Internet World 1996
remarks by Bill Gates , Tuesday, April 30, 1996

MR. GATES: Good morning. Well, it's a great pleasure to be able to participate, even though I'm not there in person. In the future, I think we'll conduct this across the Internet and not have to worry about satellite scheduling, but maybe it'll be a year or two before we have the right video quality.

What I want to talk about this morning is bringing together the PC, the Intranet and the Internet, and paint a picture of how we see this evolving over the next couple of years. The Internet phenomenon is really unbelievable. It's the most fantastic thing to happen in the world of computing since the original PC. The level of investment, the speed of innovation, all of this is really accelerating achieving the original vision -- making it easy for everyone to get at information.

The impact here, of course, goes way beyond the computer industry itself. It really is the revolution in communications that deserves to be compared with advances like television or radio or the telephone. In many ways, this brings so much more to commerce, to education, to people reaching out and finding other people's common interests, that I think it's even more important than those previous revolutions.

Now, what we're doing here is we're building this on things that have come before. People often ask, "Why did it happen just in the last couple of years?" Well, I think many elements came together. The most critical is the reduction in the price of communications -- the use of optic fiber, advanced switchers, more and more asynchronous transfer modes really allowed us to get prices to the point where this idea of the single global network, where you're not tariffed depending on how long you use it or the distance that you're connected at, that became possible. And so we saw a critical-mass phenomenon -- as soon as we had enough content out there, we got enough users that then it made sense to get more and more content, which led to more and more users.

If you think of the Internet six months ago and think of it today, the delta's been pretty fantastic. And I think that kind of rapid improvement will continue. And companies are beginning to wonder, how do those pieces fit together? They've made large investments over the last decade in buying personal computers, connecting them together with the local area network, and providing productivity software to their users so that people can offer content, view data in very rich ways, and even personalize the way they go about doing their planning and decision-making processes. And what we need to do is take that PC revolution and this Internet revolution and make sure those two absolutely work together.

Now, all of the Internet standards -- and there are so many acronyms here, it's pretty unbelievable -- all of those need to be embedded deeply in the system. They need to be part of the standard working set of the operating system. And so whenever you're looking at help files or whenever you're navigating around trying to find information, you have the code that implements these standards as part of the working set.

I think the proper combination of taking those investments that are out there and taking what's going on today, and keeping them coupled together so we have one rich set of standards for distributed programming, for audio, for video, for security, for authoring -- by bringing those together, we can really achieve the goal of information at your fingertips.

Now, there's a lot of different scenarios, and each one of these has incredible business potential. If you can take decision support within a company and make that more effective, the leverage is dramatic. If you can take business-to-business transactions and allow those to proceed without paperwork and with constant information flow, the leverage there is incredible.

Companies spend a lot of money today publishing information, putting out paper brochures or answering phone calls. And by making that information easy to access on the Web, the business can do a better job, get the information out there faster, and do it with far lower costs.

We're already seeing on our Web site today that in our developer area, a typical user goes in and spends over an hour browsing that area. And we've already seen the demand for paper collateral go down, because now people find it so easy to go out there and get information. Even the technical support calls we see in that area have gone down, and we think that's directly attributable to the investment we've made in that Web site.

Group collaboration is something that will become very strong on the Web with the ability to not only have audio, but also to share applications. For example, if you're negotiating a contract, you'll be able to use whatever your Windows application is, bring that up and share with somebody else on the Web and talk to them while you're editing together.

Customer communication, making sure that you make quick decisions for customers, that you bring together in one place all the information about the interaction with that customer, and make sure that everyone in the organization knows whatever the latest developments are, the value of those systems is really hard to exaggerate.

So in all of these scenarios, there has been work that predates the Internet. We have EDI for business transactions. We have great productivity software for decision support. We have lots and lots of authoring tools.

Now, being able to take that investment and those standards and bring them together with the Internet is something that can be done very rapidly. And, in fact, it won't require people to go back and make very large investments. The network connections are there. They don't need to be changed, although if people move to audio and video, they'll probably want to move up to higher speeds. The PCs are there. The software foundation is there.

In many cases, this is just a case of publishing pages to make the navigation easier than it has been before. You know, we can contrast in a company what you had to do to go and get, say, sales information just a year ago with what it would be today. A year ago you needed to know a server name, you needed to type that in, you needed to know the file names.

