Interview: Are You Ready to Rock with Michael Tiemann

An Interview with Michael Tiemann on his first ever trip to India.

Linux For You’s editor, together with technologists Robert Adkins and Alolita Sharma, met Michael Tiemann, vice president, open source affairs, Red Hat, on his first visit to India in October 2004. Tiemann, with a long string of successes — from being the 23 year old developer of the GNU C++ compiler, to co-founder of a very successful software company Cygnus Solutions, is today part of Red Hat’s founding team and eager to push the benefits of open source to the global frontier. In this wide-ranging interview, Tiemann shares his vision, enthusiasm and entrepreneurial spirit.

Q: What brings you to India?

MT: Well, this is my first trip to India and India is well known as an emerging player in the IT market. I have always been fascinated by major trends. And so that might explain why, even though free software was just in its infancy when I started learning about it and writing code in 1987, I saw the dynamics of the free software model being incredibly powerful. It reminded me of that famous quote from Archimedes — “give me a long enough lever and a proper fulcrum and I can move the world.” I saw that kind of dynamics with free software. When I contemplate the resources available in India and the kinds of problems that can be solved, it is a second-to-none opportunity for change.

I see India as a country, as a population, as a group of people struggling with other groups of people to all make life better for everyone. Our CEO, Matthew Szulik, met President Abdul Kalam during his visit to India in January 2004. Szulik came back and he gave a speech at Red Hat about that meeting and the Indian president’s passion to improve hundreds of millions of lives. Whatever good intentions other presidents may have, no [other] president has that opportunity. It was inspirational to me and I think, to Red Hat that we are working with commercial and public sector leaders who have that kind of vision in India.

Q: What is your mission at Red Hat?

MT: My title is Vice President, Open Source Affairs. Not many people know what a Vice President, Open Source Affairs means or does. But what I do is discuss technology policy and strategy with the public and private sector. There are two other people that have the term “affairs” in their titles in the US. Venkatesh Hariharan has that title here in Red Hat India. Tom Rabon is Vice President, Corporate Affairs and Mark Webbink is Deputy General Counsel focusing on legal affairs for open source. Rabon does public policy and strategy, Webbink does legal policy and strategy and I do technology policy and strategy. So those three form a miniature cabinet for how we understand and how we advocate and how we move forward as the world changes and we help change it.

Q: You really do have a global world view then.

MT: Absolutely.

…what I see is if you look at the population of available resources in India, despite any constraints, there will be somebody who basically says, ‘why can’t I change the world?’ Out of India’s population of 10 million people, is there not one Linus Torvalds?

Q: How do you see the official Indian approach to open source which claims to be neutral?

MT: I am still learning. This is my first visit, however, when people tell me that their policy is to be neutral, I take what I think is a responsibly skeptical approach. You know “What do you mean by neutral? What is the real benefit you are looking for?” Some people claim to be technology or policy neutral and it’s not really true. But at the same time, I don’t think that the Indian government owes us anything. I think it is our responsibility to understand what the challenges are, to understand the opportunities and then to successfully advocate for that. If the government is willing to listen and we have something to say I am sure that something will happen.

Red Hat’s responsibility as a business is to be able to transform all the successes that the entire market has into our own success in terms of profitability, and in terms of sustainability. I’m very proud of what we’ve been able to do.

Q: What is your advice to IT firms and CEOs in India who are sitting on the fence? They know there is a lot happening on the open source front but don’t know whether its time to jump in or to just wait and watch for now.

MT: The first thing is that it is important to make an informed decision. There is an incredible amount of FUD (Fear, Uncertainity and Doubt) out there and people who are telling the story badly, or telling the story wrongly. Some of these are well intentioned people and some of them do not have good intentions. Let me give you an example of how the story should be told.

A couple of weeks ago, I visited an IT services company in the US. A billion dollars a year in revenue, and US$ 50 million a year in profits. Nothing has changed for four years, they’ve been flat. They held an internal conference for all their IT people. 500 people from around the world showed up from Japan, India, Canada, US, etc. [They] came to one location and the CEO kicked off the meeting and said, “The time has come that we have to re-engineer the company and we’re going to do it with open source.” A lot of things are going to break. It’s going to be painful and things are going to change. But this company has been around for many years and every 10-12 years we go through this and when we finish it, we are stronger, we are in better positions, we’re more competitive, and we’re more profitable. They specifically identified a strategy based on lowering operating costs, because they are an IT services company. That’s one dimension.

