(Applause.)

BILL GATES: Thank you. Well, I'm pleased to have this chance to chat with you about where things are going in the future in terms of how people are going to work together using the latest tools.

During the last 25 years the original vision of Microsoft, which was about a computer on every desk as a very empowering tool, that vision has come a long ways. In fact, today in a sense we can say we take it for granted that most knowledge workers have access to electronic mail, access to browsing, that the way they create documents is very electronic.

If you talk to somebody who's coming into the workforce today about slides, they might look at you like what do you mean slides and you have to explain that there used to be these glass slides that would break and fog up and you used to have to know whether to put them upside down and backwards and all those things and that's essentially a thing of the past. Now it's PowerPoint with all of its beauty -- (laughter) -- and a different world.

So we've come a long way in getting these tools out there, but the view I have is that we've achieved well less than half of what we can achieve and will achieve in the next decade in terms of making people effective.

Well, why is that? Where are we falling short? Well, certainly we can't say today that every knowledge worker has complete visibility of the things they'd like to know that are going on inside their company: Who else might have expertise, what similar projects might have taken place in the past that they could learn from, what is the true profitability of a particular activity, what are some of the customers or people internally they work with thinking about the work that they're doing and how they could do that better.

People are living with very, very imperfect information and the overhead of gathering information and getting feedback and staying in touch on what's going on is still way too high.

And so I believe over this next decade we will make huge progress on that and, in fact, when we look at the economic productivity over the last, particularly the last 10 years, a lot of that has been fostered by the advance in these tools and I think there's a lot to be optimistic about in terms of these jobs being more effective, more interesting and also the basic economic results that we'll have because of the productivity advances here.

Knowledge worker productivity has always been tough for economists to measure. In fact, it was only late in the '90s that they finally admitted that very positive results were coming because of these tools that were letting people do things in a new way.

Now, towards the end of the '90s, the last couple of years, there was an excess of people thinking that dot-com approaches overnight would change things and that it would be startups rather than traditional companies who would come in and make the big advances.

That was overblown in a way but many of the things that were talked about then in terms of the empowerment, information sharing, allowing businesses to relate to each other in different ways, those visions are absolutely correct.

Some of the challenges that we face in terms of making it real were much tougher than people acknowledged then, and that's why we have things now like standards around XML that are just coming together in people thinking through how data is represented inside their company and their applications and how they're going to use that to achieve those visions.

So it's an exciting time, partly because over the last year and a half as that dot-com hype wore off people's expectations about what's going to come out of these tools has come down quite a bit, and so we'll be able to surprise people quite dramatically that the hard work, the tough R&D and the improvements in processes really are going to take place.

Well, why will that happen in these next 10 years? Well, one of the reasons we always have to give credit to is the advance at the hardware level itself. After all, personal computing couldn't exist if it wasn't for the miracle of the microprocessor, the fact that every 18 months or so the power of those chips more than double and the price comes down as the power goes up exponentially.

If we look at those trends and say, "Okay, how are they working for us in this next decade," we can see the pace is unbelievable, in fact, faster than ever before. And it's not just the microprocessor, although there's no limit in sight for the next decade. The kind of storage capacity we'll have goes up even faster so that today, of course, you can store everything that you type in a lifetime and you can begin to think about storing photos and databases and even videos very effectively.

The speed of optic fiber communications again is going up even faster than the speed of the processor.

The quality of these screens, the flat panel LCD screens that finally allow us to think about resolution such that you want to read long documents off of the screen, that's also come along and in the next few years will be essentially standard on desktop type devices.

The wireless approaches that let us connect these things together, so called Wi-Fi 802.11 is becoming standard built into these devices.

And so it's fair to say that every dream we have about hardware, virtually all of those are being made a reality. The performance of these Windows servers is going way beyond what the most expensive mainframe or UNIX machine would have been in the past. The peripherals, the digital cameras, the printers, all of those things are coming together.

There are a couple exceptions: The cost of broadband into the home is still very high but not into business and the cost of some of the wireless approaches are very high, but if we get this 802.11 in all of the key spaces, in businesses, hotels and homes, we'll overcome that as well; so huge opportunities.

Well, of course, Microsoft's not in that hardware side at all. What we're in is looking at those advances and saying how can software, which is the real magic, the part that we focus on, how can that magic software, together with those hardware things, do things that end users really care about. And we're spending now about five billion a year, a lot of that in pure research but a lot in classic development to build products that sort of bridge this gap, take technology and end user needs and bring those together.

