INTERVIEW: Robert Manchise, security scientist

About 50 percent of our employees do information technology work, and the rest do engineering. A lot of the first 50 percent are building IT solutions, from requirements all the way to general design, detailed design and implementation, fielding and then operation and maintenance of large systems. That includes electronic business and systems integration. The engineering side does a lot of custom development for the military. A current project is the Joint Logistics Warfare Initiative, and Anteon is the prime contractor. We're developing a prototype for the future logistics capability for the joint forces.In past years, computer systems were large mainframes with character-based screens. Then 10 years ago came the client-server model, where we had PCs and Unix servers. And now the model is the Web portal and Internet paradigm. We're designing future logistics systems around that. Yes, we do, and computer security and privacy of data just happen to be personal concerns of mine. It amazes me that after we've spent billions of dollars on computer security, we still find 14-year-olds breaking into our systems.Twenty years ago, you had a special cable going from the mainframe to your terminal, and data security was a lot easier to maintain. Now things go over the Internet, and the Internet was designed to be open. There's a major push to move into electronic commerce within civilian agencies. The Postal Service's Web-based purchasing application [] is one, and the Federal Emergency Management Agency is another for which we're building an e-commerce solution.Anteon has a large development effort with FEMA. We built a client-server system in 1996, and it is gradually being converted from client-server to a Web-based architecture. It's basically like a large insurance company application for disaster management. Not only does the system track the potential for disasters'storms in the Atlantic, snowfall in the Rocky Mountains, snowmelt in the Great Plains and forest fires in the West'but it also predicts when the disasters might occur.When they do occur, components of the system use Global Positioning System satellite data to build maps of the areas.The accounting part of the system tracks insurance claims. It has interfaces to the Treasury Department, which pays out the money. And it tries to come up with ways to keep disasters from occurring again. Let's take the second question first: How do you think the public is going to accept electronic signatures? My prediction is: very slowly. They're skeptical, and rightly so, because very few people understand it.How is the government going to implement it? I predict we're going to see a slow rollout of systems that have electronic-signature capabilities. Then, agencies are going to find that software from Vendor B may not work with software from Vendor A. The problems need to be worked out. Three things come to mind. One, wireless devices are going to become more popular and more ubiquitous. As handheld devices become smaller and more powerful'and they're doing that almost every other month now'we're going to see a lot of infrastructure built around them.Another area is computer security.The third one is something that's a concern to me, and that's the security of personal data. I think the government needs to come up with a policy on how we're going to protect that. Concerns about the amount of personal data out on the Internet are keeping people from adopting technology. The companies that are building wireless devices are building security into them.If there is no security on the wireless transmission, could I pick up the data stream from your wireless device? Absolutely.But, remember how that was a concern with cordless phones five years ago, and now it's not? That's because cordless phones these days are frequency-hopping across the spectrum and embedding data encryption. Manufacturers are doing the same things with wireless data devices. First, though Linux is considered an open-source product, we at Anteon recommend getting the Linux operating system from a company that takes the open source code and then maintains the configurations and produces a commercially available product and supports it.The most popular thing that's been around for several years is to run Web servers for Linux, such as the Apache Software Foundation's free product.We're also seeing business applications on Linux, but there are some significant downsides. Because Linux is not as broadly implemented as other operating systems, we find that not many applications run on it. The Apache Web server is an exception. There are a few office or productivity applications for word processing or e-mail, but they're nowhere near as popular as Microsoft Office.Another downside that we have found to be significant is the low level of hardware drivers. You generally have to write those drivers yourself. Industry is product-driven, so we don't build products unless we feel there is some profit in building them. We often will only build products that we can sell, but that doesn't necessarily mean it's the right thing to do.Now, how do you distinguish what should be financed by industry, independent labs or government grants to universities? That's hard to say. It depends on the state of the technology.If it weren't for the Defense Advanced Research Projects Agency, we might not have had networking as quickly as we did. So I think there's a place for government financing of pure research.