Michael Tiemann's early work with GNU software created world-leading technologies, becoming an inspiration to Linus Torvalds and an enabling technology for Linux in 1991. He continues to shape the future of open source today as chief technical officer at Red Hat. Prior to joining Red Hat, Tiemann co-founded and served as acting CTO at Cygnus Solutions, which was acquired by Red Hat in January. Tiemann sat down with Editor at Large Ed Scannell to talk about his work at Red Hat and the issues facing Linux in today's marketplace.
InfoWorld: What is your role at Red Hat as CTO?
Tiemann: My role is to basically provide both communication to users about what we are doing and what is happening in the open-source space. It is also taking information from users and from the marketplace and distilling it down so that our product and engineering teams can really understand what needs to be built.
InfoWorld: Do you get involved with the other top officers of the company in terms of formulating a business strategy?
Tiemann: Absolutely. In fact, that is one of the things that Red Hat got when they bought Cygnus Solutions. Few people realize how small Red Hat was [and] it has grown in terms of head count by more than a factor of 10. Cygnus was the first open-source company founded in 1989. We were 5 years old when Red Hat was founded. So we have been around the block more than a few times on business models and looking at what the operational issues are, what the scalability issues are as you build an open-source company. I bring to the table the knowledge from that experience of helping to formulate the business plans for Red Hat going forward.
InfoWorld: What percentage of your revenue comes from services as opposed to products?
Tiemann: On the Cygnus side, it was about 80 percent services-based. On the Red Hat side, it was completely the opposite; it was almost all products. Bringing the two companies together has actually given us something nice because there are some [end-user] companies out there that want to buy a blend. They want to have a product platform which is standard and acts as a reference platform. Red Hat has done a great job of making Red Hat Linux, the reference platform for Linux. It makes it easy for CIOs and CTOs to just worry about one variable in the Linux space. It is a very comforting thing for them. But on the other hand, they want customization, services, and support. Cygnus has built about 100-plus engineering teams who are able to deliver those kinds of services. So with the merger, people do not have to say, 'I am going to go with' a services-based model or a product-based model. Now they can get the reference platform with services on top of it.
InfoWorld: What do you think is the biggest misperception about Linux in the industry today?
Tiemann: Just this week I read a comment from an analyst that made me see red. Many of them are using the wrong yardstick by which to measure Linux. The comment was about how Linux doesn't have clustering, high availability, volume manager, and file system support. That sort of comment ignores the Red Hat-Veritas relationship, the Red Hat-Oracle relationship, and our clustering technology that we developed several years ago and that several others have reimplemented or rebranded in various ways.
InfoWorld: So it's basically overlooking the collective technical strength of the Linux community?
Tiemann: Some analysts are making asssumptions about what Linux can't do, and this completely misses the point, which is that with open-source technologies, anything can be done. I was at the Intel Developer Forum, where the big news there was the IA-64 architecture. As I wandered around looking at all the systems demonstrating this processor, every one I saw was running Red Hat Linux. Not too many were [running] Windows 2000. When you see the SGI [Silicon Graphics Inc.] running cluster machines and IBM running parallel servers, you begin to get the idea that any [user] can do this tomorrow.
InfoWorld: How important is the IBM relationship with the Linux community?
Tiemann: I think it can be very important. IBM has technical experts in a variety of areas and can offer the Linux community some leading-edge capabilities in terms of things like journaling files systems, clustering technology. And they make some pretty cool processors and silicon technology. Linux and open source is all about being best of the best, and so I am excited to see what will happen as IBM makes its copper-based and silicon-on-insulator-based high-density processors available to the Linux community.
InfoWorld: John M. Thompson, who heads up IBM's software business, aggressively supports Linux and plans to continue that support. But asked if he plans to make his Linux products open source he says, 'No, because we have to make money somehow.'
Tiemann: What I say about that is never say, 'Never.' If you went to sleep five years ago and just woke up today, and you saw that SGI was going through their most valuable operating system technology portfolio and actually open-sourcing virtually everything, then you saw what IBM was contributing in terms of putting world-class compiler technology and Java technology into the open-source community, you would really wonder if someone had put something in your drink. From our perspective, IBM stands to enjoy a huge upside, with the right kind of focus on not just Linux but with open source as well
InfoWorld: When you deal with larger end-user companies, who do you first go to inside those companies?
Tiemann: I often talk first with either the CTO, the CEO, or the vice president of engineering. Although in larger enterprise cases I might also meet with the CIO. The conversation we typically have involves projects in the $2 to $10 million range. Often the reason for these face-to-face meetings is to talk about a fundamental re-engineering that takes full advantage of open source. We are not trying to replace one kind of sweetener with another here. What we are presenting them with is, if they really want to achieve a dramatic improvement in ability to deploy infrastructure, they have to take a hard look at the open-source model.
InfoWorld: Is that your hardest sell, doing the missionary work for open source?
Tiemann: Well, here is what makes it not such a hard sell lately. People have been banging their heads against the walls with the best that proprietary companies have had to offer for the last five or 10 years. The fact that they are facing the same problems in the year 2000 that they were facing in 1990 basically informs them that life is not going to get any better with conventional technologies.
InfoWorld: Most of those running large IT shops are heads down on delivering projects on time and under budget with the technologies they have had for years. Largely, they do not seem to have the time too focus on open-source solutions.
Tiemann: That heads-down attitude was the right attitude in 1990 when the stock market was going nowhere, and it didn't really matter if you were 10 percent ahead or behind everyone else. But today, big money is moving to the people who can rapidly deploy infrastructure. Where would Cisco or AOL be if they couldn't continue their growth rate today?