The Reinvention of Novell

With the recent acquisitions of Ximian and SuSE, Novell Chief Executive Officer Jack Messman has transformed a marginalized company into a mover-and-shaker. Look out, Red Hat - and Microsoft - Novell has big plans for Linux.

Founded in 1979 as Novell Data Systems, Novell is almost as old as the personal computer itself. And like the PC, Novell has a colorful history. Started as a computer manufacturer and maker of disk operating systems, the company’s first product, ShareNet, was expensive and complicated, and sold poorly. By 1983, the first incarnation of Novell was on the verge of collapse.


Enter Ray Noorda. Noorda, a Novell investor and a sharp businessman, saw value in ShareNet’s software core — its network operating system (NOS). With Noorda leading the company, Novell switched to designing and marketing software and hardware for computer networks. The company’s original intent? Enable CP/M PCs to share an expensive hard disk.

Of course, networking is where Novell made its mark. In 1983, the company introduced NetWare, the first commercial local area network (LAN) software to implement a shared server, including such prescient features as access control, shared files, and shared printers. Wisely and uniquely, NetWare was also hardware independent, capable of running over Ethernet, ARCNet, and Token Ring.

As the use of personal computers grew throughout the 1980s, so did the fortunes of Novell. In a move that’s downright Microsoft-like, Novell even bought several network interface card manufacturers, driving hardware prices down, but NetWare market share up. By 1987, Novell dumped its hardware division to focus on software — but its plan to control the market was a phenomenal success. Companies such as Corvus, Banyan, and 3Com had been decimated, and even IBM’s LAN Server and Microsoft’s LAN Manager were in nominal use.

By 1990, Novell — nicknamed “Big Red” — had a near monopoly on computer networking, and NetWare’s Internetwork Packet Exchange (IPX) protocol was a de facto industry standard.

By 1994, the company’s annual revenue exceeded $2 billion.

But then something happened: Windows NT. Suddenly, Microsoft had a real server operating system to complement its growing monopoly on the desktop. Then something else happened: the Internet. Exacerbated by Novell’s lack of focus and an increasingly difficult to use NetWare product, Novell stumbled. Trying to regain revenue and leadership, Novell purchased UNIX from AT&T, DRDos from Digital Research, and acquired the entire WordPerfect company. Ultimately, none of those acquisitions helped resuscitate Novell’s lagging fortunes. By 1996, the acquisitions were sold off (the specifics of the sale of UNIX acquisitions are at the root of the recent SCO v. Novell suit, as you’ll understand shortly), and the second coming of Novell seemed to be coming to its end.

But perhaps the third time’s a charm. In 1997, with wishful thinking, Novell hired Sun Microsystems’ Chief Technology Officer Eric Schmidt to renovate Novell once again. Schmidt (finally) made NetWare compatible with TCP/IP, and tried to launch a consulting business, eventually acquiring large consulting firm Cambridge Technology Partners for $266 million.

But Schmidt’s impact was minimal, and Schmidt left comparatively quickly in 2001 to helm Google. Former Cambridge Technology Partners CEO Jack Messman — who oddly was Novell’s president from 1982-1983 — became Novell’s fourth potential savior.

As you’ll learn, NetWare was still in steep decline in 2003 when Novell announced that it was acquiring Linux software developers Ximian and SuSE. The architect of those acquisitions was none other than Messman, who saw great potential in wedding NetWare to industry standard Linux. Indeed, in the span of less than a year, Novell has transformed itself from a sleepy, perhaps even marginalized software vendor to a Linux leader, a direct competitor to Linux principal Red Hat, and perhaps even a worthy combatant against old foe Microsoft.

Shortly following Novell’s BrainShare 2004 conference, Linux Magazine Editor-in-Chief Martin Streicher spoke to Novell CEO Messman about the history of Novell, the case against the SCO Group, Novell’s plans for Linux, and what lay ahead for Novell’s twenty-fifth year in business.

LINUX MAGAZINE: What happened to Novell between the release of Windows for Workgroups and now?

JACK MESSMAN: Novell originally had a large market share. However — and I think the government’s cases have proven this — Microsoft began using its monopoly on the desktop to enter the server market, an area where [Novell] was strong.

