as stock prices for both Phone.com and Software.com rose sharply during trading Wednesday. Software.com, which began the day at $107.75 per share, added almost $35 per share, climbing to a close of $142.44. Phone.com, which acts as the acquiring firm in the deal, saw its shares rise almost 17 percent, from a Tuesday close of $78.06 to Wednesday's final mark of $91.13.

Cisco, whose former executive vice president Don Listwin will be CEO and president of the new firm, also improved Wednesday, from $65.50 to $67.81. The merging companies will each own about half of the combined entity, with each Software.com share being exchanged for 1.6015 Phone.com shares, accounted as a pooling-of-interests deal. The deal, approved by both companies' boards, is expected to close by the end of the current calendar year.

What Phone.com and Software.com bring to their marriage is a combination of wireless Internet smarts, centered around Phone.com's wireless application protocol (WAP) browser technology and Software.com's back-end software for combining email, messaging, and voice services. While both Phone.com and Software.com have won customers on the strength of their respective product offerings, one analyst says the merger has more to do with meeting customer needs than technology.

IN YOUR FACE

"Large carrier customers require plenty of face time," says Ted Jackson, senior analyst with financial services firm U.S. Bancorp Piper Jaffray. "They require lots of hand-holding, and want to deal with a local sales staff, not someone in Silicon Valley."

The merger, Mr. Jackson says, will create a company with 1,400 people, in 25 offices that span the globe. With a combined sales force of almost 500 people, the merger will "clearly accelerate" the companies' ability to hit the carrier market in multiple locales, including Europe, Asia, and other foreign markets.

"It's not about the technology, as much as it is about who's got the customers," Mr. Jackson notes.

Bringing Mr. Listwin on board as CEO and president is just "icing on the cake," according to Mr. Jackson. But it's a tasty frosting, given Mr. Listwin's history of helping build Cisco's presence in the carrier market. It helps that the 41-year-old Mr. Listwin, widely viewed as a potential heir apparent to Cisco CEO John Chambers, is leaving Cisco on amicable terms. That lets potential customers know that Cisco -- which held 7 percent of Software.com -- has "given its blessing" to the deal, Mr. Jackson says.

CALLING FOR GROWTH

The overall field for back-end carrier infrastructure services is widely seen as one about to explode, especially as declining service-access prices force carriers to seek new streams of revenue. With wireless phone use expected to grow exponentially worldwide as Internet access becomes more available, technologies like Phone.com's WAP-based browser products will likely grow in use. It's a phenomenon not lost on players like Nokia and Microsoft, with the latter unveiling new smart-phone and mobile-browsing technologies on Tuesday. Personal digital assistant king Palm will also vie for market share with its Palm OS and related services.

Still, some observers argue that Phone.com's WAP offering is already dated, and may not be as compelling a technology going forward.

"WAP is definitely not the right answer for the enticing consumer experience," says Roel Pieper, a longtime industry veteran who is now a leading European venture capitalist well-versed on the wireless market. What will truly drive carrier profits, Mr. Pieper says, are more innovative offerings that tie consumers directly to IP data, which may be presented graphically or via voice.

On the back-end messaging services side, Software.com has been competing with players like Critical Path, Mail.com, and USA.net, who are all vying for a share of the messaging outsourcing pie. What may help the merged companies -- at least initially -- is the fact that their customer bases have little crossover.

During a Wednesday conference call, Phone.com founder and former CEO Alain Rossman said the companies' customer lists overlap by only 7 percent, presenting "a phenomenal cross-selling opportunity." Piper Jaffray's Mr. Jackson, who covered both companies pre-merger, says the combined entity's revenues could top $500 million in fiscal 2001, a projection that does not include possible revenues from some recently announced products, like Software.com's lightweight directory access protocol (LDAP) directory server.

Mr. Listwin's presence, meanwhile, gives the combined entity -- which may be called Phone.com, Software.com, or some other name -- a bit of star power, along with an energetic boss who knows the market well. Mr. Listwin also knows Software.com well, having served on its board.

"I'm really excited to lead this combination," said Mr. Listwin in the conference call. Though he expressed regret for having to leave Cisco, Mr. Listwin said he was looking forward to joining forces with Phone.com's Mr. Rossman and John McFarlane, the CEO of Software.com, a duo he hailed as "great entrepreneurs and visionaries."

Mr. Listwin, Mr. Rossman, and the other heads will need every ounce of their vision and ability to satisfy Wall Street, which clearly expects the merger to yield Web gold. In the end, that may be their biggest challenge.

Discuss wireless technologies and trends in the ongoing Wireless discussion forum, or check out forums, video, and events at the Discussions home page.

The most recognizable entry on Mike McCue's resume is the three-year stint he served at Netscape. That doesn't mean it's also his favorite.

"Netscape was like a vacation," says the early-thirty-something Mr. McCue, who reminds us he was a CEO before Netscape, just like he's a CEO now, after Netscape. But still, while he didn't have the responsibilities and headaches he's got now, as leader of the voice-powered info-portal Tellme Networks, could his travails as vice president of technology at Netscape have really been relaxing enough to be considered a vacation?

"OK, it was a vacation -- a vacation in Bosnia," he says.

BACKSTAGE PASS

The laugh and the smile accompanying the joke is the first sign of real-person emotion from Mr. McCue, who literally has just been on stage for the better part of two hours. As one of the featured speakers on a panel at yet another Churchill Club dinner, Mr. McCue plays the startup CEO role perfectly, starting with his de rigeur Silicon Valley "important meeting" outfit of black slacks and very blue shirt (no tie or jacket required).

