Web Services Journal strives to bring the latest information regarding Web Services to you in a variety of ways. In addition to our printed journal and digital edition, we have begun a series of Point/Counterpoint sessions with leading industry executives to determine their viewpoints and issues with various facets of Web services. Here is the first in an ongoing set of conversations regarding the industry.

We spoke to Graham Glass, CEO, chief architect, and founder of The Mind Electric; Derek Ferguson, editor-in-chief of SYS-CON Media's upcoming .NET Developer's Journal and chief technology evangelist for Expand Beyond Corp.; and Mark Colan, lead e-business technology evangelist for IBM.

1. One industry commentator recently said that "Web services are like high school sex. Everyone is talking about doing it, but hardly anyone is...and those that are probably aren't doing it well." Do you agree, or have Web services on the contrary already graduated and gone to college?

Graham: Our experience so far has been that many companies are already using Web services to integrate their internal systems. This comes as no surprise, since their alternatives, such as CORBA and DCOM, are more complex and do not interoperate well. Several large customers have already deployed systems using our GLUE Web services platform. This trend will continue when additional features like security and transactions are standardized.

Derek: Well, at least I feel better about high school now! Seriously, though, I have to agree with the commentator on this one. I work on a team that probably has the best 10 developers in all of Chicago, and not only am I the only one who has ever built or used a Web Service - but I'm the only one who has any interest in doing so. If this is the level of disuse at the top of the development game, I can easily imagine the apathy and ignorance that must exist at more junior levels.

Mark: Well, I can understand why that quote made it into the media, but it really is way off the mark in terms of reflecting the true nature of Web services today! IBM has worked with many customers to help them realize their business goals by way of Web services technologies. The results speak for themselves: have a look www.ibm.com/software/jstart for real-life business case studies. In general, the work that is being done today is either using Web services as a strategy for an EAI infrastructure, which will yield flexibility, scalability, and agility in responding to changing business needs, or in improving existing business partnerships, which can help reduce the cost of doing business via new efficiencies and more timely business-critical information.

Web services interoperability is key to making Web services really work for business and that's why we think WS-I.org is so important. All these specifications around Web services are either in or headed for standards organizations and there is tremendous industry uptake of these technologies in vendor products.

Companies that make aggressive use of these technologies today will realize immediate benefits and begin to outpace their rivals in the future. They'll do this by using Web services technology that makes them more able to jump on emerging business opportunities and reduce costs. Web services make sense in good economies and in bad: you need to be able to make your company more efficient, you need to be able to put your IT efforts into differentiating your company from your competitors, and you need to be able to deal with growth opportunities when they present themselves.

2. .XML has a reputation for sometimes making even simple things difficult - how does your company's approach to Web services deal with that?

Graham: Our GLUE Web services platform allows you to create and consume Web services in pure Java, shielding developers from any details of XML, SOAP and WSDL. There's no reason why Java developers should ever have to deal with XML unless for some reason the application demands it.

Mark: I've never heard this comment made by anyone who has used XML or Web services technologies. On the contrary, our customers agree that these technologies radically simplify integration of heterogeneous environments, which is why they are so popular today in the industry.

As software development tools mature, more and more "wizards" are being created to ease and speed the creation of applications based on these technologies. In fact, it has been possible for some time for the majority of programmers to create XML applications without ever seeing angle brackets.

Of course, if the world's software was all written in one programming language for one platform, the simplest approach would be to exchange data in the native form as expressed in the programming language, and we would not have to worry about differences between processors, operating systems, or middleware. THIS IS NOT GOING TO HAPPEN! The world has never been this simple: we have to deal with integration of existing software investments while using the same software to run our businesses.

XML and Web services technologies simplify cross-platform integration in a general way by focusing on a universal data representation which can also be used to effect actions for process-level integration. By focusing on simple, standard message exchange, we benefit both by hiding the complexities of the wide variety of programming and operating system environments and by making software that, by way of loose coupling, is less likely to break because of changes to hard-coded APIs and file formats. In this way XML and Web services are really a breath of fresh air -- and people who use them breathe a sigh of relief, at last, in standards-based, platform-neutral exchanges.

Derek: .NET deals with the complexities of XML in the best way possible -- it conceals them. If you want to create a Web Service using Visual Studio .NET, you can do so without ever even seeing a single node of XML -- if that is your desire. Simply click a few buttons and write a few lines of code -- the same kind of code you would write even if you were creating software for use on a single machine -- and presto, you've built your first industry-standard Web Service. The same is true is you wish to consume a Web Service using Visual Studio .NET.