There are a lot of different scenarios that are going to change pretty dramatically even in the next three or four years and I wanted to touch on a few of those, because I think all the people in your companies are going to find these things coming at a much greater pace and more dramatic in terms of change than current expectations.

The first one really isn't a scenario at all but it's just a fundamental way that we think about the personal computer: Is my software up to date? Am I willing to put new software on it? You know, something is frustrating, it doesn't work at all or it doesn't seem to work; do I get to give feedback on that?

Historically, the whole realm here is benefited by the incredible openness, where anybody can build the hardware, anybody can build the drivers, the peripherals, the application software, but that openness has become a little bit of a disadvantage because the pieces don't always fit together.

And so when somebody has a problem, you know, in the sense they don't have any clear recourse or they can say, "Hey, this hasn't worked well for me," well we've built now into Windows, starting with XP, this idea of knowing whenever a system doesn't work, and we offer whenever that happens to report back, and about 90 percent of people choose to send those things back. And it's a very fantastic thing, because by having that feedback we can see which drivers aren't working, which applications aren't working and not only see that but work with those third parties and use the Internet then to get the improvements, the fixes to go out onto all the machines automatically.

And so instead of people running into a problem, whether it's in the security realm or something that simply doesn't work well, instead of everyone having to run into that, the first person who runs into that, which happens as early as possible, then you get that information and you do the updates on those systems before anything else takes place.

And so the fundamental way that you think about the reliability of these systems has to change and is changing because of that infrastructure, and so that people will be willing to try out new things and be more ambitious because they've always got that lifeline that people are analyzing any frustration, it's not just it doesn't work at all, any report back that says, "Hey, I didn't like the way this thing worked."

On the servers we're building in a level of redundancy so that it's run like the phone system is where even if a piece of hardware fails, it fails over to the backup hardware and so that the levels of reliability are more like the electrical system has been than like IT systems have been historically.

Now, let's move to another scenario, which is this reading scenario. It's been a Holy Grail to have people want to be able to treat information on the screen like they do on paper. Why is that? Well, let's say you're in a meeting about a project and somebody's saying, you know, this isn't going well or we should change something here. You want to be able to call up not only that top-level information but you want to be able to point to anything or particular interest and expand that, look into it and say, okay, how does that break down by geography or person or skill set and eventually get to a point where you see something that's a problem or may need to change and you want to put a note on that and mail it off to your colleague so you can start to play what change will take place.

If you have the information on paper and you see something there you can't dive in to get the detail, and if you want to give somebody feedback on it you have to go write an e-mail saying, "I was looking at this funny piece of paper; you may have one similar or mine may be out of date -- who knows -- but what I was seeing here," and you have to describe it, and then when they get the mail they don't know exactly what you're referring to. But if you do everything with digital documents, then those things change.

So why aren't things purely digital today? Well, there are many reasons. One key reason is when you come to meetings there's this protocol expression of do you bring your laptop computer or not. At Microsoft we have rules: Some meetings we allow that, some we don't, because there's a trade-off there in terms of the screen getting in the way and somebody typing.

Now, bringing a tablet to a meeting, that's always been allowed and so part of the idea was, okay, how do we get so you have digital documents on a tablet-like device?

We also studied reading and said, okay, why do people print things out? Well, it turns out if a document is very long, because the screen is in a fixed position it's fatiguing to read something more than four or five pages, and even subconsciously without knowing why you like to print it out and then have it in your hand so you can hold and subconsciously you're changing your viewpoint so your neck muscles aren't fatigued and you can have so-called immersive reading.

So the only way to get this is to have better display of text, better resolution displays but also something you hold in your hand. And all those hardware advances I've talked about are giving us now a device that's small and light and it's a tablet-like device, where you not only can read but also you have a pen where you can write notes and have handwriting recognition.

So these tablet-like devices come out later this year. People like Compaq, Acer and many others will have these. They won't cost any more than a current portable PC. In fact, just think of it as having a portable machine where you can detach the keyboard and take the display and be able to work with it in this fashion. So some meetings you'll take it as a tablet, some meetings you'll take it with a keyboard.

What this means in terms of looking at documents, writing notes on documents, sharing things about documents is very dramatic in terms of how people share and work with information.

Another scenario we think is pretty exciting, and this is even more ambitious, is the idea of making it easy to have meetings where you have remote participants or you help facilitate the meeting or if somebody's not in the meeting and later they want to see what was discussed, both the video and the sound have been indexed in a way they can search that or somebody can send them just a little link that they click on to see the part of the meeting that's relevant to them.