And Mr. McCue had a message, yes he did, and if you were at the dinner his message was as hard to miss as traffic on Highway 101. Tellme is the power of the Internet, available over the phone. Right now that's not a statement transferable to profits, but it could be the next big thing. Did Mr. McCue remind you that you could pick up the phone, call Tellme's toll-free number, and order a cab -- which would be dispatched via the Internet? Yes he did, more than once. More than twice.

Meanwhile, the audience was sitting there thinking: There are cabs in Silicon Valley?

Voice, in case you haven't heard, is the next step in the evolution of Internet access. In some ways, it makes sense -- if cell phones and other handheld devices are soon going to surpass PCs as our main interface to the Internet (as all the fashionable analysts are telling us now), it seems reasonable that picking up your phone and saying "I'd like a taxi" would be preferable to hitting the 8 key, then the 2 key, then the 9 key (twice) then the 4 key (three times), which is what you would have to do to spell "taxi" on a Web-enabled phone.

OK, so sometimes voice is better. Got it.

That simple notion is the main part of the business thinking behind operations like Tellme, which use voice-recognition software to help you find Internet-based information with simple voice commands. Tellme officially launched this week, after a brief "trial" period. And in the official press release -- a story about taxis!

Yes, it's a long way from being profitable or seamless, but at the recent dinner Mr. McCue was not going to be trapped into technical minutiae, though he certainly could explain that part of it, too. In front of the Churchill Club crowd -- which was heavy on the sales and marketing, hold the geeks -- Mr. McCue was all about staying on message, repeating the taxi thing ad nauseum.

TRACTION FRICTION

At one point, Mr. McCue seemed to be battling Keen.com CEO Karl Jacob to see who could say the word "traction" most in a sentence. Never mind that such obvious marketspeak was making even that audience wince, visibly. On that night, staying on message, repeating a theme until it's all you can remember, was Mr. McCue's job as CEO. AT & T does not give you $60 million (as they did in a recent investment in Tellme) to ramble on about how many servers it takes to run voice-recognition software.

Instead, Mr. McCue talked about how Tellme is like Moviefone ("we consider them proof of concept"), how it's great for grandma to find sports scores (he lost us on that one), how it's "the ultimate delivery vehicle for the power of the Internet."

"It's about hearing a song on the radio, and being able to call Tellme and buy it," Mr. McCue says. Simple.

But for two hours it was just a little bit too much message to take, even for a panel discussion that was supposed to be fact-free, since it's so future-looking. But afterward, when he let down his close-cropped (and graying) hair a bit, Mr. McCue showed that at heart he's a real person, one who says he's having a blast.

"I've always known I wanted to be a CEO, wanted to lead a company," he says. "It's what I enjoy doing."

HUMMING THE CEO TUNE

In fact, Mr. McCue was a CEO before -- of a company called Paper Software, which was just successful enough to be acquired by Netscape in 1996, mainly for its brainpower. And when Netscape cofounder Marc Andreessen started focusing more on the business end of things, the title of cool geek at Netscape briefly fell into Mr. McCue's lap, or at least onto the passenger seat of his red Humvee, an accoutrement Netscape's PR force played up to the hilt.

So maybe if it was a vacation in Bosnia, at least Mr. McCue had some wheels that could take a beating.

But by January of 1999, Mr. McCue had had enough of Netscape, and left to found Tellme, where he could focus on the future again. At Tellme, "he is responsible for the overall leadership and direction of the company" (as directly quoted from the company's Web page). There's no doubt: Mr. McCue is in charge of his new firm.

That's why the smiles, the handshakes, the earnest exchange of business cards come so easily after his talk -- Mr. McCue has no agenda now, he's just Mike, telling other people about his cool new company. The one that's not part of AOL, the one that didn't go down in flames at the hands of Microsoft.

Sure it's a startup, and Tellme has a tough road ahead of it, trying to live up to the hype that's partially created by appearances like these. But Mr. McCue seems ready, eager for the challenge. We probably don't have to tell him about the hard work ahead. It's not a vacation. But it's not Bosnia either.

Discuss today's column in the Big Fish column discussion. Or check out forums, video, and events at the Discussions home page.

To truly help the Internet economy, Mr. Schmidt needs to leave Novell behind and find a new cause to champion, a new technology to help bring to reality. We don't have the answer as to what that new technology is, but we're not worried. Mr. Schmidt is very smart, and he'll no doubt find something that most of us haven't thought of yet. But first, he needs to leave Novell, which is at present a millstone around his neck.

Mr. Schmidt, of course, will certainly beg to differ, and perhaps he'll even take offense at the notion that things aren't working out at Novell. When asked about the recent nose-dive of Novell's stock, he responded with a patient shrug of his shoulders and optimistic thoughts about winning in the long run.

IT GOES UP, IT GOES DOWN

"I learned from Scott [McNealy, his old boss at Sun Microsystems] that there are good times and bad times for stock prices," Mr. Schmidt said, in a private conversation after speaking at a recent Churchill Club dinner. In the current state of the market, Mr. Schmidt notes, failing to make your numbers during a financial quarter is asking for a whipping.

"If your numbers go up, the stock price keeps going up," he noted. "But if your numbers go down, your stock goes way, way down."

Mr. Schmidt's personal stock is a perennial blue chip. His resume is impressive, including stints at Xerox PARC and Bell Labs, which preceded his well-received 14-year run at Sun, where his most-recognized accomplishment was pushing Java to a winning position. Now, however, he's like the NBA star who finds himself playing for the Los Angeles Clippers: His individual skills are respected, but everyone in the league knows he's not going to win a championship there anytime soon.