A second dimension is improving developer efficiency. And what they have discovered and what many IT organizations would admit if they were just a little bit truthful, is that they internally have all these stovepipes where an application takes on a life of its own and before you know it they have 50 ways of doing the same thing. This is very inefficient and it also leads to inflexibility. When they want to offer customers new solutions, they can’t because half of the solutions are in one stovepipe, and the other half are in another. They can’t combine.

And this CEO believes, and I agree, that the open source development model, when properly practiced, leads to the kind of modularity, the kind of code reuse and sharing, the kind of interoperability that will dramatically improve their internal operational efficiency. Now there is a third piece which the CEO did not even know about, which I believe is also possible and which I will be promoting not only to him but to everybody. That is the work that Eric von Hippel, professor, Massachusetts Institute of Technology’s Sloan School of Management, pioneered with “user-driven innovation”. I believe that open source provides a superior way to change the customer relationship from “vendor-customer” into “collaborators” so that innovation happens where innovation happens best — not limited to “if it doesn’t come from the vendor you can’t do it”.

It’s my belief that solving the TCO problem, solving the developer efficiency problem, and making progress in user-driven innovation is the hat trick which will define IT competitiveness for the next 10 years. Then something else will happen.

Q: In the example you cited, the CEO was of a services firm. What if the CEO is of a product firm like Adobe?

MT: Adobe is a very interesting story. Adobe came up with Postscript. They filed to become public and they said in their prospectus “we’re going to make Postscipt standard open and publish it”. They went public and then they started changing their mind. One of my co-founders of Cygnus went to the CEO of Adobe, and reminded him, “You know, you said in your prospectus that you would do this as an open standard.” John Warnock, Warnock admitted, “I guess we did”. If you Google this story you will find that Warnock has turned it into a success story about how, if Adobe had not published Postscript, other people would have challenged and destroyed the standard, etc.

If you look at that story it is the counter-proof of Sun’s Java story. So whether by design, by accident or by some combination of the two, Adobe created and sustained its market through openness even as a product company.

The design of the Postscript language permitted arbitrary participation and arbitrary collaboration. I think there is an increasing trend. I would suspect that across all the projects on SourceForge, if you look at which ones began as open source and which ones began as proprietary and then opened up for whatever reason, that a growing percentage [shows] that people are deciding that open source is a suitable transformation strategy for a product. You might say, well gosh, these companies are dumping their orphans to the care of the world. But let us not forget the great contribution that some orphans have made to the care of the world. I think that it proves that ultimately open source is a more sustainable model because the reverse case, people starting with open source and then taking it proprietary, like VA Linux…

Q: or Sistina?

MT: Yes, but we fixed that problem But in the case of VA Linux, I would say, they went from a company with the potential to change the world, to yet another product company.

…the Indian president’s passion to improve hundreds of millions of lives, a billion lives. Whatever good intentions other presidents may have, no [other] president has that opportunity. It was inspirational to me and to Red Hat that we are working with commercial and public sector leaders who have that kind of vision in India.

Q: But you will have a very hard time convincing any Indian product company of that story.

MT: Rewind 10 years, go to America, same story. I have not changed my position in 17 years on open source and free software. I used to play ultimate frisbee, a very fun and very fluid game. My position was called “deep” which meant it, my job was to go and catch the long pass. There was one crucial time where I just did not stretch quite far enough. My teammate said to me, “Michael, if you want to get in the air you have to let go of the ground”. Ever since I’ve been able to stretch a little bit farther. But here’s the good news. The good news is that everyday, life is telling them “if you want to get into the air, you have to let go of the ground.” It just requires that one time that they hear it.

Q: But even Red Hat might not remain in business very long if it was only an open source product company, it would not scale.