If the files had moved, then you were probably going to pick up the phone and call somebody. If you were confused about the format of those files, likewise you'd probably pick up the phone. Or you just wouldn't access that information in an electronic way. You'd say, "Well, I'll wait till the printout comes around," or just try and make the decision without having that information.

Now, by having pages that are very self-descriptive and let you navigate through the network without even knowing what servers are involved, without even knowing if you're going halfway around the world to get that information, we turn decision support into a few clicks of the mouse. Now, we have to make sure to retain the rich views, pivot tables, recalculation, graphing and all of those things that are down at the lead notes as you do this navigation, and we'll do this by integrating together productivity software with these Internet standards.

So let's take a look at what all the components are that come together to make this happen. As I said, on the hardware side, there's not much new here -- higher connectivity speeds, but that's about it. The pieces I'm showing here are software pieces. Information access -- which is the software that runs on the local machine, often thought of as just the browser but becoming far broader than that. Information management -- the code that runs up on the server, which is turning more and more into not just static pages served up with HTTP, but rather a connection protocol that allows you to connect up to data stored in a variety of forms. And in particular, it lets you access data that's stored in relational data bases.

And so, from the store point of view, you have the richness of the query speed, the on-line backup, the integrity that comes with that rich store. And yet you can access all of that simply by having a client that has browser-type capability. And so the picture on the server is more than just files and the HTTP server. In fact, that's becoming richer and richer all the time.

Now, at the bottom I'm showing the authoring pieces. And, in fact, when you author content, there's quite a spectrum between simply creating text or animation all the way to writing lots and lots of code. And part of the beauty of the Web standards, particularly how HTML is being extended now with things like our active architecture and our add-in capabilities, is that code and content work together seamlessly. So when you see content, the fact that code is there animating it and coming up with the latest information, the person who's browsing doesn't have to think about that.

However, the actual creation process will probably have specialization, where the people who write the code and the people who actually do the content will be able to be experts in their own areas, and simply to use drag-and-drop capabilities to take components and include those into the content. And so we have content authoring tools and application development tools that allow this integration to take place.

Now, binding this all together, there are some common infrastructure pieces that I want to talk about, because it's there that we see rapid advances in the standards, particularly in directory and security. And those are fundamental to the scenarios that we talk about.

Now let's focus on the client. Internet browsers are by far the fastest-growing piece of software that I've ever heard of. They are significantly larger today than a typical productivity application, something like a spreadsheet or a word processor. And every three or four months, whether it's a richer run-time or better multimedia support, there seems to be a lot going on.

In some ways, the browser is becoming almost an operating system itself. People talk about print management, memory management, even local store formats coming into the browser. And so in a little bit, if we're not careful, we'll basically have a whole new operating system with all the things that go along with that, sitting on top of existing operating systems if we don't do the integration in the right way.

We also want to avoid the notion that the only way to browse the Internet is by using the browser itself. The browser, in fact, should hide behind the scenes. It should be there to render extended HTML pages, but if what you want on your screen is a spread sheet-type display, if you want a project management display or any type of display that an application can make, you should have that flexibility. And all of the capability of going to your favorite location or going backwards and forwards, that should be completely independent of whether the page is being rendered by the browser or by some other application or object manager that happens to run in your system.

Now, one of the things we're bringing to this is the idea of common navigation so that the information you store locally on your hard disk can be navigated in exactly the same way you browse the Intranet or Internet. In the past, the way you browsed the local information has been fairly primitive. We just moved up to long file names and so-called shortcuts to let you navigate around. But now, with these directory pages being HTML pages, you can have a lot of description and even active control that, as you browse around, give you hints on where you might want to go.

So by taking HTML rendering, building that into the shell and allowing all these directories to have HTML, then you have a seamless experience. You're not seeing something different when you're running locally versus when you go out onto the Web. Likewise, the same kind of search power is there, the kind of hierarchy that we've seen on the Internet with things like Yahoo, the kind of search capability we've seen on the Internet with things like AltaVista, all of that can apply to local storage and corporate-type storage.

Now, to make this work, we have to have hyperlinking standards, so that if you go from, say, a spreadsheet to a browser page to a word document to a data base, the applications will present those links and record the history in exactly the same way. And so that hyperlinking is being built into the Windows operating system and being adopted by all of the applications.

I want to emphasize the notion of personalization. I don't think in the Web, going out a year or two, that everybody going to a site will be seeing the same thing. If you've been in and seen the data before and you've set up a view of your own, that should be preserved. Depending on the types of preferences you have, the type of customer you have, that display should be generated for you. And that applies to information no matter where it is. The views are going to get richer and richer.