In fact, the cost of storing the video of these meetings or even having a cheaper camera, 360-degree camera, is such now that this will be commonplace in the same way that PowerPoint is commonplace for presentations today. You'll just put a little ring camera, which is 360 degrees, in the room and record the meeting and people will be able to see it later. In fact, people who aren't in the meeting can just go to the parts of the meeting that they care about, so it's more efficient.

We also have a technique where you can play the meeting back at about twice the speed that it really took place, and so you'll have this huge incentive not to go to the meeting -- (laughter) -- because it's much faster in retrospect than it was in the real thing. And, you know, people will think twice about, okay, meetings are about collaboration, information sharing; if all you want to do is listen, then maybe you don't need to do that.

Communications is another fascinating thing. You know, we all love e-mail and we all hate e-mail. We have multiple addresses that, you know, if you're sitting in your office concentrating on something and your machine says there's a new e-mail, the temptation to go read it is very high and it's probably e-mail that somebody left their lights on in the parking lot and then you try and remember, you know, what was I working on, and the extreme is junk mail but also any mail you get can tend to be an interruption and it's true for your mobile phone ringing, it's true for all these ways that we communicate. Instant messaging now is just a new flavor of this that comes in and you have to decide how you want to work with it.

Well, what we should really have is we should have one address. We shouldn't have a phone number, a fax number, a portable number, multiple e-mail addresses. We should have one address that we're willing to give out and then software works on our behalf to decide when we should be notified, when we should be interrupted. And so depending who it's from, whether we're in a meeting, what it's about, how urgent they think it is, all those factors, it comes to us when it's appropriate. If it's truly junk mail, that's never. If it's something of a personal nature that's not at all urgent, maybe when you log in at home that night you see things like that, or if it's urgent it comes to you immediately.

So this idea of having your communications all pulled together and the important things coming to you is a very important advance. Say there's a project that you're monitoring; the idea that you have to go out and look and say, okay, did something go wrong, did something go wrong, that information should be brought to you in your unified inbox and prioritized appropriately instead of your having to go out and look. Even simple things like did a flight time change or did a schedule change; if so, that information ought to be brought to you. And so this is really a software issue to work with the user and understand their profile and prioritize things on their behalf.

Another big scenario is the way business is done. This term e-business, of course, has been used very heavily in the last three or four years. E-business is a real thing. The idea of matching different buyers and sellers and taking complex transactions, allowing those to work, the benefit of this is unbelievable.

Now, this is where these XML standards come in, where different industries take things like a patient record or a financial record and decide how they should look so you can exchange the information both inside a company between applications and workers and across company boundaries.

And today e-business is not taking place. It's still phone calls, e-mails, faxes and only working with the people you've traditionally worked with, mostly working inside the company itself. The information between these different systems is not well coordinated with each other and XML promises to fix that.

So we've bet our company on XML the same way we bet a decade ago on graphics interface and there's been great progress around this. We're rebuilding all of our software -- Windows, SQL, Office, Project -- around XML as the key data type for this easy, easy exchange. And so a lot of these applications will really come into play over the next several years and you'll think of that as a new approach and a much more efficient approach.

The final scenario I'd pick is knowledge worker productivity and this is a fascinating one because people say to us, "Well, when it comes to the individual worker, creating a document, creating, say, a word processing document, haven't you done everything that they need to do? Isn't that it?" Well, the fact is nobody works in isolation and so the idea of how you share with other people, get them to review those things, today it's very, very complex. Information sharing is very imperfect. People are moving around files and trying to synchronize things and you'll get different versions of things in e-mail from different people and it's hard to put those together.

So a huge focus of this .NET strategy is about the information sharing, making it easy for people to work together. Whether that's business intelligence forecasting, which is where the spreadsheet is going, or working together on documents or working together on schedules for key projects that are going on, the big thrust is drawing everybody in so you can see the same information and work together in a rich way.

And so this really mirrors the evolution we've gone through with all the different productivity products, particularly what we've done with Microsoft Project. This is a product that we started doing about 1990 and it was interesting when we first had the discussion people came in and said, "Well, people don't use computers to do project management." And I said, "Well, they will." And somebody said, "Well, the size of the market is really small and it's just these big construction firms and you really don't want to do this." And we said, "Well, in every category we've gone in it hasn't been about the share of the existing way of doing this thing, it's been about creating a tool that allows us to really take what had been paper and pencil approaches without the same coordination and change that on the desktop."

