Return to: Articles Index
Microsoft.NET will transform software from shrink-wrapped CDs to Internet-based services. You've probably heard about .NET - but do you get it? If you're confused, you're not alone. Here we'll give you an idea of what it is, how it could affect you, and what some industry observers think about it.
You've probably heard about Microsoft's ambitious Internet strategy, known as ".NET" - but do you get it? If you're confused, you're not alone.
Microsoft.NET will transform software from shrink-wrapped CDs to Internet-based services. Why is this necessary? Ruth Anne Lorentzen, Microsoft's general manager of marketing and business development for .NET services, said individual Web sites or devices connected to the Internet are yielding to what Microsoft calls "constellations of computers, devices and services that work together to deliver broader, richer solutions." To tap into this trend, Microsoft has imagined .NET as an all-encompassing development and deployment model.
This radical strategy uses two Web technologies to link Microsoft software and devices, and transform applications and data into Web-based services. XML (Extensible Markup Language) is a standard Web language (akin to HTML) used to label data, and SOAP (Simple Object Access Protocol) provides a standard way to tap Web services.
Four major components comprise the initiative:
The key service being bandied about is "HailStorm," a set of consumer Web services providing authentication, calendar updates, event notification and the ability to tailor transactions to personal preferences. HailStorm will use Microsoft's Hotmail and Passport.NET, a popular MSN authentication service.
Several .NET ready products and services are already available:
Some of .NET's more advanced features will not be available until the release of the Windows 2002 (formerly "Whistler") server OS, due early next year.
Most .NET components are aimed at software and system developers who will be creating .NET services for their market or organization. For the IT administrator charged with rolling out and supporting applications, however, .NET may have two major benefits.
Tim Landgrave, founder and CEO of Vobix Corp., an applications provider, said .NET will simplify application deployment by letting admins copy applications into destination directories without having to run setup and registry programs. Another benefit is the ability to run multiple versions of applications side by side.
"In the past, running multiple versions of the same application or even two different applications with shared functionality could result in DLL hell," Landgrave said. "This happened because two applications needed different versions of the same DLL and couldn't be installed at the same time because of limitations in the Windows registry." (DLL stands for Dynamic Link Library, executable code or data that Windows uses to provide services such as a distributed filing system to applications.)
Tom Manter, research director, Windows NT/2000 Technologies & Platforms for the Aberdeen Group, a market research firm, added that IT administrators will be able to create highly manageable applications that can be delivered quickly. "These applications will all support XML, a key component of the .NET framework for delivering data," he said.
There are some caveats. Dick Sullivan, IBM vice president of solutions and integrated marketing, said some IT administrators may be reluctant to deploy .NET. Many large organizations have three or four operating systems, and Linux, in particular, is becoming very popular. Market research firm IDC reported that Linux captured 27 percent of server-OS new license shipments in 1999-2000. IBM recently announced it will spend more than $1 billion on Linux this year because the company wants the technology, which is less expensive than Windows (and not controlled by a single vendor), to become a computing standard.
"Linux creates an alternative platform and more and more Web services are going to be implemented using Linux because it provides better cross-platform capability," Sullivan said. "So some large companies may implement .NET, but others will gravitate to other platforms like Linux and Solaris."
Nonetheless, .NET partners are signing on, including American Express, Click Commerce, eBay, Expedia.com and Groove Networks. Expedia indicates that .NET will allow customers to use their existing buddy list and universal calendar to better coordinate travel planning with friends, family and colleagues. Groove Networks CEO Ray Ozzie (who created Lotus Notes) said HailStorm "will complement the capabilities of emerging, edge-based peer computing applications such as Groove."
To use .Net, network administrators and engineers will need to learn more about how to optimize and manage networks with multiple applications running on multiple boxes at multiple layers of an application. Much of this knowledge won't even be developed until a large enough body of applications using the technology is deployed.
"It would behoove any admin or engineer to learn the basics of the transports (HTTP), protocols (SOAP) and languages (XML) that will combine to form the next generation of applications, not only from Microsoft, but also from IBM, Oracle and Sun Microsystems," Landgrave said. Indeed, .NET isn't entirely unique in its goal to unify access to and communications among applications and devices. What's new is its scope.
In the near term, however, the primary market for .NET will be developers for large corporations, service providers and Internet consumers who need to build large, distributed applications that can scale to thousands of users. "Internet consumers will benefit from a rapid increase in the number and quality of Internet-accessible applications brought to market by the more than three million Windows developers who will migrate to .NET," Landgrave said.
Starbucks Coffee Co. provides an example of how .NET could help IT professionals deploy and manage network services. Seeking to improve network and site management, and reduce management and new feature development costs, Starbucks is now using Commerce Server 2000. Starbucks reports that .NET has improved profiling, targeting and analysis capabilities.
But several hurdles loom on the horizon.
IBM's Sullivan is concerned that .NET will become an "all or nothing" situation. "Different components of the foundation are so tightly woven they have to be implemented together," he said. "Large enterprises won't implement .NET if it won't support devices like the Palm or other operating systems besides Windows. It will require more flexibility."
Landgrave disagrees, however, stating that IT professionals can implement pieces of the .NET platform. "Many will build next-generation network services applications designed to run beside and integrate with their legacy mainframe and client-server applications," he said. "It will take several years for applications to be rewritten to function in a network-services-only configuration. And in many cases it's not necessary to rewrite them."
Some people also have privacy concerns about .NET. Microsoft may have an uphill battle in convincing businesses that their customers' data will be safe on the company's servers, especially in light of recent security lapses.
Microsoft maintains that .NET protect privacy because users will be in control of their data - they will dictate which people and applications can access or alter particular pieces of their personal data, Lorentzen said. The company has also rolled out security technologies such as XML messaging interfaces on various building block services to enable auditing and accountability, and digital rights management for content authorization.
So is .NET a clear, feasible vision or a hallucination? What are the chances that IT leaders will buy into it? And what might give them pause?
Sullivan said the vision is correct - "the devil is in the details and implementation. IT professionals are savvy enough to know they can't afford to get locked in to a platform. They're very careful about rolling out new technologies, so I think it will take awhile to integrate .NET into existing applications."
Aberdeen's Manter adds, however, that many organizations already have a large Microsoft presence. ".NET is one feasible vision of the future of computing," he said. "IT leaders still have time before they need to decide how much they will adopt .NET. The only thing that may give them pause is if a better alternative comes along or they determine there isn't enough value in .NET to justify a change."
Landgrave said .NET is a vision that will take several years to complete. If companies take advantage of the features as they became available and plan their architecture to accommodate future advances, they will find the vision to be both clear and feasible. If they think they can force fit all of their current applications into a set of .NET applications in the next 12 months, "they are hallucinating," he said. "But to be fair to Microsoft, they would be hallucinating if they tried to move to anyone's network services strategy in the next 12 months. Much of the technology and the standards we need to create truly interoperable services won't be implemented or agreed upon for at least another 12-18 months."
Microsoft is betting the company on .NET. "[Their] real challenge is for the IT buyer to get excited about an initiative that will take several years before it delivers clear value," said Manter. "The excitement today is really coming from the development community, which is important. Adopting .NET-ready hardware and software today does not impact your existing architecture - it simply positions companies to take advantage of .NET when they are ready."
Neal Leavitt is president of Leavitt Communications, a Fallbrook, CA-based international marketing communications company with affiliate offices in Paris, France and Hamburg, Germany. Contact him at firstname.lastname@example.org.
Return to: Articles Index