In fact, to give an example, if you're looking at a set of messages, you'll want to be able to, on an ad hoc basis, filter that by subject or priority, and that dynamic generation does require code. And now, with the new distributed architecture, that code can either run on the client, which is very important for example, in the portable machine scenario, or can run on the server, which is very important in cases where you don't want to download any client code so you just ask the server to do the work.

The balance here is tricky because if you're not careful, you can overload the server. You can go back to a time-sharing type situation where you're not getting great responsiveness. But the architectural flexibility allows us to use the client, where that makes sense, or to use the server where that's important.

Now let's talk about the content offering tools. The variety here is really quite amazing. All of these tools, whether it's paint tools, movie editing tools or animation tools, all of them are being enabled through the Web. They are being set up so they can create controls that execute inside the HTML pages, and have the same type of standards for navigation that is being used for the pure text.

Now, when we're talking about content offering, it's important to include in this the ability to collaborate, allowing people to simultaneously do editing and to look at edits and see who made those changes. Also, the idea of versioning is very, very important here. You want to be able to say, "What did our Web site look like two weeks ago," and have a history that's preserved so you can roll back and roll forward through those kinds of changes.

Now, the world of paper has some disadvantages, but one advantage is that preserving a record of history is very easy, and on the Web today we don't have that easy preservation that often would be very, very valuable.

Now, as you build Web documents, you're going to find that the number of pages in the links start to create a unique management problem. There are a number of new products coming out, including the Front Page product we have, that are starting to give people different views to make sure all the links tie together, or the ability to see how those links will be navigable as you're editing those documents.

And so we talk about the idea of Web documents as a set of not only pages, but files in any format that have these links, and the editing interfaces that make it easy to see what's going on to make sure there's no problem. You'll see a lot of evolution there. That's become very, very important.

A lot of times when people think of the Web, they're thinking about content read-only behavior and just looking at pages. For the broad Internet, that's a high percentage of the behavior. But particularly as you move into the Intranet, and even in some cases for the Internet offering is crucial and you want to give people great flexibility there. People are creating content, including drawing on some of the content that's already out there and pulling that together. Not only linking it, but embedding it in, putting their own annotations on pages, and today the Web doesn't have great standards for version control of annotation and I'm sure those will come along as part of the new protocols.

As we look at code, here we can have the best of both worlds. We can take existing code and make that work and we can have new code, whether it's written with object extensions to C++ or with new approaches like Java. We see Java as a great development. We think the number of languages that will be popular for application development will continue to be fairly large. Some people will be using Cobol even five and ten years from now. C will continue to be very important as will Visual Basic. Java will be a first-class language that will fit into the same integrated development environment.

Now, one way to implement this is to have real-time compilation, what we call just-in-time compilation for interpretive data. It's not just Java that can create interpretive byte codes; in fact, we've been doing that with Visual Basic for a long time. So for any of these languages where you use the byte codes to get density and process independence, you can have on-the-fly compilation to gain back some of the speed.

Now, you'll never gain back nearly as much as when you have a full-blown compiler, but for all of these languages, if you're willing to bind to a particular processor, you'll have that option. In fact, with the way code is stored up on servers on the Internet, you'll be able to dynamically determine if the client has one of the processor types you've done a full compilation for, and if so, you can bring down the control that works that way. If it's some other type, then fine, you download the byte codes and those go through the just-in-time compilation process, providing adequate speed, but not as good as the full compilation.

Now, the Web finally has people looking hard at distributed development. There's been a lot of talk over the years about distributed objects, but here, it really becomes necessary because you want to bind some of the controls that the user is using to navigate around. You want to bind those up to application objects up on the server, which are then bound to rich stores like relational databases. And so, the standards there are moving forward very, very quickly supporting all the different languages.

A critical element for application development is the idea of code signing. We're not going to be able to restrain code to be mathematically safe in the sense that it can never write to a file or can never change any kind of permanent storage, because if you have those restrictions, then you can't do what the user wants you to do.

The way to have the best of both worlds here is to have code signed. You can indicate who you trust, either vendors or developers within your shop so that if they've signed it you're willing to have that code run on your machine. And so, that takes a step beyond where the Internet is today where if you're downloading arbitrary code to render an object or a new version of a browser application, you're not really sure if it's been tampered with. And so, the code signing piece will come in and be critical, and we're certainly pushing that in the standards committee as part of the infrastructure.