At the recent Churchill Club dinner, Mr. Schmidt was paired with former FCC Chairman Reed Hundt in a wide-ranging panel discussion about the future of the networked world. A perfect foil for the boisterous Mr. Hundt, Mr. Schmidt turned on his trademark dry wit while showing that his thoughts and interests aren't entirely monopolized by Novell's market struggles.

INTERNET OPTIMIST

Mostly, Mr. Schmidt is an Internet optimist, one who is clearly excited by the promise of an always-on, readily available broadband network. "There will be all sorts of new client devices, ones we haven't even thought of yet," Mr. Schmidt predicted. "When the network is not only a utility, but always on and reliable... it will change the way we think."

Since taking over as Novell's CEO in April 1997, Mr. Schmidt has been trying to change the way Novell thinks, making it more Internet-centric and less a Microsoft competitor. But even Mr. Schmidt lapsed into the familiar competitive pigeonhole Novell has been in since the early '90s, when he first blamed (and later retracted) part of Novell's recent fiscal woes on competition from Windows 2000.

By championing technologies like Digitalme, a network-based identity service, Mr. Schmidt has tried to move Novell into the Internet age, but it's hard to break from the NetWare versus Windows mold. Although Novell isn't going away anytime soon -- even with its problems, the company will still sell more than a billion dollars' worth of software this year -- it's suffering from the same problems that also plague Lotus, Computer Associates, and other past software leaders: It's hard to move a big, installed base forward at Internet speed.

SEEING THINGS IN 3D

Mr. Schmidt needs to be freed from Novell's limitations, so he can pursue radical ideas like 3D hologram presentations. When asked how long it might take the industry to support ubiquitous hologram delivery, Mr. Schmidt didn't just guess, he cited some microprocessor and networking statistics and predicted full hologram support in 10 to 15 years.

This is where we need Mr. Schmidt -- pushing for holograms, instead of trying to revive Novell's reseller base. During his appearance at the Churchill Club dinner, Mr. Schmidt mentioned that he was reading a lot about railroads lately and was struck by several comparisons between the advent of railroads and the current state of the Internet industry. Some towns that were bypassed by the railroads, Mr. Schmidt noted, are still bankrupt areas now, some 150 years later. The Internet, he believes, might have the same kind of make-or-break power going forward.

Mr. Schmidt may not like the comparison, but we'll take the poetic leap and compare Novell to a dusty town that once was booming but has now been bypassed by the railroads. And we'd hate to see Mr. Schmidt left behind as the Internet train goes running down the track.

Discuss today's column in the Big Fish column discussion. Or check out forums, video, and events at the Discussions home page.

On display at PC Expo in New York this week was the newest version of Cisco's IP Phone, the latest instrument in the company's war to eliminate the dominant office phone, the private branch exchange (PBX). While Cisco doesn't actually plan to make a killing selling phones, it does stand to benefit from any ideas that can help eliminate the need for separate data and voice networks in corporate installations.

As a proof of concept device, the Cisco IP 7960 phone is an impressive piece of hardware. Crafted out of dark and light shades of gray plastic, the device also sports a small LCD screen that might be used for simple email, text voice mail, or other telephone-related applications. The phone is directly attached to an Ethernet network and does not need to be linked to a separate PC to work.

ROOTING FOR ROUTERS

On the back end, the phone has a built-in four-port switch that allows it to attach to other devices that might need phone access, like a fax machine or another phone. The server software that makes the phones work on an IP network is contained in a sealed-box, Microsoft Windows NT-based PC.

According to company representatives in Cisco's booth, the phones themselves carry a list price of $500, "but you can probably get a discount from any of our reseller partners," the booth rep said. The software is priced at $125 per user.

While IP phones probably won't make a huge contribution to Cisco's bottom line, by making it real the company hopes to encourage third-party development in the voice-over-IP arena. The bigger opportunity for Cisco, of course, is the potential of converting installed PBX systems into pure IP networks that will require more Cisco routers and switches.

FORGET THE WIRES

Cisco was also showing off its newly acquired Aironet wireless local area network (LAN) technology at the PC Expo booth, a product Cisco landed when it purchased Aironet back in March for $799 million. According to booth representatives, the technology supports Ethernet connection speeds of up to 11 Mbps for distances up to 100 feet between the base station and the user's PC.

The base station, a device about the size of a small cigar box, is priced at $1,200 per unit; wireless antenna PC cards are $245 each. Each base station can support between 20 and 30 users, the representatives said.

Wireless LAN technology has long been a holy grail of sorts for IT managers, who would love nothing better than to give up the headaches of stringing, configuring, and maintaining office wiring plans. But the field overall has been stymied by performance issues and a lack of standards that could ensure interoperability between, say, different vendors' antenna cards and disparate base-station devices.

One group trying to eliminate some of that confusion is the Wireless Ethernet Compatibility Alliance, which is supporting a third-party testing site. There, devices can be checked to see if they comply with the Institute of Electrical and Electronics Engineers's 802.11B standard for high-band, wireless Ethernet connections.

Discuss networking, communications, and optical technologies and trends in the Network Talk discussion forum. Or check out forums, video, and events at the Discussions home page.

Maybe those of us in the media are just moving at a speed slower than Internet time, a feeling that always hits when you're trying to take notes at the same pace as Mr. Andreessen's rapid-fire speech pattern. But last time we checked, Mr. Andreessen was a featured guest of the plaintiffs' antitrust case against Microsoft -- a lawsuit that can be almost entirely boiled down to an incrimination of Microsoft's market destruction of Netscape, Mr. Andreessen's former corporate home.

