Joe Belfiore Talks Windows XP Media Center Edition

A look at the design and development of Microsoft's new multimedia interface for PCs

In this exclusive interview with Joe Belfiore, General Manager, User Experience, at Microsoft's Windows eHome Division, we discuss how Windows XP Media Center Edition--code-named "Freestyle"--evolved over time from concept to product. Joe has been an instrumental part of Microsoft's user experience efforts since the first version of what became Windows NT, and his background includes user interface work on Windows 95, Internet Explorer 3 and 4, Windows 98, Windows 2000, Windows XP, and the new Media Center experience in XP Media Center Edition.

I first met Joe at the Windows XP Beta 2 Reviewer's Workshop in February 2001, but remembered him from an IE 3 alpha demonstration in early 1996, when he was working on Microsoft's then-nascent attempts to combine the Windows shell with IE, providing users with a consistent experience between the two navigation paradigms. What struck me most about Joe at the time was his honest excitement and enthusiasm for the work he was doing at Microsoft, and now that I've gotten to know him a bit better since then, I can say that he hasn't changed a bit. Joe is one of those instantly likeable guys, someone I look forward to meeting with.

Since the completion of Windows XP, Joe has moved on to the eHome Division, and in his current position, he helped guide the development of Microsoft's latest XP version, Media Center Edition. I discussed the design and development of Media Center with Joe just after the product's code was finalized. Here's what he had to say.

Paul Thurrott (PT): Before we get started, let's talk a bit about your background. What have you worked on at Microsoft?

Joe Belfiore (JB): I was hired by Microsoft in August 1990, right out of school as a CS major. My first job was actually as a Program Manager for OS/2 2.0, though that only lasted about a week. Then the whole Microsoft/IBM split happened and I was reassigned. I became the first-ever user interface manager for NT, and stayed in that job for a few years. Then, right before Windows NT shipped, I moved over to the Windows team. I became the Lead Program Manager for the Windows 93 user interface, which became known as "Chicago," and then as Windows 95.

After Windows 95 shipped, my team and I began working on the Internet Explorer user interface. I became the Group Program Manager for the Internet Explorer 4 integrated shell, which was released both to the Web and as part of Windows 98. The same folks who had been doing the Windows shell did the IE 3 and 4 user interfaces. The idea was that we would combine the Web browser with the Windows shell, so that users could navigate from folders to Web pages and back again with a consistent interface. Then, I worked on Windows 2000, and finally Windows XP, as Product Unit Manager for the Windows user experience.


PT: So how would you describe your influence over the design of these products?

JB: Well, I'm not a graphic designer, and I don't draw pictures very well. What I do is focus on the concepts and how we communicate them to users. I've always believed that effective implementation is the key to helping users.

Let me give you a couple of examples. In Windows XP, I pushed hard for us to do friendly English descriptions of tasks in the My Pictures and My Music folders. At Microsoft, the goal was always about making people more efficient. But this led to too many easily accessible but visually indiscernible icons in toolbars all over the place. Because the PC appeals to more people now, I felt that we needed to find way to communicate capabilities to users in a more conversational way.

Another example is wireless networking. You have to remember that most of the people who work at Microsoft on Windows are sophisticated computer hobbyists. We have to be disciplined, and understand that many of our users are different from us. With wireless, we could have taken the typical route to adding these capabilities, which is to make it possible, add lots of options, and then stop there. It would have worked, but the real value is when you keep going and make it invisible to the user. I try to make the engineers remember that the people who use the products aren't necessarily like us. We need to go the distance to make it easier.


PT: What made you decide to join the eHome team?

JB: I had been working on the core Windows user interface, including IE, for about 10 years by the time XP was nearing completion, and it was just time to do something a little different. XP was literally the completion of everything I had worked on at Microsoft, the combination of the NT kernel, high performance, rock-solid stability, and the friendliness and compatibility of Windows 95. Finishing that work was important to me. But throughout the development of XP, I started to get interested in tinkering with the PC as an entertainment device. In my basement, I had set up LCD projector like the ones you'd see in a conference room at work, and hooked it up to a PC. I used the device to watch DVD movies, and listen to music. It worked, but the PC obviously wasn't really designed to be a media consumption device, like a stereo or VCR. There was no remote control, and the text was hard to read on-screen.

During the development of XP, we had people over to watch the Super Bowl in hi-definition (HD) via the PC, and we were able to get it working, but you had to be engineer to make it happen. I became really interested in that scenario, though, and thought, the PC could be great at this. It's powerful and has the support of innovative third parties, and maybe we could just add capabilities to Windows that would let it work with a remote control, and expose the TV and DVD stuff. It seemed pretty compelling.

Right around that time, the eHome group had started up, so I talked to [Microsoft Corporate Vice President, eHome Division] Mike Toutonghi about making a change. Maybe I could do a different type of user interface work for a while. At the time, the eHome division was experimenting with a strategy of building a PC-architected dedicated device, similar to the Xbox, but aimed at digital media experiences. It would have used the PC architecture, to ensure cheap components and a standard software development environment, and be dedicated to scenarios like music, photos, and movies. But when I joined on, the strategy had changed, because there is now an increasing demand to use the PC in those scenarios. And the openness of the PC business model meant that we could easily add the capabilities we needed, and provide users with more flexibility. We could put the PC in a home office and use it to watch TV, put it in a child's room where it could be used as PC, TV, or stereo, or stick it in the basement with LCD projector. So that represented a shift in strategy.

If you think about the price of PC hardware today, it's easy to see that you can add capabilities and make a device that does everything. It was a perfect fit for me, and it was my opinion for the way to do things as well.