I mentioned that the server is a key part of this picture. We are going to have lots and lots of code running on the server. We're not going to go to the extreme where you always run everything on the server because that has some serious drawbacks in terms of latency and cost overload. If you try and put all your computer power on the server, you'll find that there's non-linear costs to trying to create that capacity compared to distributing it out to individual machines.

However, the server is the best place to keep data stored and it's the place where you can just go ahead and generate HTML so all you have to have is a browser to get at this information. The variety of storers up there -- message storers, media server-type storers, bulletin boards -- all those different things can easily be connected up so that the browser thinks that it's just navigating links. It's just passing URLs and getting down through HTTP the HTML pages.

So the Web server knows how to map those into different files, but it works through a connector to get at the other information. And there are now standards for what those connector APIs look like and how that makes it easy to write these applications.

Some of the great demonstrations now are showing that with just a few minutes work, anybody who has a browser can go into a relational database and navigate the information in a way that's as rich as any tool provided in the past. So that kind of reporting and display logic that used to be on the client has now moved up to the server. And so, we're taking tools like Access and moving that code up there so it runs in that fashion.

Now, common infrastructure is crucial here. We need to have the notion of a directory. There are some directory standards for the Internet today, but they don't go far enough. We're very glad to see people coalescing around LDAT, but we think for richer object types, more properties and better ways of browsing those things, we're going to have to go a lot further.

Within the Windows family we have started to use the directory for all applications, so you don't have a different one for the database than you have for the mail system, which is different from the one you have for protecting files or printing. All of that is brought together in one place with this common, very rich user object. And so, we think the directory plays a very key role.

We also think that encryption and signing will be very important. One of the applications I mentioned is code signing, but that's just for authentication to know who's at the other end of the connection is very, very important. Another big step we're taking here is publishing what we call a Crypto API. That's part of the 4.0 version of NT that ships this summer. We also have it on Windows 95, as well as on the Mac and Unix.

Now, this Crypto API allows people to plug in different types of encryption and yet lets the application use those APIs to be able to get at anything that's underneath. This is a very important step. We can't have encryption married to just the browser. We have to have it as a common service that can be used.

As we move to the next version of the Crypto API, a big step there will be making them so that you can store certificates in a way that's application independent. All the different applications need to be able to get at these certificates, and so the certificates can be used to determine, say, what your banking privileges are, as well as what your file-sharing privileges are or mail privileges are.

It's clear that the X-509 certificates and that format will be a key part of this, and again, the standards committees are a critical place to work all of this out. We take these standards and we build them into the Windows family, and so applications developers find it very straight-forward to use these things.

In the case of transactions, there's been great progress there. The SET protocols that many companies worked together to come up with will be in products that come out this summer. We'll have those in what we call our commerce server and our merchant server that's up and running on NT. All the other participants are doing their implementations and there will be lots of testing for interoperability.

We really have gotten to the point where people should feel comfortable about using a credit card across the Internet as they do, say, to make a purchase across the telephone.

Well, now to show how a lot of these pieces can come together, I'm going to ask my helpers who are down there at the show, Craig Fiebig and Steve Guggenheimer to come on stage and give you a look at a scenario we think shows some of the potential here.

MR. FIEBIG: Thanks, Bill. I'm Craig Fiebig. I'm the group product manager for Internet servers at Microsoft, and this is Steve Guggenheimer, product manager for Internet Explorer. What we wanted to do was to talk about some of the technologies that Bill has just described, specifically with respect to what customers are giving us feedback on in terms of the directions that they think that we need to take our products.

From the client side, we see that there are classes of problems that need to be addressed and need to be addressed in a relatively consistent, but also powerful way. It comes down to four basic capabilities that you need to be able to view the information. You think of that as browsing. But perhaps more interestingly, once you're viewing the information, what capability do you have to do analysis on the data that's been presented to you. And then obviously, you may want to edit it or you may be in the mode to create that information.

The sources of the information itself may be indeterminate at the outset. You can't assume that you're going to be tagging or looking for information of a specific data type in a specific data storer. So as you're creating or analyzing or editing information, you may be looking at things that are coming from a messaging or groupware system or a sequel database back end or a Web server.

And a couple of the pieces that I think are becoming increasingly important are this ability to work within the context of information that may be stored on your existing host, or many computer systems, as well as the wide range of document files that exist on local area networks today. This needs to occur within the context of the infrastructure that Bill just described.

