Name: David Smith Age: 34 Place of Birth: Boston, MA Position: Co-director of Worldwide Software Research, International Data Corp. (IDC)
Time in Current Position: 1 month (3 years at IDC) Years in the Industry: 14 Car He Drives: 1993 Honda Accord
Reading Habits: "I tend to read magazines and newspapers and short things. The books I read are computer industry-oriented books."
Pet Open Systems Peeve: "Too many people calling me up asking me to define it."
At 34, David Smith is already one of the industry's most widely quoted and trusted UNIX analysts. Since joining the technology research firm International Data Corp. (IDC) in Framingham, MA, three years ago as a research director, he has moved to director of the company's UNIX research program and then, a month ago, to co-director of worldwide software research.
It's a job Smith likes and plans to keep. "I spend a lot of time finding out what vendors are up to, trying to put two and two together to see if it adds up to four, and if it doesn't, flagging it and trying to figure out why," Smith says. "I do a lot more strategy and technology consulting than quantitative analysis. I do a lot of analysis about whether what one vendor tells me jibes with what another vendor tells me. I try to figure out their strategy."
Since his most recent promotion, he's had to spend more time as a manager. "Unfortunately, it isn't nearly as much fun as being an analyst," he notes.
Smith grew up in the Boston area and received his B.A. from Clark University in Worcester, MA, in applied mathematics. After receiving his introduction to computers at Clark, he went on to Boston University, where he earned his master's in computer engineering.
"What got me interested in computers was probably the vast power that seemed to be behind the glass wall, that I was able to use to do things that were quite a bunch of drudgery otherwise," he says. "I was in the sciences and math and saw that computers could help solve things." But his first computer experience didn't get his blood flowing. "I was using punch cards in a Fortran programming class and I disliked it so much that I didn't go near a computer again for at least two years-until they had interactive terminals," he remembers. "Then the interaction was the thing that really got me interested. You could sit down at a terminal, type commands and get a response back immediately and not worry about the operators and the punch cards and the delay. It was quite an enjoyable experience."
Smith's first experience with UNIX came in graduate school when he was taking an operating system class. "The first versions of Berkeley UNIX were starting to be made available to universities," he recalls. "At BU it was the system of choice because the alternative was an IBM mainframe. UNIX really became the system of choice for me quite quickly."
After college, Smith worked for a time as a programmer for a small time-sharing organization in Boston. Then he joined Digital Equipment Corp., a move that helped define his career. "I wanted to work at DEC and no other place," he says. "I lived near there and had worked with their equipment. It was the place to be, and it was an incredibly great environment to work in for a long time. Actually, my first operating system experience was not with UNIX but with a DEC operating system. I saw a lot of similarities once I got exposed to UNIX a couple of years later."
Smith stayed at DEC for eight years, where he was an applications engineer, worked in technical support, field support, marketing and then competitive analysis. It was in his job as a technical support person-in some cases converting software from one vendor's equipment to another's-that he attended his first UniForum trade show, an event that he remembers as triggering his concentration on UNIX. "It was in 1985 in Dallas," Smith recalls, "right around the time Sun introduced NFS [Network File System]." Sun implemented NFS in 1984, allowing different systems to share files with each other.
"I think I started to realize the movement that UNIX was going to have and the impact that it was going to have on the industry. That's when I decided I really wanted to learn a lot about it and be part of that movement. It was the exposure to the excitement in the industry-and not just the technical power but the marketing potential, the opportunities for myself and for vendors, and what it could really do for users. NFS was a very exciting technology, and when I saw that, I instantly knew that it was going to be a huge success. It was then that I really decided I wanted to pursue UNIX in a great way."
From that point on, Smith began campaigning for UNIX at DEC, with only partial success. "I focused on it when I had opportunities and I became very much a part of the internal UNIX community," he says. "I worked a lot back in those days to get people to say the "U word" at Digital. It had its moments of difficulty and it had its moments of great satisfaction, because for some time, Digital did make some progress with UNIX. It had to do with getting some people to see the light on the advantages of open systems, because there were a lot of people at Digital at the time who didn't get it. And I tried to open the eyes of a lot of people to the benefits of it." In addition, Smith worked for the adoption of the Transmission Control Protocol/Internet Protocol (TCP/IP) standard over the Open Systems Interconnection (OSI) networking protocol.
"I think in a lot of ways Digital doesn't get enough credit for their stance on open systems." Smith says. "If you think about some of the things they were saying in the '80s about open systems not being UNIX and about it being about standard interfaces and interoperability, those are a lot of things that the rest of the industry has come around to believe. What they were not very good at is communicating open systems, marketing it, and positioning it. That's unfortunate because I think they had a lot of the right ideas and they always had very technically elegant products." He adds, "DEC gave me a great opportunity to learn the industry and to learn technology. It's a real shame to see what's happening to them now."
It was as a competitive analyst at DEC that Smith realized he wanted to take a new career direction. "It's a real eye-opening position," he says. "It allows you to look at your product and your competition's products objectively. It allows you to look at where you're doing well and where you're not. It just generally opens your eyes to the business world much more. When you're a competitive analyst at a company, you look at information with basically two goals. When you're selling products, you want to say how great they are, even though they might not be, and when you're telling the engineers to build better products, you're telling them how bad the products are, no matter whether they are or not."
Because of the experience, "I really got excited about looking at what's going on in the industry and doing technology analysis-being an independent analyst. It always looked like a really fun job." After two years as a competitive analyst for DEC, he moved to IDC.
Smith finds his biggest joys in operating system analysis. "I really get into operating system technology-the features, the competitive environment, shrink-wrapped operating systems, the Windows NT vs. UNIX debate, standardization, and unification."
About NT, Microsoft's distributed operating system introduced last year, Smith has a caution for UNIX vendors. "Sometimes the UNIX vendors are a little complacent in looking at the risk posed by NT and Microsoft in general," he warns. "Just because NT didn't take off the way a lot of people predicted doesn't mean that it's dead. It tends to be the same people who predicted that it would take over the world who are now saying that it's dead. And I think that in both cases, they're wrong. NT was never in a position to take over the world a year and a half ago when everybody was predicting it. And it's also not in a position to just give up and die right now. I think it's going to be around for awhile and I think it's going to give UNIX a run for its money over the long term."
UNIX unification is going in the right direction, Smith says, even though the moves of the last two years may have come too late. And he advises against waiting for big trade shows before making announcements. "There seems to be this mad scramble to make open systems announcements right around UniForum and UNIX Expo, then they go out and do their own thing. They need to do more than announce. There needs to be a little more cooperation, more long term strategy and actual delivery of things that are promised when they are promised."
COSE, the Common Open Software Environment, "was a great idea, but it focused too much on the desktop, and desktop UNIX is a niche market. The real market for UNIX is on servers, and they didn't address the types of issues that need to be addressed in the server area-for example, systems management, which is very much on the minds of most users today. No vendor has really stepped up and delivered what the clients are looking for."
One of Smith's favorite current topics is an analysis of whether the software industry and the computer industry in general is consolidating or diversifying. "I'm leaning toward the theory that for it to continue to diversify, it has to consolidate a little more first," he says. "It's kind of like the Big Bang theory of the universe creation. In the beginning, there was the Big Bang and the universe was expanding, and we don't know whether it's going to continue to expand forever or whether at some point it's going to contract, go all the way back to the Big Bang, and start expanding again."