And so we started that but the limitation was it was very much at the individual level, and so the way the information sharing worked that improved with e-mail and things like that, but I'd say it's only with the work that we're coming out with this year that we can say that the way people work together in the enterprise, the way that you coordinate and share things, that we finally have that as a very core element of how this activity works. And you'll see that not just in Microsoft Project but in all the things we do because it's a key part of the theme of the .NET strategy -- all the devices working together, all the different people working together.

So what does this map onto here? Well, the people who have been using Project have been the project managers, people who have that skill set and are in the central role of helping to coordinate things. When it comes to executives or team members or the people who keep the list of resources up to date, they really haven't been drawn into use the software. In fact, in many ways that's been a problem because their view of things is based on sending e-mail or some other tool that they're working with. And so you have this flow of information between the project managers and those other people has been very, very manual.

And so as we move to the idea of this enterprise project management and allow people to think about these things as centralized pools of information that are easy to look at and view in a very personalized way, although it's all coordinated, that required a major rethink of the products and how we do the software.

And I will say that the Project team by getting out in front of this has had a really good influence on how we're thinking about all the different software products that we're doing. They come out with this in the middle of the year in June.

I will highlight the role of our partners in helping us with this, really thinking through the scenarios and making sure we're in touch with how people use this thing, people like KPMG and P-Cubed, really a great list of partners that have joined in.

Real quickly I want to give you a look at what I'm talking about of different people coming in and being involved, and so I'll ask Chris Capossela to come up and give us a quick look at that. Welcome, Chris.

(Applause.)

CHRIS CAPOSSELA: Well, good afternoon, everyone. I thought that the best way to sort of show our excitement around Project 2002 was to take the next ten minutes and visit three different users in a fictitious manufacturing company called Champion Zone. We're going to start off with a resource manager and see how they can easily manage their entire pool of resources and the skills that those resources have and then switch hats and look at the project manager and finally wrap up with the executive who's interested in managing their portfolio of projects.

What you see on your screens here is just Internet Explorer pointed to the Project Server 2002. So I see a log-on screen and I'm going to log in as the resource manager. Her name's Toni. And you'll quickly see that Toni's not using Project for Windows here; she's just working with her browser and she's got a nice portal screen here that welcomes her and tells her that she doesn't have any updates or new tasks or new issues to track. Obviously, that can change as time goes by, but here she can work with projects and tasks and resources, as well as documents and issues.

Let's go ahead and take a look at what we call the resource pool. I've clicked on resources here and we see the names of the resources and their standard rates and their overtime rates and the last time I as a resource manager modified their skills.

This is just one of many, many views that Toni has the ability to create in looking at her resources.

Let me change to another view that shows some of the skill sets of my people. I see the resource name. I also see that we're actually tracking the department that each resource works in. That's going to be important to us later. And I also show that we're actually tracking one skill level for all of our people. So this person right here, Aaron, is a VB developer. The person above it is in management. Adam is a systems' analyst.

I have tools here that let me filter and just take a look at all my VB developers, and I can select a whole bunch of these folks and click "view availability" and now Toni with a couple clicks of the mouse is going to be able to see how her entire development pool of VB developers is allocated over the coming months or the coming weeks. We're seeing this across all the projects that these resources are working on. So again one of the big benefits of Project Server is being able to store everything in one place and be able to slice and dice the data and really get an understanding of how over-allocated are these people or how underutilized are these people.

Let's click back to the resource center and I want to point out one other column right over here. You notice this says "generic." Many of these people are actually real people -- Aaron and Adam and Alan -- right now I'm sorting by alphabetical order. But you'll also notice an accounting resource, which is listed as a generic resource. Let me filter and just take a look at all our generics and you can see that we've defined a generic C programming resource, a generic LAN engineer, a generic legal engineer. This allows us to build generic resources that have a very particular set of skills and then later essentially decide whether or not we can take on a new project by looking at assigning generic to them, figuring out what kind of revenue forecasting and cost forecasting they're going to have, and I don't have to actually assign real people yet. When we decide that we want to take the project on, sure enough we can use these skills and these generic resources and replace them with real people in our enterprise pool.

Let's go ahead and close the browser and let's switch hats now. I'm actually going to launch project and log in as Steve Masters. He's our project manager, not our resource manager. Steve is quickly going to create a new file from an enterprise template. This is a file that helps me create a new product for Champion Zone. Enterprise templates are again another benefit of having the server where you can put everything in one place and everyone gets the benefit of that single storage location.