All of these pieces have to be in place for us to be able to develop solutions that have the capability to integrate what we think of as the Intranet with the Internet. Our proposal is that Microsoft will seek to continue to innovate around the products that we have, driving some of these capabilities into the families of products that you see here on the slide.

We work inside the concept of a family of products, but we make sure that that family is a complete family of products and that it consists of best of breed applications throughout the suite itself. On the desktop, we're going to look at how we're using the operating system in conjunction with Office and Front Page to create a Web site working with some of the data storers who we have listed on the right-hand side.

The way that we'll do that is we'll use a scenario from Windermere Realty. Windermere is one of the largest real estate agencies in the Pacific Northwest, and they have a functioning Web site today at www.windermere.com that is built on Microsoft technology. What we're going to do is give you an example of how we've extended that current capability. We have a couple of different scenarios that we'll look at from internal users and someone who's managing the site, accessing information, using the Internet information server as the point of consistency to be able to obtain data from a variety of back-end data stores and then also demonstrate how we would incorporate that with their requirement that customers have access to similar information.

So, with that, I'll turn it over to Steve Guggenheimer, and he'll take us through the scenarios.

MR. GUGGENHEIMER: Great. Thanks, Craig.

So what I want to do today, as Craig said, is take what Windermere has today and expand upon it, leveraging their existing tools and technologies along with some of our existing and new technologies. I've broken it into three scenarios.

The first thing I want to do is be a real estate agent for Windermere and show how I can browse around and get my work done over our company's Intranet directly from within the browser.

After that, I want to put on the hat of the office manager, where I'm responsible for both creating content and managing the content for our Web site, and managing our local area network.

And lastly I want to take a look at the site, as an external user would see it, and show how some of the Active X technologies have been employed to make the site more useful and more interactive.

And so what you can see here on the screen right now is the Windermere Web site. And I just wanted to show you this is where we are going to come back to. And I want to build to this from the inside out.

So let me close this down. Let me put on my real estate agent hat, and let me go to work. The first thing I do when I go to work each day is I check my mail. I look at my in box. Today is a good day -- I only have one message -- so that makes it fairly straightforward. And what I've got is an open page saying the new regional sales report is on the Windermere Internet site in the "What's New" section. And he's actually created a link for me that I can click on, which will take me straight to that page on the Web site using the Internet Explorer. So the nice thing is Dave have to do anything to create that link. He just typed "HTTP" and the address, and Exchange was smart enough to create the link automatically.

Now, once I'm here he said go to the "What's New" section. So I'll go ahead and click on "What's New" and let it go out and browse for a server. Now, we're on our Internet, which is all set up back here in the back. So you know what we have is a full set with Internet information server, an Exchange back in and a Sequel back end. And we'll go out to the regional sales support, as he suggested.

Now, what we are going to do here -- we're using Internet Explorer 3.0, which supports Active X technology. And one of the things it enables is for me to seamlessly integrate document types with the browser. So for those applications that write the document object spec, I can seamlessly integrate those within the browser. So, for our company, that means our office manager doesn't have to go and save all the spreadsheets they have as HTML, but rather I can browse seamlessly to them from within my browser.

And sure enough you see here our Excel spreadsheet. It's got the monthly report and I can use all the rich analysis capabilities associated with Excel right from within the browser. And I can see that, by "Sales Rep," I'm having an okay month. But let me compare. Let me use visit tables to drag the "Sales Rep" down here, and take a look at how I'm doing relative to everyone. And I see that I'm not doing so well. I better go out and sell more houses.

So with that what I'll do is I'll navigate backwards. And, again, I've got the nice blending of the Internet Explorer and Excel, so I can seamlessly navigate backwards. I'm not going to save any changes. And with that I'm going to head back to the beginning, because I think I need to work on selling property. So I'll go out to my property listings area and bring down the price of one of my houses. I know I've had one on the market for quite a while now, and maybe if I reduce the price I can sell it more quickly.

Now, what's happened is I've gone out, and it's looked at my user name, and it's gone to our Sequel database and pulled up all the listings I have automatically. And we see the exclusive Broadhurst home is listed for $439,000. Maybe if I drop it $10,000 I can stimulate the sale a little better. So, again, I am now working on our Sequel database, but I'm doing it all from within forms on the browser. So it's much easier for me. I never had to open Sequel or know anything about it.