On the other hand, Visual Studio .NET provides complete flexibility for developers who are more advanced. If you really want to delve into the internals of how the SOAP messages are being constructed, parsed, or interpreted, .NET provides you with all of the raw XML tools that you need in order to go about performing this work in the most productive fashion possible.

3. Back in April*, Daniel Sholler of Meta Group claimed that "Global 2000 organizations will have heterogeneous application environments indefinitely, but .NET share will increase to 30 percent of enterprise development projects as J2EE use stabilizes at 40 percent by 2004." Do you agree?

Derek: I think Mr. Sholler has seriously under-estimated the appeal of .NET in trying to predict its growth, so I have to disagree on this point. I suspect that, within the next 12 months, you are going to see a dramatic shift in favor of .NET over J2EE for a couple of reasons. To begin with, .NET is a much easier tool than J2EE for most developers to use in creating and/or consuming Web Services. For more information on this, please see my previous answer.

More important, perhaps, is the fact that .NET builds upon the massive, and extremely well-entrenched current base of Microsoft technologies. On the one hand, you have to consider the fact that .NET is completely compatible with COM -- which means that, where developers choosing to migrate to Java will have to abandon a great deal of their legacy Windows code, developers embracing .NET have an almost completely painless migration path provided for them.

On the other hand, .NET supports languages -- such as Visual C++ and Visual Basic -- which are already used by many times more developers than Java. The average Visual Basic developer can become proficient enough in VB.NET to be productive within a matter of weeks, which is a far cry from how long it would take them to retrain as Java developers. Furthermore, once they have learned either VB.NET or C#, they can then leverage their skills in building any kind of web or device-based application via the miracle of ASP.NET and the .NET Compact Framework. This is infinitely superior to the severely limited subsets of Java (such as MIDP) that are available for building device-based applications with Java or the awkward "tag" extensions (such as Java Server Pages) that attempt to leverage Java for use in constructing Web Sites.

By 2004, I look for .NET usage to stand at better than 50% of all organizations and J2EE to have fallen to less than 25%.

Graham: I think that the Microsoft market share will be higher if they aggressively port the CLR and .NET framework to other operating systems. There doesn't seem to be any technical or financial reason that they couldn't do this, since they've already put a lot of time and effort into supporting a multiplatform runtime and it could be sold with a similar licensing fee to their current windows platform. Their mantra would then become "develop on Windows, deploy everywhere."

Mark: Let's face it, it's completely naïve to think that any one programming language, operating system, database, or packaged application is going to exist to the exclusion of competitors. It might be nice for those vendors to dream that everyone will use their platforms, but it just isn't going to happen. GET REAL!

Organizations will continue to have heterogeneous application platforms, because we have always had that in the past, and it is neither cost-effective nor safe to rip-and-replace code, learn a new programming language, and move to a new environment unless there is a very good ROI. Instead, we need a way of simplifying the integration of our existing software investments. Java, XML, and Web services technologies are a potent combination that make integration easy, fast, and reliable, and provide a solution that offers agility, flexibility, and cost savings in the bargain.

J2EE has the advantage of being a robust, mature, and proven platform for implementing e-business because it has been released and used in the real world for more than five years. It has a wealth of features that help programmers be more productive, and it allows companies to easily move between computer platforms as their business grows. It uses one well-considered object model that hides a lot of complexities in an elegant and powerful fashion. And it advanced quickly to a position of popularity because of contributions of substantial technologies from IBM, not to mention many other companies in the industry. The next major release of the J2EE platform, version 1.3, is now supported by IBM software.

The benefits of J2EE are obvious to our customers, and IBM is confident that its use will continue to enjoy rapid growth, because it is a proven approach that can work with any hardware or operating system. How well .NET will fare when the server product is finally released and supported is anyone's guess. But in any case the numbers you cite are questionable in one respect: they add up only to 70%, and it's hard to imagine that other integrating platform options, even added together, will ever have 30% market share.

While it is interesting to compare the take up of J2EE and .NET, it is important to note that we can talk about Web services without mentioning either one of them. .NET and J2EE are platform choices. If I build my IT systems under the assumption that I will only talk to partners or customers that use .NET or J2EE, I will miss out on a lot of business. If you use Web services correctly, it simply does not matter if you are invoking a Web service application that was built using J2EE, or .NET, or COBOL, Perl, Python, and anything else, for that matter. IBM loves Java, as everybody knows, and thinks it is an excellent choice for building and invoking Web services. Remember, though that Web services as a business connection technology transcends any particular programming language or environment.