Next big hurdle a matter of semantics for Web services

The current tough times of IT are forcing an unusual consensus of common sense among vendors performing in the Web services show, but will it last?

It's not everyday you see representatives from bitter rivals Microsoft and Sun Microsystems hug in public. It's just not done. Yet, in a rare display of orchestrated solidarity, IT vendors are more than keen at the moment to be to be seen to rally around the single flag of Web services, and exchanging the occasional hug. Sure, there are still a few issues to sort out they all agree, but the fundamentals have been established -- and they work. Or at least that was the message coming out of last week's "Evolve 2002", the Australian Web services show-and-tell conference, which the Australian R&D think tank Distributed Systems Technology Centre (DSTC) staged.

In theory, Web services are the next big thing when it comes to making a real difference in the enterprise computing world. The sell is that organisations will be able to deploy their information resources, products and services anywhere in the world in a seamless and efficient manner, thus reducing both the overheads and obstacles to doing business. Business will all be done over the Web, at any place, any time on any thing -- be it authorising air tickets or stock trades to ordering dinner or booking the car in for a service.

Twin nasties

To date, the nastiest hurdle to overcome in making Web services perform in the wild has been the twin-headed beast of security and interoperability. At Evolve 2002 there appeared to be a clearly visible consensus among those presenting that while there was still a lot of programming work to be done, solutions to both security and interoperability issues existed at proof of concept stage.

Delegates heard from Kelvin Lawrence, IBM's CTO for dynamic e-business technology, that Big Blue and Microsoft have combined to put out a Web services security roadmap. Lawrence said both Microsoft and IBM jointly describe this map as "a comprehensive Web service security model that supports, integrates and unifies several popular security models, mechanisms and technologies (including both symmetric and public key technologies) in a way that enables a variety of systems to interoperate securely in a platform- and language-neutral manner".

In theory this means that, at least until product and code is released, tested and evaluated, the Web services security issue has a solution. Or at least the vendors seem to think so, which makes for a pleasant change to their usual squabbling.

Interoperability for applications and platforms also appears well on its way to becoming a functional reality, according to the vendors present at Evolve 2002. Although there is still much work to do and code to cut, it will happen and it must happen.

As if to prove the point, a representative selection of six stakeholders took to the stage all at once to take questions from the floor. They were: Janet Daley, head of communications for the World Wide Web Consortium (W3C); Jim Green, CTO WebMethods; Kelvin Lawrence, IBM's CTO for dynamic e-business technology; Simon Phipps, chief technical evangelist for Sun Microsytems UK; Sinisa Zimek, director of technology architecture, SAP, and last but not least Paul Cotton, program manager XML standards, Microsoft Canada. It was such a convincing sight that one delegate was heard to compare it to "the last supper". Then came the questions.

"Why was http the best way to go in terms of protocol?" asked Sun's Simon Phipps, using a written question solicited from the audience.

"It's everywhere, it's here to stay and it's a foundation," said W3C's Daley.

"HTTP is first amongst equals -- SOAP [Simple Object access Protocol] will let you do the bindings to other protocols; so if there are better protocols you can do that," replied Microsoft's Cotton.

"There are lots of ways to invent a better protocol," offered WebMethods' Green, clearly undaunted by the premise of the question. "It's not the best, but pragmatically everyone will want to use port 80: no one will want to reset firewalls, so people will use it regardless," Green continued.

"What are the missing pieces for Web services to become the ultimate future application?" asked Sun's Phipps.

"Just because Web services is cool, it doesn't mean that the rules of physics or robust computing don't apply. Even when you have interoperability and security you will still have issues with 'semantic meaning'. If you can solve that, you will have a huge leap forward." Green went on to explain that the meaning of one record or name within an application (current, future or legacy) may not mean the same thing on another application, citing the example of something that calls itself a "contact".

Microsoft's Paul Cotton's answer to the question: "Plumbing and customer pressure for interoperability."

To WC3's Daley the missing pieces are: "More code less hype. Web Services needs to be about serving Fortune 200 companies as much as mid-size institutions and it has to be scalable."

"You need a some general policy framework that is common and agreed upon," said IBM's Lawrence.

"Where does the cost model fit in?" asked a member of the audience.

Silence.

"I don't know, that's a good question," offered Microsoft's Cotton.

More silence.

So far so good. The vendor consensus machine appeared to be working well, even with the occasional qualification. Just a few more technical questions, a little more hugging, and the panel could call it a day and retire some well earned beers in the hotel lobby.

Somehow, out of the morass of technical detail and acronyms, came strangely familiar sounds.

"By the way, we ship more operating system products than you two put together," responded IBM's Lawrence to some sort of veiled technical jibes from Sun and Microsoft.

"And you're proud of that?" retorted Sun evangelist Phipps.

"It's called heritage!" countered Lawrence, as a sound very similar to that of a pencil being snapped in half came from the direction of Daley. Things were getting back to normal.

Outside, after final hugs for the cameras by Sun and Microsoft, IBM's Lawrence was more than comfortable to admit that the current tough times for the information technology industry was part of the force driving vendors together.

IBM conservatively expects to have Web services deployments happening by early 2004 -- and the vision is that the great vendor love-in will continue indefinitely, semantics or otherwise. In Australia, the Australian Bureau of Statistics is understood to have laid the foundations for a Web services system for its business process management with parts expected to be operational by early next year.

What's in a name?

Semantics is still an issue according to Kelvin Lawrence, IBM's CTO for dynamic e-business technology, but, like others before it, not an insurmountable one if cooperation continues.

Attempting to explain the intricacies of what happens with semantics and Web services, a local delegate offered a local insight to a local journalist.

"Look, it's like beer. In New South Wales you have schooners and middies. In Victoria you have pots, but a pot is [the same size as] a middy, but you don't have pots in NSW. But you do have them in Queensland. But now a twistie is also a middy. Then you have longnecks and stubbies.