Let's go ahead and bring the price down to $429,000 and update the listing. and we'll bring it up again. Let's go to the editor and property listings. We know the scenario: We want to reduce the price of this home. Let's bring it down to $429,000 again. And we'll go ahead down to the bottom, update the listing, and click here to return, so we'll head back in. And I'll refresh this page, and we'll notice that not only was I able to update the Sequel database from within my browser within a form, but Sequel is smart enough to know to redraw -- when I redraw the HTML page, to update it. So it can dynamically create HTML pages on the fly, so that I can make changes as an agent to my listing, without having to go tell the admin or whoever to go update the HTML page.

MR. FEBIG: Now, can you change everybody else's listings?

MR. GUGGENHEIMER: No. That's a good question. When I logged onto the computer, it used NT security to authenticate me and let me go through. So I only have access to the pages and the information I should have access to, and nobody else's Sequel database information.

So with that there is one other thing I need to do. I know that when we make changes we need to let everybody else in the office know we've done this. Now, the easiest way to do this is to go ahead and go out to the public folders area on our Exchange server and post a note. But instead of opening up Exchange, I'm simply going to go to the public folders, again from within my browser. We have used the Web connector for Exchange to enable me to get to all the public folders from within my browser. So that enables me to seamlessly go from Sequel data out to Exchange data.

Let's go to the changes to the listings area. I see there are several postings there now, and I'm going to add one. That allows me to let everybody know that I've updated the price on the Windermere home. So we'll give this a second. Okay, so this is from Steve, and it's price change, and it's now 429K. So I'll go ahead and send the post, check the page, and there we see the price has already been updated.

So I think we've seen a pretty good idea of how I can seamlessly navigate across our Intranet to document type, to Sequel storage information, and to Exchange information, all from within the browser, and all seamlessly to me as an end user.

So with that let's go ahead and switch over to my other machine. Now I want to put on the hat of the administrator for the office. My job in this role is to update and create content for our Web site, manage the Web site, and manage our local area network. Since we're running obviously off the NT server, it's a fairly straightforward task, and I can do it all myself.

The first thing I want to do is bring up Word. And I've been working on a document that I want to go ahead and create as an HTML file for all of our users. So what we've been able to do is use the rich formatting capabilities of Word in the creation of documents. But now I have the capability to save them out as HTML files. (Phone rings.) Hmm, incoming call from Internet World. All right, well, I happen to have conferencing open, so let's go ahead and accept the call and see who it is.

MR. GATES: Steve, it's Bill up here.

MR. GUGGENHEIMER: Hi, Bill.

MR. GATES: I thought I'd help you out editing that document.

MR. GUGGENHEIMER: Okay, well, let me go ahead and share that out for you, Bill, and then I'll minimize the screen, and you can take control.

MR. GATES: So what we are doing is running across the Internet here, and it's very easy for me to show Steve how you should do this right. In fact, what I am going to do here is go ahead and use Word, insert a little picture. I think we have a little moving van out there that would look good in this document. Yeah. And I don't think that is quite big enough. Let me just grab that, see if I can't make it a little bigger.

MR. GUGGENHEIMER: I think you've got it in the server. That's probably as far as we can go right now. (Laughter.)

MR. GATES: You don't want me to draw on that spot for you? (Laughter, applause.)

MR. GUGGENHEIMER: Let me go in and take control back. Go in and close him down. (Laughter). Okay. (Laughter.) So what we've just shown, though, is a fairly important use of the Active X technologies bringing the best of the PC and the Web together. I've shared out my Word document from here, and shared it with Bill, who has been able to take control and edit this document from his desktop. He didn't have to have Word open. We can do this with multiple people, and we can do this all using standards, so other vendors can tie in.

So it's a very powerful piece of technology for Bill to be able, from Redmond, take over control of my document and help me make it look a little better. I think though actually what Bill was trying to do was just grow it out a little bit there. (Laughter, applause.) And with that I want to keep moving forward and actually save this document now as an HTML file as opposed to a document.

So now all of our office applications still have their rich formatting capabilities and editing tools, but we can also save as HTML files as well as their native files. And I'm going to cancel that part and go ahead and save this page for now. I've saved this page. You know, what I want to do is go integrate it with my Web site. So I'll minimize this and bring up the Front Page Explorer, the new member of our office family, and this allows me to put all the pieces of our Web site together. If I make changes to the name of a particular file, it updates to the links. If I want to get the view or the tree view of the piece I can easily navigate out and see what's going on.