Microsoft also withheld [information], and that prevented us from porting WordPerfect to the desktop. At the time, people who were using WordPerfect had a choice: they could move to Windows and leave WordPerfect behind, or they could keep WordPerfect and not move to Windows. History shows that users moved to Windows.

That was the first major hit.

The second hit was Novell’s decision to stay with IPX rather than switch to TCP/IP. Novell believed — and perhaps even does to this day — that IPX was faster, more scalable, and more robust. But the industry had moved to TCP/IP as the de facto standard, leaving Novell behind.

LM: Eric Schmidt and Netware 5 changed that?

MESSMAN: Yes. In 1997, Eric Schmidt became CEO and the first thing he did was make Netware compatible with IP. Netware 5 supported the Internet protocols and dropped IPX.

[Eric] also announced a few new products. One of the new products was ZenWorks, which manages Windows and UNIX desktops. It does downloads, it keeps track of where hardware is, what software is running, that kind of resource management. ZenWorks has grown to be a significant product for Novell, doing about $125 million a year.

Then [Eric] noticed that Novell had a number of highly technical products that were difficult to sell. He saw that [Novell] needed to get into the solutions business, providing integrated business solutions that used Novell products. His approach was to start a consulting operation to develop those solutions.

However, [Eric attempted to start that division] at the height of the dot-com boom when consulting firms were highly priced and in high demand. After looking at the prices of firms, he decided that he had to grow [Novell's operation] organically. He tried that for a year, but wasn’t able to grow it very fast. So, he ended up with 300 people, which wasn’t critical mass.

A year later, Eric decided to take another look at the economics and found that prices of consulting firms had come down as the dot com bust was starting to happen. He looked again at Cambridge Technology Partners, where I was CEO, and acquired it. [Eric] combined the companies and assigned their management to me. [Schmidt then moved to Google.]

So, that put [Novell] in the consulting business, and we started developing solutions that would pull our products through. We had to change a number of things in the company to be able to do that. We had to train the sales force to sell solutions rather than products.

LM: How did Linux enter the picture?

MESSMAN: About the time I joined Novell, it was clear that NetWare was in a long decline — four or five years worth of declines. In any quarter, NetWare sales were declining between 12 and 15 percent over the previous year.

Now, we did have a large installed base. When we asked people still using NetWare why they liked it, they said, “Because it just works.” NetWare was a very stable product, very highly scalable, and offered high performance. But we were slowly losing the installed base because of Microsoft’s aggressive actions.

Over time, Microsoft has been able to take the unique features of Novell’s products and incorporate them into Windows. And, of course, when they do that, they offer those features for free. And while some of the features they put into the Windows products were not as robust and high performance as Novell products, they were good enough. The “good enough” strategy seems to work for Microsoft.

Thus, faced with a steadily declining NetWare base, I tried to figure out how to get off that. Linux was taking off. So, in January 2003, we decided to take the networking services that were inside of the NetWare operating system — file, print, collaboration, messaging, directory, and a few others — and put them on a Linux kernel.

We went to some of our hardware vendors and asked for feedback on the strategy. They thought it was a super idea, because they were looking for things to cause people to buy Linux hardware. The hardware vendors saw two impediments to boosting sales: first, there was a lack of applications that would work on top of the Linux kernel; and second, there wasn’t anyone out there providing service, technical support, and assistance to make CIOs sleep well at night.

In April 2003, during our annual Brainshare conference, we announced that we would put our networking services on top of the Linux kernel.

LM: You acquired Ximian just five months later. How did that happen?

MESSMAN: One of the things that has always hurt us competing against Microsoft was their ability to tightly integrate what happened on the server with what happened on the desktop. To compete more effectively, we wanted a desktop product. Shortly after Brainshare 2003, we became aware that Ximian was available.

The acquisition of Ximian added four products to our line: a desktop produced based on GNOME; Evolution, which connects a Linux client to an Exchange back-end; Red Carpet, which is the equivalent of Novell’s ZenWorks, but for Linux desktops; and Mono, which allows a .NET developer to developer code on Linux.

LM: And then you turned around and bought SuSE in November. Quite a shopping spree.

MESSMAN: Shortly after we closed the acquisition of Ximian, the people at Ximian kept coming to us and saying, “[Novell] needs to get into the Linux distribution business.” Indeed, we were finding that we were losing business, because many of the people we were trying to sell the Ximian products to had already bought Red Hat. [Novell] didn’t have a distribution, so we were making our software work on Red Hat as well.

