Linux Chameleon: An Interview with Dirk Hohndel

Dirk Hohndel, SuSE's CTO, shares his thoughts on the differences between the

Linux Chameleon
SuSE’s Dirk Hohndel is a Hacker Turned Business Executive
by Robert McMillan
Interview Opener

Dirk Hohndel is a rare creature. A respected hacker who began contributing to Linux eight weeks after Linus Torvalds first announced the project, he now makes his living as one of the top executives in Germany’s largest Linux company, SuSE A.G. Though it’s been a year and a half since he last contributed code to the XFree86 project — the job that brought him fame within the Linux community — Hohndel says he’s found that you don’t have to be a hacker to contribute to Linux. As CTO for Nuremberg-based SuSE, he’s found a few interesting things for his staff of more than 100 developers to do. Apart from working on XFree86, SuSE developers have been instrumental in creating the ReiserFS journaling filesystem, developing the K Desktop Environment, and helping IBM do its port of Linux to the S/390 mainframe. Hohndel had dinner recently with Linux Magazine’s Robert McMillan to discuss the differences between the U.S. and German Linux markets, crazy IPOs, and the global impact of Linux.

Linux Magazine: SuSE doesn’t get a lot of attention in the United States, but it’s actually one of the largest Linux companies. How many people work there?

Dirk Hohndel: About 600.

LM: Is the company profitable?

DH: We’re currently not profitable, but we expect to be so soon.

LM: 600 employees — how does that break down? I assume most of your employees are in Germany?

DH: Yes, most of the people who work for us are in Europe and specifically in Germany. We have about 60 percent technical people, the rest are in administrative, sales, marketing, strategic alliances, etc.

LM: How many people have you got in the United States?

DH: We have about 15 in the United States.

LM: You downsized your U.S. operations a while ago. How many were there before that happened?

DH: There were 45.

LM: Why did you reduce your American operations? The United States is a big market for Linux.

DH: Well, we had prepared ourselves in the United States for the same type of market that we have in Germany, where commercial support and professional services are a very good revenue chain. This has not materialized in the United States. Not for us, and looking at many of the competitors, not for them. A year and a half ago, the investment community told you to, “Grow, grow, grow as fast as you can.” Today, the investment community only asks you, “When are you going to be profitable?” So if you have an important part of your business that doesn’t generate the revenues that you are expecting, you have to refocus.

LM: So you basically weren’t making any money from your U.S. operations?

DH: No, we weren’t making much money out of local support and professional services. We are doing fine selling our SuSE Linux products, however. An interesting thing is that it’s much more cost-effective to do installation support, business support, and professional services for the United States from Germany. It costs less money to fly somebody from Frankfurt to New York than to fly someone from San Francisco to New York. So, bringing consultants over here for different projects is trivial, and we do that. We have completed a few projects over here. We have 24/7 support available for business customers, and 24/7 is 24/7 all over the world. So what we have done is just transferred execution of our services business from the Oakland office to Germany.

Insane Public Offerings


LM: How has the Linux market changed since the days of, “Grow, grow, grow?”

DH: Well, 18 months ago we had this — what did [Federal Reserve Chairman Alan] Greenspan call it? — “irrational exuberance,” in the stock market. At one point Red Hat was worth $26 billion. Well, obviously the hype was totally out of proportion, and the expectation that this essentially would be like the money-printing business never had any relationship to the facts.

Fundamentally, what changed is that people shifted their focus; this is not only the case in the Linux market, but for all of the technology market. People changed their focus from dreams, theoretical prospects, and fantasies to real life facts — to business. We saw Linux as just a normal business from the beginning. So this is about revenues, about getting customers — happy customers. This is about just creating another business. This is not about some magical way you can print money. The concept of losing 10 times your revenues is outrageous to me. Look at the whole dot-com space, look at the whole technology industry; the focus has shifted from fantasy to facts. And this has caused the contraction of share prices.

LM: Why didn’t SuSE go public? Didn’t you wonder how you were going to compete with all of that capital?

