“Web services are infinitely variable”

While Web services technologies and concepts are permeating almost every industry, there are still a lot of misconceptions about the potential of these technologies. Dion Wiggins, research director at Gartner, in an e-mail interview with Srikanth R P, discusses the potential of Web services, how it is affecting modern software development and key issues affecting its adoption

While the concept of Web services is technologically sound, it has been hampered by standardisation for Web services ingredients such as messaging and security. What do you think is the way forward?

The beauty of Web services today is its simplicity, but as anyone who follows the software market would know—simplicity inevitably leads to formidable complexity. And so it is in the case of Web services standards. Vendors (and enterprises) are hard at work adding additional layers to the existing Web services stack to address perceived (and real) issues such as security, transaction management, user interface development, collaborative and peer-to-peer environments, business-to-business (B2B) interactions and more. The emerging stack comes in multiple flavours depending on the vendor, industry association and standards.

There will be recurring attempts to build an entire stack of Web services standards that might satisfy every requirement that an enterprise might foresee, and without exception, these attempts will fail due to the vastness of their scope. ebXML might be one such example. More importantly, Web services standards need to fit within a larger framework that can support comprehensive enterprise requirements. One such framework is depicted here (See The emerging standards stack). At the Gartner Summit India 2003 to be held between July 16-17, we will advise our clients on how to use the depicted e-business standard framework to measure the comprehensiveness of a vendor’s support to standards. But from a customer point of view, do not expect any one vendor to provide everything.

There has been a lot of hype about Web services but there have been very few examples of practical
deployments. What do you think is the scenario with respect to commercial
deployments?

Businesses must be forgiven for their scepticism when IT optimists trumpet the benefits of any given innovation, but avoid acknowledging the inevitable organisational and technical challenges that technology brings as baggage. ‘The next big thing’ is now a derided term as often as it is a promise of innovation.

But this understandable attitude has also had a surprising side effect. This time, ‘the next big thing’—the Web services revolution in the continuum of technology evolution—has, at its heart, the realistic possibility that it will bring fewer challenges than any previous generation. Simplicity is both the Web services’ concept’s promise and strength. Businesses that ignore its potential, or decide to sit out its early stages, will find themselves outpaced by rivals that take advantage of Web services to improve their agility and even to transform themselves into new kinds of enterprises. Because of their inherent ease of use, dynamism and flexibility, Web services will permeate business from the executive suite to the IS ‘clean room’. Enterprises of all sizes will find that Web services offer a more cost-effective way to perform agilely in all environments. In terms of success, Gartner expects that by the end of 2005, new licenses for software that uses Web services standards will represent $21 billion in sales.

What do you think is the scenario in India with respect to understanding of the concept of ‘Web services’?

In India, Web services are not well understood and one of the key attempts of my Web services session at the Gartner Summit India 2003 is to clear the misconceptions and understand the truth within the hype. While there have been some integration projects behind the firewall for some enterprises, bandwidth constraints and lack of understanding have limited their exposure. In contrast to this, many Indian vendors are undertaking Web services projects for their clients from abroad.

What do you think are the opportunities for Indian software service players in the Web services arena?

One factor that has a positive outlook for outsourcing to India is the changing nature of the work itself. By the year 2006, service-oriented architecture (SOA) will be at least partially adopted in more than 60 percent of new, large and systematically oriented application development projects. The proliferation of Web services and SOA is causing software to be developed in smaller units that are easier to map to business processes. These smaller units are also ideal for an offshore environment. Larger projects are harder to manage and even more difficult in an offshore model. Smaller projects using service-oriented development of applications (SODA) are easier to manage, of lower risk and will deliver better value over a shorter timeframe as businesses begin to make the move to the real-time enterprise. With this move to SODA both technologists and business people are talking, working with and understanding processes better. Communication between all parties is in terms of processes and sub-processes, more accurately mapping business needs. Through 2006, service-oriented development will change the way software is built, packaged and sold by more than 80 percent of ISVs. Quite simply, it is becoming easier to outsource than ever before.

How do you see the impact of Web services for enterprise users?

The future of software development and the delivery of business logic are tied to the architectural changes inherent in service orientation and Web services models. Enterprises should begin to experiment with Web services through the use of portals and expect their vendors and internal development initiatives to shift toward service-based development in 2003. Service-oriented architecture (SOA) is redrawing the map of software creation and deployment. From the process of identifying inflows and outflows and demanding that what goes in-between be represented in a standard way, SOA and services-oriented development of applications are making possible a new model of assembly and reuse. The new model is simultaneously more restrictive and more freeing. Project development must consider not only how to assemble and reuse software elements (independent of its code) but also how a project fits into a complex ecology of other software elements in the same enterprise, other elements in other enterprises, and a stretching continuum of past and future projects. Web services are the medium that surround and define this ecology. The standards-based connection culture of these services is accelerating the process of logic distribution, which, in turn, is creating a new model for software creation and distribution. Because software is the basis of how business efforts are fulfilled, the impact on business will be profound.

In your view, what are the applications / application areas wherein the concept of Web services would have a profound impact?

Portals first, and then thick clients, will dominate Web services consumption through 2006. Portals and portal servers will account for more than 60 percent of Web services’ consumer platforms through 2004. High-fidelity clients will account for 50 percent of Web services consumer platforms by 2006. Significant amount of time has been spent on the question of what Web services look like, as though the services can be isolated and welded to a presentational element. The point of Web services is that the reverse is true. Web services are infinitely variable, based on what user or programmatic interface they are linked to. Initially, the shape of these interfaces will usually be a portal element (often through constructs known as portlets), which have no integrative power but serve to present valuable information in a familiar format and a handy location. Such a model for presentation circumvents the thorny issue of how to provide a programmatic interface at two ends of a data transaction, which is equally intelligible to both. Instead, the user simply views data that was previously inaccessible or, at the very least, inconveniently situated in its own system.