So how could Mr. Andreessen suddenly partner up with the company he once hated with a very public passion? Was there a large cash payout? An invite to Bill Gates' royal palace? Two large cash payouts?

SILVER PIECES

"We got an incredibly good deal," Mr. Andreessen said in response to the sellout question, laughing. The truth is, he said, that Microsoft's announced strategy "dovetails nicely" with Loudcloud's vision of providing software applications as a service. And thankfully, Mr. Andreessen hasn't totally sold his objectivity completely down the river, as his further comments proved.

For example, his take on the XML-based details of Microsoft's new development tools strategy is that the plan is a replacement for the company's Distributed Component Object Model (DCOM) application-development project of a couple years ago.

"They [Microsoft] probably wouldn't agree with me, but their new plan basically replaces DCOM," Mr. Andreessen said. "It also replaces CORBA [the open-systems Common Object Request Broker Architecture effort]," Mr. Andreessen said, claiming that "none of that stuff ever worked."

But, if history is our guide, won't Microsoft simply try to steer the XML standards development in a direction that locks in Microsoft proprietary code? Remember, at the company's famed "Internet Day" in 1995, Microsoft said it would license Java -- and look what happened with that effort. Wasn't Mr. Andreessen aware that Microsoft doesn't have a good track record with standards?

"Truly, I'm shocked to hear this," he says, laughing again.

OUR CLIENTS MADE US DO IT

Mr. Andreessen says Loudcloud's customer base has been asking for Microsoft-based services from Loudcloud. "Our customers have a lot of Microsoft expertise and software knowledge in-house," he said. "We [at Loudcloud] are very excited to have Microsoft people instruct us on their products."

Sure, but how was the relationship cemented? After all, wasn't there once a framed photo of Mr. Gates on a Netscape office wall to remind its programmers that he was the mortal enemy? Did Mr. Andreessen and Mr. Gates bury the hatchet? Was there a face-to-face meeting?

"We never did put that meeting together," says Mr. Andreessen, giggling a bit. Will such a meeting take place anytime soon? "We don't have such a meeting set up," Mr. Andreessen admits, snickering.

Seriously, Mr. Andreessen says Loudcloud has a good relationship with Microsoft executives Rick Belluzzo, group vice president for the company's consumer group (and formerly CEO of SGI), and Jeff Raikes, group vice president for product support. "We got to know them when we were setting up the business," Mr. Andreessen says.

THE WAY IT SHOULD BE

And the business of Loudcloud -- delivering software as a service -- is more important to Mr. Andreessen now than any lingering emotions he might have surrounding Internet browsers.

"In retrospect, the software industry itself is a big fluke," he says. "If there had been a high-speed network all along, companies never would have sold software as a product, and they never would have expected people to install it and maintain it themselves. It's a pretty big change to go to services, but it's one we believe in very deeply."

Now that sounds like some deep thinking -- and not a Microsoft commercial. Although his enemy is now his partner, Mr. Andreessen hasn't lost the sense of humor that's made him an engaging personality throughout his short but storied career.

"We are very definitely having fun," he says of Loudcloud, which now has 270 employees and just closed a $120 million round of financing. "We're moving at hyperspeed."

One final question: Why has Mr. Andreessen opted for the title of "working chairman" instead of CEO? Is there some underlying reason for the semantics?

"It's the cool new job title," he says, claiming the last laugh. "All the kids have it."

Discuss today's column in the Big Fish column discussion. Or check out forums, video, and events at the Discussions home page.

Of course, we have no firm statistics to rest this opinion on. It's a gut feeling, based on one short meeting with the man. That type of deduction, no doubt, wouldn't sit well with Mr. Zollar, who seems to weigh all positions carefully before taking an educated stand. It's just this type of steady hand that's needed now at Lotus, as the company seeks to right its ship after weathering a perfect storm in its wheelhouse late last year.

The reason for Mr. Zollar's ascension was the necessary departure of former CEO Jeff Papows, whose penchant for personal motivational tales of a somewhat fictitious nature earned him some unwanted attention from the Wall Street Journal last year. While Mr. Papows hung on at Lotus's helm until early this year, the clock timing the end of his stay atop IBM's most-visible software division truly started ticking the moment the embarrassing story appeared. Add to that sexual harassment complaints and additional allegations of made-up rйsumй details, and Mr. Papows's fate was sealed.

After Mr. Papows tendered his resignation in January, Mr. Zollar, a 23-year IBM veteran, was tapped to take over Lotus's helm on February 1. While he's no stranger to public speaking or large crowds, Mr. Zollar admits that his first official moment as Lotus's leader was something new.

OVER THE RAINBOW

"We were at Lotusphere 2000 [Lotus's user conference], and I was sitting in the front row when they introduced me as CEO," Mr. Zollar recalls. "A whole row of photographers moved over and almost blinded me with the flashbulbs. All I could think was, 'Toto, we're not in Kansas anymore.'"

Where Mr. Zollar landed was at the top of a somewhat fantasy company, a wholly owned division of IBM that retains the pomp and circumstance of the independent firm it once was. While a stand-alone Lotus would still be one of the top software companies in the world, its IBM ties and its PC local area network proprietary-technology heritage hurts it when compared with the buzz factor allotted to, say, pure Internet or open-source software concerns. Though Lotus has focused its strategies on the Internet, it's not perceived as being on the cutting edge of the technology knife.

Mr. Zollar, of course, will hear nothing of talk that paints Lotus as dead in the water. In San Francisco last week to address a Lotus developers' conference on business-to-business Internet matters, Mr. Zollar in a private interview showed himself to be well-versed in important buzzword-bingo topics like knowledge management, distance learning, and something about collaborative economics that we didn't quite catch.