So let's go to the "Moving Considerations" page, and let me open up the "Moving Considerations" document so that I can edit the file we're working on. The file we were working on had to do with movers, and movers in our area, that people can contact when they're ready to make a move. Okay, so we see the image on the bottom here. What I'd like to do is link to the new page. But as opposed to creating a text link, I'd like to make the image around that entire moving truck. So within Front Page it has a very powerful editor that allows me, for example, to create an image map here and link it to our browse outdoor site, to the page we were working on earlier.

So let me go down to "Images" -- excuse me, to the Windermere site -- and my "Moving Specialist" page -- click "Okay." Click "Okay." And now we've got a link there. So when I go anywhere on that site it will link, and I can write mouse click on this particular image and follow the link forward to our new page. So now we've integrated it with what we're having. And there's the document I was working on.

Now, there a couple of things I forgot to do. I want to let everyone know that this is for local area only. So let me go in quickly, using our editor, click the "Table" cells and go to two rows instead of three, click "Okay," and let everybody know this is local area only. Let me make sure everybody can see that. So I'll bring it up one font size, give it a little bit of a color so it stands out. Hit "Okay." Go ahead and center it real quick like -- center this. And now I think we've got all the components we need. Let's go ahead and take a look at this out on the Web site itself.

Let's go back to our home page -- and this is where we started before. And let me go into the selling a home section. Now, throughout this site we've used Active X technologies to make it a little bit more animated and interesting. And the first thing you'll notice on the top of this page as it fills in is some Active X animation which I built using Powerpoint because all of our new tools will support the creation of Active X animation and controls.

Okay, so the page is animated. Let me go into the moving area. And we see here's the page we were just looking at -- all nicely integrated. There's our moving truck. It's got a nice large area in the slide, and if I click on that it takes me to the new page, seamlessly integrated, that we've just created. So we've gone ahead and used all of our office tools to not just create documentation, but edit it and add it to our system.

MR. FIEBIG: So it seems like I was able to create a fairly sophisticated Web page without any understanding of HTML at all.

MR. GUGGENHEIMER: That's true. One last thing I need to do as the office admin, before I forget, is we've got a new user starting on Monday, and I want to go ahead and manage the NT server, but I want to do it in a new fashion. I figure that since we are in Santa Clara, what better machine to manage my NT server with than this one, and what better tool to do it with than the Internet Explorer from Macintosh. So this is my Macintosh machine running the new browser. I'm going to go ahead and open the Internet Explorer. And what I want to show you is the new Web-based administration tools for NT server.

MR. FIEBIG: So you're going to manage security on a Windows NT server from a Macintosh through a browser?

MR. GUGGENHEIMER:Yes.

MR. FIEBIG: And maintain NT security?

MR. GUGGENHEIMER: You've got it. So the first thing I want to do -- I'm the admin for the network -- I want to go into account management. And this is the Web-based administrator for NT which allows me from anywhere on the Internet, as long as I have the proper access, to get to and control the functions I would normally do at the server itself. It makes remote management a lot simpler.

So you'll notice as soon as I click on it says, "NT Admin Test," and I actually have to log-in. So it preserves the investment we have in our NT security, but allows me to administrate remotely from anywhere within the organization. We'll click "Okay." And now what I want to do is again I want to add a new user. And you'll see that the user interface has some of the components we're familiar with. For example, if I want to add a new user in the user manager area, I can simply create -- let's say we want to add Fred, Fred Smith -- a nice generic name. We'll go ahead and create that entry. And then we'll head back to the original page, refresh this page, and we'll see that Fred has now been added to our list. So I've been able to remotely administer the NT server from somewhere on the Internet, and in this case using any browser I want to from any machine I want to.

With that, one last key click -- I think we've done a pretty good job of creating and managing our Internet, or Intranet. I want to show our Internet site, okay? So as an end user coming to Windermere, we've worked and used some of the Active X technologies to make the site more useful and more interactive.

So let me bring up the default HTML again, which is our Windermere site. Let me go ahead and refresh this page. And again in this case I've used the Active X animation control to make the page a little more engaging when we first come to it. Okay, now it sounds like a friendly real estate company. Why don't I go look for the properties for sale. Now, I hear you're looking for a new home -- in the Redmond area maybe?

MR. FIEBIG: Yeah, I need to get closer to the office.