So, we had the idea of doing our own distribution. We were going to call it Cleanex, a play on words and a reaction to the SCO Group’s claims that Linux was contaminated with UNIX code. We assembled a consortium of leading vendors — large vendors that I won’t identify, hardware and software vendors — and met in Chicago with the purpose of developing a third distribution [Red Hat and SuSE, being the other two.] But the vendors said, “Look, Novell is obviously capable, but we don’t want three distributions. We only want two.” They didn’t want just Red Hat, because Red Hat might then start acting like Microsoft. They didn’t want three because of the expense required to support so many distributions. Ultimately, the companies at that meeting — the so-called Chicago 7 — recommended that Novell buy SuSE.

Well, Novell had no idea that SuSE was available, but many of the vendors did, because they were investors at SuSE and were aware that the venture capitalists wanted some liquidity. SuSE was available, and after some healthy competitive bidding by Sun and even Red Hat, Novell paid $210 million for SuSE.

SuSE was doing about $40 million in revenue at the time. While we paid five times revenues, [Novell's] market cap went upward by approximately $3 billion.

I think the reason for that was that people saw the logic behind our strategy. The acquisition gives Novell a technology that’s consistent, that has the same code base from the desktop, to the server, to the mainframe. And we’re the only people who can say that.

Of course, Red Hat just recently announced that they’re going to do a desktop product. While they’ll match us on the bottom end, we still have server software that they don’t. The services that have been sitting on top of NetWare all these years will work on top of SuSE Linux.

LM: Red Hat is a formidable, established Linux vendor. How do plan to differentiate Novell?

MESSMAN: While Novell may be new to Linux, Novell has impressive credentials in operating systems. We’ve doing operating systems for twenty years or so, and our customer base has always appreciated that we know what we’re doing. The same expertise now applies to Linux.

We have a worldwide technical support organization, and we’ve trained all of them on SuSE Linux. We now have 650 technical support people trained on SuSE Linux — that’s more people than Red Hat’s got people. We have an education channel, and we created the value-added reseller (VAR) channel back in the early 1980s. Right now, Linux is largely a channel opportunity. We believe that small businesses can take advantage of Linux technology and our VARs are in an excellent position to call upon them. We don’t believe that we can directly reach all of the customers who want to buy Linux. We have a direct sales force, but we need the channel to really gain leverage. Based on the response we’ve gotten since acquiring SuSE, I’d say the channel really seizes the opportunity as well.

Novell is also offering indemnification for its Linux customers.

LM: What is Novell’s current position on SCO’s claims?

MESSMAN: Novell sold UNIX Systems Laboratories to the SCO Group back in 1995. All we sold them was the right to conduct a business based on the UNIX copyrights and patents that we still own. We didn’t give [SCO] — we didn’t sell them — the copyrights, although they claim that we sold them the copyrights.

Novell also has a license to use UNIX inside Novell and at our customers. So, if there is any UNIX in Linux — and we don’t believe there is, because we own the code and we are able to look at it — but if there is any UNIX in Linux, then Novell has the right to use it.

In turn, we offer our customers indemnification. There are several ways you can indemnify people: one is to have a huge balance sheet — we have $750 million in cash — or you have some legal basis on which you can provide that indemnification, and we have that as well.

LM: What do you think of Microsoft’s Longhorn?

MESSMAN: Some people are saying that they’re going to turn Longhorn into “Shorthorn” so they can get it out earlier.

I think what Microsoft is trying to do is make customers more dependent on their software. They’re going to try and make many of the things that are in the open source community unusable by locking people into [Microsoft's] proprietary environment.

For example, Microsoft has said that they’re trying make the Longhorn desktop just like going to a browser: easy and transparent to the user. But in true Microsoft fashion, it would create dependencies on that environment, so that moving to a tool like Mozilla would be very difficult.

I think it’s Microsoft’s strategy to [architect Longhorn] to lock people into the Microsoft environment, therefore prolonging their hold over the desktop.

And, I think they’re trying to get to a common code base, so that the interoperability between the desktop and the server and maybe even the mainframe is better.

Our feeling is that it’s too early to tell someone some of these things, but we believe that whatever emerges in the future should be open and standards-based and without a single vendor controlling the API and the process. Longhorn appears to be heading in the opposite direction, totally against what the open source movement is all about.

