Product Unfocus

Vendors that put customers and services at the center, rather than products, are favored by product reviewers this year

By Michael L. Gonzales, Nelson King, Stewart McKie, Mark Smith

YEAR IN REVIEWS EXCERPTS FROM x.0 PRODUCT RELEASE REVIEWS IN 2000

Microsoft SQL Server 2000 SQL Server 2000 has radically changed the face of enterprise database applications by including many new features and services that formerly could be added only by expensive third-party products that too often require experts to run them. - Geoffrey Hollander (upcoming issue) ProcessModel 4.0 The main advantage of this tool compared to its competitors is that ProcessModel seamlessly combines flowcharting and process modeling with realtime animation and simulation of process executions. - Mark M. Davydov, Oct. 20, 2000 SilverStream 3.0 By its almost seamless integration of HTML and Java2 Enterprise Edition plus the integrated application server, SilverStream goes a long way toward making complex, multitier Web applications less difficult to develop. - Nelson King, Sept. 29, 2000 ESRI Spatial Analyst 2.0 ESRI's supplemental products, Splus and 3D Analyst, should be standard parts of Spatial Analyst. Coupling these strengths with tools like ModelBuilder creates an unparalleled, formidable product, the clear industry leader and benchmark for any future competitors. To offer anything short of that is to play Russian roulette. - Michael L. Gonzales, Aug. 18, 2000 WorkWise Data Agent Server Of course you can achieve much of what DAS does using database triggers and stored procedures or any of a number of scheduling products. But DAS just makes it easier by bundling everything together. - Stewart McKie, June 5, 2000 Radiant Logic RadiantOne 1.0 In some ways, using RadiantOne is easier than trying to describe it - often true with innovative products. RadiantOne should become a must-have program for network and database administrators working on applications in an enterprise environment with multiple heterogeneous directories. - Nelson King, March 1, 2000

The year 2000 will stand out as the year the IT industry seriously began to make the transition from a product-based to a service-based economy. The trend began in the late 1990s with self service via corporate intranets to employees and via collaborative extranets to business partners. Then a growing number of application service providers (ASPs) made it possible to rent shrink-wrapped applications hosted on their Web servers. And now it seems clear that everybody from emperors of shrink-wrap like Microsoft to emerging dot-com startups are focusing on software-as-service as they reposition themselves as Internet business service providers (BSPs).

From Product to Service

For software vendors, the advantages of software-as-service are clear: Minimized production and distribution costs, a more manageable revenue stream, and a closer connection to customers with realtime feedback. The model also obviously benefits customers: Minimized acquisition and installation costs, one-click user signup, more predictable cost-cashflow, and the end of resource-draining product-upgrade cycles. IT departments can focus on more value-added activities and users get access to applications on an anytime, anywhere basis.

But it may not be desirable, practicable, or possible to switch every application product to an application service. Which is why there will be a transition phase during which software users and vendors will assume that business solutions will not be delivered either as a product or a service but as a loosely coupled combination of both, whereby internally installed, packaged software is combined with externally sourced Web services. Microsoft is reorganizing around this vision through its .Net initiative, and ERP vendors, such as SAP with its MySAP.com offering, are already raising the kimono on how this future of loose coupling could look.

The rapidly expanding range of XML-based document schemas will provide the document-level APIs for cross-application conversations and collaborations. The development and use of Web services will be given a shot in the arm by two recently announced initiatives: universal description, discovery, and integration (UDDI) and Web services description language (WSDL).

One area that will benefit greatly from UDDI is supply chain management via the Internet, making the assembly of "virtual" supply chains a lot easier. Eventually UDDI may be able to partially or wholly automate finding and connecting to services programmatically, something that will be assisted by the development of WSDL.

If service providers make use of WSDL as a means to request their service or respond with a service deliverable, the whole concept of loosely coupled applications could well become a reality.

In the transition from products to services, the year 2001 will prove to be a watershed year. Packaged software will begin to disappear from the shelves; licenses will become subscriptions; software vendors will continue to morph into service providers; and IT departments will face up to a new challenge: how to assemble and manage a new generation of loosely coupled business solutions.

Stewart McKie (www.cfoinfo.com) is an independent e-business consultant and technology writer.

---

Pulled Into the .Net

Some years there just don't seem to be any clear favorite products or companies. Most of 2000 seemed that way to me. Then along came Microsoft's September announcement of its .Net initiative and I've been pulled kicking and screaming to consider it the most significant development this year.

I'm sure it's not very popular to write favorably about Microsoft and .Net. For obvious reasons, industry folk tend to dance around anything Microsoft does these days. For another, most people would say that .Net is vaporware, and who could be serious about vaporware?

Regardless, I've come to think it is myopic - no, stupid - not to consider .Net important. For one thing, it validates the ASP model. It pushes XML into the very forefront of Internet development. It also validates most of the concepts behind Java. It puts a very large company on a path that by no means is ensured of success; in fact might turn out to be totally in the wrong direction. By the same token, simply because it is Microsoft, it might speed the Internet in a direction it was already inclined to go.

Microsoft needs to do a better job of explaining what .Net is; but in a nutshell it's a framework for developing applications and services that plugs just about every product made by Microsoft into an Internet-centric system. Instead of selling packaged products, Microsoft (and partners) will move toward licensing or renting applications and services delivered via the Internet - the ASP model. The technology that keeps the whole thing together will be XML (in its many incarnations) as the protocol and data language, and a universal runtime engine (shades of the Java Virtual Machine) that will work across programming languages and support a much-improved third version of Microsoft's Component Object Model (COM).

.Net is full of sleepers, my favorite being the new programming language C# (C-sharp), which will get attention soon enough. This isn't just Microsoft getting back at Sun Microsystems' Java; it's also a worthy successor to C and C++. The person most responsible for it is Anders Hejlsberg, who was also responsible for Borland's extremely successful TurboPascal and Delphi. From what I've seen, C# does it again. It incorporates many Java-like features but retains some advanced C++ features that programmers will appreciate.

I know about Microsoft delivery schedules - as tenuous as methane over a swamp. Microsoft says some elements of .Net won't be complete until 2002, or later. Such long-range projections make everybody skeptical. However, from what I've used myself and heard from the developer community, the pieces of .Net that are currently almost ready are surprisingly robust. This is especially true of components in Visual Studio .Net. The emphasis on developer tools is a major reason to believe that Microsoft is willing to build .Net the right way, from the ground up.

Microsoft has always had a penchant for the grand design. Like IBM before it, it's constantly attempting to explain to the world some kind of Hegelian master plan that will bind all its vast undertakings into a single architecture. I say good luck, but Microsoft has a lot of guts to take on such a fundamental change in direction as .Net. It'll get my respect if and when the company pulls it off, but for now I must admit it's the most interesting and important move I've seen in 2000.

Stewart McKie (www.cfoinfo.com) is an independent e-business consultant and technology writer.

Nelson King (nelsonking@earthlink.net) has written nine books on database application programming and spends much of his time in the trenches of enterprise software development.