You'll notice right away that we've got a whole bunch of tasks in this new product template and all of these tasks have generic resources assigned to them: marketing research, process engineering, legal, accounting, et cetera. So this gives me a really nice way to model how much this project is going to cost without assigning real people to it.

When it's time for me to fill this project, when I actually decide we're going to go forward with this, we now can give you great tools like the Resource Substitution Wizard to basically let you easily staff the project, let's only staff this new Champion Zone project. I could actually do this across many, many projects at the same time if I wanted to. And this screen lets me actually choose the resources that we want to apply to it. So I'm going to scroll down and pick the resources that are located in Washington. So just like resources can have lots and lots of skills, we can track locations, we can track languages. Really it's up to the user or the program management office to define what different things you want to track.

Now when I hit "next" Project is going to go off and look at all of those resources and say, "Hey, a lot of these folks in Washington are working on other projects as well; when you decide to pull people in, do you want to include what they're doing on those projects so that you don't borrow from one set of projects for another?" I also have the ability to set the priority of these projects so that project can be smart about trading one project off for another when allocating these resources.

I'm going to go ahead and hit "next" and now I'm going to actually update the project with the results of the wizard. Notice if you keep your eye down here at the generic resources when I hit "next" here it's going to go ahead and do the work and lo and behold we've replaced all those generic with actual real people and now their availability has been taken with all of the tasks in this project.

When I'm ready to actually publish this project, when I'm done tweaking with my resources, I have the ability to go off and send that off to other people. They'll automatically get an e-mail notification in their Inbox that says, "Hey, you have a whole bunch of new tasks; click here on this link." When they click on the link up comes the browser. They're now able to log in to Project Server and see all of the tasks.

At that point they can actually click on a task and say, "Hey, you know what, I don't think I can do this." Maybe some personal thing has come up: I need to take vacation or an emergency has come up, but I need to go somewhere else.

At that point the project manager is going to get notified and they actually have the ability to use some more sophisticated tools, something that we call let's build our team from the enterprise pool where they can say, "Okay, I'm interested in taking this person and swapping them out for somebody else who has the same skills and the same availability."

So on the left hand side of this dialogue box you'll see that we've got 253 resources in the pool that I have permission to work with. On the right hand side you see all of the people who are already working on the project itself.

I'm going to choose Francesca and I'm going to click the "match" button. And what that does is filter our 253 down to 7. There are only seven people in the pool that have the exact same skills as Francesca. Now I have a high level of confidence that I can pick one of these and replace Francesca with this new person.

I'm skipping over a lot of the power of this tool in terms of further restricting and filtering the pool based on availability or custom filters that I build, but for the purposes of this demo I'm going to choose Wendy and hit "replace" and how when I hit "okay" sure enough Wendy has automatically been substituted for all of Francesca's tasks. She'll get the next notification when I publish this off.

So it's very powerful to have that single Project Server with all of our resources in it, all of our projects in it, all of our views in it and different users interact with it differently.

I'm going to switch hats one last time and I'm going to log in as Mindy Martin. Mindy is our executive. She's interested in a much simpler view of the data in the Project Server. She doesn't really want to get into the nitty-gritty. She wants to see things at a very high level. So you see that when she logs in she has a slightly simpler user interface than our previous user.

I'm going to go ahead and click on "projects" and very quickly Mindy can see that all of the projects that are within her domain, and in this case I've grouped this data by project manager, so she can see that Barbara has three projects, Linda has three and Steve Masters has a few, and she can see from these stoplights which projects are in trouble or which projects are doing well.

I've set this up for these stoplights basically indicating how she's doing on her schedule but these are very customizable so you can basically set these up so that they're telling you about costs or resource allocation or any of the different fields that are available for the product.

At this point as an executive I actually do have the ability to drill into one of these projects and see very detailed info, but more likely than not I'm actually going to want to do some further analysis at a high level, not at the project level, but I want to use what we call the portfolio analyzer to really slice and dice the data that's stored in the Project Server, which is built on top of SQL Server, which gives us a lot of our power here.

So I've quickly brought up a view that shows some of Steve Master's resources -- you can see the view right there -- and it shows Alan and Andrew and Brad and the blue bars are the actual work that each one of these resources has. The purple bar is the baseline work, what they thought they were going to have, and the actual green area here is the availability. So you can quickly see that Brad Sutton, when looking at all of Steve's resources across all of the projects, Brad's the one who's really in trouble.

