What Web Services Are NOT (4/4) - exploring XML
What Web Services Are NOT
Now let's talk about directories. Do NOT buy into the story that there will be one (or a few) universal directories listing all available Web services and that you will be able to go and pick your favorite one to interact with an enterprise. While this may work for toy applications like checking the weather, this is far from reality for applications like the inventory management example. In fact, most enterprises may not want open communication into their infrastructure available in a public directory. So while the concept is appealing, unfortunately it works practically only for consumer focused simple services.
Just having a Web service interface to an application does NOT make it robust, secure, or scaleable. You still need to implement the application's business logic on a robust platform that provides transaction safety, clustering support, failover and handles multi-user concurrency. Once again, true enterprise platforms will enable this to be done without burdening the application creator with additional coding responsibility or architectural planning. Hence, the application creator needs to focus only on the business logic and leave the underlying plumbing to be taken care of automatically by the platform. (For whitepapers on how this can be done, see the "resources" section of www.instantis.com.)
With every software company claiming to be a Web services player the landscape has already become quite confusing in terms of the real underlying technology, and real business value that Web services can provide. To reiterate, the real business benefit is in creating an efficient way for interactions and transactions to take place either internal to organizations or between business partners wherever the motivation for doing such business already exists. Web services are not going to create new business rationale, rather, they will be enablers for implementation. As for real technology breakthroughs, robust, scalable platforms that offer a "native" and simple way to create and deploy Web applications and seamlessly expose them as Web services provide the greatest value - both near-term, as well as for "future proofing" for the enterprise.
About the author
Sriram Rajaraman is the Vice President of Engineering at Instantis and brings 15 years of experience building large-scale systems software to the company. Prior to Instantis, he was Chief Architect of a key component of IBM's Websphere Application Server at IBM in Pittsburgh. He led a team to develop the administration infrastructure for Websphere. Prior to that, he was at Transarc, where he was Chief Architect of the TXSeries application server product. Before joining Transarc, he was a member of a team developing an object-oriented operating system at Intel Corporation in Portland, Oregon. Sriram has a M.S. in Computer Science from the University of Wisconsin at Madison and a B.S. degree in Electrical Engineering from the Indian Institute of Technology in Madras.
Produced by Michael Claßen
Created: Feb 16, 2002
Revised: Feb 16, 2002