MR. GUGGENHEIMER: That's it. You need to work a little more. So let me do a search on Redmond in Washington. Now, the nice thing is this particular page ties back into our Sequel database. Though you don't have access to change anything, we can still use that database for both our internal customers as well as our external customers. And you'll notice that, when I do the query, up comes our broader cell and even reflects the price change. So we've really been able to leverage our investment in our BackOffice for both the Intranet users and the Internet users.

Let's take a look at this house. Now, here we've used the Active Movie control, which allows me to play AVI files, MPEG, Quick Time, stream formats of any form within a Web page. And you'll see here we've got a new format that takes audio and sinks in pictures at a regular time.

So what do you think of the house?

MR. FIEBIG: I think it's a great house, but I'm waiting to see how the Internet gets the price down to something I can afford. (Laughter.)

MR. GUGGENHEIMER: Okay. Now, that particular image is done at a 28.8 speed, so it's set up to be a real useful format that people can use today on their existing sites. So I see that a loan amount's already put in there. Now again we're using Active X technology. Here we have an Active X script-based calculator, which is a loan calculator. I've done it using VB script. I could have created it using Java script. Let me go ahead and give you a 10 percent down payment and do the calculations.

Now, what's happening is it calculated a monthly payment of $3,151, but instead of going to the server each time to get a query on the information, it's actually doing the calculation live, real-time, because this is a real calculator using VB code and simple buttons to make this happen.

In fact, let's make a change. Let's say you're a really good negotiator. You can bring it down to $329,000. You're a marketing guy; you can do that. Bring it down, 10 percent down payment still calculated, and we see your payments are only $2,417 a month.

MR. FIEBIG: Great job.

MR. GUGGENHEIMER: So we've calculated that you can't afford the house.

MR. FIEBIG: Yeah, you could help me get that $100,000 loan.

MR. GUGGENHEIMER: Okay, so, in fact, I think we've covered all of our pieces. We've gone ahead, and as an employee of Windermere, managed and worked on our daily tasks from within the Intranet. We've then gone out and been the administrator for our company, created and managed content for the web site as well as managed our local area network remotely from within the Web -- from within the Internet, and then, lastly, we've used Active X technologies to create a useful and engaging site for our customers.

I'll give it back to you, Craig.

MR. FIEBIG: I think what'd I'd like to do now is turn it back over to Bill and have him take us through some of the innovations that are coming to the market today.

Bill? (Laughter, applause.)

MR. GATES: Well, thanks, guys. It's always great to see the beta software do its thing.

Well, some of the announcements that we're making today relate to evolving how these pieces work together. The Internet Explorer Version 3.0, we've got that out in developers' hands. Now we go to a formal beta in the next month. and we'll have that out in final form in the June time frame.

We are shipping a new version of the Internet Explorer for Windows 3.x and for the Mac, and these allow people to do rich development, including frames and other new things, across all the different platforms. We have a lot of Internet service providers that are supporting the Explorer, and we're announcing how we're taking the job of virtual machines, including this just-in-time compiler, and shipping that with all of the Windows machines.

The Front Page editor -- AT&T has announced that they're including this as part of their Web posting service, and so we expect a lot of people to be able to take advantage of that. We're also shipping the BackOffice update, which includes all the different server components -- the Internet information server, a new version of SQL, and, of course, the recently shipped Exchange server.

You can expect to see rapid announcements from us every couple weeks, there's a lot that we're doing with other companies, and with the standards committees, and it's really a faster pace than has ever been seen before.

The key is getting these building blocks put together. What's being focused on here is not something that's going to take a lot of time or even a lot of investment for companies to take advantage of. There's only a small change to how they think of navigating information and making it easily available across networks for their PC users, and so that's why people talk about the Intranet being a major focus for the next couple of years.

The wonderful thing is that the work people do on the Intranet is exactly what prepares them for the Internet. This prepares them to not only start to provide information to customers, but also to start to do transactions that way and be able to reach out in a fashion that was never possible before.

The number of pieces is quite large. It's tools innovations, server innovations, browser and client innovations and operating system innovations. All of those have to work together with standards.

So the key points are that it's a great time to see everybody embrace and extend the Internet, not throw out what was there before but take it and be able to use it. Integration is absolutely fundamental to make all of this work. There's no way users are going to learn all these protocols and acronyms, and we have to hide that from them. We have to hide the idea of even installing software. They have to think of it as all browsing, even though behind the scenes there's code moving down to the local machine and being verified.

And so the Internet and Intranet are really the keys to solving this entire information sharing puzzle that we've been working on for the last decade, and so it's a great opportunity for the industry to come together in the way that it has.

Thank you.