DH: See, this is an argument from the wrong side. In my book, the company has the task of creating business by creating solutions for the customers. In order to do that, the company needs capital. And how the company gets that capital, how much it invests, how quickly it grows, are derived issues. The main issue is where’s the customer, what are his needs, where’s the business? From that point of view, if and when you go public is really a secondary issue. There are a lot of companies that have been founded solely with the purpose of going public as soon as humanly possible.

Being a private company has its advantages. SuSE never had the desperate need to go public. There have been a lot of rumors about SuSE’s wanting to go public. Every single one of them caught us by surprise. Going public was never the core of our business and it isn’t today. And yes, we will go public eventually. But it’s not a big issue, quite frankly. The big issue is to create a sustainable business. There are so many companies that have gone public, and you can count the days until they go out of business.

LM: Do you think that all those crazy valuations could actually end up hurting Linux?

DH: Yes. We had a lot of good people with good ideas but with no business experience whatsoever being handed way too much money. And that created a lot of frustration and a lot of false hope in a sense. I think it would have been much better if those businesses had grown slowly. SuSE was profitable for the first six years of its existence. Then in 1998, we made the conscious decision to say, “Okay, we need to accelerate our growth and risk losing money and make an investment to grow quicker.” But this process of making do for six years and creating a profitable business is very educational. I mean, if you are able to have a business with losses that are 10 times revenues…I believe this was damaging to Linux.

LM: So what are the results of this? Are you finding that CIOs are skeptical about Linux companies, including your own, because they don’t think you’ll be around in a few years?

DH: Yes. We see skepticism on the viability of the business model, because so many people have seen the balance sheets of companies that basically were in the business of burning the VC money. And some of them went away. And this is hurting us. And of course, there is the perception that the Linux business is a very immature business, is a business done by kids, and is a business that is not focused on customer needs. I call this the “we have been funded” craze. You have these six-person companies that have the 40 by 40 foot booth at a very expensive trade show, and you wonder where this comes from. Maybe this is just the humorless German in me, but I think that a little more realism would have been good, would have been very helpful. But I guess many people who hear me say that will say that I am just envious of companies that were better funded than we are.

Supporting Growth

LM: You said that support is important to SuSE’s business model. A lot of people have criticized support-based business models because they are difficult to grow into large businesses. A business model like Microsoft’s, where you license software and put it in a box, scales very well. Are you concerned about SuSE’s ability to grow?

DH: It’s a very different business. I understand the fact that the Microsoft business model is very attractive. The proprietary model — being the monopolist, being able to scale, and having operating profits of about 50 percent — just won’t happen in the open source space. On the other hand, it is very possible to scale a supporting-services business by automating pieces of what you do. That’s the one end. The other end is using the knowledge that you gain by doing that to create new lines of revenue.

Here’s a very simple example: our e-mail server product, which we sell successfully. It encapsulates projects we’ve done before; we have done a number of service projects that involved setting up an e-mail server. We took that knowledge, commodified it, and sold it as a shrink-wrap product. And this is a process that is obviously repeatable. So this is another way of scaling a service and support business. I realize that if you just look at the paper, the classical proprietary software model scales much better. But quite frankly, I believe that the days of the proprietary software model in the IT infrastructure are over.

The Teutonic Twist

LM: How is the German market different from the American market?

DH: In the sense that brick and mortar companies, both large and small, are using Linux in production environments for real-life projects and using it as the core of their business. SuSE has customers ranging from some of the largest financial institutions down to companies with three or four people. We have companies, like a medium size meat and sausage manufacturer, who have done a B2B solution with us. So I’m talking about real live projects, where people are paying you to solve their problems. This is where Germany is just miles ahead of the U.S. market.

LM: Why is that?

DH: There are a lot of things to consider here. There are a lot of pieces that came together. And yes, SuSE is a part of that. We have created this market. We have talked to the customers. Ask Compaq and IBM how they see the global market, and you will hear from both that Germany is unique in this market. And SuSE being there, working with customers, is part of that. But it’s just one part of this. There are many other aspects. One is something that I mentioned before. IT purchases in Germany are much more fact and issue-driven than marketing-driven. It’s not about who has the glossier brochure; it’s about who can solve the problem. And Germany has traditionally been a market that is very open to Microsoft alternatives.

