A .NET My Services Scenario

< BACK  NEXT >

[oR]

The best way to see how an application might use .NET My Services is to look at a scenario. Suppose, for example, that a theater Web site wishes to offer a Web services based application. This application might let a customer book seats using a custom client and then send an invitation to one or more of his friends inviting them to join him for the show. Figure 8-5 shows how the application might use .NET My Services to accomplish this.

Figure 8-5. A Web services based application that allows booking theater seats might exploit .NET My Services in many different ways.

How Can Microsoft Guarantee the Security of .NET My Services Data?

The .NET My Services servers and the information they contain will surely be tempting targets for hackers. Furthermore, Microsoft's history in building secure Web software is far from perfect, as the repeated problems with IIS have shown. How can the company guarantee that the data we place in .NET My Services will be secure?

I don't think it can. In fact, I don't believe Microsoft or any vendor can absolutely guarantee that its Internet-accessible servers will be hacker-proof. Once again, though, it's a business issue: If hackers are able to compromise the data .NET My Services contains, its customers will desert in droves, and .NET My Services will die. Guaranteeing complete security is probably impossible, but Microsoft certainly has a very strong incentive to protect its .NET My Services servers it's life and death for this business.

It's also worth pointing out that every Internet-accessible Web service will face this problem. In fact, many of the issues that have been raised by .NET My Services are actually generic concerns that must be addressed by any organization that wishes to provide public Web services. .NET My Services is the first highly visible application of this idea, so it is the canary in the coal mine for public Web services. Other organizations that hope to mine this technology vein should pay close attention.

Applications can use .NET My Services in a variety of ways

Suppose that Mark wishes to book seats and have the application automatically invite Bob to join him. The process begins with Mark's system logging in from his home computer to Passport (step 1), and then requesting a ticket to the booking application (step 2). Next, Mark uses Web services based client software to convey his request to the booking application (step 3). This request includes Mark's Kerberos ticket, which the booking application uses to authenticate him and the application he's using. Mark also conveys to the booking application his wish to invite Bob to join him at the theater

Responding to Mark's request, the booking application reserves two seats and then uses .NET My Services to convey Mark's invitation to Bob. This process begins with the application acquiring tickets to the services it will access (step 4). It then uses Mark's .NET Services data to locate and access Mark's .NET Contacts service (step 5). The booking application looks up Bob in Mark's contact list and extracts Bob's PUID, which is stored with his name (step 6). The application next accesses Bob's .NET Services data (step 7) to find Bob's .NET Alerts service and adds an alert message to Bob's .NET Alerts data (step 8) containing Mark's invitation.

Once this is done, .NET My Services uses Bob's .NET Presence information to locate a device Bob is near. In this case, that device is Bob's home computer, so the alert containing the invitation is sent there (step 9). How this alert appears to Bob depends on the software available on the device on which he receives it. In Windows XP, for instance, it can be delivered via a pop-up on the screen (sometimes referred to as toast because of the toaster-like way the message appears from the bottom of the screen).

This example could be significantly more complicated

Other services could also be used in this scenario. For example, the booking application could access Mark's .NET Wallet service to charge for the tickets, and it could also write the theater date into both Mark's and Bob's .NET Calendar service. The main thing to note in this example, however, is how much .NET My Services did on behalf of the booking application. First, this application used .NET My Services to authenticate Mark, its client. Then, once Mark had supplied Bob's name, the application relied on .NET My Services to locate Bob and send him the invitation. Providing these services independently, especially the location and asynchronous notification capabilities, would likely be a daunting task for the developers of the booking application. And if every application did these things for itself, the result would be an extraordinary duplication of effort. Just as with an operating system, having a common way to perform common tasks and store common information makes sense.

What's the Killer App for .NET My Services?

Nobody buys Windows for its own sake. Windows is a platform for applications, and it's those applications that make us choose it over another operating system. Similarly, .NET My Services is a platform. Without attractive applications, people won't pay for it. So what's the killer app for .NET My Services? What software built on this platform will be so valuable that a large number of customers will rush to sign up?

Nobody knows. Early tests have suggested that the ability to overlay the calendars of different individuals and organizations might be an attractive possibility. Imagine being able to display the free/busy days of every colleague you'd like to have at a particular meeting, or of every member of your family you'd like to join you on holidays, on one screen. Suppose your favorite sports team or the local symphony posted its schedule in a .NET Calendar, then let you overlay those dates with your own .NET Calendar.

Or, because more and more people have multiple client devices computers at home and at work, mobile phones, PDAs, and others perhaps using .NET My Services to synchronize address books and other information will be a service that people will pay for. Or maybe some new .NET My Services based feature in a future version of Office will be a major driver for adoption. There's no shortage of possibilities.

But should we be surprised that the killer app for this new platform isn't yet obvious? No. The killer app for previous new platforms, the thing that really drove people to buy it, wasn't commonly obvious when the platform first shipped. Who knew that Lotus 1-2-3 would be so important to the future of IBM's personal computer or that desktop publishing would motivate the masses to buy a Macintosh? Both the creators of applications and their users take time to figure out which ones really matter. For .NET My Services to succeed, some very useful applications must appear. I'm not at all surprised that we don't yet know what these will be.

< BACK  NEXT >