MT: I agree with you and let me say that same thing in a different way. The world does not owe Red Hat a living. We are in business because we identify and deliver value to customers. God did not say, “if you build a product, you will be successful.” And God did not say, “if you do open source this way you will forever be a growing company.” What happens is the market says, here’s how far we can go, and here’s how far we’d like to go, and if you have any intelligent ideas about how we can get from here to there, tell me in 15 minutes. The market drives it. And Red Hat has changed a lot. When Red Hat went public our gross margins were 48 per cent as an open source “product company”.

In our most recent public earnings call, our blended margins were 80 per cent, and our enterprise subscription gross margins were over 90 per cent. Would you rather be 48 per cent gross margin or would you rather be say 80 or 90 per cent? The reason we improved is not because we raised prices. The reason we improved is because we found a better way to deliver more value to customers. It turned out, in our case, that there were many more customers who received much more value having a long-term, defined, supportable enterprise product than those who wanted to buy a short shelf-life retail offer. Go figure.

Q: Some people contend that Red Hat is going to be the next Microsoft. What is your reaction?

MT: The first thing that I would say is that Red Hat represents choice. And we represent choice in a number of different ways. But it’s all tied to that central notion of open source and our belief that “choice equals freedom equals power”.

So what we build are solutions that give customers choice. If they choose to download and choose to support themselves, that’s a choice. We do not run around and sue people who are supporting themselves. We do of course protect our trademarks but, as far as what kind of relationship they want with their hardware vendor and whether they want to have a direct relationship with our company, it’s their choice.

We are beginning a lot of discussions with systems integrators and with value added resellers. I’m sure those relationships will expand so that even more choice exists. But fundamentally, we believe that choice has value. In fact, at a Harvard conference, I ran up to an economist and started talking about the value of option values. And she said, “Absolutely right, take my class.” Everybody’s selling! In the case of Microsoft, what they are offering is a choice of one. You look at the cost of exit. It’s clearly, with them, a one choice world.

So if we are able to achieve great margins while offering choice and freedom, I think that’s the new generation of technology company. And just in time. Look at the technology challenges that India is facing. Look at the technology challenges that the whole world is facing in terms of access and in terms of information. This revolution can’t come too soon.

…Some people [in government] claim to be technology or policy neutral and it’s not really true. [While] I don’t think that the Indian government owes us anything, if they are willing to listen and we have something to say I am sure that something will happen.

Q: Often companies like Red Hat and other MNCs come into India and pitch the wrong products and services because they do not realize the level of infrastructure in developing economies drives the market toward the desktop instead of advanced server solutions. How does Red Hat adjust to the ground realities of countries like India?

MT: So here is my “Law of the Hacker”, which is that “with a big enough crowd there is at least one”. And crowds get pretty big in India. When I talk strategy with people, sometimes I talk with CIOs of the most conservative companies. Many of these companies are intellectual deserts: conformity, bureaucracy, process, and all the rest of that stuff. The CIO sits and says, “Michael, I would love to solve this problem, these viruses are killing us. I’ve got row after row of infrastructure that I just can’t do anything with. What should I do?” And I basically would say, “There are two things you can do. The most positive thing you can do is to let people know that they won’t get into trouble for being successful. And the second thing you can do is just don’t fire the guy when you find out he has been successful with Linux.”

When we went to Wall Street, we had a series of four meetings with one of the largest investment banks on Wall Street. At the first meeting we sat down and talked about our Linux strategy. And they said well, it all sounds well and good, but it will never work in this firm. The Law of the Hacker: One of the guys in the back started sort of smiling and said, “We already have it runni
ng on 75 servers.” And the manager said, “What?”. “You know all those file and print servers in building X, they’re all running Linux.” And he was a bit surprised.

Then we had a second meeting which included some of the same players plus a higher level manager. The previous manager introduced us and said, “Well, I guess we do run about 75 of their servers, but nothing really in the enterprise.” Again somebody in the back of the room was smiling and said, “Actually it’s more like 150. And now we’re also doing ‘blah, blah, blah’”.

We had a third meeting with subsequently higher level managers. The conversation went, “Aside from the 150 servers it’s really not in the firm. Er, That number is now 250.” At the fourth meeting it was 475, at which point they just said your next meeting will be with the CIO and CTO and we’ll get this thing done. If they had fired the person who had done the 75, maybe there would still be 150, but they wouldn’t talk about it.