LM: Yes, wasn’t Germany the country that embraced [Microsoft DOS alternative] DR-DOS?

DH: Yes, and Germany was the number one OS/2 market. Even today, half of the financial industry uses OS/2. Germany was the biggest DR-DOS market. Germany has always been a very strong Linux market, one of the biggest VAX markets. So the openness to non-Microsoft operating systems in Germany has traditionally been very high.

LM: Is that because of a conscious anti-Microsoft decision on the part of the Germans?

DH: No, not at all. I think this is very much…I’m not a social historian and this is a more appropriate question for those guys. But from my understanding, this has always been very much driven by finding solutions to problems, by looking at the technical expertise, looking at quality — very basic values. I don’t think there is the slightest bit of ABM (Anything But Microsoft) there. Not at all. The key factor driving companies to use Linux today is the desire to solve problems.

There’s always been an interest in Linux. The first Linux conference was in Heidelberg in 1994. The two most influential German technical magazines, c’t and iX, started going on about Linux very early. Things like this helped in creating the first customers; they talked to other people, so other people got interested.

LM: How far behind do you think the U.S. is regarding the adoption of Linux by enterprise companies?

DH: I have been saying it’s 18 months. I can’t scientifically prove that. It’s just my feeling. I just look where we were in Germany 18 months ago, with the beginning of the professional services business, and where we are here today in the United States.

Made Everywhere

LM: One of the things that’s always struck me about Linux is the fact that it was created outside of America. When you look at the list of Linux contributors, they come from everywhere. It’s really amazing. Do you think that Linux is becoming more or less U.S.-centric?

DH: The U.S. has always been a very attractive immigration country. So that is a very natural process. This is not specific to Linux. I believe that the U.S. contribution is very big, and it’s increasing as many U.S. companies — be it IBM, HP, SGI, Intel, or AMD — start to have Linux people on staff. On the other hand, the countries that are on the threshold between being real developing countries and being second world — the cross-over countries — most of South America, South Africa, parts of Asia — in these countries, the interest in Linux is phenomenal. And there’s a very good social reason for that.

In an open source world, the price of doing IT projects is mainly defined by your local prices. If you do a project based on a proprietary software component and are in a country with very low salaries, with very low economic strength, then the overriding part of the cost is the license for NT. If you’re doing a Linux project, then the overriding factor of cost is your local economic cycle. So if you’re investing in a Linux project, most of the money stays in your economy, because the people you’re paying are going to go buy something, so you’re creating an economic upswing.

This is a wonderful way for these cross-over countries to jump-start their IT businesses, because suddenly, companies with very little funding can be part of leading edge technology, working on the standard operating system on the Internet. So we’re creating a tremendous economy outside of the Western world. And this is why we see so many more developers from Brazil, Chile, from South America, from India, from Russia. We see a lot of people that in the pre-open source days would never have had a chance to be part of the IT community.

LM: Are there partners or companies in particular that leap to mind when you think of people that have been able to leverage this?

DH: There are lots and lots of small companies in those countries. If you go to a Linux conference in Brazil and talk to the people there, every other person owns their own company or is in a small, 10-person company that does projects around Linux. Some of them sell hardware; they buy cheap, off-the-shelf hardware and create mail servers or Web server applications. Some of them provide local support or training. You see a gazillion of these companies popping up and creating small business epicenters around the world.

So there is an enormous amount of small businesses being created. And there is an enormous amount of development being done. This is not on the multi-billion dollar level, but on the day-to-day level of human beings.

LM: Do you think that will ultimately have an effect back on America and on the IT industry in Europe?

DH: Of course. Because it will create markets where otherwise there would not be markets. It will create economic strength. At the end of the day, this creates customers in some shape or form. These computers are going to run on some chips. And yes, these will most likely be low-end computers, but still, this creates a market — be it a market for used equipment from the U.S., be it a market for software, for services. I see this as one of the greatest aspects of open source.

Robert McMillan is editor at large with Linux Magazine. He can be reached at bob@linux-mag.com.

Comments are closed.