Again, I've set up a view that lets me focus right on Brad Sutton's workload. I can come in here and quickly see that Brad Sutton's work in January and February looks okay but, boy, March, April, May, June, those look very bad, and I can get into more granular detail and extend this out over Q3, Q4 and really see that things are bad for Brad for quite some time. (Laughter.) Not a good thing.

As an executive, I kind of have the ability to model different scenarios, because I have lots of projects that I'm responsible for. They have different priorities. I might want to trade off one project for another.

So I'm going to open up what we call the "portfolio modeler" and I'm going to click on a model that I've already built called the "Steve Masters' Project." When I do this, it's going to open up a model that has all of Steve's projects in it and it's going to show me a particular view of my resource allocation. So you can we're looking at basketball, soccer and hockey and you can see that over this time period things are looking pretty bleak for the resources that are on this project.

Well, as an executive I may feel like these folks are working on the most important projects that are in my portfolio so I really want to raise the bar here and bring in some additional folks and trade off against other projects that for me are less important.

So I'm going to bring up what we call the toolbox, which is a very powerful set of functionality here. This toolbox basically let's me say, "Hey, the basketball project is my top priority. I'm going to change it to priority one versus all the others." It lets me specify that I'm in charge not just of the resources in the West district, but I actually am willing to talk to my peers and bring in other people from the entire United States, because these three are really important. And this down here lets me specify the scheduling constraints that I want to use in this model. I can keep the current start and end dates and see what just adding people will do to the project itself. I can reassign resources within a particular project to see if there's any room there. But in this case at the executive level what I'm really interested in doing is building a model across many projects, across many resources and then applying all of the resources in the model to this. So I'm going to select that, USA folks, and let's apply them to the basketball project.

And now Project is basically doing some modeling for us and it's shifting resources around. It's not actually making any changes to the actual project themselves; this is purely a snapshot of data that I'm playing with. I have basically improved things a little bit for the basketball project. I might want to do that exact same model and apply that to the soccer ball project, maybe give it a lower priority if I think I want to trade that one off, and we'll see that as we add resources the finish date on that one moves out but, boy, the basketball one is looking better.

And like any good demo, if I go ahead and click on the hockey one and click "allow resources to move around within the model" and I hit "apply" we're going to see lots of green, which is obviously kind of what we're after.

I want to make it really clear that this modeler is just a tool for the executives. They're not actually changing the schedule. From here they have the ability to look at the constraints that they've just placed on all of these different projects and all of these resources and share those constraints with the project managers who are actually going to get the directions from Mindy, "Hey, we are going to prioritize these three over the others or this one over the others; here's what I've played around with. Now I want you to go figure out how do you make this happen using these constraints."

So it's a really quick look at some of the functionality we've put into the product for those there users. I could go on and on, as you know. We're super excited about what our partners and our customers are doing. In addition to the product, I'm learning a lot about what NASA is doing here in Orlando. They're doing some amazing things to extend what I've shown you here with a bunch of custom work as well.

So we're excited to get this product out the door in June and start getting all that great feedback from you guys. Thank you very much.

Thanks, Bill.

(Applause.)

BILL GATES: Thanks, Chris.

So I hope you got a sense that this whole philosophy of information availability, connecting everyone together has a pretty dramatic effect on the value people get out of these software advances. There are many elements that need to come together for our industry to realize these things rapidly. The standards, particularly there I'd highlight XML and how things are being connected up with that. In the last few months the announcement of a group called Web Services Interop, WSI, which is IBM and Microsoft and some other people, I'd really highlight that as a big milestone where people are going to get this rich interoperability, even as they mix and match different software stacks. So all the business thinker needs to make sure goes on is that their applications are built using these XML approaches and then the different industry software tools can be chosen according to whichever one is best for that project.

Another thing that's important to us is that in doing this new work the way you look at information, whether you're offline using a PC, whether you're connected up to the server, whether you're working between companies, but all those things come together and they don't bring in new vulnerabilities in terms of the information security, that you really know who the user is, what their rights are and that's all being handled in the appropriate way.

So there are big investments taking place in that. This .NET thing is a profound change for all the work we're doing, but we think it's also very exciting because it maps down into these real scenarios.

For us, we'll just keep investing in the software. We've got a great feedback loop, which is both the kind of conversations we have with people and now that very direct monitoring of the information they're willing to share with us on how they're using their systems, what's going on there, and it really is neat that we've got revolutionary advances coming both for the business and for the nature of these jobs, that allowing people to be more effective in using information can make them far more creative as well as productive.

So we look forward to taking on this challenge and working with all of you. Thank you.

(Applause.)