This is the “trickle up” theory. Look at my own story. Nobody told me to do the GNU C++ compiler. Nobody told that me I could, but nobody told me I couldn’t. Well, theoretically, the GPL said I could. The GPL said welcome, if you want to hack it, may the source be with you. It was an absolutely fantastic experience. I single handedly brought to life that compiler. When I gave a talk at ATT Bell Labs, which is where C++ came from, there were 35 people in the audience who were directly working on ATT’s native code compiler. They had been working on it for a period of 5 years and had never released it. And here comes this 23 year old, who put his first release out single-handedly in 6 months. Now, of course, Richard M Stallman did a lot of work, but imagine what it felt like to me as an individual that with a little bit of hard work, 18 hours a day, 7 days a week for 6 months I could send an announcement to a public mailing list, and say “please download this and send me bugs.” I got a bug back almost instantly. That’s what kept me going. So what I see is if you look at the population of available resources in India, despite any constraints, there will be somebody who basically says, “why can’t I change the world?”

With the availability of Fedora for example and certainly with your magazine generously publishing the occasional Fedora CD, we’re providing close to 4 CDs of source code ready to change the world. Out of India’s population of 10 million people, is there not one Linus Torvalds?

Q: Any message for the developer community? What’s going through my mind is like a guitar player standing on stage asking, “Are you ready to ROCK?”. And waiting for the developers to yell back. If all I can hear is a cricket, then that means I’m in the wrong auditorium. I know the developers are there somewhere.

MT: We will know that the developer community has really come of age when not only have we identified the great developer that everybody respects. But we also have the egotistical one, and we have the one that is mostly wrong, and we have the one who is shilling for somebody else. You know, the whole mix. That is just the way the world is. I’m willing to suffer many fools to be able to talk to one wise person.

Q: What is your message to Sun? Sun’s current rhetoric seems very divisive of the open source community.

MT: To say the same thing in a different way. Sun is externalising their own schizophrenia to the open source community. That kind of contagiousness is not necessarily good. I would say two things to Sun. The first thing I would say is — IBM, HP, Dell are all proving that you can sell billions of dollars of hardware with a good open source partner and a good open source strategy. I know there are customers that would be happy to buy Sun if they had a better story with respect to their own Linux strategy. The second thing I would say is — if they want to fly they need to let go of the ground.

Q: In the Indian scenario, where resources are limited and an engineer might be the only bread earner of the family, letting go of the ground could be scary.

MT: You are making an argument of averages without looking at the true population. At an average level, you are absolutely right. The average engineer cannot afford to make that leap. But when you look at the standard deviation across the whole population, I guarantee that there is an engineer who could walk home that night and say let’s do it, let’s rock! And their family would support them.

…if we [Red Hat] are able to achieve great margins while offering choice and freedom, I think that’s the new generation of technology company. And just in time. Look at the technology challenges that India is facing. This revolution can’t come too soon.

Q: And that’s the engineer who will become the next Linus Torvalds from India.

MT: That’s right.

Q: Red Hat is sometimes perceived, rightly or wrongly, especially after its withdrawal from the desktop retail market when it introduced Fedora, as being more interested in corporate success than in community well-being. Can you address that concern?

MT: If we gave a wrong impression last year when we launched Fedora, shame on us. We have been doing everything that we can, to better represent what we are trying to do. For example, look at Stateless Linux which we announced about 3 weeks ago as a way of providing secure managed desktops, through re-architecting everything from the Linux install process to application resources, etc. That received very positive responses from the community. We got about a dozen people from the community saying “I’ve been trying to solve this problem too, here’s how I’ve been doing it, what can you reuse?” So we are thrilled. I think the way we improved the perception of Red Hat was – we stuck to our goals and we decided that it was going to be our patience and dedication that would achieve the results, not marketing spin. Denials in the press don’t really help. It’s taken about a year to transform that community skepticism into trust, but we are very committed and I think that every day more developers are willing to acknowledge that whatever motives they ascribe to Red Hat, the commitment and the outcomes are consistent with their values and therefore, in their eyes, Red Hat is being a good community member. And if we are making a profit and we are growing, good for us.