"We're going to stay focused on growth," Mr. Zollar says of Lotus. "New markets are forming, and there's tremendous opportunity for us to add value."

BIG. BLUE. DIFFERENT.

Boring and safe statements, perhaps, but kind of what you want to hear from someone who has to compete with Microsoft in the messaging and collaboration arena. When it comes to matters software, Mr. Zollar has some serious credentials, all earned within the corporate confines of IBM. But according to Mr. Zollar, those walls are incredibly expandable.

"There's a mythology around a single culture existing in IBM," says Mr. Zollar, when asked if he's too true-blue to lead Lotus. "I guarantee you, the IBM research labs in Almaden [California] are different than a sales office in Paris."

With time spent in many different IBM technology disciplines -- including the DB2 database division, the software development tools division, and network management software -- Mr. Zollar's experience can't be painted with a single brush stroke.

"I like to think of myself as a veteran of the software industry, and not just of IBM," he says.

HONEST TO A FAULT

All too soon our time with Mr. Zollar is over, and he has to catch a car to the airport, back to Cambridge, the latest home in his IBM career stops. Maybe in the future we'll have more time to explore some of the more interesting details of his life, including what it's like to be a 46-year-old African-American executive at IBM, heading up a software firm that regularly deals with some of the biggest corporations in the world as customers.

We'd certainly like to hear him expound a bit more on his penchant for being honest with customers, to the point of even admitting that Lotus may not have all the answers to their software needs.

"We can't afford to pull the wool over our customers' eyes," he says, directly addressing a question about products but also nimbly answering any lingering doubts about whether Lotus has learned from Mr. Papows's actions. "We have to be forthright in that dimension, have open and honest dialogue," Mr. Zollar says. "You're always better off if you're honest with the customer."

And Lotus, for now, seems lucky to have Mr. Zollar at the helm of a division he wants to keep unique -- as unique as any part of IBM can be. Given his solid history, both in software and at IBM, it looks like he'll get some time at least to try.

"There are some differences [between Lotus and IBM]," Mr. Zollar says. "But our responsibility is to continue to build this established brand, to return value to IBM shareholders. I know where there are real synergies to be mined."

Discuss today's column in the Big Fish column discussion. Or check out forums, video, and events at the Discussions home page.

So what, you ask. Why shouldn't Microsoft push its own products? Well, the company claims to be standards-focused, and much of the discussion on Thursday will be couched in that context.

But look at Tuesday's announcement that Microsoft, Compaq Computer, and Intel will invest a total of $45 million in Digital Island to help fund that company's streaming media infrastructure. It's being viewed by most as a straightforward funding and technology deal. But it could also be seen as classic Microsoft: using dollars and collaboration to lock in support for its technology.

While Microsoft can't be faulted for wanting to further its (and its stockholders') interests, its history suggests that neither can the company be taken at its word when it proclaims that standards, especially in regard to the Internet, take priority over proprietary Microsoft technology.

SAME AS THE OLD BOSS

Coincidentally, Tuesday also saw the judge in the Microsoft antitrust case ask the Supreme Court to hear the case, a legal fast-track maneuver. Surprisingly, the judge also granted Microsoft's request that his mandated behavioral modifications be stayed pending appeal. So Microsoft basically gets carte blanche to carry on, for better or for worse.

That probably means more deals like the one with Digital Island, whose new infrastructure equipment will support only media created with Microsoft's formats, a subtle but effective way to win converts. Though Digital Island stressed that it remains agnostic to all media formats -- indeed, the company has a similar internal infrastructure, based mainly on Sun Microsystems servers, that runs content created using Realnetworks's tools -- real standards support would have seen it try to merge the technologies, rather than create separate, warring infrastructures. It's just more Windows everywhere, a strategy that's likely to continue for the foreseeable future, no matter what happens with its current legal entanglements.

That's why we should view Thursday's pronouncements, in whatever form they arrive, with an extremely critical eye. While we can expect Microsoft to pay plenty of lip service to server-based Internet interoperability (be prepared for tales of how Redmond is now the champion of Internet standards activity), the follow-through will most likely attempt to make Microsoft server software as much a technology lock-in as Windows is today on the desktop.

Much of Thursday's event is expected to focus on Microsoft's work with eXtensible Markup Language, or XML, technology, and how XML will connect Microsoft applications with other systems across the Internet. While XML is ostensibly a standard, history buffs will remember that Microsoft once also declared it would license and support Java, and we all know how that effort turned out.

NO FRIEND TO STANDARDS

What's perhaps more instructive is to look at the company's current battle with Realnetworks and Apple Computer in the streaming-media space, which can perhaps be simplistically compared to the legendary "browser wars" between Microsoft and Netscape. In the latter, Microsoft used its dominant desktop marketing might (as well as some licensing and bundling deals that the Justice Department cited as examples of Microsoft's abuse of its monopoly powers) to win market share, which certainly played the biggest part in making Netscape takeover bait.

Is the company using the same tactic to win in streaming media? While its deal with Digital Island is not exclusionary, it's obvious that Microsoft will continue to push its technology with whatever means are possible, including payouts and partnerships. Mind you, there's nothing necessarily wrong with this approach; but it's hardly the kind of coцperation that got the Internet built, and that should be remembered when Microsoft tries to pitch itself as an active participant in standards development.

The only standards Microsoft is interested in are the de facto ones it can control -- like Windows, which became a "standard" in no small part because Microsoft choked off all other alternatives, either through successful competition or by illegal use of monopoly powers, depending on whose lawyers you believe.