LM: Speaking of Microsoft, can the Linux desktop compete with Windows?

MESSMAN: When you talk about the desktop, there’s really more than one kind of desktop. You have the “fat client,” as it’s called, which most of us have on our desktop. Then you have the “thin client,” which is really a dumb terminal connected to a server or a mainframe. Then you have various permutations and combinations of that.

I think our desktop strategy will yield a product in each of those areas. Obviously, we believe that much of the software that works on the Microsoft desktop will ultimately work on the Linux desktop. But we also believe there are applications that are more process-driven where you don’t need a lot of CPU horsepower. There, a stripped-down desktop can really save on expenses. I think we’re going to be a lot more flexible than Microsoft has been in that regard.

LM: Novell and Microsoft were once arch-competitors. Is Novell headed down that path again?

MESSMAN: Some people say, “Aren’t you really taking on Microsoft again?” They say, “You know, you learned with WordPerfect that you shouldn’t do that, and you learned with DRDOS that you shouldn’t do that.”

My answer to that is no, we’re not taking Microsoft on — the open source community is taking Microsoft on. Open source is 900,000 developers and growing that know what they like and work when they want to work. It’s the guerrillas taking on the gorilla.

The general feeling is that the average experience of an open source programmer is around 11 years, and some of them are very brilliant. Because they work on stuff they like, they do a better job. The open source community is going to be a new way of developing software that we haven’t really fully explored yet. However, we see Novell as being one of the companies that can help our clients take advantage of it.

I think you’re going to see more and more corporations that develop software internally take an open source approach. And there’s far more software developed inside companies than there is by dedicated, commercial software vendors.

LM: Do you see Novell releasing more products as open source?

MESSMAN: Perhaps, but the other thing to keep in mind is that proprietary code and open source code can coexist. All you have to do is look at what’s happened to MySQL. MySQL develops the code, but also makes it available to the open source community for free use, feedback, and enhancements. Additionally, they offer a commercial license to the same code.

Our view is that we can sell open source and proprietary code in mixed environments and not have a problem doing that. People have the misconception that open source means free, and that Novell is trying to sell free software. Well, we’re not trying to sell free software. We sell software and services that allow you to use the free code. The distributions that we’ve put together have a lot of added value in them. Sure, you can get all that technology off the Internet if you want to, but we stitch it together, we clean it up, we make sure that the 3,700 or so packages all work and don’t have any licensing problems. For that, you pay us a fee. And then, of course, we support it with technical services.

We think that the open source community and the proprietary software community can coexist, and it isn’t the problem, the red flag, that Microsoft tried to raise early on with regard to Linux.

LM: Looking forward, what do you see for Novell?

MESSMAN: The result of all this is that Novell has done some very good things in the last few years. We’ve positioned ourselves into very fast growth markets, including Linux and identity management, where the latter is the result of spending many years developing our own directory technology. We were working on eDirectory [Novell's directory service for managing identities and secure access] nine years ago, before people understood what directories could do for you. Now that we have the Internet, the need to be able to control what people do when they come to your firewall is a very important thing. If you run a business and a large customer comes to your site, you might display a large inventory. However, a small customer might only see a fraction of that inventory. We’ve had a number of directory services for some time, but no one knew what they were for. Now, those services are in high demand.

We’re also into web services. We bought a company in early 2002 called Silver Stream Software that produces a tool set that allows you to integrate applications and construct web services very easily. Without rewriting any database or any application, the tools let you repurpose all of your databases and all of your applications. It’s a visual environment, and we think it’s about to take off.

Finally, the major area that we’re in is resource management, which is managing desktops — what software is there, what hardware is there, and optimizing the process of maintaining all that. We believe the whole area of “on demand computing” or “autonomic computing” or “utility computing,” whatever you want to call it, starts at the desktop and works up from there.

So, we’ve positioned ourselves in some fast growth segments. We’ve stabilized our core business. We believe that we’ve started to turn the sales of NetWare around. We believe that Linux is facilitating the turnaround.

And, I think with an improving IT market, you’ll see a rejuvenated Novell. Our shareholders believe that.

Jack Messman is the Chairman of the Board and Chief Executive Officer of Novell. Martin Streicher is Editor-in-Chief of Linux Magazine and last interviewed kernel developer Andrew Morton in September 2003.

Comments are closed.