PT: So you've got this PC-based device now. But could this have been done without Windows XP?

JB: Not a chance, no way. First of all, Windows XP is reliable and rock solid, and that foundation is necessary to give consumers the feel of a consumer electronics device. We were able to make the Media Center simpler because XP takes care of a lot of media chores, things like getting pictures from a digital camera onto the PC, or copying music and creating playlists. Now, using the remote you can focus on enjoying media, not managing it. If those features weren't in XP, it would have resulted in a more complex Media Center.


PT: How did your experience with the Windows team prepared you for eHome?

JB: It helped in two ways. First, in building the product, we wanted the Media Center experience feel like a natural, integrated part of XP, and not just be slapped on. So we spent a lot of time and energy on making the Media Center work with the underlying system. So you can use the remote with the welcome screen to click on your name and logon. An XP-style balloon tip appears in the system tray to let you know that a show is being recording in the background as you're working in XP. From a technical and design perspective, knowing how windows was architected has helped to do this in a natural way.

Secondly, a part of the culture of the Windows team is about creating software that enables third parties to add value. So we built Media Center in a way that would allow third parties to enable great new scenarios for users. We've worked with a variety of hardware component vendors, PC makers that will create exciting form factors, and software vendors, which can make games and other experiences that work with the remote and are accessible from the Media Center. That philosophy will make the Media Center more valuable to the industry and end users.


PT: What was your greatest challenge in bringing this product to life?

JB: The greatest challenge was taking the complexity inherent in the new technologies we created, and from all the partners we worked with, and integrating it in a seamless way. We worked with a very sophisticated and complex set of technologies--the user interface is done in DirectX, most of the Media Center code was written in state-of-the-art C#, and we worked with a ton of partners, all of whom have their own code--and integrated it all together in an attractive, simple and straightforward package

From a technical standpoint, Media Center user interface functionality is almost entirely written in C# managed code, on top of native Win32 and DirectX Windows XP components. These operating system components render video and draw fluid animations smoothly on the screen at 60 frames per second, with hardware acceleration and MPEG decoding provided by 3rd parties. Getting all these technology components to work together well was our biggest challenge.


PT: Now, one of the controversial aspects to Media Center is the packaging. Media Center is only available as part of a new Media Center PC, which today is available only from a single vendor, Hewlett-Packard, in the US. Why did you decide to package Media Center this way, and not as a free download, or possibly a Plus!-style package with the necessary hardware add-ons?

JB: Really, we were hoping to accomplish two things. First, people should have an easy, great experience that delivers on all these scenarios. So you need a remote control, TV tuner card, DVD playback, compatibility with cable or satellite TV, and a variety of other hardware and software components, all of which must work together. By requiring a new PC for the first version, we felt we could consumers a better package that ready to go, off the shelf. It should be easy to buy and set up. We want people to be successful when they use Media Center, not frustrated.

Secondly, we are really defining a new class of PCs that is more powerful and media savvy--not just another inexpensive PC. We hope the Media Center PC will define a new standard for media capabilities, and enable new innovation in media on personal computers. For example, IEEE-1394 is a great connectivity standard, but many PCs still don't have it, and therefore there is less third-party innovation. With Media Center Edition, we are defining a new category of PCs with a more compelling standard for their capabilities, including more powerful processors, better graphics, DVD, 1394, and so on.

We did those things in the first version, and made it a PC-only product. But as the category gets going, we hope to make it available to people on their existing PCs. But we need PCs to be more powerful and capable out of the box first.


PT: What was your development team's greatest focus throughout the product development cycle?

JB: We wanted to make it simple, easy to use, convenient, and friendly, and make the Media Center feel as good as any consumer electronics device. We were very disciplined about simplicity. It was a key focus for us when we designed the Media Center interface.


PT: What do you think are the greatest opportunities in Media Center for the industry?

JB: There are several big areas here. Hardware vendors can create new value with various new industrial designs in Media Center PCs, that consumers can look at and choose from. Hopefully we will see lots of innovation there, with PCs designed for different rooms, quieter PCs, and terrific remotes. There are all kinds of great things companies can do to make compelling PCs.

There are also millions of scenarios for software and services. As an example, we're going to post a version of Solitaire on the Web, as a PowerToy, that you can use with a remote. Those types of games are an example, as is Internet Radio, video on demand, and applications that interact with home security systems. There are a lot of possibilities there.


PT: What has the feedback been like from beta testers?

JB: A lot of people are surprised by the value of using the PC with the remote. In general, testers are finding that they enjoy doing things, like watching TV, while working. One thing that surprised us was the number of people that have TV sets in same room as their PC. A lot of these people configured Media Center to show up on the TV set as well as the PC monitor, [so they could watch movies or TV through the new interface]. And we've gotten a lot of suggestions for improving Media Center going forward as well, of course.

Regarding testing, we have hundreds of people we sent preconfigured PCs, and several hundred at Microsoft that received take home kits for their existing home PCs, probably about a thousand overall. The feedback has been great.


PT: Where do you see Media Center and PC/digital media integration going from here?

JB: One of the most significant things over the next two years is that we're going to upgrade Media Center to support every TV in the house. What that will mean is that all of your digital pictures, music, and videos will be available from any TV in the house. Suddenly, the power of the PC is projected throughout the house, and I think that will be very, very compelling.

In the short term, we will build on the core digital media experiences that are in the product today. People want to browse the Web, listen to Internet and local radio, and so on. We look at requests from users and testers, and we'll put those features in future versions. And of course third parties will add those kinds of features in as well.


PT: Thanks Joe, it's always great speaking with you.

JB: Thanks!