No amount of talk about standards, interoperability, and coexistence with other computing platforms can change that. Tuesday's agreement with Digital Island shows that the "next generation" of Microsoft will continue to put its money -- and focus -- behind deals that put Windows first.

Discuss this article in the Public Companies in the News discussion, or visit the Discussions home page.

Granted, our visit came on the very day they announced the company, so it's forgivable that nobody was at the receptionist desk to meet us. In fact, there isn't really much left of the desk in the entryway, which looks like a casualty of a half-finished remodeling job. Aside from the laser-printed piece of paper taped to the door (with "Packet Design" printed on it), a couple of battery-packs from cordless drills stand silent sentry, recharging.

Beyond an inner door, we're greeted by nothing but a big open space, with a pile of office chairs pushed against one wall, still in their protective shrink-wrap. A cable conduit or two hangs from the ceiling, spliced open, wires ready to be connected. It's like the room is saying, "insert startup company here." In one corner office, Ms. Estrin and Mr. Carrico are huddled together, hunched over a computer, getting ready to do just that. And they've never seemed happier.

FOREVER ENTREPRENEURS

This, after all, is the place for this duo -- a pair with roots in the very beginnings of Silicon Valley and the Internet. Ms. Estrin, perhaps the better-known of the couple, spent time in the '70s at Stanford as one of Vint Cerf's "elves," the tinkerers and programmers who put together the Internet's building-block protocol, TCP/IP. After meeting and becoming an item while working at chip manufacturer Zilog, Ms. Estrin and Mr. Carrico have spent most of their careers building companies from the ground up, a process that used to involve hard work and perseverance.

"It's much different now than it was then," says Mr. Carrico, remembering the details of the couple's first startup, a networking hardware concern called Bridge Communications (which was later sold to 3Com). "It took us 10 months just to find the money to start the company. Then, we spent four years building the business before we went public. When we did, we had $75 million in revenues, and were getting 60 percent margins on sales. And with all that, we only got a valuation of $85 million."

Now, especially in networking, the game has changed considerably. Forced by market pressures to keep sales at a breakneck pace, networking leaders like Cisco Systems, Nortel Networks, and Lucent Technologies have become voracious consumers of talent and ideas, using their inflated stock prices to scoop up nascent companies at lofty valuations.

CASHING IN

Precept Software, the last startup Mr. Carrico and Ms. Estrin founded, was part of this tale. It was scooped up by Cisco in April of 1998 for $84 million. Although the founders didn't necessarily want to sell, it was truly a deal they couldn't pass up, for the good of the company and its investors. However, it also meant that the lifelong entrepreneurs had to spend a bit of vesting time in the belly of the Borg, a situation they describe in only the politest of terms.

"Cisco offered us a view of a different level of breadth," says Ms. Estrin, who became Cisco's chief technology officer after Precept was bought. While she fulfilled the obligations of her job, Ms. Estrin never seemed truly comfortable as a spokesperson at a large company. She seemed to chafe a bit at having to deal with the inevitable handlers and minions that come with such a position, and appeared intellectually hamstrung by having to hew to the corporate line when giving big-picture strategic speeches at major trade shows.

Clearly, both Ms. Estrin and Mr. Carrico are more comfortable at the face-to-face level, where both their passion for technology and their business smarts are in full play. Although their profitable exploits have made them both well-known and well-off, neither seeks the spotlight. Instead, they prefer to spend time mucking around in the gray zones of networking, figuring out solutions to the tough problems that take time to tackle.

TIME TO THINK

"Nobody has time to think anymore," Ms. Estrin says, noting that networking startups these days seem to focus on point-product strategies, the better to float a quick IPO or to be acquired. That market reality is why the couple structured Packet Design the way they did -- as a technology-driven company that will spin out ideas as licensing deals or as completely new companies -- so that they don't have to ever again give up the fun of being at a startup.

"We didn't want to start another product company, because we didn't want to be in the position nine months down the road of being forced to sell the company and start all over again," says Mr. Carrico, whose wife politely corrects him by suggesting that "encouraged" should replace "forced" in his sentence. She also jokes at his expense, noting her husband's penchant for staying in the background. "We said if we get bought again, Bill has to be CTO," she says -- a line that produces laughter from both.

Too soon, the lunch and the meeting is over, and Mr. Carrico is quickly back in his office, while Ms. Estrin escorts her guests out through the empty main room. "The cubes are coming in tomorrow," she says with an air of anticipation and excitement for what comes next.

Discuss today's column in the Big Fish column discussion, or visit the Discussions home page.

Even if the U.S. District Court's judgment to split Microsoft up into two separate companies sticks, it's unlikely that the company will lose much of its clout in technology or the public markets anytime soon. In a statement released just moments after the ruling, company chairman Bill Gates said that Microsoft will fight the ruling tooth and nail and will file legal appeals as soon as possible, which if successful might make Wednesday's high drama and histrionics a moot point.

While the court's proposed conduct remedies may give rivals and partners a slight bargaining edge in future dealings with the software giant, Microsoft's huge installed base of Windows users and developers -- and its already-stated plans to move in directions strategically different than those that brought about the trial in the first place -- means that any remaking of the high-tech landscape will still be viewed, by many, through a Microsoft-developed window.

THE FIRST GAVEL

The ruling, delivered by U.S. District Court Judge Thomas Penfield Jackson after the close of markets Wednesday, puts an end to the trial phase of the case, which began when the Justice Department sued Microsoft way back in October of 1997. While the judgment gives Microsoft four months to break up the company (into an applications-focused company and an operating-system concern), the judge also called for an immediate imposition of conduct remedies that analysts say may have a more lasting impact than splitting up the company's employees and assets.

