Mindful of the past and optimistic about the future, Bill Gates opened the Comdex trade show in Las Vegas this year with an overview of Microsoft's efforts to deal with some of today's most-pressing IT problems. In his Nov. 16 speech, Microsoft's chairman and chief software architect talked about anti-spam, security, and search software, and then something new: seamless computing. It was Gates' twentieth consecutive year at Comdex--a milestone he recognized by displaying a slide from 1983 that appeared, as it did back then, backwards. At the time, Gates' father ran the slide projector for his son's presentation. "Now, when we look at what's limiting us, my belief is that the constraints we have to get rid of now, the seams that hold us back, are of a different nature than in the previous two decades," Gates said last week. "The constraints now are much more about pure software challenges." One day after his keynote, Gates talked with InformationWeek senior editor-at-large John Foley about the issues facing Microsoft and its customers. Following is a transcript of the interview.

InformationWeek: You introduced the idea of "seamless computing" in your speech last night. It sounded like some concepts that we've been hearing about for some time. What's new about seamless computing? And what do your customers need to do differently to reach that goal?

Gates: Certainly, the goal is a pretty dramatic change in how IT departments think about customizing software and being able to visualize what's going on, either in terms of business processes or just IT operations state. The current model, where you have to write a lot of code for fairly minor differences between two businesses--it works, [but] it's expensive, it makes it hard to take updates from the main provider, and it makes it hard to have visualizations of what's going on with these different processes. So the ultimate goal of being able to model these things, and have those models be active even as you're doing the development and after the software is deployed, is a pretty dramatic change in how IT departments think of customization.

Now, we'll approach this in evolutionary steps, just like the idea of the automatic data center [where] the software is moving things around between the different servers and automatically finding capacity if it's necessary. That vision will take most of this decade to realize, but even what we've done to date in terms of making it easy to provision systems, whether it's for testing or load balancing--what we already did in Windows Server 2003 is pretty good. So, what's new about seamless computing that people wouldn't have heard before last night? I don't think they would have heard that all the problems that need to be solved are now software problems. I don't think they would have heard that we're using the Web-services architecture to solve problems that they would think of as very different problems. [For example], how you secure information as you're doing collaborative things across the corporate boundary. We're going to use Web services to do that. How you monitor applications to make sure they're working properly and not have this completely separate infrastructure of different management software. That's a use of Web services. How you make--as you move between PCs or your phone and PCs, different devices--all the information show up without a lot of extra work. That's using the Web-services architecture. So Microsoft, in the same way we pushed one approach for the graphical interface, pushed the .Net platform for doing Web-type applications, now we're taking this Web-services architecture and saying it can, in a nice evolutionary way, solve some of these tough problems.

I wanted to be very concrete in the speech. The only thing I showed that was at all futuristic was the "Stuff I've Seen" technology that has influenced the design of Longhorn. Everything else I showed--SMS 2003, ISA [Server] 2004--either is in the market or very soon to be in the market, and that's part of this paradigm I talk about, where we have to make advances like this new version of [Systems Management Server] in order to free up resources and budgets in IT departments so that they're willing to say, "OK, we'll try putting the wireless network everywhere. We'll try the Tablet PCs. We'll try building these new Web-services applications." And we also have to get them to try to feel comfortable with all the security issues: Is their firewall complete enough? Is the way they do updating--are we making it clear enough what they should do and making that simple for them? There are a lot of things that they view as key basics that, [if they] could be solved, it would really open their minds to the ambitious approaches that seamless computing will make available to them in the years ahead.

InformationWeek: On the issue of security, in recent months, Microsoft has talked a lot about what it's doing and what it plans to do. Is there anything more that you're doing that you haven't yet talked about that goes beyond the issue of software updates, firewalls, quality control, because this seems to be a front-of-mind issue among business customers. We're wondering if there's more to it than that.

Gates: There's a lot of pieces. It turns out, going back a few years, E-mail was a problem. And then we, in Outlook and Exchange, showed people how to not propagate executables through E-mail, and E-mail by and large hasn't been a problem since then, because people did that fix. Now it's very clear that some types of updates--and we need to make it dead simple and dead clear which ones--and an ability to audit your firewalls, those would have prevented any of the problems that have come along. And so it's important that we provide [the necessary tools]. We have [Software Update Services] 2.0 that's a subset of [Systems Management Server]. Ninety percent of our enterprise customers license SMS. Many of them were waiting for this new version to do broad deployments of SMS, and now that's out there.

There are other things like moving to smart cards for certain critical access--Microsoft has made itself a model for that--[and] changing the basic mail protocol so you can verify that mail isn't fake, that it really comes from who it appears to come from. There's things that have to be done at the routing level to really make it easier to prevent flooding attacks, where Microsoft and Cisco and the standards groups have to work together. So this is going to be an ongoing thing. We had some great success. The E-mail as a vector thing was brought down dramatically. Now these things that relate to patching, those will be brought down to really the noise level. But password attacks or spoof-mail attacks, there are things like that [where] we have to make sure we shut the door before those cause a problem for people. And giving them tools so they can know [if they're] following the policies. That's all pre-Longhorn stuff. That's all stuff that has such a priority we don't wait until some big release of the OS in order to do it.