"People have been focusing on the breakup, but any such structural remedy is at best two years away," says Chris Le Tocq, research director for personal and distributed technologies with Gartner Group. According to Mr. Le Tocq, some of the conduct remedies proposed in the judgment -- specifically, ones dealing with Microsoft's ability to negotiate contracts that favor the use of its technologies -- could force the company to change some of its haggling habits and to change contracts already signed.

A proposed ban on exclusive dealing, for instance, could jeopardize future dealings as well as agreements Microsoft already has in place, including deals like last year's $5 billion payout to AT&T that ties Ma Bell into using Windows CE on its TV set-top boxes.

"Under that remedy, the AT & T deal goes away," Mr. Le Tocq says. Another remedy would prohibit Microsoft from dictating what the opening screen of its power-up looks like, a facet that could allow for greater flexibility and innovation by PC manufacturers and other licensees of the operating system. Of course, Microsoft is expected to challenge vigorously both the remedies and the proposed company breakup, and will likely ask for a stay of any part of the judgment until it can file its first appeals.

SPLITSVILLE

At the core of the judgment is the proposal to break Microsoft up into two separate entities, one with the Windows operating system components and the other with application- and Internet-related businesses. While the Justice Department and state attorneys general seemed to think the divestiture was a necessary step to take, few people in the industry could guess how it might actually be implemented.

"I just don't know how it works, it's really mind-boggling," said Ruthann Quindlen, a general partner at Institutional Venture Partners, about the proposed split. "It's going to be interesting to see what's going to happen with the leadership [of the companies]," she said.

The stock market reacted mildly before the decision, ultimately sending Microsoft up 88 cents, or 1.26 percent. But even Microsoft competitors seemed to care less about the ruling than about the ongoing battle with Microsoft products.

"We're not preoccupied with whether there are two Microsofts or 20," said David Brotherton, a spokesperson for Realnetworks, a fierce competitor with Microsoft in the interactive media space. "We're focused on our own business model, and not losing sleep over what happens [in court]."

There's even some question over what will happen next on the legal front, since there are procedures that could allow for the Supreme Court to review the case, skipping over the usual procession to the U.S. Circuit Court of Appeals for the District of Columbia. Mr. Gates, however, says Microsoft views the ruling as "the beginning of a new chapter in this case," and says he's confident the company will win on appeal.

FULL SPEED AHEAD

Even in the worst-case scenario, Microsoft may not be a total loser, since the company already has started putting the balance of its marketing and development efforts behind its applications business, which has much higher profit margins than the company's operating systems divisions. Since its proposed Next Generation Windows Services appear to rely more on the Internet than Windows per se, Gartner's Mr. Le Tocq notes that an applications-only Microsoft may end up being even more of a powerhouse than the current firm.

"Everyone thinks the split-up is equal, but it's not a 50-50 deal," Mr. Le Tocq says. "It's more like carving off a chunk."

Ms. Quindlen, who was the software analyst at investment banker Deutsche Bank Alex. Brown when that firm took Microsoft public in 1986, says that whatever the case's final outcome, neither side has a claim to any high moral ground.

While calling the government's stance "inappropriate," Ms. Quindlen also noted that Microsoft caused some of its own problems by not knowing how to quit while it was ahead.

"It wasn't enough for them to win; other people had to lose," she says of Microsoft's modus operandi. "If they had not had that extra twinge, I think they just would have been viewed as a very strong, competitive company," she says. "But because they were over the line, they ended up making incredible enemies. With a worldview like that, this [antitrust case] was kind of inevitable."

Julie Landry contributed to this story.

Discuss this story in the Microsoft discussion forum, or visit the Discussions home page.

With the worldwide boom in Internet commerce and traffic, demand for space in such "data centers" -- which act as central hosting and access-connection facilities for Web servers and router equipment -- far outstrips the current supply. But speakers at the Capacity Wholesale Market Global 2000 conference here this week made it clear that a shakeout in the co-location market is not far off, with carriers facing off against independent suppliers in a battle to provide the best access platform between the long-haul backbones and the local business and consumer network loops.

THE PLACE TO BE

Rod Matthews, service provider Global Crossing's president and COO for Europe, says that the huge growth in Internet applications -- a list that includes Web-site hosting, email delivery, and electronic-commerce services -- has driven the demand for co-location facilities, which can put a company's servers as close to the Internet backbone as possible. A number of co-location facilities have been built here in Docklands, a factor that has helped make this city the world's biggest hub of Internet traffic, an honor it recently wrested away from New York.

Typically, co-location facilities include one or more high-speed connections to the Internet backbone, and also provide client equipment with secure, fail-safe physical plant conditions as well as network management services, usually at a cost savings over what clients could achieve by themselves.

Gail Smith, president of Europe operations for IP backbone-services provider Level 3 Communications, says that many co-location customers are Internet service providers (ISPs) or application service providers (ASPs) who may spend between 25 and 50 percent of their expenses on bandwidth. Since these customers can sometimes save as much as 50 percent on access costs by putting their equipment into a co-location center, it's no surprise that Ms. Smith and Level 3 see co-location as a potential $50 billion global market in the near future, a guesstimate that others in the market think might even be too low.

"That's the reason why every man and his dog is talking about co-location services," says Mr. Matthews, whose company runs ten such facilities in the U.S., with more planned for Europe, adding to its already-running facility in London's Docklands.

CLOSE TO THE (BACK)BONE

The question consumers of co-location space will soon be facing is whether to locate with an independent provider, such as market pioneer Exodus, or directly with a carrier, such as a regional telecom giant or a newer backbone provider like Global Crossing or Global Telesystems (GTS), a leading provider of broadband network services across Europe.

While some European carriers are building such centers in the U.S. simply to provide better global channels between U.S. content providers and European consumers, others say the facilities are needed to provide accurate service level agreements (SLAs), the quality contracts written between providers and bandwidth customers. Brian Thompson, chairman and CEO of GTS, says having data centers on the network allows GTS to ensure the quality of its network connections.

"Having devices on the network is critical," says Mr. Thompson, whose company opened its latest co-location facility in Docklands the week that began May 29. "With a data center on our network, we can guarantee performance -- it's under our control," he says. "That's where an Exodus is at a disadvantage."

NOT LOCKED IN

But players on the independent side say that having the ability to connect to multiple backbone carriers makes their co-location facilities more flexible and protected against proprietary pressures.

Guy Willner, CEO of IXEurope, says his company is dedicated to building carrier-neutral co-location facilities.

"Our customers like the comfort zone," says Mr. Willner, whose company has two centers in the London area and has leased space for another in Paris. "If they don't like one carrier, they can always change."

Mr. Willner also wonders how receptive some vendor-based co-location centers might be to managing or accepting equipment from competitors. "Will customers really be able to go into Intel's hub with their Sun Microsystems and Hewlett-Packard machines?" he wonders. "That's another reason why I think carrier, ISP, and manufacturer-neutral hubs will be in demand."

Currently, co-location centers of any flavor are filling up faster than they can be built, a factor that hasn't been lost on the venture-capital and banking community, which is doing all it can to help fund co-location center construction, Mr. Willner says. For the near-term future, that should ensure more competition on all sides before the shakeout begins.

"To those guys, the co-location business is the next best thing since sliced bread," Mr. Willner says.

Discuss bandwidth, hosting, and Net services trends in the ongoing Bandwidth and Hosting discussion forum, or visit the Discussions home page.

"I'm attracted to this role," says Mr. Lang, who works for one of the few business lines where Cisco is, indeed, playing catch-up in the marketplace. Although he's a youthful 35 years old (turning 36 in June), Mr. Lang is a veteran of many networking seasons, and he says he'll enjoy helping Cisco battle the telecom equipment-provider giants like Lucent Technologies and Nortel Networks.

"I'm a bit of an iconoclast, a bit of a rabble-rouser," says Mr. Lang. "It's easier to do that when you're in the underdog role."

READY FOR MY CLOSE-UP

Not many underdogs, however, get tapped to give keynote speeches at major industry trade shows, like Mr. Lang did earlier this month. Although some attendees at the Networld+Interop event (the networking industry's equivalent of Comdex) in Las Vegas earlier this month may have been surprised to see Mr. Lang on stage instead of Cisco's charismatic, spotlight-friendly CEO, John Chambers. The topic and style of Mr. Lang's speech were tailored specifically to Cisco's carrier customers.

Although he claims to have spoken before large crowds before, Mr. Lang admitted he was "a little bit nervous" before his Networld+Interop keynote, mainly because his presentation included three walk-on appearances by Cisco service provider customers.

"There was just so much that had to come together, I'm glad it came off as well as it did," said Mr. Lang of his keynote. Although the customer presentations were predictable -- they were of the soft-sell variety, where the focus is on the application and not the underlying Cisco products -- Mr. Lang injected a needed bit of humor into the proceedings by showing some Letterman-esque video bits he'd taped the day before.

Walking around the Las Vegas Convention Center with a camera crew in tow, Mr. Lang did some impromptu man-on-the-street interviews with showgoers, often getting amusing responses to his deadpan questions.

IS THAT YOUR FINAL ANSWER?

"My secret plan is to be a TV star," Mr. Lang says. "If Regis Philbin ever wants to quit hosting Who Wants to be a Millionaire, I'll be ready."

Mr. Lang's non-entertainment career path took an upward turn in 1991, when he left a string of telecom-related jobs (with AT&T as well as the regional New England telephone company) to join Cisco, then a true underdog. In his first turn with Cisco, Mr. Lang eventually worked his way up to become the product manager for Asynchronous Transfer Mode equipment, an important network-backbone technology.

In 1996, Mr. Lang left Cisco for a startup called Ipsilon Networks, which gained considerable attention for its plan to build switches that theoretically would route network traffic faster and more efficiently than Cisco's well-established routers. Although Mr. Lang says the trade press overplayed the sometimes-contentious war of words between Ipsilon and Cisco ("They seized on the whole David vs. Goliath thing," Mr. Lang says now), Ipsilon eventually failed to make a dent in the market, and ended up being purchased by Nokia in late 1997 for about $120 million.

YOU CAN COME HOME AGAIN

"It [Ipsilon's technology] was really more a feature than a company," Mr. Lang says, also shrugging off the idea that eyebrows were raised when his next job after Ipsilon was back with Cisco, his former employer's main competitor.

"I thought that after leaving Ipsilon, I'd do another startup," Mr. Lang says. "The funny thing is, two of the three companies I talked to at the time eventually ended up being acquired by Cisco. It seems there was a certain inevitability that I'd end up back here."

A dinner with Don Listwin, Cisco executive vice president, convinced Mr. Lang to come back and lead the company's nascent service-provider marketing effort, a job he says combines the best of both worlds.

"You get the fun of the startup rocketship ride, plus you have tremendous resources," says Mr. Lang, who hosted interviewers after his speech in a spacious suite in the Las Vegas Hilton. Truly not the digs of an underdog, but certainly those of a competitor.

Discuss today's column in the Big Fish column discussion, or visit the Discussions home page.